news.mit.edu.rss.xml - sfeed_tests - sfeed tests and RSS and Atom files
(HTM) git clone git://git.codemadness.org/sfeed_tests
(DIR) Log
(DIR) Files
(DIR) Refs
(DIR) README
(DIR) LICENSE
---
news.mit.edu.rss.xml (399489B)
---
1 <?xml version="1.0" encoding="utf-8"?>
2 <rss xmlns:atom="http://www.w3.org/2005/Atom" xmlns:dc="http://purl.org/dc/elements/1.1/" xmlns:media="http://search.yahoo.com/mrss/" xmlns:content="http://purl.org/rss/1.0/modules/content/" version="2.0" xml:base="https://news.mit.edu">
3 <channel>
4 <title>MIT News</title>
5 <link>https://news.mit.edu/rss/research</link>
6 <atom:link href="https://news.mit.edu/rss/research" rel="self" type="application/rss+xml"/>
7 <description>MIT News is dedicated to communicating to the media and the public the news and achievements of the students, faculty, staff and the greater MIT community.</description>
8 <language>en</language>
9
10 <lastBuildDate>Tue, 27 Oct 2020 11:00:00 -0400</lastBuildDate>
11 <item>
12 <title>Study helps explain why motivation to learn declines with age</title>
13 <link>https://news.mit.edu/2020/why-learn-motivate-age-decline-1027</link>
14 <description>Research on mice suggests aging affects a brain circuit critical for learning to make some types of decisions.</description>
15 <pubDate>Tue, 27 Oct 2020 11:00:00 -0400</pubDate>
16 <guid isPermaLink="true">https://news.mit.edu/2020/why-learn-motivate-age-decline-1027</guid>
17 <dc:creator>Anne Trafton | MIT News Office</dc:creator>
18 <content:encoded><p>As people age, they often lose their motivation to learn new things or engage in everyday activities. In a study of mice, MIT neuroscientists have now identified a brain circuit that is critical for maintaining this kind of motivation.</p>
19
20 <p>This circuit is particularly important for learning to make decisions that require evaluating the cost and reward that come with a particular action. The researchers showed that they could boost older mice’s motivation to engage in this type of learning by reactivating this circuit, and they could also decrease motivation by suppressing the circuit.</p>
21
22 <p>“As we age, it’s harder to have a get-up-and-go attitude toward things,” says Ann Graybiel, an Institute Professor at MIT and member of the McGovern Institute for Brain Research. “This get-up-and-go, or engagement, is important for our social well-being and for learning — it’s tough to learn if you aren’t attending and engaged.”</p>
23
24 <p>Graybiel is the senior author of the study, which appears today in <em>Cell</em>. The paper’s lead authors are Alexander Friedman, a former MIT research scientist who is now an assistant professor at the University of Texas at El Paso, and Emily Hueske, an MIT research scientist.</p>
25
26 <p><strong>Evaluating cost and benefit</strong></p>
27
28 <p>The striatum is part of the basal ganglia — a collection of brain centers linked to habit formation, control of voluntary movement, emotion, and addiction. For several decades, Graybiel’s lab has been studying clusters of cells called striosomes, which are distributed throughout the striatum. Graybiel discovered striosomes many years ago, but their function had remained mysterious, in part because they are so small and deep within the brain that it is difficult to image them with functional magnetic resonance imaging (fMRI).</p>
29
30 <p>In recent years, Friedman, Graybiel, and colleagues including MIT research fellow Ken-ichi Amemori have discovered that striosomes <a href="https://news.mit.edu/2015/brain-circuit-controls-decisions-causing-anxiety-0528">play an important role</a> in a type of decision-making known as approach-avoidance conflict. These decisions involve choosing whether to take the good with the bad — or to avoid both — when given options that have both positive and negative elements. An example of this kind of decision is having to choose whether to take a job that pays more but forces a move away from family and friends. Such decisions often provoke great anxiety.</p>
31
32 <p>In a <a href="https://news.mit.edu/2016/neural-connections-linked-emotional-decision-making-0919">related study</a>, Graybiel’s lab found that striosomes connect to cells of the substantia nigra, one of the brain’s major dopamine-producing centers. These studies led the researchers to hypothesize that striosomes may be acting as a gatekeeper that absorbs sensory and emotional information coming from the cortex and integrates it to produce a decision on how to act. These actions can then be invigorated by the dopamine-producing cells.</p>
33
34 <p>The researchers later discovered that chronic stress has a major impact on this circuit and on this kind of emotional decision-making. In a <a href="https://news.mit.edu/2017/stress-can-lead-risky-decisions-1116">2017 study</a> performed in rats and mice, they showed that stressed animals were far more likely to choose high-risk, high-payoff options, but that they could block this effect by manipulating the circuit.</p>
35
36 <p>In the new <em>Cell</em> study, the researchers set out to investigate what happens in striosomes as mice learn how to make these kinds of decisions. To do that, they measured and analyzed the activity of striosomes as mice learned to choose between positive and negative outcomes.</p>
37
38 <p>During the experiments, the mice heard two different tones, one of which was accompanied by a reward (sugar water), and another that was paired with a mildly aversive stimulus (bright light). The mice gradually learned that if they licked a spout more when they heard the first tone, they would get more of the sugar water, and if they licked less during the second, the light would not be as bright.</p>
39
40 <p>Learning to perform this kind of task requires assigning value to each cost and each reward. The researchers found that as the mice learned the task, striosomes showed higher activity than other parts of the striatum, and that this activity correlated with the mice’s behavioral responses to both of the tones. This suggests that striosomes could be critical for assigning subjective value to a particular outcome.</p>
41
42 <p>“In order to survive, in order to do whatever you are doing, you constantly need to be able to learn. You need to learn what is good for you, and what is bad for you,” Friedman says.</p>
43
44 <p>“A person, or this case a mouse, may value a reward so highly that the risk of experiencing a possible cost is overwhelmed, while another may wish to avoid the cost to the exclusion of all rewards. And these may result in reward-driven learning in some and cost-driven learning in others,” Hueske says.</p>
45
46 <p>The researchers found that inhibitory neurons that relay signals from the prefrontal cortex help striosomes to enhance their signal-to-noise ratio, which helps to generate the strong signals that are seen when the mice evaluate a high-cost or high-reward option.</p>
47
48 <p><strong>Loss of motivation</strong></p>
49
50 <p>Next, the researchers found that in older mice (between 13 and 21 months, roughly equivalent to people in their 60s and older), the mice’s engagement in learning this type of cost-benefit analysis went down. At the same time, their striosomal activity declined compared to that of younger mice. The researchers found a similar loss of motivation in a mouse model of Huntington’s disease, a neurodegenerative disorder that affects the striatum and its striosomes.</p>
51
52 <p>When the researchers used genetically targeted drugs to boost activity in the striosomes, they found that the mice became more engaged in performance of the task. Conversely, suppressing striosomal activity led to disengagement.</p>
53
54 <p>In addition to normal age-related decline, many mental health disorders can skew the ability to evaluate the costs and rewards of an action, from anxiety and depression to conditions such as PTSD. For example, a depressed person may undervalue potentially rewarding experiences, while someone suffering from addiction may overvalue drugs but undervalue things like their job or their family.</p>
55
56 <p>The researchers are now working on possible drug treatments that could stimulate this circuit, and they suggest that training patients to enhance activity in this circuit through biofeedback could offer another potential way to improve their cost-benefit evaluations.</p>
57
58 <p>“If you could pinpoint a mechanism which is underlying the subjective evaluation of reward and cost, and use a modern technique that could manipulate it, either psychiatrically or with biofeedback, patients may be able to activate their circuits correctly,” Friedman says.</p>
59
60 <p>The research was funded by the CHDI Foundation, the Saks Kavanaugh Foundation, the National Institutes of Health, the Nancy Lurie Marks Family Foundation, the Bachmann-Strauss Dystonia and Parkinson’s Foundation, the William N. and Bernice E. Bumpus Foundation, the Simons Center for the Social Brain, the Kristin R. Pressman and Jessica J. Pourian ’13 Fund, Michael Stiefel, and Robert Buxton.</p>
61 </content:encoded>
62 <media:content url="https://news.mit.edu/sites/default/files/styles/news_article__cover_image__original/public/images/202010/MIT-Aging-Motivation-01.gif?itok=mt_HGA2Y" medium="image" type="image/jpeg" width="390" height="260">
63 <media:description type="plain">“As we age, it's harder to have a get-up-and-go attitude toward things,” says study author Ann Graybiel.</media:description>
64 <media:credit>Image: Christine Daniloff, MIT</media:credit>
65 </media:content>
66 </item>
67 <item>
68 <title>Silencing gene expression to cure complex diseases</title>
69 <link>https://news.mit.edu/2020/immuneering-gene-expression-1026</link>
70 <description>Immuneering uses bioinformatics to develop new medicines while also helping large pharmaceutical companies improve their treatments.</description>
71 <pubDate>Mon, 26 Oct 2020 00:00:00 -0400</pubDate>
72 <guid isPermaLink="true">https://news.mit.edu/2020/immuneering-gene-expression-1026</guid>
73 <dc:creator>Zach Winn | MIT News Office</dc:creator>
74 <content:encoded><p>Many people think of new medicines as bullets, and in the pharmaceutical industry, frequently used terms like “targets” and “hits” reinforce that idea. Immuneering co-founder and CEO Ben Zeskind ’03, PhD ’06 prefers a different analogy.</p>
75
76 <p>His company, which specializes in bioinformatics and computational biology, sees many effective drugs more like noise-canceling headphones.</p>
77
78 <p>Rather than focusing on the DNA and proteins involved in a disease, Immuneering focuses on disease-associated gene signaling and expression data. The company is trying to cancel out those signals like a pair of headphones blocks out unwanted background noise.</p>
79
80 <p>The approach is guided by Immuneering’s decade-plus of experience helping large pharmaceutical companies understand the biological mechanisms behind some of their most successful medicines.</p>
81
82 <p>“We started noticing some common patterns in terms of how these very successful drugs were working, and eventually we realized we could use these insights to create a platform that would let us identify new medicine,” Zeskind says. “[The idea is] to not just make existing medicines work better but also to create entirely new medicines that work better than anything that has come before.”</p>
83
84 <p>In keeping with that idea, Immuneering is currently developing a bold pipeline of drugs aimed at some of the most deadly forms of cancer, in addition to other complex diseases that have proven difficult to treat, like Alzheimer’s. The company’s lead drug candidate, which targets a protein signaling pathway associated with many human cancers, will begin clinical trials within the year.</p>
85
86 <p>It’s the first of what Immuneering hopes will be a number of clinical trials enabled by what the company calls its “disease-canceling technology,” which analyzes the gene expression data of diseases and uses computational models to identify small-molecule compounds likely to bind to disease pathways and silence them.</p>
87
88 <p>“Our most advanced candidates go after the RAS-RAF-MEK [protein] pathway,” Zeskind explains. “This is a pathway that’s activated in about half of all human cancers. This pathway is incredibly important in a number of the most serious cancers: pancreatic, colorectal, melanoma, lung cancer — a lot of the cancers that have proven tougher to go after. We believe this is one of the largest unsolved problems in human cancer.”</p>
89
90 <p><strong>A good foundation</strong></p>
91
92 <p>As an undergraduate, Zeskind participated in the MIT $100K Entrepreneurship Competition (the $50K back then) and helped organize some of the MIT Enterprise Forum’s events around entrepreneurship.</p>
93
94 <p>“MIT has a unique culture around entrepreneurship,” Zeskind says. “There aren’t many organizations that encourage it and celebrate it the way MIT does. Also, the philosophy of the biological engineering department, of taking problems in biology and analyzing them quantitatively and systematically using principles of engineering, that philosophy really drives our company today.”</p>
95
96 <p>Although his PhD didn’t focus on bioinformatics, Zeskind’s coursework did involve some computational analysis and offered a primer on oncology. One course in particular, taught by Doug Lauffenburger, the Ford Professor of Biological Engineering, Chemical Engineering, and Biology, resonated with him. The class tasked students with uncovering some of the mechanisms of the interleukin-2 (IL-2) protein, a molecule found in the immune system that’s known to severely limit tumor growth in a small percentage of people with certain cancers.</p>
97
98 <p>After Zeskind earned his MBA at Harvard Business School in 2008, he returned to MIT’s campus to talk to Lauffenburger about his idea for a company that would decipher the reasons for IL-2’s success in certain patients. Lauffenburger would go on to join Immuneering’s advisory board.</p>
99
100 <p>Of course, due to the financial crisis of 2007-08, that proved to be difficult timing for launching a startup. Without easy access to capital, Zeskind approached pharmaceutical companies to show them some of the insights his team had gained on IL-2. The companies weren’t interested in IL-2, but they were intrigued by Immuneering’s process for uncovering the way it worked.</p>
101
102 <p>“At first we thought, ‘We just spent a year figuring out IL-2 and now we have to start from scratch,’” Zeskind recalls. “But then we realized it would be easier the second time around, and that was a real turning point because we realized the company wasn’t about that specific medicine, it was about using data to figure out mechanism.”</p>
103
104 <p>In one of the company’s first projects, Immuneering uncovered some of the mechanisms behind an early cancer immunotherapy developed by Bristol-Myers Squibb. In another, they studied the workings of Teva Pharmaceuticals’ drug for multiple sclerosis.</p>
105
106 <p>As Immuneering continued working on successful drugs, they began to notice some counterintuitive patterns.</p>
107
108 <p>“A lot of the conventional wisdom is to focus on DNA,” Zeskind says. “But what we saw over and over across many different projects was that transcriptomics, or which genes are turned on when — something you measure through RNA levels — was the thing that was most frequently informative about how a drug was working. That ran counter to conventional wisdom.”</p>
109
110 <p>In 2018, as Immuneering continued helping companies appreciate that idea in drugs that were already working, it decided to start developing medicines designed from the start to go after disease signals.</p>
111
112 <p>Today the company has drug pipelines focused around oncology, immune-oncology, and neuroscience. Zeskind says its disease-canceling technology allows Immuneering to launch new drug programs about twice as fast and with about half the capital as other drug development programs.</p>
113
114 <p>“As long as we have a good gene-expression signature from human patient data for a particular disease, we’ll find targets and biological insights that let us go after them in new ways,” he says. “It’s a systematic, quantitative, efficient way to get those biological insights compared to a more traditional process, which involves a lot of trial and error.”</p>
115
116 <p><strong>An inspired path</strong></p>
117
118 <p>Even as Immuneering advances its drug pipelines, its bioinformatics services business continues to grow. Zeskind attributes that success to the company’s employees, about half of which are MIT alumni — the continuation of trend that began in the early days of the company, when Immuneering was mostly made up of recent MIT PhD graduates and postdocs.</p>
119
120 <p>“We were sort of the Navy Seals of bioinformatics, if you will,” Zeskind says. “We’d come in with a small but incredibly well-trained team that knew how to make the most of the data they had available.”</p>
121
122 <p>In fact, it’s not lost on Zeskind that his analogy of drugs as noise-canceling headphones has a distinctively MIT spin: He was inspired by longtime MIT professor and Bose Corporation founder Amar Bose.</p>
123
124 <p>And Zeskind’s attraction to MIT came long before he ever stepped foot on campus. Growing up, his father, Dale Zeskind ’76, SM ’76, encouraged Ben and his sister Julie ’01, SM ’02 to attend MIT.</p>
125
126 <p>Unfortunately, Dale passed away recently after a battle with cancer. But his influence, which included helping to spark a passion for entrepreneurship in his son, is still being felt. Other members of Immuneering’s small team have also lost parents to cancer, adding a personal touch to the work they do every day.</p>
127
128 <p>“Especially in the early days, people were taking more risk [joining us over] a large pharma company, but they were having a bigger impact,” Zeskind says. “It’s all about the work: looking at these successful drugs and figuring out why they’re better and seeing if we can improve them.”</p>
129
130 <p>Indeed, even as Immuneering’s business model has evolved over the last 12 years, the company has never wavered in its larger mission.</p>
131
132 <p>“There’s been a ton of great progress in medicine, but when someone gets a cancer diagnosis, it’s still, more likely than not, very bad news,” Zeskind says. “It’s a real unsolved problem. So by taking a counterintuitive approach and using data, we’re really focused on bringing forward medicines that can have the kind of durable responses that inspired us all those years ago with IL-2. We’re really excited about the impact the medicines we’re developing are going to have.”</p>
133 </content:encoded>
134 <media:content url="https://news.mit.edu/sites/default/files/styles/news_article__cover_image__original/public/images/202010/MIT-Immuneering-01-PRESS.jpg?itok=OJSnYqdH" medium="image" type="image/jpeg" width="390" height="260">
135 <media:description type="plain">Immuneering develops drugs to bind to disease pathways and silence them.</media:description>
136 <media:credit>Image: courtesy of Immuneering</media:credit>
137 </media:content>
138 </item>
139 <item>
140 <title>Yogesh Surendranath wants to decarbonize our energy systems</title>
141 <link>https://news.mit.edu/2020/yogesh-surendranath-energy-1023</link>
142 <description>By developing novel electrochemical reactions, he hopes to find new ways to generate energy and reduce greenhouse gases.</description>
143 <pubDate>Fri, 23 Oct 2020 00:00:00 -0400</pubDate>
144 <guid isPermaLink="true">https://news.mit.edu/2020/yogesh-surendranath-energy-1023</guid>
145 <dc:creator>Anne Trafton | MIT News Office</dc:creator>
146 <content:encoded><p>Electricity plays many roles in our lives, from lighting our homes to powering the technology and appliances we rely on every day. Electricity can also have a major impact at the molecular scale, by powering chemical reactions that generate useful products.</p>
147
148 <p>Working at <a href="https://news.mit.edu/2015/inexpensive-new-catalysts-fine-tuned-0916">that molecular level</a>, MIT chemistry professor Yogesh Surendranath harnesses electricity to rearrange chemical bonds. The electrochemical reactions he is developing hold potential for processes such as splitting water into hydrogen fuel, creating more efficient fuel cells, and converting waste products like carbon dioxide into useful fuels.</p>
149
150 <p>“All of our research is about decarbonizing the energy ecosystem,” says Surendranath, who recently earned tenure in MIT’s Department of Chemistry and serves as the associate director of the Carbon Capture, Utilization, and Storage Center, one of the Low-Carbon Energy Centers run by the MIT Energy Initiative (MITEI).</p>
151
152 <p>Although his work has many applications in improving energy efficiency, most of the research projects in Surendranath’s group have grown out of the lab’s fundamental interest in exploring, at a <a href="http://news.mit.edu/2019/thermodynamic-electron-proton-reactions-predict-0502">molecular level</a>, the chemical reactions that occur between the surface of an electrode and a liquid.</p>
153
154 <p>“Our goal is to uncover the key rate-limiting processes and the key steps in the reaction mechanism that give rise to one product over another, so that we can, in a rational way, control a material's properties so that it can most selectively and efficiently carry out the overall reaction,” he says.</p>
155
156 <p><strong>Energy conversion</strong></p>
157
158 <p>Born in Bangalore, India, Surendranath moved to Kent, Ohio, with his parents when he was 3 years old. Bangalore and Kent happen to have the world’s leading centers for studying liquid crystal materials, the field that Surendranath’s father, an organic chemist, specialized in.</p>
159
160 <p>“My dad would often take me to the laboratory, and although my parents encouraged me to pursue medicine, I think my interest in science and chemistry probably was sparked at an early age, by those experiences,” Surendranath recalls.</p>
161
162 <p>Although he was interested in all of the sciences, he narrowed his focus after taking his first college chemistry class at the University of Virginia, with a professor named Dean Harman. He decided on a double major in chemistry and physics and ended up doing research in Harman’s inorganic chemistry lab.</p>
163
164 <p>After graduating from UVA, Surendranath came to MIT for graduate school, where his thesis advisor was then-MIT professor Daniel Nocera. With Nocera, he explored <a href="http://news.mit.edu/2011/artificial-leaf-0930">using electricity to split water</a> as a way of renewably generating hydrogen. Surendranath’s PhD research focused on developing methods to catalyze the half of the reaction that extracts oxygen gas from water.</p>
165
166 <p>He got even more involved in catalyst development while doing a postdoctoral fellowship at the University of California at Berkeley. There, he became interested in nanomaterials and the reactions that occur at the interfaces between solid catalysts and liquids.</p>
167
168 <p>“That interface is where a lot of the key processes that are involved in energy conversion occur in electrochemical technologies like batteries, electrolyzers, and fuel cells,” he says.</p>
169
170 <p>In 2013, Surendranath returned to MIT to join the faculty, at a time when many other junior faculty members were being hired.</p>
171
172 <p>“One of the most attractive features of the department is its balanced composition of early career and senior faculty. This has created a nurturing and vibrant atmosphere that is highly collaborative,” he says. “But more than anything else, it was the phenomenal students at MIT that drew me back. Their intensity and enthusiasm is what drives the science.”</p>
173
174 <p><strong>Fuel decarbonization</strong></p>
175
176 <p>Among the many electrochemical reactions that Surendranath’s lab is trying to optimize is the <a href="http://news.mit.edu/2016/greenhouse-gas-into-gasoline-1115">conversion of carbon dioxide</a> to simple chemical fuels such as carbon monoxide, ethylene, or other hydrocarbons. Another project focuses on <a href="http://news.mit.edu/2017/new-way-harness-wasted-methane-1017">converting methane</a> that is burned off from oil wells into liquid fuels such as methanol.</p>
177
178 <p>“For both of those areas, the idea is to convert carbon dioxide and low-carbon feedstocks into commodity chemicals and fuels. These technologies are essential for decarbonizing the chemistry and fuels sector,” Surendranath says.</p>
179
180 <p>Other projects include improving the efficiency of catalysts used for water electrolysis and fuel cells, and for producing <a href="http://news.mit.edu/2019/mit-process-could-make-hydrogen-peroxide-available-remote-places-1023">hydrogen peroxide</a> (a versatile disinfectant). Many of those projects have grown out of his students’ eagerness to chase after difficult problems and follow up on unexpected findings, Surendranath says.</p>
181
182 <p>“The true joy of my time here, in addition to the science, has been about seeing students that I've mentored grow and mature to become independent scientists and thought leaders, and then to go off and launch their own independent careers, whether it be in industry or in academia,” he says. “That role as a mentor to the next generation of scientists in my field has been extraordinarily rewarding.”</p>
183
184 <p>Although they take their work seriously, Surendranath and his students like to <a href="https://news.mit.edu/2019/chemistry-bonds-quirky-researchers-hard-working-surendranath-lab-1226">keep the mood light</a> in their lab. He often brings mangoes, coconuts, and other exotic fruits in to share, and enjoys flying stunt kites — a type of kite that has multiple lines, allowing them to perform acrobatic maneuvers such as figure eights. He can also occasionally be seen making balloon animals or blowing extremely large soap bubbles.</p>
185
186 <p>“My group has really cultivated an extraordinarily positive, collaborative, uplifting environment where we go after really hard problems, and we have a lot of fun along the way,” Surendranath says. “I feel blessed to work with people who have invested so much in the research effort and have built a culture that is such a pleasure to work in every day.”</p>
187 </content:encoded>
188 <media:content url="https://news.mit.edu/sites/default/files/styles/news_article__cover_image__original/public/images/202010/MIT_Yogesh_Surendranath_SL-press.jpg?itok=JycVnNL8" medium="image" type="image/jpeg" width="390" height="260">
189 <media:description type="plain">“The true joy of my time here, in addition to the science, has been about seeing students that I've mentored grow and mature to become independent scientists and thought leaders, and then to go off and launch their own independent careers, whether it be in industry or in academia,” Yogesh Surendranath says.</media:description>
190 <media:credit>Photo: Gretchen Ertl</media:credit>
191 </media:content>
192 </item>
193 <item>
194 <title>A wearable sensor to help ALS patients communicate</title>
195 <link>https://news.mit.edu/2020/sensor-als-communicate-1022</link>
196 <description>Researchers have designed a skin-like device that can measure small facial movements in patients who have lost the ability to speak.</description>
197 <pubDate>Thu, 22 Oct 2020 11:00:00 -0400</pubDate>
198 <guid isPermaLink="true">https://news.mit.edu/2020/sensor-als-communicate-1022</guid>
199 <dc:creator>Anne Trafton | MIT News Office</dc:creator>
200 <content:encoded><p>People with amyotrophic lateral sclerosis (ALS) suffer from a gradual decline in their ability to control their muscles. As a result, they often lose the ability to speak, making it difficult to communicate with others.</p>
201
202 <p>A team of MIT researchers has now designed a stretchable, skin-like device that can be attached to a patient’s face and can measure small movements such as a twitch or a smile. Using this approach, patients could communicate a variety of sentiments, such as “I love you” or “I’m hungry,” with small movements that are measured and interpreted by the device.</p>
203
204 <p>The researchers hope that their new device would allow patients to communicate in a more natural way, without having to deal with bulky equipment. The wearable sensor is thin and can be camouflaged with makeup to match any skin tone, making it unobtrusive.</p>
205
206 <p>“Not only are our devices malleable, soft, disposable, and light, they’re also visually invisible,” says Canan Dagdeviren, the LG Electronics Career Development Assistant Professor of Media Arts and Sciences at MIT and the leader of the research team. “You can camouflage it and nobody would think that you have something on your skin.”</p>
207
208 <p>The researchers tested the initial version of their device in two ALS patients (one female and one male, for <a href="https://conformabledecoders.media.mit.edu/diversitystatement.html">gender balance</a>) and showed that it could accurately distinguish three different facial expressions — smile, open mouth, and pursed lips.</p>
209 <img alt="als" data-align="center" data-entity-type="file" data-entity-uuid="b762f998-fc75-4bc6-9689-444eedb12342" src="/sites/default/files/images/inline/ALS-communication-1.gif" />
210 <p>MIT graduate student Farita Tasnim and former research scientist Tao Sun are the lead authors of the study, which appears today in <em>Nature Biomedical Engineering</em>. Other MIT authors are undergraduate Rachel McIntosh, postdoc Dana Solav,&nbsp;research scientist Lin Zhang, and senior lab manager David Sadat. Yuandong Gu of the A*STAR Institute of Microelectronics in Singapore and Nikta Amiri, Mostafa Tavakkoli Anbarani, and M. Amin Karami of the University of Buffalo are also authors.</p>
211
212 <p><strong>A skin-like sensor</strong></p>
213
214 <p>Dagdeviren’s lab, the <a href="https://conformabledecoders.media.mit.edu/">Conformable Decoders</a> group, specializes in developing conformable (flexible and stretchable) electronic devices that can adhere to the body for a variety of medical applications. She became interested in working on ways to help patients with neuromuscular disorders communicate after meeting Stephen Hawking in 2016, when the world-renowned physicist visited Harvard University and Dagdeviren was a junior fellow in Harvard’s Society of Fellows.</p>
215
216 <p>Hawking, who passed away in 2018, suffered from a slow-progressing form of ALS. He was able to communicate using an infrared sensor that could detect twitches of his cheek, which moved a cursor across rows and columns of letters. While effective, this process could be time-consuming and required bulky equipment.</p>
217
218 <p>Other ALS patients use similar devices that measure the electrical activity of the nerves that control the facial muscles. However, this approach also requires cumbersome equipment, and it is not always accurate.</p>
219
220 <p>“These devices are very hard, planar, and boxy, and reliability is a big issue. You may not get consistent results, even from the same patients within the same day,” Dagdeviren says.</p>
221
222 <p>Most ALS patients also eventually lose the ability to control their limbs, so typing is not a viable strategy to help them communicate. The MIT team set out to design a wearable interface that patients could use to communicate in a more natural way, without the bulky equipment required by current technologies.</p>
223
224 <p>The device they created consists of four piezoelectric sensors embedded in a thin silicone film. The sensors, which are made of aluminum nitride, can detect mechanical deformation of the skin and convert it into an electric voltage that can be easily measured. All of these components are easy to mass-produce, so the researchers estimate that each device would cost around $10.</p>
225
226 <p>The researchers used a process called digital imaging correlation on healthy volunteers to help them select the most useful locations to place the sensor. They painted a random black-and-white speckle pattern on the face and then took many images of the area with multiple cameras as the subjects performed facial motions such as smiling, twitching the cheek, or mouthing the shape of certain letters. The images were processed by software that analyzes how the small dots move in relation to each other, to determine the amount of strain experienced in a single area.</p>
227
228 <p>“We had subjects doing different motions, and we created strain maps of each part of the face,” McIntosh says. “Then we looked at our strain maps and determined where on the face we were seeing a correct strain level for our device, and determined that that was an appropriate place to put the device for our trials.”</p>
229
230 <p>The researchers also used the measurements of skin deformations to train a machine-learning algorithm to distinguish between a smile, open mouth, and pursed lips. Using this algorithm, they tested the devices with two ALS patients, and were able to achieve about 75 percent accuracy in distinguishing between these different movements. The accuracy rate in healthy subjects was 87 percent.</p>
231
232 <p>“The continuous monitoring of facial motions plays a key role in nonverbal communications for patients with neuromuscular disorders. Currently, the mainstream approach is camera tracking, which presents a challenge for continuous, portable usage,” says Takao Someya, a professor of electrical engineering and information systems and dean of the School of Engineering at the University of Tokyo, who was not involved in the study. “The authors have successfully developed thin, wearable, piezoelectric sensors that can reliably decode facial strains and predict facial kinematics.”</p>
233
234 <p><strong>Enhanced communication</strong></p>
235
236 <p>Based on these detectable facial movements, a library of phrases or words could be created to correspond to different combinations of movements, the researchers say.</p>
237
238 <p>“We can create customizable messages based on the movements that you can do,” Dagdeviren says. “You can technically create thousands of messages that right now no other technology is available to do. It all depends on your library configuration, which can be designed for a particular patient or group of patients.”</p>
239
240 <p>The information from the sensor is sent to a handheld processing unit, which analyzes it using the algorithm that the researchers trained to distinguish between facial movements. In the current prototype, this unit is wired to the sensor, but the connection could also be made wireless for easier use, the researchers say.</p>
241 <p>The researchers have filed for a patent on this technology and they now plan to test it with additional patients. In addition to helping patients communicate, the device could also be used to track the progression of a patient’s disease, or to measure whether treatments they are receiving are having any effect, the researchers say.</p>
242
243 <p>“There are a lot of clinical trials that are testing whether or not a particular treatment is effective for reversing ALS,” Tasnim says. “Instead of just relying on the patients to report that they feel better or they feel stronger, this device could give a quantitative measure to track the effectiveness.”</p>
244
245 <p>The research was funded by the MIT Media Lab Consortium, the National Science Foundation, and the National Institute of Biomedical Imaging and Bioengineering.</p>
246 </content:encoded>
247 <media:content url="https://news.mit.edu/sites/default/files/styles/news_article__cover_image__original/public/images/202010/MIT-ALS-Communication-01-PRESS.jpg?itok=XW-qgfQB" medium="image" type="image/jpeg" width="390" height="260">
248 <media:description type="plain">All images: David Sadat</media:description>
249 <media:credit>Image: David Sadat</media:credit>
250 </media:content>
251 </item>
252 <item>
253 <title>Autonomous boats could be your next ride </title>
254 <link>https://news.mit.edu/2020/autonomous-boats-could-be-your-next-ride-1026</link>
255 <description>Five years in the making, MIT’s autonomous floating vessels get a size upgrade and learn a new way to communicate aboard the waters. </description>
256 <pubDate>Mon, 26 Oct 2020 09:00:00 -0400</pubDate>
257 <guid isPermaLink="true">https://news.mit.edu/2020/autonomous-boats-could-be-your-next-ride-1026</guid>
258 <dc:creator>Rachel Gordon | MIT CSAIL</dc:creator>
259 <content:encoded><p>The feverish race to produce the shiniest, safest, speediest self-driving car has spilled over into our wheelchairs, scooters, and even golf carts. Recently, there’s been movement from land to sea, as marine autonomy stands to change the canals of our cities, with the potential to deliver goods and services and collect waste across our waterways.&nbsp;</p>
260
261 <p>In an update to a five-year project from MIT’s Computer Science and Artificial Intelligence Laboratory (CSAIL) and the Senseable City Lab, researchers have been developing the world's first fleet of autonomous boats for the City of Amsterdam, the Netherlands, and have recently added a new, larger vessel to the group: “Roboat II.” Now sitting at 2 meters long, which is roughly a “Covid-friendly” 6 feet, the new robotic boat is capable of carrying passengers.</p>
262 <p>Alongside <a href="https://www.ams-institute.org/">the Amsterdam Institute for Advanced Metropolitan Solutions</a>, the team also created navigation and control algorithms to update the communication and collaboration among the boats.&nbsp;</p>
263
264 <p>“Roboat II navigates autonomously using algorithms similar to those used by self-driving cars, but now adapted for water,” says MIT Professor Daniela Rus, a senior author on a new paper about Roboat and the director of CSAIL. “We’re developing fleets of Roboats that can deliver people and goods, and connect with other Roboats to form a range of autonomous platforms to enable water activities.”&nbsp;</p>
265
266 <p>Self-driving boats have been able to transport small items for years, but adding human passengers has felt somewhat intangible due to the current size of the vessels. Roboat II is the “half-scale” boat in the growing body of work, and joins the previously developed quarter-scale Roboat, which is 1 meter long. The third installment, which is under construction in Amsterdam and is considered to be “full scale,” is 4 meters long and aims to carry anywhere from four to six passengers.&nbsp;</p>
267
268 <p>Aided by powerful algorithms, Roboat II autonomously navigated the canals of Amsterdam for three hours collecting data, and returned back to its start location with an error margin of only 0.17 meters, or fewer&nbsp;than 7 inches.&nbsp;</p>
269
270 <p>“The development of an autonomous boat system capable of accurate mapping, robust control, and human transport is a crucial step towards having the system implemented in the full-scale Roboat,” says senior postdoc Wei Wang, lead author on a new paper about Roboat II. “We also hope it will eventually be implemented in other boats in order to make them autonomous.”</p>
271
272 <p>Wang wrote the paper alongside<em> </em>MIT Senseable City Lab postdoc Tixiao Shan, research fellow Pietro Leoni, postdoc David Fernandez-Gutierrez, research fellow Drew Meyers, and MIT professors Carlo Ratti and Daniela Rus. The work was supported by a grant from the Amsterdam Institute for Advanced Metropolitan Solutions in the Netherlands. A paper on Roboat II will be virtually presented at the International Conference on Intelligent Robots and Systems.&nbsp;</p>
273
274 <p>To coordinate communication among the boats, another team from MIT CSAIL and Senseable City Lab, also led by Wang, came up with a new control strategy for robot coordination.&nbsp;</p>
275
276 <p>With the intent of self-assembling into connected, multi-unit trains — with distant homage to children’s train sets&nbsp;— “collective transport” takes a different path to complete various tasks. The system uses a distributed controller, which is a collection of <a href="https://controlstation.com/what-is-a-distributed-control-system/">sensors, controllers, and associated computers</a> distributed throughout a system), and a strategy inspired by how a colony of ants can transport food without communication. Specifically, there’s no direct communication among the connected robots — only one leader knows the destination. The leader initiates movement to the destination, and then the other robots can estimate the intention of the leader, and align their movements accordingly.&nbsp;</p>
277
278 <p>“Current cooperative algorithms have rarely considered dynamic systems on the water,” says Ratti, the Senseable City Lab director. “Cooperative transport, using a team of water vehicles, poses unique challenges not encountered in aerial or ground vehicles. For example, inertia and load of the vehicles become more significant factors that make the system harder to control. Our study investigates the cooperative control of the surface vehicles and validates the algorithm on that.”&nbsp;</p>
279
280 <p>The team tested their control method on two scenarios: one where three robots are connected in a series, and another where three robots are connected in parallel. The results showed that the coordinated group was able to track various trajectories and orientations in both configurations, and that the magnitudes of the followers’ forces positively contributed to the group — indicating that the follower robots helped the leader.&nbsp;</p>
281
282 <p>Wang wrote a paper about collective transport alongside Stanford University PhD student Zijian Wang, MIT postdoc Luis Mateos, MIT researcher Kuan Wei Huang, Stanford Assistant Professor Mac Schwager, Ratti, and Rus.&nbsp;</p>
283
284 <p><strong>Roboat II</strong></p>
285
286 <p>In 2016, MIT researchers <a href="http://news.mit.edu/2016/autonomous-fleet-amsterdam-roboat-0919">tested</a> a prototype that could move “forward, backward, and laterally along a pre-programmed path in the canals.” Three years later, the team’s robots were updated to “shapeshift” by <a href="https://news.mit.edu/2019/roboats-autonomous-connect-assemble-0829">autonomously disconnecting and reassembling</a> into a variety of configurations.&nbsp;</p>
287
288 <p>Now, Roboat II has scaled up to explore transportation tasks, aided by updated research. These include a new algorithm for Simultaneous Localization and Mapping (SLAM), a model-based optimal controller called nonlinear model predictive controller, and an optimization-based state estimator, called moving horizon estimation.&nbsp;</p>
289
290 <p>Here’s how it works: When a passenger pickup task is required from a user at a specific position, the system coordinator will assign the task to an unoccupied boat that’s closest to the passenger. As Roboat II picks up the passenger, it will create a feasible path to the desired destination, based on the current traffic conditions.&nbsp;</p>
291
292 <p>Then, Roboat II, which weighs more than 50 kilograms, will start to localize itself by running the SLAM algorithm and utilizing <a href="https://oceanservice.noaa.gov/facts/lidar.html">lidar</a> and GPS sensors, as well as an inertial measurement unit for localization, pose, and velocity. The controller then tracks the reference trajectories from the planner, which updates the path to avoid obstacles that are detected to avoid potential collisions.&nbsp;&nbsp;</p>
293
294 <p>The team notes that the improvements in their control algorithms have made the obstacles feel like less of a giant iceberg since their last update; the SLAM algorithm provides a higher localization accuracy for Roboat, and allows for online mapping during navigation, which they didn’t have in previous iterations.&nbsp;</p>
295
296 <p>Increasing the size of Roboat also required a larger area to conduct the experiments, which began in the MIT pools and subsequently moved to the Charles River, which cuts through Boston and Cambridge, Massachusetts.</p>
297
298 <p>While navigating the congested roads of cities alike can lead drivers to feel trapped in a maze, canals largely avoid this. Nevertheless, tricky scenarios in the waterways can still emerge. Given that, the team is working on developing more efficient planning algorithms to let the vessel handle more complicated scenarios, by applying active object detection and identification to improve Roboat’s understanding of its environment. The team plans to estimate disturbances such as currents and waves, to further improve the tracking performance in more noisy waters.&nbsp;</p>
299
300 <p>“All of these expected developments will be incorporated into the first prototype of the full-scale Roboat and tested in the canals of the City of Amsterdam,” says Rus.&nbsp;</p>
301
302 <p><strong>Collective transport </strong></p>
303
304 <p>Making our intuitive abilities a reality for machines has been the persistent intention since the birth of the field, from straightforward commands for picking up items to the nuances of organizing in a group.&nbsp;</p>
305
306 <p>One of the main goals of the project is enabling self-assembly to complete the aforementioned tasks of collecting waste, delivering items, and transporting people in the canals — but controlling this movement on the water has been a challenging obstacle. Communication in robotics can often be unstable or have delays, which may worsen the robot coordination.&nbsp;</p>
307
308 <p>Many control algorithms for this collective transport require direct communication, the relative positions in the group, and the destination of the task — but the team’s new algorithm simply needs one robot to know<em> </em>the desired trajectory and orientation.&nbsp;</p>
309
310 <p>Normally, the distributed controller running on each robot requires the velocity information of the connected structure (represented by the velocity of the center of the structure), but this requires that each robot knows the relative position to the center of the structure. In the team’s algorithm, they don’t need the relative position, and each robot simply uses its local velocity instead of the velocity of the center of the structure.</p>
311
312 <p>When the leader initiates the movement to the destination, the other robots can therefore estimate the intention of the leader and align their movements. The leader can also steer the rest of the robots by adjusting its input, without any communication between any two robots.&nbsp;</p>
313
314 <p>In the future, the team plans to use machine learning to estimate (online) the key parameters of the robots. They’re also aiming to explore adaptive controllers that allow for dynamic change to the structure when objects are placed on the boat. Eventually, the boats will also be extended&nbsp;to outdoor water environments, where large disturbances such as currents and waves exist.</p>
315 </content:encoded>
316 <media:content url="https://news.mit.edu/sites/default/files/styles/news_article__cover_image__original/public/images/202010/MIT-Roboat%20II-Close.jpg?itok=2CROnQzf" medium="image" type="image/jpeg" width="390" height="260">
317 <media:description type="plain">The latest version of MIT's autonomous boat is now 2 meters long and capable of carrying passengers. </media:description>
318 <media:credit>Photo courtesy of the researchers.</media:credit>
319 </media:content>
320 </item>
321 <item>
322 <title>What are the odds your vote will not count?</title>
323 <link>https://news.mit.edu/2020/odds-mail-vote-not-count-1019</link>
324 <description>MIT professor’s study quantifies how many mail-in ballots became “lost votes” in the 2016 U.S. federal election.</description>
325 <pubDate>Mon, 19 Oct 2020 00:00:00 -0400</pubDate>
326 <guid isPermaLink="true">https://news.mit.edu/2020/odds-mail-vote-not-count-1019</guid>
327 <dc:creator>Peter Dizikes | MIT News Office</dc:creator>
328 <content:encoded><p><em>This is part 2 of a two-part </em>MIT News<em> series on voting research and the 2020 election. <a href="https://news.mit.edu/2020/votes-counted-after-election-1015">Part 1</a> focuses on shifts in post-Election Day vote tallies.</em></p>
329
330 <p>In elections, every vote counts. Or should count. But a new study by an MIT professor indicates that in the 2016 U.S. general election, 4 percent of all mail-in ballots were not counted — about 1.4 million votes, or 1 percent of all votes cast, signaling a significant problem that could grow in 2020.</p>
331
332 <p>The study quantifies the range of reasons for this, including late-arriving ballots, problems with ballot signatures and envelopes, and improperly marked ballots, among other things.</p>
333
334 <p>“Mail ballots tend to have more mistakes on them,” says Charles Stewart, a professor in MIT’s Department of Political Science and author of a paper detailing the study, which looks at data from all 50 U.S. states.</p>
335
336 <p>Voting by mail — the same thing as absentee voting — will probably be more prevalent than ever in 2020, as voters seek to avoid crowds at polling places during the Covid-19 pandemic.</p>
337
338 <p>As the study suggests, states that have more experience with mail-in voting tend to have a slightly lower percentage of lost votes. Thus the 2020 election could feature an unusually high percentage of lost mail-in voting attempts, and the odds of your mail-in ballot counting may vary a bit, depending on where you live.</p>
339
340 <p>“The likelihood of a vote being lost by mail is, in part, determined by how the state feels about that,” says Stewart, who is the Kenan Sahin Distinguished Professor of Political Science and head of the MIT Election Data and Science Lab. “States can put more or less effort into ensuring that voters don’t make mistakes. … There are different mail-ballot regimes, they handle the ballots differently, they operate under different philosophies of what mail balloting is supposed to achieve, and who bears the risk of mail balloting.”</p>
341
342 <p>The paper, “Reconsidering Lost Votes by Mail,” appears as a working paper on the Social Science Research Network, and will be published by the <em>Harvard Data Science Review</em>.</p>
343
344 <p><strong>Check your work</strong></p>
345
346 <p>The concept of “lost votes” was first studied comprehensively by the Caltech/MIT Voting Technology Project (VTP) following the contested 2000 U.S. presidential election. The VTP concluded that of 107 million votes cast in 2000 — of all kinds, not just mail-in voting — between 4 million and 6 million went unrecorded. The federal Help America Vote Act of 2003 (HAVA) subsequently reduced that number to between 2 million and 3 million.</p>
347
348 <p>The current paper extends that line of analysis to absentee votes, and updates a 2010 Stewart study. Overall, there are three main types of problems with mail-in votes: postal issues, procedural problems involving things like signatures and ballot envelopes, and vote-scanning problems.</p>
349
350 <p>In the first case, about 1.1 percent of all mail-in votes are lost because of problems during the mailing process — from unfilled absentee ballot requests to the return of those ballots. Some of those lost votes represent election-administration errors, not postal issues. Stewart does not think recent reductions in U.S. Postal Service capacity will necessarily change that, although many experts are urging voters to mail in their ballots promptly.</p>
351
352 <p>“Postal service problems, literally the ballot not arriving, the ballot arriving late, getting lost in the office, that’s one source,” Stewart says. “But it’s probably the least important source of loss, despite all the controversy about the postal service.”</p>
353
354 <p>Secondly, votes can also be lost when voters handle the process incorrectly: They fail to sign ballots, are judged to have submitted mismatched signatures, or do not use the ballot’s safety envelope, among other things. About 1.5 percent of mail-in votes suffer from these problems, Stewart estimates.</p>
355
356 <p>“The voter can make a mistake in the certification process,” Stewarts says. “They don’t sign the envelope where they’re supposed to, they don’t seal it properly … there are all sorts of things that lead to rejected ballots.” Still, Stewart observes, “Election offices could be less persnickety about technical issues.”</p>
357
358 <p>The third type of problem, comprising 1.5 percent of all attempts at absentee voting, occurs when scanning machines in polling places reject ballots.</p>
359
360 <p>“The scanning problems, nobody really talks about because it’s the most abstract, but I think it may be the most important,” Stewart says.</p>
361
362 <p>This category includes voter mistakes that could be corrected in person, but lead to rejection on absentee ballots. When people “overvote,” selecting too many candidates, scanning machines catch the errors — and HAVA mandates that in-person voters can re-do the ballot.</p>
363
364 <p>“If you overvote, there’s a requirement in federal law that the ballot be kicked back to you,” Stewart says about in-person voting. “If you undervote, there’s not a requirement, but many states will kick back the ballot [to voters]. But if you do that and drop your ballot in the mailbox, there’s nobody to kick the ballot back to you.”</p>
365
366 <p>One frequent type of overvote happens when voters redundantly add their chosen candidate’s name to the write-in line, Stewart says: “The most common reason for overvotes is people will fill in the bubble for their candidate, and then they’ll go down to the bottom and write in the name of their candidate.”</p>
367
368 <p>There are other ways a voter can foul up a ballot as well.</p>
369
370 <p>“It could be, if you’re making choices and put your pencil down next to every name, that could be picked up as a vote by the scanners,” Stewart says. “There are things you just don’t think about that could go wrong.”</p>
371
372 <p><strong>The geography of lost votes</strong></p>
373
374 <p>To conduct the study, Stewart used a variety of data sources, including U.S. Postal Service on-time rates, the Survey of the Performance of American Elections, the Cooperative Congressional Election Study, and the Current Population Survey of the U.S. Census Bureau.</p>
375
376 <p>One finding of the study is that the percentage of lost mail-in votes is lower in states that lean more heavily on absentee balloting overall. It is 3.5 percent in states that conduct their elections almost completely by mail (Colorado, Oregon, and Washington) and in those that keep a permanent absentee ballot list (Arizona, California, Hawaii, Montana, and Utah, plus Washington, D.C.). But the lost votes percentage for mail-in ballots is higher, at 4.4 percent, in states that honor absentee ballot requests with no excuse needed, and it’s 4.9 percent in states that require an excuse for absentee balloting.</p>
377
378 <p>That suggests both that voters become more proficient when they have more experience at mail-in voting, and that states may process mail ballots more effectively when it becomes routine for them. Stewart, for one, believes that election officials do an exceptional job overall.</p>
379
380 <p>“I’m very sanguine about the integrity of the process, from what I know about election officials,” Stewart says. Still, he acknowledges, absentee voting can be a tricky process, and a significant number of votes may be lost in 2020.</p>
381
382 <p>“That’s why we have a lot of voter education going on right now,” Stewart says.</p>
383 </content:encoded>
384 <media:content url="https://news.mit.edu/sites/default/files/styles/news_article__cover_image__original/public/images/202010/MIT-Lost-Votes-01-PRESS.jpg?itok=1dwFDvy-" medium="image" type="image/jpeg" width="390" height="260">
385 <media:credit>Image: Christine Daniloff, MIT</media:credit>
386 </media:content>
387 </item>
388 <item>
389 <title>A controllable membrane to pull carbon dioxide out of exhaust streams</title>
390 <link>https://news.mit.edu/2020/membrane-carbon-dioxide-exhaust-1016</link>
391 <description>Electrically switchable system could continuously separate gases without the need for moving parts or wasted space.</description>
392 <pubDate>Fri, 16 Oct 2020 14:00:00 -0400</pubDate>
393 <guid isPermaLink="true">https://news.mit.edu/2020/membrane-carbon-dioxide-exhaust-1016</guid>
394 <dc:creator>David L. Chandler | MIT News Office</dc:creator>
395 <content:encoded><p>A new system developed by chemical engineers at MIT could provide a way of continuously removing carbon dioxide from a stream of waste gases, or even from the air. The key component is an electrochemically assisted membrane whose permeability to gas can be switched on and off at will, using no moving parts and relatively little energy.</p>
396
397 <p>The membranes themselves, made of anodized aluminum oxide, have a honeycomb-like structure made up of hexagonal openings that allow gas molecules to flow in and out when in the open state. However, gas passage can be blocked when a thin layer of metal is electrically deposited to cover the pores of the membrane. The work is described today in the journal <em>Science Advances</em>, in a paper by Professor T. Alan Hatton, postdoc Yayuan Liu, and four others.</p>
398
399 <p>This new “gas gating” mechanism could be applied to the continuous removal of carbon dioxide from a range of industrial exhaust streams and from ambient air, the team says. They have built a proof-of-concept device to show this process in action.</p>
400
401 <p>The device uses a redox-active carbon-absorbing material, sandwiched between two switchable gas gating membranes. The sorbent and the gating membranes are in close contact with each other and are immersed in an organic electrolyte to provide a medium for zinc ions to shuttle back and forth. These two gating membranes can be opened or closed electrically by switching the polarity of a voltage between them, causing ions of zinc to shuttle from one side to the other. The ions simultaneously block one side, by forming a metallic film over it, while opening the other, by dissolving its film away.</p>
402
403 <p>When the sorbent layer is open to the side where the waste gases are flowing by, the material readily soaks up carbon dioxide until it reaches its capacity. The voltage can then be switched to block off the feed side and open up the other side, where a concentrated stream of nearly pure carbon dioxide is released.</p>
404
405 <p>By building a system with alternating sections of membrane that operate in opposite phases, the system would allow for continuous operation in a setting such as an industrial scrubber. At any one time, half of the sections would be absorbing the gas while the other half would be releasing it.</p>
406
407 <p>“That means that you have a feed stream coming into the system at one end and the product stream leaving from the other in an ostensibly continuous operation,” Hatton says. “This approach avoids many process issues” that would be involved in a traditional multicolumn system, in which adsorption beds alternately need to be shut down, purged, and then regenerated, before being exposed again to the feed gas to begin the next adsorption cycle. In the new system, the purging steps are not required, and the steps all occur cleanly within the unit itself.</p>
408
409 <p>The researchers’ key innovation was using electroplating as a way to open and close the pores in a material. Along the way the team had tried a variety of other approaches to reversibly close pores in a membrane material, such as using tiny magnetic spheres that could be positioned to block funnel-shaped openings, but these other methods didn’t prove to be efficient enough. Metal thin films can be particularly effective as gas barriers, and the ultrathin layer used in the new system requires a minimal amount of the zinc material, which is abundant and inexpensive.</p>
410
411 <p>“It makes a very uniform coating layer with a minimum amount of materials,” Liu says. One significant advantage of the electroplating method is that once the condition is changed, whether in the open or closed position, it requires no energy input to maintain that state. Energy is only required to switch back again.</p>
412
413 <p>Potentially, such a system could make an important contribution toward limiting emissions of greenhouse gases into the atmosphere, and even direct-air capture of carbon dioxide that has already been emitted.</p>
414
415 <p>While the team’s initial focus was on the challenge of separating carbon dioxide from a stream of gases, the system could actually be adapted to a wide variety of chemical separation and purification processes, Hatton says.</p>
416
417 <p>“We’re pretty excited about the gating mechanism. I think we can use it in a variety of applications, in different configurations,” he says. “Maybe in microfluidic devices, or maybe we could use it to control the gas composition for a chemical reaction. There are many different possibilities.”</p>
418
419 <p>The research team included graduate student Chun-Man Chow, postdoc Katherine Phillips, and recent graduates Miao Wang PhD ’20 and Sahag Voskian PhD ’19. This work was supported by ExxonMobil&nbsp;through the MIT Energy Initiative.</p>
420 </content:encoded>
421 <media:content url="https://news.mit.edu/sites/default/files/styles/news_article__cover_image__original/public/images/202010/MIT-Gated-Membranes-01-PRESS.jpg?itok=-_WyytS2" medium="image" type="image/jpeg" width="390" height="260">
422 <media:description type="plain">On the right is a porous anodized aluminum oxide membrane. The left side shows the same membrane after coating it with a thin layer of gold, making the membrane conductive for electrochemical gas gating.</media:description>
423 <media:credit>Image: Felice Frankel</media:credit>
424 </media:content>
425 </item>
426 <item>
427 <title>A global collaboration to move artificial intelligence principles to practice</title>
428 <link>https://news.mit.edu/2020/global-collaboration-moving-ai-principles-to-practice-1019</link>
429 <description>Convened by the MIT Schwarzman College of Computing, the AI Policy Forum will develop frameworks and tools for governments and companies to implement concrete policies.</description>
430 <pubDate>Mon, 19 Oct 2020 13:50:00 -0400</pubDate>
431 <guid isPermaLink="true">https://news.mit.edu/2020/global-collaboration-moving-ai-principles-to-practice-1019</guid>
432 <dc:creator>MIT Schwarzman College of Computing</dc:creator>
433 <content:encoded><p>Today, artificial intelligence — and the computing systems that underlie it — are more than just matters of technology; they are matters of state and society, of governance and the public interest.&nbsp;The choices that technologists, policymakers, and communities make in the next few years will shape the relationship between machines and humans for decades to come.</p>
434
435 <p>The rapidly increasing applicability of AI has prompted a number of organizations to develop high-level principles on social and ethical issues such as privacy, fairness, bias, transparency, and accountability. Building on those broader principles, the <a href="https://aipolicyforum.mit.edu/" target="_blank">AI Policy Forum</a>, a global effort convened by the MIT Stephen A. Schwarzman College of Computing, will provide an overarching policy framework and tools for governments and companies to implement in concrete ways.</p>
436
437 <p>“Our goal is to help policymakers in making practical decisions about AI policy,” says Daniel Huttenlocher, dean of the MIT Schwarzman College of Computing. “We are not trying to develop another set of principles around AI, several of which already exist, but rather provide context and guidelines specific to a field of use of AI to help policymakers around the world with implementation.”</p>
438
439 <p>“Moving beyond principles means understanding trade-offs and identifying the technical tools and the policy levers to address them. We created the college to examine and address these types of issues, but this can’t be a siloed effort. We need for this to be a global collaboration and engage scientists, technologists, policymakers, and business leaders,” says MIT Provost Martin Schmidt. “This is a challenging and complex process for which we need all hands-on deck.”</p>
440
441 <p>The AI Policy Forum is designed as a yearlong process. Activities associated with this effort will be distinguished by their focus on tangible outcomes — their engagement with key government officials at the local, national, and international level charged with designing those public policies, and their deep technical grounding in the latest advances in the science of AI.&nbsp;The measure of success will be whether these efforts have bridged the gap between these communities, translated principled agreement into actionable outcomes, and helped create the conditions for deeper trust between humans and machines.</p>
442
443 <p>The global collaboration will begin in late 2020 and early 2021 with a series of AI Policy Forum Task Forces, chaired by MIT researchers and bringing together the world’s leading technical and policy experts on some of the most pressing issues of AI policy, starting with AI in finance and mobility. Further task forces throughout 2021 will convene more communities of practice with the shared aim of designing the next chapter of AI: one that both delivers on AI’s innovative potential and responds to society’s needs.</p>
444
445 <p>Each task force will produce results that inform concrete public policies and frameworks for the next chapter of AI, and help define the roles that the academic and business communities, civil society, and governments will need to play in making it a reality. Research from the task forces will feed into the development of the AI Policy Framework, a dynamic assessment tool that will help governments gauge their own progress on AI policy-making goals and guide application of best practices appropriate to their own national priorities.</p>
446
447 <p>On May 6–7, 2021, MIT will host — most likely online — the first AI Policy Forum Summit, a two-day collaborative gathering to discuss the progress of the task forces towards equipping high-level decision-makers with a deeper understanding of the tools at their disposal — and trade-offs to be made — to produce better public policy around AI, and better AI systems with concern for public policy. Then, in fall 2021, a follow-on event at MIT will bring together leaders from across sectors and countries and, built atop the leading research from the task forces, the forum will provide a focal point for work to move from AI principles to AI practice, and serve as a springboard to global efforts to design the future of AI.</p>
448 </content:encoded>
449 <media:content url="https://news.mit.edu/sites/default/files/styles/news_article__cover_image__original/public/images/202010/MIT-AIPF-global-AI.jpg?itok=79NXW2Xw" medium="image" type="image/jpeg" width="390" height="260">
450 <media:credit>Image: Pete Linforth/Pixabay</media:credit>
451 </media:content>
452 </item>
453 <item>
454 <title>Neural pathway crucial to successful rapid object recognition in primates</title>
455 <link>https://news.mit.edu/2020/neural-pathway-crucial-successful-rapid-object-recognition-primates-1020</link>
456 <description>Recurrent processing via prefrontal cortex, necessary for quick visual object processing in primates, provides a key insight for developing brain-like artificial intelligence.</description>
457 <pubDate>Tue, 20 Oct 2020 15:45:00 -0400</pubDate>
458 <guid isPermaLink="true">https://news.mit.edu/2020/neural-pathway-crucial-successful-rapid-object-recognition-primates-1020</guid>
459 <dc:creator>Alison Gold | School of Science</dc:creator>
460 <content:encoded><p>MIT researchers have identified a brain pathway critical in enabling primates to effortlessly identify objects in their field of vision. The findings enrich existing models of the neural circuitry involved in visual perception and help to further unravel the computational code for solving object recognition in the primate brain.</p>
461
462 <p>Led by Kohitij Kar, a postdoc at the McGovern Institute for Brain Research and Department of Brain and Cognitive Sciences, the study looked at an area called the ventrolateral prefrontal cortex (vlPFC), which sends feedback signals to the inferior temporal (IT) cortex via a network of neurons. The main goal of this study was to test how the back-and-forth information processing of this circuitry — that is, this recurrent neural network — is essential to rapid object identification in primates.</p>
463
464 <p>The current study, published in <em>Neuron</em> and available via open access, is a followup to prior work published by Kar and James DiCarlo, the Peter de Florez Professor of Neuroscience, the head of MIT’s Department of Brain and Cognitive Sciences, and an investigator in the McGovern Institute and the Center for Brains, Minds, and Machines.</p>
465
466 <p><strong>Monkey versus machine </strong></p>
467
468 <p>In 2019, Kar, DiCarlo, and colleagues identified that primates must use some recurrent circuits during rapid object recognition. Monkey subjects in that study were able to identify objects more accurately than engineered “feed-forward” computational models, called deep convolutional neural networks, that lacked recurrent circuitry.</p>
469
470 <p>Interestingly, specific images for which models performed poorly compared to monkeys in object identification, also took longer to be solved in the monkeys’ brains — suggesting that the additional time might be due to recurrent processing in the brain. Based on the 2019 study, it remained unclear, though, exactly which recurrent circuits were responsible for the delayed information boost in the IT cortex. That’s where the current study picks up.</p>
471
472 <p>“In this new study, we wanted to find out: Where are these recurrent signals in IT coming from?” Kar says. “Which areas reciprocally connected to IT, are functionally the most critical part of this recurrent circuit?”</p>
473
474 <p>To determine this, researchers used a pharmacological agent to temporarily block the activity in parts of the vlPFC in macaques while they engaged in an object discrimination task. During these tasks, monkeys viewed images that contained an object, such as an apple, a car, or a dog; then, researchers used eye tracking to determine if the monkeys could correctly indicate what object they had previously viewed when given two object choices.</p>
475
476 <p>“We observed that if you use pharmacological agents to partially inactivate the vlPFC, then both the monkeys’ behavior and IT cortex activity deteriorates, but more so for certain specific images. These images were the same ones we identified in the previous study — ones that were poorly solved by ‘feed-forward’ models and took longer to be solved in the monkey’s IT cortex,” says Kar.</p>
477
478 <p>“These results provide evidence that this recurrently connected network is critical for rapid object recognition, the behavior we're studying. Now, we have a better understanding of how the full circuit is laid out, and what are the key underlying neural components of this behavior.”</p>
479
480 <p>The full study, entitled “Fast recurrent processing via ventrolateral prefrontal cortex is needed by the primate ventral stream for robust core visual object recognition,” will run in print Jan. 6, 2021.</p>
481
482 <p>“This study demonstrates the importance of prefrontal cortical circuits in automatically boosting object recognition performance in a very particular way,” DiCarlo says. “These results were obtained in nonhuman primates and thus are highly likely to also be relevant to human vision.”</p>
483
484 <p>The present study makes clear the integral role of the recurrent connections between the vlPFC and the primate ventral visual cortex during rapid object recognition. The results will be helpful to researchers designing future studies that aim to develop accurate models of the brain, and to researchers who seek to develop more human-like artificial intelligence.</p>
485 </content:encoded>
486 <media:content url="https://news.mit.edu/sites/default/files/styles/news_article__cover_image__original/public/images/202010/Ko-DiCarlo-Neuron-edited_0.jpg?itok=afkjUs-p" medium="image" type="image/jpeg" width="390" height="260">
487 <media:description type="plain">MIT researchers used an object recognition task (e.g., recognizing that there is a “bird” and not an “elephant” in the shown image) in studying the role of feedback from the primate ventrolateral prefrontal cortex (vlPFC) to the inferior temporal (IT) cortex via a network of neurons. In primate brains, temporally blocking the vlPFC (green shaded area) disrupts the recurrent neural network comprising vlPFC and IT, inducing specific deficits and implicating its role in rapid object identification. </media:description>
488 <media:credit>Image courtesy of Kohitij Kar; brain image adapted from SciDraw.</media:credit>
489 </media:content>
490 </item>
491 <item>
492 <title>AI Cures: data-driven clinical solutions for Covid-19 </title>
493 <link>https://news.mit.edu/2020/ai-cures-data-driven-clinical-solutions-covid-19-1027</link>
494 <description>MIT conference illustrates technologies developed in response to the pandemic and new opportunities for AI solutions for clinical management. </description>
495 <pubDate>Tue, 27 Oct 2020 11:55:00 -0400</pubDate>
496 <guid isPermaLink="true">https://news.mit.edu/2020/ai-cures-data-driven-clinical-solutions-covid-19-1027</guid>
497 <dc:creator>Terri Park | MIT Schwarzman College of Computing</dc:creator>
498 <content:encoded><p>Modern health care has been reinvigorated by the widespread adoption of artificial intelligence. From speeding image analysis for radiology to advancing precision medicine for personalized care, AI has countless applications, but can it rise to the challenge in the fight against Covid-19?</p>
499
500 <p>Researchers from the Abdul Latif Jameel Clinic for Machine Learning in Health (Jameel Clinic), now housed within the MIT Stephen A. Schwarzman College of Computing, say the ongoing public health crisis provides ample opportunities for leveraging AI technologies, such as accelerating the search for effective therapeutics and drugs that can treat the disease, and are actively working to translate this potential to success.</p>
501
502 <p><strong>AI Cures</strong></p>
503
504 <p>When Covid-19 began to spread worldwide, Jameel Clinic’s community of machine learning and life science researchers redirected their work and began exploring how they can collaborate on the search for solutions by tapping into their collective knowledge and expertise. The ensuing discussions led to the launch of <a href="https://www.aicures.mit.edu" target="_blank">AI Cures</a>, an initiative dedicated to developing machine learning methods for finding promising antiviral molecules for Covid-19 and other emerging pathogens, and to lower the barrier for people from varied backgrounds to get involved by inviting them to contribute to the effort.</p>
505
506 <p>As part of the mission of AI Cures to have broad impact and engagement, Jameel Clinic brought together researchers, clinicians, and public health specialists for a conference focused on the development of AI algorithms for the clinical management of Covid-19 patients, early detection and monitoring of the disease, preventing future outbreaks, and ways in which these technologies have been utilized in patient care.</p>
507
508 <p><strong>Data-driven clinical solutions</strong></p>
509
510 <p>On Sept. 29, over 650 people representing 50 countries and 70 organizations logged on from around the globe for the virtual AI Cures Conference: Data-driven Clinical Solutions for Covid-19.</p>
511
512 <p>In welcoming the audience, Daniel Huttenlocher, dean of the MIT Schwarzman College of Computing, remarked that “AI in health care is moving beyond the use of computing as just simple tools, to capabilities that really aid in the processes of discovery, diagnosis, and care. The potential for AI-accelerated discovery is particularly relevant in times such as these.”</p>
513
514 <p>Attendees heard from 14 other speakers, including MIT researchers, on technologies they developed over the past six months in response to the pandemic — from epidemiological models created using clinical data to predict the risk of both infection and death for individual patients, to a wireless device that allows doctors to monitor Covid-19 patients from a distance, to a machine learning model that pinpoints patients at risk for intubation before they crash.</p>
515
516 <p>James Collins, the Termeer Professor of Medical Engineering and Science in MIT’s Institute for Medical Engineering and Science (IMES) and Department of Biological Engineering, and faculty co-lead of life sciences for Jameel Clinic, gave the first talk of the day on harnessing synthetic biology to develop diagnostics to address Covid-19 and how his lab is using deep learning to enhance the design of such systems. Collins and his team are utilizing AI techniques to create a set of algorithms to effectively predict the efficacy of RNA-based sensors. The sensors, first developed in 2014 to detect the Ebola virus and later tailored for the Zika virus in 2016, were designed and optimized for a Covid-19 diagnostic, and related CRISPR-based biosensors are being used in a mask developed in Collins’ lab that produces a detectable signal when a person with the virus breathes, coughs, or sneezes.</p>
517
518 <p>While AI has proven to be an effective tool in health care, a model requires good data for it to be valuable and useful. With Covid-19 being a new disease, limited amounts of information are available to researchers, and in order to advance even more efforts to combat the virus, Collins notes that “we need to put in place and secure the resources to generate and collect large amounts of well-characterized data to train deep learning models. At present we generally don’t have such large datasets. In the system we developed, our dataset consists of about 91,000 RNA elements, which is currently the largest available for RNA synthetic biology, but it should be larger and expanded to many more different sensors.”</p>
519
520 <p>Offering perspective from the clinical side, Constance Lehman, a professor at Harvard Medical School (HMS), discussed the ways in which she’s implementing AI tools in her work as director of breast imaging at Massachusetts General Hospital (MGH). In collaboration with Regina Barzilay, the Delta Electronics Professor of Electrical Engineering and Computer Science and faculty co-lead of AI for Jameel Clinic, Lehman designs machine learning models to aid in breast cancer detection, which became a critical tool when mammography screenings were put on hold during the emergency stay-at-home-order issued in Massachusetts last March. By the time screenings reopened in May, around 15,000 mammograms had been cancelled. MGH is gradually rescheduling patients using a model developed by Lehman and Barzilay to help ease the process. “We took those women that had been diverted from screening and ranked them by their AI risk models and we reached out to them, inviting them back in.”</p>
521
522 <p>However, according to Lehman, many are choosing to opt out of screening and, in particular, fewer women of color are returning. “There are many determinants of who returns for screening. Social determinants can swamp all of our best, most scientific evidence-based approaches to effective and equitable health care. We’re delighted that our risk model is equally predictive across races, but I am dismayed to see that we are screening more white women than women of color during these times. Those are social determinants, which we are working very hard on.”</p>
523
524 <p>The conference culminated in a panel discussion with those who are at the front line of the pandemic. The panelists — Gabriella Antici, founder of the Protea Institute in Brazil; Rajesh Gandhi, a professor at HMS and an infectious disease physician at MGH; Guillermo Torre, a professor of cardiology and president of TEC Salud in Mexico; and Karen Wong, data science unit lead for the Covid-19 clinical team at the U.S. Centers for Disease Control and Prevention — shared their experiences in handling the crisis and had an open conversation with Barzilay, the panel’s moderator, on the limitations of AI and what is currently not being addressed.</p>
525
526 <p>“Those from the AI community like myself are always asking ourselves if we are solving the right problems,” says Barzilay. “We hope to come up with new ideas for AI solutions and what we can do in the future to help.”</p>
527
528 <p>Gandhi offered that “we need more refined and sophisticated approaches to deciding when to use different drugs and how to use them in combination.” He also suggested that integrating physiologic data could be useful in considering how to treat individual patients from different age ranges exhibiting a variety of Covid-19 symptoms, from mild to severe.</p>
529
530 <p>In her closing remarks, Barzilay expressed hope that the conference “illustrates the types of problems that we need to be addressing on the AI side” and notes that Jameel Clinic will widely share any new data they obtain so that everyone can benefit to help patients suffering from Covid-19.</p>
531
532 <p>The event was the first in a pair of conferences that took place as part of the AI Cures initiative. The next event, <a href="https://www.aicures.mit.edu/drugdiscoveryconference">AI Cures Drug Discovery Conference</a>, which will focus on cutting-edge AI approaches in this area developed by MIT researchers and their collaborators, will be held virtually on Oct. 30.</p>
533
534 <p>AI Cures: Data-driven Clinical Solutions was organized by Jameel Clinic, MIT Schwarzman College of Computing, and Institute for Medical Engineering and Sciences. Additional support was provided by the Patrick J. McGovern Foundation.</p>
535 </content:encoded>
536 <media:content url="https://news.mit.edu/sites/default/files/styles/news_article__cover_image__original/public/images/202010/AI-Cures-Conference-panel.jpg?itok=LsoI8I7a" medium="image" type="image/jpeg" width="390" height="260">
537 <media:description type="plain">Panelists discuss open problems in Covid-19 patient care and new opportunities for AI solutions. Left to right and top to bottom: Regina Barzilay, MIT; Gabriella Antici, Protea Institute; Rajesh Gandhi, MGH; Karen Wong, CDC; and Guillermo Torre, TEC Salud.</media:description>
538 </media:content>
539 </item>
540 <item>
541 <title>Machine learning uncovers potential new TB drugs</title>
542 <link>https://news.mit.edu/2020/gaussian-machine-learning-tb-drug-1015</link>
543 <description>Computational method for screening drug compounds can help predict which ones will work best against tuberculosis or other diseases.</description>
544 <pubDate>Thu, 15 Oct 2020 11:00:00 -0400</pubDate>
545 <guid isPermaLink="true">https://news.mit.edu/2020/gaussian-machine-learning-tb-drug-1015</guid>
546 <dc:creator>Anne Trafton | MIT News Office</dc:creator>
547 <content:encoded><p>Machine learning is a computational tool used by many biologists to analyze huge amounts of data, helping them to identify potential new drugs. MIT researchers have now incorporated a new feature into these types of machine-learning algorithms, improving their prediction-making ability.</p>
548
549 <p>Using this new approach, which allows computer models to account for uncertainty in the data they’re analyzing, the MIT team identified several promising compounds that target a protein required by the bacteria that cause tuberculosis.</p>
550
551 <p>This method, which has previously been used by computer scientists but has not taken off in biology, could also prove useful in protein design and many other fields of biology, says Bonnie Berger, the Simons Professor of Mathematics and head of the Computation and Biology group in MIT’s Computer Science and Artificial Intelligence Laboratory (CSAIL).</p>
552
553 <p>“This technique is part of a known subfield of machine learning, but people have not brought it to biology,” Berger says. “This is a paradigm shift, and is absolutely how biological exploration should be done.”</p>
554
555 <p>Berger and Bryan Bryson, an assistant professor of biological engineering at MIT and a member of the Ragon Institute of MGH, MIT, and Harvard, are the senior authors of the study, which appears today in <em>Cell Systems</em>. MIT graduate student Brian Hie is the paper’s lead author.</p>
556
557 <p><strong>Better predictions</strong></p>
558
559 <p>Machine learning is a type of computer modeling in which an algorithm learns to make predictions based on data that it has already seen. In recent years, biologists have begun using machine learning to scour huge databases of potential drug compounds to find molecules that interact with particular targets.</p>
560
561 <p>One limitation of this method is that while the algorithms perform well when the data they’re analyzing are similar to the data they were trained on, they’re not very good at evaluating molecules that are very different from the ones they have already seen.</p>
562
563 <p>To overcome that, the researchers used a technique called Gaussian process to assign uncertainty values to the data that the algorithms are trained on. That way, when the models are analyzing the training data, they also take into account how reliable those predictions are.</p>
564
565 <p>For example, if the data going into the model predict how strongly a particular molecule binds to a target protein, as well as the uncertainty of those predictions, the model can use that information to make predictions for protein-target interactions that it hasn’t seen before. The model also estimates the certainty of its own predictions. When analyzing new data, the model’s predictions may have lower certainty for molecules that are very different from the training data. Researchers can use that information to help them decide which molecules to test experimentally.</p>
566
567 <p>Another advantage of this approach is that the algorithm requires only a small amount of training data. In this study, the MIT team trained the model with a dataset of 72 small molecules and their interactions with more than 400 proteins called protein kinases. They were then able to use this algorithm to analyze nearly 11,000 small molecules, which they took from the ZINC database, a publicly available repository that contains millions of chemical compounds. Many of these molecules were very different from those in the training data.</p>
568
569 <p>Using this approach, the researchers were able to identify molecules with very strong predicted binding affinities for the protein kinases they put into the model. These included three human kinases, as well as one kinase found in <em>Mycobacterium tuberculosis</em>. That kinase, PknB, is critical for the bacteria to survive, but is not targeted by any frontline TB antibiotics.</p>
570
571 <p>The researchers then experimentally tested some of their top hits to see how well they actually bind to their targets, and found that the model’s predictions were very accurate. Among the molecules that the model assigned the highest certainty, about 90 percent proved to be true hits — much higher than the 30 to 40 percent hit rate of existing machine learning models used for drug screens.</p>
572
573 <p>The researchers also used the same training data to train a traditional machine-learning algorithm, which does not incorporate uncertainty, and then had it analyze the same 11,000 molecule library. “Without uncertainty, the model just gets horribly confused and it proposes very weird chemical structures as interacting with the kinases,” Hie says.</p>
574
575 <p>The researchers then took some of their most promising PknB inhibitors and tested them against <em>Mycobacterium tuberculosis</em> grown in bacterial culture media, and found that they inhibited bacterial growth. The inhibitors also worked in human immune cells infected with the bacterium.</p>
576
577 <p><strong>A good starting point</strong></p>
578
579 <p>Another important element of this approach is that once the researchers get additional experimental data, they can add it to the model and retrain it, further improving the predictions. Even a small amount of data can help the model get better, the researchers say.</p>
580
581 <p>“You don’t really need very large data sets on each iteration,” Hie says. “You can just retrain the model with maybe 10 new examples, which is something that a biologist can easily generate.”</p>
582
583 <p>This study is the first in many years to propose new molecules that can target PknB, and should give drug developers a good starting point to try to develop drugs that target the kinase, Bryson says. “We’ve now provided them with some new leads beyond what has been already published,” he says.</p>
584
585 <p>The researchers also showed that they could use this same type of machine learning to boost the fluorescent output of a green fluorescent protein, which is commonly used to label molecules inside living cells. It could also be applied to many other types of biological studies, says Berger, who is now using it to analyze mutations that drive tumor development.</p>
586
587 <p>The research was funded by the U.S. Department of Defense through the National Defense Science and Engineering Graduate Fellowship; the National Institutes of Health; the Ragon Institute of MGH, MIT, and Harvard’ and MIT’s Department of Biological Engineering.</p>
588 </content:encoded>
589 <media:content url="https://news.mit.edu/sites/default/files/styles/news_article__cover_image__original/public/images/202010/MIT-MachineLearning-01-Press.jpg?itok=mb5iRL9o" medium="image" type="image/jpeg" width="390" height="260">
590 <media:description type="plain">Using a machine-learning approach that incorporates uncertainty, MIT researchers identified several promising compounds that target a protein required for the survival of the bacteria that cause tuberculosis.</media:description>
591 <media:credit>Image: MIT News</media:credit>
592 </media:content>
593 </item>
594 <item>
595 <title>How many votes will be counted after election night?</title>
596 <link>https://news.mit.edu/2020/votes-counted-after-election-1015</link>
597 <description>Study measures the “blue shift” from absentee and provisional ballots, underscores uncertainties of 2020 vote.</description>
598 <pubDate>Thu, 15 Oct 2020 00:00:00 -0400</pubDate>
599 <guid isPermaLink="true">https://news.mit.edu/2020/votes-counted-after-election-1015</guid>
600 <dc:creator>Peter Dizikes | MIT News Office</dc:creator>
601 <content:encoded><p><em>This is part 1 of a two-part </em>MIT News<em> series on voting research and the 2020 election.</em></p>
602
603 <p>When you watch election returns on Nov. 3, keep this in mind: In some U.S. states, it will take days to count all the ballots, and the winner might only be clear later, rather than sooner.</p>
604
605 <p>Four straight U.S. presidential elections have featured a “blue shift,” in which the post-Election Day ballot count helped the Democratic Party candidate gain ground on the Republican nominee. And the GOP’s Richard Nixon twice enjoyed a “red shift” from post-Election Day vote counting.</p>
606
607 <p>A study co-authored by an MIT political scientist quantifies this effect by state, analyzes its causes, and shows why the 2020 election might indeed be decided after Nov. 3.</p>
608
609 <p>“It’s one of the reasons people are bracing for a bit of a rocky ride after the polls close,” says Charles Stewart, a professor in MIT’s Department of Political Science and co-author of a paper detailing the study’s results.</p>
610
611 <p>As the study shows, a growing share of votes since 1992 have been counted after Election Day; in 2016, it was about 10 percent of all votes. The use of provisional ballots and absentee ballots is the leading driver of this trend. Last time out, Hillary Clinton’s national popular-vote margin increased by 0.30 percentage points due to votes counted after Election Day.</p>
612
613 <p>Moreover, the Covid-19 pandemic seems likely to generate more absentee voting than ever. These factors have led many political commentators to speculate that President Donald Trump, who has been mostly critical of mail-in voting, might verbally claim victory on election night despite an unfinished vote count. This may especially be relevant to Pennsylvania and Michigan, which have little or no advance counting of mail-in ballots.</p>
614
615 <p>However, Stewart notes, we do not know what will unfold. Fully 42 states start counting absentee ballots before Election Day, and if voters return mail-in ballots unusually quickly, some absentee vote counts might wrap up routinely. In that case, “There could be more [issues] with Election Day voting than mail voting,” Stewart says.</p>
616
617 <p>Additionally, Stewart says, if Democrats are particularly focused on sending in absentee ballots early, “We could have a red shift in 2020 in some of these states, if Democratic ballots [have] already been scanned and preloaded, and if Republican ballots are the last ones, which will get counted on Wednesday or Thursday.”</p>
618
619 <p>The paper, “Explaining the Blue Shift in Election Canvassing,” is co-authored by Stewart, the Kenan Sahin Distinguished Professor of Political Science at MIT, and Edward Foley, the Charles W. Ebersold and Florence Whitcomb Ebersold Chair in Constitutional Law, and director of the election law program at Moritz College of Law at Ohio State University. It appeared this summer in the <em>Journal of Political Institutions and Political Economy</em>.</p>
620
621 <p><strong>Why more votes are counted later …</strong></p>
622
623 <p>To conduct the research, Foley and Stewart examined all presidential elections since 1948. First, to gain an overall sense of the size of the post-Election Day vote count, they compared the vote tabulations appearing in <em>The New York Times</em> on the Thursday after every Election Day with the eventual vote totals (using Dave Leip’s Election Atlas as the source for final results).</p>
624
625 <p>From 1948 through 1956, the number of votes counted after Election Day was higher than it is now, above 10 percent, which the researchers attribute to the slower forms of communication (and thus vote reporting) of the time. That number generally stayed under 5 percent for a few decades but ticked up in 1992 and again starting in 2004.</p>
626
627 <p>Two main factors likely account for this growth: greater use of provisional ballots and more mail-in voting (also known as absentee voting). In the first case, the Help America Vote Act (HAVA), passed by the U.S. Congress in 2002, modernized voting equipment and required all states to issue provisional ballots to voters.</p>
628
629 <p>Provisional ballots allow people whose registration is challenged at the polls to vote anyway; their ballot is evaluated again after Election Day. Prior to 2002, only half of the states used provisional ballots. In 2016, about 2.5 million provisional ballots were cast; about 1.7 million of those were fully or partially counted, with around 800,000 provisional ballots being rejected.</p>
630
631 <p>At the same time, voting by mail has grown in popularity. Using the federal Election Assistance Commission’s Election Administration and Voting Survey (EAVS) and U.S. Census Bureau data, Foley and Stewart conclude in their paper that “there is a correlation between the number of provisional and mail ballots that must be processed by a state’s election officials and the number of overtime votes” — that is, those counted after Election Day.</p>
632
633 <p>“The reforms after the 2000 election routinized some of these dynamics,” Stewart says, while state-level changes “removed [the need for] excuses for voting by mail.”</p>
634
635 <p><strong>… and why has the shift been blue?</strong></p>
636
637 <p>Still, if an increasing number of votes are counted after Election Day, why has that boosted the Democratic Party candidate? The post-Election Day vote count generated a 0.12 percentage point shift in the national popular vote in favor of John Kerry in 2004, a 0.35 point shift for Barack Obama in 2008, and a 0.39 point shift for Obama in 2012, before Clinton’s 0.30 point gain in 2016.</p>
638
639 <p>One explanation, which Foley and Stewart detail in the paper, is that Democrats are more likely to cast provisional ballots. In the Cooperative Congressional Election Study of 2016, they note, 60.1 percent of respondents who said they had cast a provisional ballot identified as Democrats, whereas only 47.8 percent of those who did not cast provisional ballots identified as Democrats.</p>
640
641 <p>Digging into state-level data, the scholars find the same pattern. In North Carolina, which has the most extensive public data about provisional ballots of any state, 39 percent of voters casting a provisional ballot in 2016 were Democrats, although just 34.6 percent of the state electorate consisted of Democrats.</p>
642
643 <p>But why are Democrats casting more provisional ballots in the first place? One reason, the scholars suggest, is that new voter registrations since 2000 have tended to favor the Democratic Party; many challenges that lead to provisional ballots being cast are due to either new voter registration records that are not reflected at the polls, or changes of address.</p>
644
645 <p>Stewart suggests another reason, though, which stems from the campaign side of politics.</p>
646
647 <p>“Starting in 2008, I think something else happened,” he says. “The Obama campaign recognized the strategic opportunity in some states to lock down the Democratic vote early, so that the election-day get-out-the-vote effort could be more [focused] and less costly. And ever since then Democratic [Party] strategists, more so than Republican [Party] strategists, have looked to mail balloting as a way of getting their votes in.”</p>
648
649 <p>Certainly the blue shift has not been constant. Nixon enjoyed a red shift of 0.20 percentage points in 1960, while narrowly losing to John F. Kennedy, and then a smaller red shift while winning in 1968.</p>
650
651 <p><strong>Eyes on the Midwest</strong></p>
652
653 <p>As Foley and Stewart also detail in the paper, states vary widely in how quickly they process votes. Florida starts counting absentee votes 22 days before the election. Conversely, Pennsylvania and Michigan, key states Trump won narrowly in 2016, have just implemented no-excuses absentee voting — but Pennsylvania will not start processing mail-in ballots until Election Day. Michigan will start processing mail-in ballots — taking them out of their envelopes, marking names off the voter list, and more — the day before the election and will feed them into vote-scanning machines on Election Day.</p>
654
655 <p>Another factor is whether states count absentee ballots that are postmarked by Election Day but arrive later. In 2016 in Washington, which uses entirely mail-in ballots, 31.3 percent of votes were counted after Election Day. In Oregon, which also is a vote-by-mail state, that figure was just 6.0 percent. Why? Washington allows ballots to be counted if they are received five days after Election Day, while in Oregon, ballots must be received by Election Day.</p>
656
657 <p>Those states are not likely to tip the outcome of the 2020 presidential election, and historically some of the biggest post-Election Day shifts have not, either. The single biggest shift the researchers found for any state in the 1948-2016 time period was a 6.9 percentage point shift for George Wallace in his home state of Alabama in 1968, but Nixon won the state anyway.</p>
658
659 <p>Still, in a few places, a relatively small shift could change the state and national results.</p>
660
661 <p>“When you do the math, you’re not talking about big [numbers of] votes,” Stewart says. “It’s going to be outcome-determinative only under a narrow range of conditions. It’s a game of inches.”</p>
662 </content:encoded>
663 <media:content url="https://news.mit.edu/sites/default/files/styles/news_article__cover_image__original/public/images/202010/MIT-Blue-Shift-01-Press.jpg?itok=XA__gNxd" medium="image" type="image/jpeg" width="390" height="260">
664 <media:description type="plain">New research looks at how post-Election Day ballot counts have affected the outcomes of past elections. MIT Professor Charles Stewart and colleagues quantify this “blue shift” or “red shift” effect by state, analyze its causes, and show why the 2020 election might be decided after Nov. 3.</media:description>
665 <media:credit>Image: MIT News</media:credit>
666 </media:content>
667 </item>
668 <item>
669 <title>Electronic design tool morphs interactive objects</title>
670 <link>https://news.mit.edu/2020/morphsensor-morphs-interactive-objects-1022</link>
671 <description>MorphSensor lets users digitally model an object’s form and electronic function in one integrated space.</description>
672 <pubDate>Thu, 22 Oct 2020 09:00:00 -0400</pubDate>
673 <guid isPermaLink="true">https://news.mit.edu/2020/morphsensor-morphs-interactive-objects-1022</guid>
674 <dc:creator>Rachel Gordon | MIT CSAIL</dc:creator>
675 <content:encoded><p>We’ve come a long way since the first 3D-printed item came to us by way of an eye wash cup, to now being able to rapidly fabricate things like <a href="https://amfg.ai/2019/05/28/7-exciting-examples-of-3d-printing-in-the-automotive-industry/">car parts</a>, <a href="https://www.sculpteo.com/en/3d-learning-hub/best-articles-about-3d-printing/3d-printed-instruments/">musical instruments</a>, and even biological <a href="https://www.medicaldevice-network.com/features/3d-printing-in-the-medical-field-applications/">tissues and organoids</a>.&nbsp;</p>
676
677 <p>While much of these objects can be freely designed and quickly made, the addition of electronics to embed things like sensors, chips, and tags usually requires that you design both separately, making it difficult to create items where the added functions are easily integrated with the form.&nbsp;</p>
678
679 <p>Now, a 3D design environment from MIT’s <a href="https://www.csail.mit.edu/">Computer Science and Artificial Intelligence Laboratory</a> (CSAIL) lets users iterate an object’s shape and electronic function in one cohesive space, to add existing sensors to early-stage prototypes.</p>
680
681 <p>The team tested the system, called MorphSensor, by modeling an N95 mask with a humidity sensor, a temperature-sensing ring, and glasses that monitor light absorption to protect eye health.</p>
682 <p>MorphSensor automatically converts electronic designs into 3D models, and then lets users iterate on the geometry and manipulate active sensing parts. This might look like a 2D image of a pair of AirPods and a sensor template, where a person could edit the design until the sensor is embedded, printed, and taped onto the item.&nbsp;</p>
683
684 <p>To test the effectiveness of MorphSensor, the researchers created an evaluation based on standard industrial assembly and testing procedures. The data showed that MorphSensor could match the off-the-shelf sensor modules with small error margins, for both the analog and digital sensors.</p>
685
686 <p>“MorphSensor fits into my long-term vision of something called ‘rapid function prototyping’, with the objective to create interactive objects where the functions are directly integrated with the form and fabricated in one go, even for non-expert users,” says CSAIL PhD student Junyi Zhu, lead author on a new paper about the project. “This offers the promise that, when prototyping, the object form could follow its designated function, and the function could adapt to its physical form.”&nbsp;</p>
687
688 <p><strong>MorphSensor in action&nbsp;</strong></p>
689
690 <p>Imagine being able to have your own design lab where, instead of needing to buy new items, you could cost-effectively update your own items using a single system for both design and hardware.&nbsp;</p>
691
692 <p>For example, let’s say you want to update your face mask to monitor surrounding air quality. Using MorphSensor, users would first design or import the 3D face mask model and sensor modules from either MorphSensor's database or online open-sourced files. The system would then generate a 3D model with individual electronic components (with airwires connected between them) and color-coding to highlight the active sensing components.&nbsp;&nbsp;</p>
693
694 <p>Designers can then drag and drop the electronic components directly onto the face mask, and rotate them based on design needs. As a final step, users draw physical wires onto the design where they want them to appear, using the system’s guidance to connect the circuit.&nbsp;</p>
695
696 <p>Once satisfied with the design, the "morphed sensor" can be rapidly fabricated using an&nbsp;inkjet printer and conductive tape, so it can be adhered to the object. Users can also outsource the design to a professional fabrication house.&nbsp;&nbsp;</p>
697
698 <p>To test their system, the team iterated on EarPods for sleep tracking, which only took 45 minutes to design and fabricate. They also updated a “weather-aware” ring to provide weather advice, by integrating a temperature sensor with the ring geometry. In addition,&nbsp;they manipulated an N95 mask to monitor its substrate contamination, enabling it to alert its user when the mask needs to be replaced.</p>
699
700 <p>In its current form, MorphSensor helps designers maintain connectivity of the circuit at all times, by highlighting which components contribute to the actual sensing. However, the team notes it would be beneficial to expand this set of support tools even further, where future versions could potentially merge electrical logic of multiple sensor modules together to eliminate redundant components and circuits and save space (or preserve the object form).&nbsp;</p>
701
702 <p>Zhu wrote the paper alongside MIT graduate student Yunyi Zhu; undergraduates Jiaming Cui, Leon Cheng, Jackson Snowden, and Mark Chounlakone; postdoc Michael Wessely; and Professor Stefanie Mueller. The team will virtually present their paper at the ACM User Interface Software and Technology Symposium.&nbsp;</p>
703
704 <p>This material is based upon work supported by the National Science Foundation.</p>
705 </content:encoded>
706 <media:content url="https://news.mit.edu/sites/default/files/styles/news_article__cover_image__original/public/images/202010/MIT-MorphSensor-Glasses1.png?itok=ZobkZeV_" medium="image" type="image/jpeg" width="390" height="260">
707 <media:description type="plain">An MIT team used MorphSensor to design multiple applications, including a pair of glasses that monitor light absorption to protect eye health.</media:description>
708 <media:credit>Photo courtesy of the researchers.</media:credit>
709 </media:content>
710 </item>
711 <item>
712 <title>Solar-powered system extracts drinkable water from “dry” air</title>
713 <link>https://news.mit.edu/2020/solar-extracts-drinkable-water-1014</link>
714 <description>MIT engineers have made their initial design more practical, efficient, and scalable.</description>
715 <pubDate>Wed, 14 Oct 2020 11:00:00 -0400</pubDate>
716 <guid isPermaLink="true">https://news.mit.edu/2020/solar-extracts-drinkable-water-1014</guid>
717 <dc:creator>David L. Chandler | MIT News Office</dc:creator>
718 <content:encoded><p>Researchers at MIT and elsewhere have significantly boosted the output from a system that can extract drinkable water directly from the air even in dry regions, using heat from the sun or another source.</p>
719
720 <p>The system, which builds on a design <a href="https://news.mit.edu/2017/MOF-device-harvests-fresh-water-from-air-0414">initially developed</a> three years ago at MIT by members of the same team, brings the process closer to something that could become a practical water source for remote regions with limited access to water and electricity. The findings are described today in the journal <em>Joule</em>, in <a href="https://www.cell.com/joule/fulltext/S2542-4351(20)30444-X" target="_blank">a paper</a> by Professor Evelyn Wang, who is head of MIT’s Department of Mechanical Engineering; graduate student Alina LaPotin; and six others at MIT and in Korea and Utah.</p>
721
722 <p>The earlier device <a href="https://news.mit.edu/2018/field-tests-device-harvests-water-desert-air-0322">demonstrated</a> by Wang and her co-workers provided a proof of concept for the system, which harnesses a temperature difference within the device to allow an adsorbent material — which collects liquid on its surface — to draw in moisture from the air at night and release it the next day. When the material is heated by sunlight, the difference in temperature between the heated top and the shaded underside makes the water release back out of the adsorbent material. The water then gets condensed on a collection plate.</p>
723
724 <p>But that device required the use of specialized materials called metal organic frameworks, or MOFs, which are expensive and limited in supply, and the system’s water output was not sufficient for a practical system. Now, by incorporating a second stage of desorption and condensation, and by using a readily available adsorbent material, the device’s output has been significantly increased, and its scalability as a potentially widespread product is greatly improved, the researchers say.</p>
725 <p>Wang says the team felt that “It’s great to have a small prototype, but how can we get it into a more scalable form?” The new advances in design and materials have now led to progress in that direction.</p>
726
727 <p>Instead of the MOFs, the new design uses an adsorbent material called a zeolite, which in this case is composed of a microporous iron aluminophosphate. The material is widely available, stable, and has the right adsorbent properties to provide an efficient water production system based just on typical day-night temperature fluctuations and heating with sunlight.</p>
728
729 <p>The two-stage design developed by LaPotin makes clever use of the heat that is generated whenever water changes phase. The sun’s heat is collected by a solar absorber plate at the top of the box-like system and warms the zeolite, releasing the moisture the material has captured overnight. That vapor condenses on a collector plate — a process that releases heat as well. The collector plate is a copper sheet directly above and in contact with the second zeolite layer, where the heat of condensation is used to release the vapor from that subsequent layer. Droplets of water collected from each of the two layers can be funneled together into a collecting tank.</p>
730
731 <p>In the process, the overall productivity of the system, in terms of its potential liters per day per square meter of solar collecting area (LMD), is approximately doubled compared to the earlier version, though exact rates depend on local temperature variations, solar flux, and humidity levels. In the initial prototype of the new system, tested on a rooftop at MIT before the pandemic restrictions, the device produced “orders of magnitude” more total water than&nbsp;the earlier version, Wang says.</p>
732
733 <p>While similar two-stage systems have been used for other applications such as desalination, Wang says, “I think no one has really pursued this avenue” of using such a system for atmospheric water harvesting (AWH), as such technologies are known.</p>
734
735 <p>Existing AWH approaches include fog harvesting and dew harvesting, but both have significant limitations. Fog harvesting only works with 100 percent relative humidity, and is currently used only in a few coastal deserts, while dew harvesting requires energy-intensive refrigeration to provide cold surfaces for moisture to condense on — and still requires humidity of at least 50 percent, depending on the ambient temperature.</p>
736
737 <p>By contrast, the new system can work at humidity levels as low as 20 percent and requires no energy input other than sunlight or any other available source of low-grade heat.</p>
738
739 <p>LaPotin says that the key is this two-stage architecture; now that its effectiveness has been shown, people can search for even better adsorbent materials that could further drive up the production rates. The present production rate of about 0.8 liters of water per square meter per day may be adequate for some applications, but if this rate can be improved with some further fine-tuning and materials choices, this could become practical on a large scale, she says. Already, materials are in development that have an adsorption about five times greater than this particular zeolite and could lead to a corresponding increase in water output, according to Wang.</p>
740
741 <p>The team continues work on refining the materials and design of the device and adapting it to specific applications, such as a portable version for military field operations. The two-stage system could also be adapted to other kinds of water harvesting approaches that use multiple thermal cycles per day, fed by a different heat source rather than sunlight, and thus could produce higher daily outputs.</p>
742
743 <p>“This is an interesting and technologically significant work indeed,” says Guihua Yu, a professor of materials science and mechanical engineering at the University of Texas at Austin, who was not associated with this work. “It represents a powerful engineering approach for designing a dual-stage AWH device to achieve higher water production yield, marking a step closer toward practical solar-driven water production,” he says.</p>
744
745 <p>Yu adds that “Technically, it is beautiful that one could reuse the heat released simply by this dual-stage design, to better confine the solar energy in the water harvesting system to improve energy efficiency and daily water productivity. Future research lies in improving this prototype system with low cost components and simple configuration with minimized heat loss.”</p>
746
747 <p>The research team includes Yang Zhong, Lenan Zhang, Lin Zhao, and Arny Leroy at MIT; Hyunho Kim at the Korea Institute of Science and Technology; and Sameer Rao at the University of Utah. The work was supported by the Abdul Latif Jameel Water and Food Systems Lab (J-WAFS) at MIT.</p>
748 </content:encoded>
749 <media:content url="https://news.mit.edu/sites/default/files/styles/news_article__cover_image__original/public/images/202010/MIT-Water-Harvester-01-PRESS.jpg?itok=cA9GoTKe" medium="image" type="image/jpeg" width="390" height="260">
750 <media:description type="plain">A prototype of the new two-stage water harvesting system (center right), was tested on an MIT rooftop. The device, which was connected to a laptop for data collection and was mounted at an angle to face the sun, has a black solar collecting plate at the top, and the water it produced flowed into two tubes at bottom. </media:description>
751 <media:credit>Image: Alina LaPotin</media:credit>
752 </media:content>
753 </item>
754 <item>
755 <title>Technique recovers lost single-cell RNA-sequencing information</title>
756 <link>https://news.mit.edu/2020/sc-rna-sequencing-fidelity-1013</link>
757 <description>Boosting the efficiency of single-cell RNA-sequencing helps reveal subtle differences between healthy and dysfunctional cells.</description>
758 <pubDate>Tue, 13 Oct 2020 11:39:00 -0400</pubDate>
759 <guid isPermaLink="true">https://news.mit.edu/2020/sc-rna-sequencing-fidelity-1013</guid>
760 <dc:creator>Anne Trafton | MIT News Office</dc:creator>
761 <content:encoded><p>Sequencing RNA from individual cells can reveal a great deal of information about what those cells are doing in the body. MIT researchers have now greatly boosted the amount of information gleaned from each of those cells, by modifying the commonly used Seq-Well technique.</p>
762
763 <p>With their new approach, the MIT team could extract 10 times as much information from each cell in a sample. This increase should enable scientists to learn much more about the genes that are expressed in each cell, and help them to discover subtle but critical differences between healthy and dysfunctional cells.</p>
764
765 <p>“It’s become clear that these technologies have transformative potential for understanding complex biological systems. If we look across a range of different datasets, we can really understand the landscape of health and disease, and that can give us information as to what therapeutic strategies we might employ,” says Alex K. Shalek, an associate professor of chemistry, a core member of the Institute for Medical Engineering and Science (IMES), and an extramural member of the Koch Institute for Integrative Cancer Research at MIT. He is also a member of the Ragon Institute of MGH, MIT and Harvard and an institute member of the Broad Institute.</p>
766
767 <p>In a study appearing this week in <em>Immunity</em>, the research team demonstrated the power of this technique by analyzing approximately 40,000 cells from patients with five different skin diseases. Their analysis of immune cells and other cell types revealed many differences between the five diseases, as well as some common features.</p>
768
769 <p>“This is by no means an exhaustive compendium, but it’s a first step toward understanding the spectrum of inflammatory phenotypes, not just within immune cells, but also within other skin cell types,” says Travis Hughes, an MD/PhD student in the Harvard-MIT Program in Health Sciences and Technology and one of the lead authors of the paper.</p>
770
771 <p>Shalek and J. Christopher Love, the Raymond A. and Helen E. St. Laurent Professor of Chemical Engineering and a member of the Koch Institute and Ragon Institute, are the senior authors of the study. MIT graduate student Marc Wadsworth and former postdoc Todd Gierahn are co-lead authors of the paper with Hughes.</p>
772
773 <p><strong>Recapturing information </strong></p>
774
775 <p>A few years ago, Shalek, Love, and their colleagues developed a method called <a href="https://news.mit.edu/2017/making-single-cell-rna-sequencing-widely-available-0213">Seq-Well</a>, which can rapidly sequence RNA from many single cells at once. This technique, like other high-throughput approaches, doesn’t pick up as much information per cell as some slower, more expensive methods for sequencing RNA. In their current study, the researchers set out to recapture some of the information that the original version was missing.</p>
776
777 <p>“If you really want to resolve features that distinguish diseases, you need a higher level of resolution than what’s been possible,” Love says. “If you think of cells as packets of information, being able to measure that information more faithfully gives much better insights into what cell populations you might want to target for drug treatments, or, from a diagnostic standpoint, which ones you should monitor.”</p>
778
779 <p>To try to recover that additional information, the researchers focused on one step where they knew that data were&nbsp;being lost. In that step, cDNA molecules, which are copies of the RNA transcripts from each cell, are amplified through a process called polymerase chain reaction (PCR). This amplification is necessary to get enough copies of the DNA for sequencing. Not all cDNA was getting amplified, however. To boost the number of molecules that made it past this step, the researchers changed how they tagged the cDNA with a second “primer” sequence, making it easier for PCR enzymes to amplify these molecules.</p>
780
781 <p>Using this technique, the researchers showed they could generate much more information per cell. They saw a fivefold increase in the number of genes that could be detected, and a tenfold increase in the number of RNA transcripts recovered per cell. This extra information about important genes, such as those encoding cytokines, receptors found on cell surfaces, and transcription factors, allows the researchers to identify subtle differences between cells.</p>
782
783 <p>“We were able to vastly improve the amount of per cell information content with a really simple molecular biology trick, which was easy to incorporate into the existing workflow,” Hughes says.</p>
784
785 <p><strong>Signatures of disease</strong></p>
786
787 <p>Using this technique, the researchers analyzed 19 patient skin biopsies, representing five different skin diseases — psoriasis, acne, leprosy, alopecia areata (an autoimmune disease that causes hair loss), and granuloma annulare (a chronic degenerative skin disorder). They uncovered some similarities between disorders — for example, similar populations of inflammatory T cells appeared active in both leprosy and granuloma annulare.</p>
788
789 <p>They also uncovered some features that were unique to a particular disease. In cells from several psoriasis patients, they found that cells called keratinocytes express genes that allow them to proliferate and drive the inflammation seen in that disease.</p>
790
791 <p>The data generated in this study should also offer a valuable resource to other researchers who want to delve deeper into the biological differences between the cell types studied.</p>
792
793 <p>“You never know what you’re going to want to use these datasets for, but there’s a tremendous opportunity in having measured everything,” Shalek says. “In the future, when we need to repurpose them and think about particular surface receptors, ligands, proteases, or other genes, we will have all that information at our fingertips.”</p>
794
795 <p>The technique could also be applied to many other diseases and cell types, the researchers say. They have begun using it to study cancer and infectious diseases such as tuberculosis, malaria, HIV, and Ebola, and they are also using it to analyze immune cells involved in food allergies. They have also made the new technique available to other researchers who want to use it or adapt the underlying approach for their own single-cell studies.</p>
796
797 <p>The research was funded by the Koch Institute Support (core) Grant from the National Institutes of Health, the Bridge Project of the Koch Institute and the Dana-Farber/Harvard Cancer Center, the Food Allergy Science Initiative at the Broad Institute, the National Institutes of Health, a Beckman Young Investigator Award, a Sloan Research Fellowship in Chemistry, the Pew-Stewart Scholar Award, and the Bill and Melinda Gates Foundation.</p>
798 </content:encoded>
799 <media:content url="https://news.mit.edu/sites/default/files/styles/news_article__cover_image__original/public/images/202010/MIT-FasterSequencing-01-press.jpg?itok=JYZ5-FYg" medium="image" type="image/jpeg" width="390" height="260">
800 <media:description type="plain">MIT researchers have greatly boosted the amount of information that can be obtained using Seq-Well, a technique for rapidly sequencing RNA from single cells. This advance should enable scientists to learn much more about the critical genes that are expressed in each cell, and help them to discover subtle differences between healthy and dysfunctional cells for designing new preventions and cures. This image illustrates the improved resolution, right, using the new technique. </media:description>
801 <media:credit>Courtesy of the researchers. Edited by MIT News.</media:credit>
802 </media:content>
803 </item>
804 <item>
805 <title>To make mini-organs grow faster, give them a squeeze</title>
806 <link>https://news.mit.edu/2020/cell-crowding-organs-grow-1013</link>
807 <description>Study finds that compressing cells, and crowding their contents, can coax them to grow and divide.</description>
808 <pubDate>Tue, 13 Oct 2020 11:30:00 -0400</pubDate>
809 <guid isPermaLink="true">https://news.mit.edu/2020/cell-crowding-organs-grow-1013</guid>
810 <dc:creator>Jennifer Chu | MIT News Office</dc:creator>
811 <content:encoded><p>The closer people are physically to one another, the higher the chance for exchange, of things like ideas, information, and even infection. Now researchers at MIT and Boston Children’s Hospital have found that, even in the microscopic environment within a single cell, physical crowding increases the chance for interactions, in a way that can significantly alter a cell’s health and development.</p>
812
813 <p>In a paper published today in the journal <em>Cell Stem Cell</em>, the researchers have shown that physically squeezing cells, and crowding their contents, can trigger cells to grow and divide faster than they normally would.</p>
814
815 <p>While squeezing something to make it grow may sound counterintuitive, the team has an explanation: Squeezing acts to wring water out of a cell. With less water to swim in, proteins and other cell constituents are packed closer together. And when certain proteins are brought in close proximity, they can trigger cell signaling and activate genes within the cell.</p>
816
817 <p>In their new study, the scientists found that squeezing intestinal cells triggered proteins to cluster along a specific signaling pathway, which can help cells maintain their stem-cell state, an undifferentiated state in which they can quickly grow and divide into more specialized cells. Ming Guo, associate professor of mechanical engineering at MIT, says that if cells can simply be squeezed to promote their “stemness,” they can then be directed to quickly build up miniature organs, such as artificial intestines or colons, which could then be used as platforms to understand organ function and test drug candidates for various diseases, and even as transplants for regenerative medicine.</p>
818
819 <p>Guo’s co-authors are lead author Yiwei Li, Jiliang Hu, and Qirong Lin from MIT, and Maorong Chen, Ren Sheng, and Xi He of Boston Children’s Hospital.</p>
820
821 <p><strong>Packed in</strong></p>
822
823 <p>To study squeezing’s effect on cells, the researchers mixed various cell types in solutions that solidified as rubbery slabs of hydrogel. To squeeze the cells, they placed weights on the hydrogel’s surface, in the form of either a quarter or a dime.</p>
824
825 <p>“We wanted to achieve a significant amount of cell size change, and those two weights can compress the cell by something like 10 to 30 percent of their total volume,” Guo explains.</p>
826
827 <p>The team used a confocal microscope to measure in 3D how individual cells’ shapes changed as each sample was compressed. As they expected, the cells shrank with pressure. But did squeezing also affect the cell’s contents? To answer this, the researchers first looked to see whether a cell’s water content changed. If squeezing acts to wring water out of a cell, the researchers reasoned that the cells should be less hydrated, and stiffer as a result.</p>
828
829 <p>They measured the stiffness of cells before and after weights were applied, using optical tweezers, a laser-based technique that Guo’s lab has employed for years to study interactions within cells, and found that indeed, cells stiffened with pressure. They also saw that there was less movement within cells that were squeezed, suggesting that their contents were more packed than usual.</p>
830
831 <p>Next, they looked at whether there were changes in the interactions between certain proteins in the cells, in response to cells being squeezed. They focused on several proteins that are known to trigger Wnt/β-catenin signaling, which is involved in cell growth and maintenance of “stemness.”</p>
832
833 <p>“In general, this pathway is known to make a cell more like a stem cell,” Guo says. “If you change this pathway’s activity, how cancer progresses and how embryos develop have been shown to be very different. So we thought we could use this pathway to demonstrate how cell crowding is important.”</p>
834
835 <p><strong>A “refreshing” path</strong></p>
836
837 <p>To see whether cell squeezing affects the Wnt pathway, and how fast a cell grows, the researchers grew small organoids — miniature organs, and in this case, clusters of cells that were collected from the intestines of mice.</p>
838
839 <p>“The Wnt pathway is particularly important in the colon,” Guo says, pointing out that the cells that line the human intestine are constantly being replenished. The Wnt pathway, he says, is essential for maintaining intestinal stem cells, generating new cells, and “refreshing” the intestinal lining. &nbsp;</p>
840
841 <p>He and his colleagues grew intestinal organoids, each measuring about half a millimeter, in several Petri dishes, then “squeezed” the organoids by infusing the dishes with polymers. This influx of polymers increased the osmotic pressure surrounding each organoid and forced water out of their cells. The team observed that as a result, specific proteins involved in activating the Wnt pathway were packed closer together, and were more likely to cluster to turn on the pathway and its growth-regulating genes.</p>
842
843 <p>The upshot: Those organoids that were squeezed actually grew larger and more quickly, with more stem cells on their surface than those that were not squeezed.</p>
844
845 <p>“The difference was very obvious,” Guo says. “Whenever you apply pressure, the organoids grow even bigger, with a lot more stem cells.”</p>
846
847 <p>He says the results demonstrate how squeezing can affect an organoid’s growth. The findings also show that a cell’s behavior can change depending on the amount of water that it contains.</p>
848
849 <p>“This is very general and broad, and the potential impact is profound, that cells can simply tune how much water they have to tune their biological consequences,” Guo says.</p>
850
851 <p>Going forward, he and his colleagues plan to explore cell squeezing as a way to speed up the growth of artificial organs that scientists may use to test new, personalized drugs.</p>
852
853 <p>“I could take my own cells and transfect them to make stem cells that can then be developed into a lung or intestinal organoid that would mimic my own organs,” Guo says. “I could then apply different pressures to make organoids of different size, then try different drugs. I imagine there would be a lot of possibilities.”</p>
854
855 <p>This research is supported, in part, by the National Cancer Institute and the Alfred P. Sloan Foundation.</p>
856 </content:encoded>
857 <media:content url="https://news.mit.edu/sites/default/files/styles/news_article__cover_image__original/public/images/202010/MIT-Cell-Crowding-01-Press.jpg?itok=evlOyQZq" medium="image" type="image/jpeg" width="390" height="260">
858 <media:description type="plain">In this image, the cell division marker Ki67 shows that the number of dividing cells in organoids increases under compression, as seen in the bottom row, during three passages.</media:description>
859 <media:credit>Credit: Yiwei Li</media:credit>
860 </media:content>
861 </item>
862 <item>
863 <title>Translating lost languages using machine learning</title>
864 <link>https://news.mit.edu/2020/translating-lost-languages-using-machine-learning-1021</link>
865 <description>System developed at MIT CSAIL aims to help linguists decipher languages that have been lost to history.</description>
866 <pubDate>Wed, 21 Oct 2020 00:00:00 -0400</pubDate>
867 <guid isPermaLink="true">https://news.mit.edu/2020/translating-lost-languages-using-machine-learning-1021</guid>
868 <dc:creator>Adam Conner-Simons | MIT CSAIL</dc:creator>
869 <content:encoded><p>Recent research suggests that most languages that have ever existed <a href="https://www.uh.edu/engines/epi2723.htm">are no longer spoken</a>. Dozens of these dead languages are also <a href="https://en.wikipedia.org/wiki/Undeciphered_writing_systems#Proto-writing" target="_blank">considered to be lost, or “undeciphered”</a> — that is, we don’t know enough about their grammar, vocabulary, or syntax to be able to actually understand their texts.</p>
870
871 <p>Lost languages are more than a mere academic curiosity; without them, we miss an entire body of knowledge about the people who spoke them. Unfortunately, most of them have such minimal records that scientists can’t decipher them by using machine-translation algorithms like Google Translate. Some don’t have a well-researched “relative” language to be compared to, and often lack traditional dividers like white space and punctuation. (To illustrate, imaginetryingtodecipheraforeignlanguagewrittenlikethis.)</p>
872
873 <p>However, researchers at MIT’s <a href="http://csail.mit.edu">Computer Science and Artificial Intelligence Laboratory</a> (CSAIL) recently made a major development in this area: a new system that has been shown to be able to automatically decipher a lost language, without needing advanced knowledge of its relation to other languages. They also showed that their system can itself determine relationships between languages, and they used it to corroborate recent scholarship suggesting that the language of Iberian is not actually related to Basque.</p>
874
875 <p>The team’s ultimate goal is for the system to be able to decipher lost languages that have eluded linguists for decades, using just a few thousand words.</p>
876
877 <p>Spearheaded by MIT Professor Regina Barzilay, the system relies on several principles grounded in insights from historical linguistics, such as the fact that languages generally only evolve in certain predictable ways. For instance, while a given language rarely adds or deletes an entire sound, certain sound substitutions are likely to occur. A word with a “p” in the parent language may change into a “b” in the descendant language, but changing to a “k” is less likely due to the significant pronunciation gap.</p>
878
879 <p>By incorporating these and other linguistic constraints, Barzilay and MIT PhD student Jiaming Luo <a href="http://people.csail.mit.edu/j_luo/assets/publications/DecipherUnsegmented.pdf" target="_blank">developed a decipherment algorithm</a> that can handle the vast space of possible transformations and the scarcity of a guiding signal in the input. The algorithm learns to embed language sounds into a multidimensional space where differences in pronunciation are reflected in the distance between corresponding vectors. This design enables them to capture pertinent patterns of language change and express them as computational constraints. The resulting model can segment words in an ancient language and map them to counterparts in a related language.&nbsp;&nbsp;</p>
880
881 <p>The project builds on <a href="https://arxiv.org/abs/1906.06718" target="_blank">a paper Barzilay and Luo wrote last year</a> that deciphered the dead languages of Ugaritic and Linear B, the latter of which had previously taken decades for humans to decode. However, a key difference with that project was that the team knew that these languages were related to early forms of Hebrew and Greek, respectively.</p>
882
883 <p>With the new system, the relationship between languages is inferred by the algorithm. This question is one of the biggest challenges in decipherment. In the case of Linear B, it took several decades to discover the correct known descendant. For Iberian, the scholars still cannot agree on the related language: Some argue for Basque, while others refute this hypothesis and claim that Iberian doesn’t relate to any known language.&nbsp;</p>
884
885 <p>The proposed algorithm can assess the proximity between two languages; in fact, when tested on known languages, it can even accurately identify language families. The team applied their algorithm to Iberian considering Basque, as well as less-likely candidates from Romance, Germanic, Turkic, and Uralic families. While Basque and Latin were closer to Iberian than other languages, they were still too different to be considered related.&nbsp;</p>
886
887 <p>In future work, the team hopes to expand their work beyond the act of connecting texts to related words in a known language — an approach referred to as “cognate-based decipherment.” This paradigm assumes that such a known language exists, but the example of Iberian shows that this is not always the case. The team’s new approach would involve identifying semantic meaning of the words, even if they don’t know how to read them.&nbsp;</p>
888
889 <p>“For instance, we may identify all the references to people or locations in the document which can then be further investigated in light of the known historical evidence,” says Barzilay. “These methods of ‘entity recognition’ are commonly used in various text processing applications today and are highly accurate, but the key research question is whether the task is feasible without any training data in the ancient language.”&nbsp; &nbsp; &nbsp; .</p>
890
891 <p>The project was supported, in part, by the Intelligence Advanced Research Projects Activity (IARPA).</p>
892 </content:encoded>
893 <media:content url="https://news.mit.edu/sites/default/files/styles/news_article__cover_image__original/public/images/202010/Ugaritic-language.jpg?itok=KrlkJapf" medium="image" type="image/jpeg" width="390" height="260">
894 <media:description type="plain">In developing a machine learning system to help decipher lost languages, MIT researchers studied the language of Ugaritic, which is related to Hebrew and has previously been analyzed and deciphered by linguists. </media:description>
895 <media:credit>Photo: S.R.K. Branavan</media:credit>
896 </media:content>
897 </item>
898 <item>
899 <title>How a worm may yield insights into the gut-brain relationship</title>
900 <link>https://news.mit.edu/2020/how-worm-could-yield-gut-brain-insights-guerrin-madan-1027</link>
901 <description>Gurrein Madan, brain and cognitive sciences graduate student and MathWorks Fellow, studies gut–brain signaling with implications for human health.</description>
902 <pubDate>Tue, 27 Oct 2020 11:30:00 -0400</pubDate>
903 <guid isPermaLink="true">https://news.mit.edu/2020/how-worm-could-yield-gut-brain-insights-guerrin-madan-1027</guid>
904 <dc:creator>Alison Gold | School of Science</dc:creator>
905 <content:encoded><p>The naked eye can barely spot the transparent nematodes at the center of PhD student Gurrein Madan’s neuroscience research. While <em>C. elegans</em> worms may initially seem an unassuming test subject for a graduate student who investigates the intricacies of gut-brain signaling, many of the genes found in <em>C. elegans</em> have counterparts in the human brain. Gurrein’s research could yield new insights into the gut-brain relationship, which may have practical health implications for humans.</p>
906
907 <p>Gurrein works in the lab of Steven Flavell, the Lister Brothers Career Development Assistant Professor in the Picower Institute for Learning and Memory and Department of Brain and Cognitive Sciences at MIT. There, researchers address some of neuroscience’s most essential questions, using <em>C. elegans</em> as a model. The lab centers around understanding how neuromodulatory systems — such as those cellular systems that release, and are stimulated by, serotonin — affect animal behavior. The millimeter-long <em>C. elegans</em> are an ideal model for this work because their nervous system, with just 302 neurons, has been well-characterized: “It is the only animal on the planet where there is a full blueprint of how all of its brain cells are wired together,” says Flavell. Combined with cutting-edge genetic and neural imaging technologies, the nematode model affords mechanistic studies of behavior from the scale of molecules to the whole brain.</p>
908
909 <p>Gurrein was recently named one of this year’s School of Science MathWorks Fellows. The fellowship is a one-year renewable opportunity for graduate students in the School of Science who use the software MATLAB to make impressive strides in their research. Funding for the fellowship is provided with support from MathWorks, founded by its president, John N. Little ’78. MATLAB is used extensively by faculty, students, and researchers across the world — and MIT’s campus — to develop algorithms, computations, and simulations.</p>
910
911 <p>Gurrein’s project specifically looks at the neurons that line the <em>C. elegan</em>s gut. These enteric neurons detect food and respond to changes in the animal’s nutritional state while receiving feedback from other parts of the brain. Gurrein <a href="https://www.sciencedirect.com/science/article/pii/S0092867418315150?via%3Dihub" target="_blank">studies the class of neurons that release serotonin</a>, which has a profound influence on the animal’s feeding behavior.</p>
912
913 <p>“Currently, we are investigating what receptors expressed in these enteric neurons regulate the neurons’ response to food, as well as to feedback from the rest of the nervous system,” Gurrein says. “By using genetics and neural imaging techniques, we attempt to uncover new molecular players involved in gut-brain signaling.”</p>
914
915 <p>Dysregulation of gut-brain signaling has been linked to psychiatric disorders, such as depression and autism spectrum disorder, in humans. Diverse molecules, including neurotransmitters and inflammatory molecules, mediate the two-way communication between the gut and the brain. However, the specific pathways behind this relationship are not well understood. Gurrein hopes to <a href="https://www.biorxiv.org/content/10.1101/2020.02.04.934547v3" target="_blank">uncover more about the signaling mechanisms</a> driving the connection.</p>
916
917 <p>“Much of our understanding of the fundamental pathways that control animal development and function comes from studies that originated in <em>C. elegans</em>, where basic genetic pathways were rapidly discovered,” says Flavell. “Lo and behold, in humans, the same pathways control the same cellular processes. Many of these pathways have then become targets for drug development to treat human disease.”</p>
918
919 <p>Using MATLAB at nearly every step of her research — from data collection and processing to analysis — Gurrein was an excellent candidate for the MathWorks fellowship. “I was excited to apply for two reasons. First, the fellowship was open to international students. Typically, international students are ineligible to apply to most fellowships out there. Second, MATLAB serves as the critical platform for comprehensively handling my data,” Gurrein says.</p>
920
921 <p>Gurrein grew up in Amritsar in northwest India. Early in high school, Gurrein was placed in the sciences track, and upon graduation traveled to the United States for her undergraduate degree. During her sophomore year at Swarthmore College, she began research in a neurobiology lab and quickly realized how much she enjoyed the process of conducting scientific research. Moreover, she found the interdisciplinary nature of the neuroscience field exciting. After graduating with a BA in neuroscience in 2017, she immediately began her PhD at MIT.&nbsp;</p>
922
923 <p>“I really like the innovative aspect of a PhD,” Gurrein says. “We are trained to expand the limits of what is known in our fields by being persistent, constantly troubleshooting, and coming up with new approaches to probe a question. I was initially considering medical school, but my research experiences led me to think that a PhD was probably a better fit for me.”</p>
924
925 <p>Gurrein’s colleagues noticed her enthusiasm for scientific discovery immediately. Flavell says she impressed him “right out of the gate.” Within her first six months in the lab, Gurrein was instrumental in designing new experiments, conducting behavioral assays, and making notable discoveries that made their way into publications.</p>
926
927 <p>“Gurrein has an enormous amount of drive and energy, always trying her best to make impactful discoveries,” says Flavell. “We have meetings once a week, and she always brings a critical eye to her own work, thinking about her datasets, what they mean, and how they give rise to new research directions. She wants to make sure the data she gets is really convincing and setting her on a path to making a true discovery.”</p>
928
929 <p>“The general topic of how the gut is influencing the brain is a relatively new field,” Gurrein says. “I think there is a lot of space for novel, exciting contributions.”</p>
930 </content:encoded>
931 <media:content url="https://news.mit.edu/sites/default/files/styles/news_article__cover_image__original/public/images/202010/Gurrein-Madan-a1.jpg?itok=nygAY_dW" medium="image" type="image/jpeg" width="390" height="260">
932 <media:description type="plain">Brain and cognitive sciences PhD student Gurrein Madan was awarded a MathWorks Fellowship for using MATLAB to investigate the gut-brain connection. </media:description>
933 <media:credit>Photo: Steph Stevens</media:credit>
934 </media:content>
935 </item>
936 <item>
937 <title>Leveraging a 3D printer “defect” to create a new quasi-textile</title>
938 <link>https://news.mit.edu/2020/defextiles-leveraging-3d-printer-defect-to-create-quasi-textiles-1020</link>
939 <description>Tulle-like DefeXtiles can be 3D printed with no custom software or hardware.</description>
940 <pubDate>Mon, 26 Oct 2020 13:55:00 -0400</pubDate>
941 <guid isPermaLink="true">https://news.mit.edu/2020/defextiles-leveraging-3d-printer-defect-to-create-quasi-textiles-1020</guid>
942 <dc:creator>Becky Ham | MIT Media Lab</dc:creator>
943 <content:encoded><p>Sometimes 3D printers mess up. They extrude too much material, or too little, or deposit material in the wrong spot. But what if this bug could be turned into a (fashionable) feature?</p>
944
945 <p>Introducing <a href="https://www.media.mit.edu/projects/defextiles/overview">DefeXtiles</a>, a tulle-like textile that MIT Media Lab graduate student Jack Forman developed by controlling a common 3D printing defect — the under-extrusion of polymer filament.</p>
946
947 <p>Forman used a standard, inexpensive 3D printer to produce sheets and complex 3D geometries with a woven-like structure based on the “glob-stretch” pattern produced by under-extrusion. Forman has printed these flexible and thin sheets into an interactive lampshade,&nbsp;full-sized skirts, a roll of fabric long enough to stretch across a baseball diamond,&nbsp;and intricately patterned lace, among other items.</p>
948 <p>Forman, who works in the Tangible Media research group with Professor Hiroshi Ishii, presented and demonstrated the DefeXtiles research on Oct. 20 at the Association for Computing Machinery Symposium on User Interface Software and Technology.&nbsp;The material may prove immediately useful for prototyping and customizing in fashion design, Forman says, but future applications also could include 3D-printed surgical mesh with tunable mechanical properties, among other items.</p>
949
950 <p>“In general, what excites me most about this work is how immediately useful it can be to many makers,” Forman says. “Unlike previous work, the fact that no custom software or hardware is needed — just a relatively cheap $250 printer, the most common type of printer used — really makes this technique accessible to millions of people.”</p>
951
952 <p>“We envision that the materials of the future will be dynamic and computational,” says Ishii. “We call it ‘Radical Atoms.’ DefeXtiles is an excellent example of Radical Atoms, a programmable matter that emulates the properties of existing materials and goes beyond. We can touch, feel, wear, and print them.”</p>
953
954 <p>Joining Forman and Ishii on the project are Computer Science and Artificial Intelligence Laboratory and Department of Electrical Engineering and Computer Science graduate student Mustafa Doga Dogan, and Hamilton Forsythe, an MIT Department of Architecture undergraduate researcher.</p>
955
956 <p><strong>Filaments to fabric</strong></p>
957
958 <p>Forman had been experimenting with 3D printing during the media arts and sciences class MAS.863 / 4.140 / 6.943 (How to Make (Almost) Anything), led by Professor Neil Gershenfeld, director of the MIT Center for Bits and Atoms. Forman's experiments were inspired by the work of a friend from his undergraduate days at Carnegie Mellon University, who used under-extruded filament to produce vases. With his first attempts at under-extruding, "I was&nbsp;annoyed because the defects produced were perfect and periodic,” he says, “but then when I started playing with it, bending it and even stretching it, I was like, ‘whoa, wait, this is a textile. It looks like it, feels likes it, bends like it, and it prints really quickly.”</p>
959
960 <p>“I brought a small&nbsp;sample to my class for show and tell, not really thinking much of it, and Professor Gershenfeld saw it and he was excited about it,” Forman adds.</p>
961
962 <p>When a 3D printer under-extrudes material, it produces periodic gaps in the deposited material. Using an inexpensive fused deposition modeling 3D printer, Forman developed an under-extruding process called “glob-stretch,” where globs of thermoplastic polymer are connected by fine strands. The process produces a flexible, stretchy textile with an apparent warp and weft like a woven fabric. Forman says it feels something like a mesh jersey fabric.</p>
963
964 <p>“Not only are these textiles thinner and faster to print than other approaches, but the complexity of demonstrated forms is also improved. With this approach we can print 3D dimensional shell forms with a normal 3D printer and no special slicer software,” says Forman. “This is exciting because there’s a lot of opportunities with 3D printing fabric, but it’s really hard for it to be easily disseminated, since a lot of it uses expensive machinery and special software or special commands that are generally specific to a printer.”</p>
965
966 <p>The new textile can be sewn, de-pleated, and heat-bonded like an iron-on patch. Forman and his colleagues have printed the textiles using many common 3D printing materials, including a conductive filament that they used to produce a lamp that can be lit and dimmed by touching pleats in the lampshade. The researchers suggest that other base materials or additives could produce textiles with magnetic or optical properties, or textiles that are more biodegradable by using algae, coffee grounds, or wood.</p>
967
968 <p>According to Scott&nbsp;Hudson, a professor&nbsp;at Carnegie Mellon University's Human-Computer Interaction Institute, Forman's work represents&nbsp;a&nbsp;very interesting addition to the expanding set of 3D-printing techniques.&nbsp;</p>
969
970 <p>“This work is particularly important because it functions within the same print process as more conventional techniques,” notes&nbsp;Hudson, who was&nbsp;not part of the study. “This&nbsp;will allow us to integrate custom 3D-printed textile components — components that can be flexible and soft — into objects, along with&nbsp;more conventional hard parts.”</p>
971
972 <p><strong>Lab @home</strong></p>
973
974 <p>When MIT closed down at the start of the Covid-19 pandemic, Forman was in the midst of preparing for the ACM symposium submission. He relocated his full lab set up to the basement of his parents’ cabin near Lake Placid, New York.</p>
975
976 <p>“It’s not a lot of large equipment, but it’s lots of little tools, pliers, filaments,” he explains. “I had to set up two 3D printers, a soldering station, a photo backdrop — just because the work is so multidisciplinary.”</p>
977
978 <p>At the cabin, “I was able to hone in and focus on the&nbsp;research while the world around me was on fire, and it was actually a really good distraction,” Forman says. “It was also interesting to be working on a project that was so tech-focused, and then look out the window and see nature and trees — the tension between the two was quite inspiring.”</p>
979
980 <p>It was an experience for his parents as well, who got to see him “at my most intense and focused, and the hardest I’ve worked,” he recalls. “I’d be going upstairs at 5 a.m. for a snack when my dad was coming down for breakfast."&nbsp;&nbsp;</p>
981
982 <p>“My parents became part of the act of creation, where I’d print something and go, ‘look at this,’” he says. “I don’t know if I’ll ever have the opportunity again to have my parents so closely touch what I do every day.”</p>
983
984 <p>One of the more unusual aspects of the project has been what to call the material. Forman and his colleagues use the term “quasi-textile” because DefeXtiles doesn’t have all the same physical qualities of a usual textile, such as a bias in both directions and degree of softness. But some skeptics have been converted when they feel the material, Forman says.</p>
985
986 <p>The experience reminds him of the famous René Magritte painting&nbsp;“The Treachery of Objects (This Is Not a Pipe),” where the illustration of a pipe prompts a discussion about whether a representation can fully encompass all of an object’s meanings. “I’m interested in the coupling between digital bits and the materials experience by computationally fabricating high-fidelity materials with controllable forms and mechanical properties,” Forman explains.</p>
987
988 <p>“It makes me think about when the reference of the thing becomes accepted as the thing,” he adds. “It’s not the decision people make, but the reasoning behind it that interests me, and finding what causes them to accept it or reject it as a textile material.”</p>
989 </content:encoded>
990 <media:content url="https://news.mit.edu/sites/default/files/styles/news_article__cover_image__original/public/images/202010/3d-printed-lamp.jpg?itok=6d9JDybx" medium="image" type="image/jpeg" width="390" height="260">
991 <media:description type="plain">In this deformation-sensing lamp, the solid supports, textile lampshade, and sensing conductive pads are created as a single print.</media:description>
992 <media:credit>Photo: Jack Forman</media:credit>
993 </media:content>
994 </item>
995 <item>
996 <title>Superconductor technology for smaller, sooner fusion</title>
997 <link>https://news.mit.edu/2020/superconductor-technology-smaller-sooner-fusion-1013</link>
998 <description>MIT-Commonwealth Fusion Systems demonstration of new superconducting cable is a key step on the high-field path to compact fusion.</description>
999 <pubDate>Tue, 13 Oct 2020 12:00:00 -0400</pubDate>
1000 <guid isPermaLink="true">https://news.mit.edu/2020/superconductor-technology-smaller-sooner-fusion-1013</guid>
1001 <dc:creator>Leda Zimmerman | Plasma Science and Fusion Center</dc:creator>
1002 <content:encoded><p>Scientists have long sought to harness fusion as an inexhaustible and carbon-free energy source. Within the past few years, groundbreaking high-temperature superconductor technology (HTS) sparked a new vision for achieving practical fusion energy. This approach, known as the high-field pathway to fusion, aims to generate fusion in compact devices on a shorter timescale and lower cost than alternative approaches.</p>
1003
1004 <p>A key technical challenge to realizing this vision, though, has been getting HTS superconductors to work in an integrated way in the development of new, high-performance superconducting magnets, which will enable higher magnetic fields than previous generations of magnets, and are central to confining and controlling plasma reactions.</p>
1005
1006 <p>Now a team led by MIT’s Plasma Science and Fusion Center (PSFC) and MIT spinout company <a href="https://cfs.energy/">Commonwealth Fusion Systems</a> (CFS), has developed and extensively tested an HTS cable technology that can be scaled and engineered into the high-performance magnets. The team’s research was published on Oct. 7 in <em>Superconductor Science and Technology</em>. Researchers included MIT assistant professor and principal investigator Zachary Hartwig; PSFC Deputy Head of Engineering Rui F. Vieira and other key PSFC technical and engineering staff; CFS Chief Science Officer Brandon Sorbom PhD ’17 and other CFS engineers; and scientists at <a href="https://home.cern/">CERN</a> in Geneva, Switzerland, and at the <a href="https://www.wgtn.ac.nz/robinson">Robinson Research Institute</a> at Victoria University of Wellington, New Zealand.&nbsp;</p>
1007
1008 <p>This development follows a recent boost to the high-field pathway, when 47 researchers from 12 institutions published <a href="https://www.cambridge.org/core/journals/journal-of-plasma-physics/collections/status-of-the-sparc-physics-basis">seven papers</a> in the <em>Journal of Plasma Physics</em>, showing that a high-field fusion device, called <a href="https://www.psfc.mit.edu/sparc">SPARC</a>, built with such magnets would produce net energy — more energy than it consumes — something never previously demonstrated.</p>
1009
1010 <p>“The cable technology for SPARC is an important piece of the puzzle as we work to accelerate the timeline of achieving fusion energy,” says Hartwig, assistant professor of nuclear science and engineering, and leader of the research team at the PSFC. “If we’re successful in what we’re doing and in other technologies, fusion energy will start to make a difference in mitigating climate change — not in 100 years, but in 10 years.”</p>
1011
1012 <p><strong>A super cable</strong></p>
1013
1014 <p>The innovative technology described in the paper is a superconducting cable that conducts electricity with no resistance or heat generation and that will not degrade under extreme mechanical, electrical, and thermal conditions. Branded VIPER (an acronymic feat that stands for Vacuum Pressure Impregnated, Insulated, Partially transposed, Extruded, and Roll-formed), it consists of commercially produced thin steel tapes coated with HTS compound — yttrium-barium-copper-oxide — that are packaged into an assembly of copper and steel components to form the cable. Cryogenic coolant, such as supercritical helium, can flow easily through the cable to remove heat and keep the cable cold even under challenging conditions.</p>
1015
1016 <p>“One of our advances was figuring out a way to solder the HTS tape inside the cable, effectively making it a monolithic structure where everything is thermally connected,” says Sorbom. Yet VIPER can also be fashioned into twists and turns, using joints to create “almost any type of geometry,” he adds. This makes the cable an ideal building material for winding into coils capable of generating and containing magnetic fields of enormous strength, such as those required to make fusion devices substantially smaller than presently envisioned net-energy fusion devices.</p>
1017
1018 <p><strong>Resilient and robust</strong></p>
1019
1020 <p>“The key thing we can do with VIPER cable is make a magnetic field two to three times stronger at the size required than the present generation of superconducting magnet technology,” Hartwig says. The magnitude of the magnetic field in tokamaks plays a strong nonlinear role in determining plasma performance. For example, fusion power density scales as magnetic field to the fourth power: Doubling the field increases fusion power by 16 times or, conversely, the same fusion output power can be achieved in a device 16 times smaller by volume.</p>
1021
1022 <p>“In the development of high field magnets for fusion, HTS cables are an essential ingredient, and they’ve been missing,” says Soren Prestemon, director of the U.S. Magnet Development Program at the Lawrence Berkeley National Laboratory, who was not involved with this research. “VIPER is a breakthrough in the area of cable architecture — arguably the first candidate to be proven viable for fusion — and will enable the critical step forward to demonstration in a fusion reactor.”&nbsp;</p>
1023
1024 <p>VIPER technology also presents a powerful approach to a particular problem in the superconducting magnet field, called a quench, “that has terrified engineers since they started building superconducting magnets,” says Hartwig. A quench is a drastic temperature increase that occurs when the cold cables can no longer conduct electrical current without any resistance. When quench occurs, instead of generating almost zero heat in the superconducting state, the electrical current generates substantial resistive heating in the cable.</p>
1025
1026 <p>“The rapid temperature rise can cause the magnet to potentially damage or destroy itself if the electrical current is not shut off,” says Hartwig.&nbsp; “We want to avoid this situation or, if not, at least know about it as quickly and certainly as possible.”</p>
1027
1028 <p>The team incorporated two types of temperature-sensing fiber optic technology developed by collaborators at CERN and Robinson Research Institute. The fibers exhibited — for the first time on full-scale HTS cables and in representative conditions of high-magnetic field fusion magnets — sensitive and high-speed detection of temperature changes along the cable to monitor for the onset of quench.</p>
1029
1030 <p>Another key result was the successful incorporation of easily fabricated, low-electrical resistance, and mechanically robust joints between VIPER cables. Superconducting joints are often complex, challenging to make, and more likely to fail than others parts of a magnet; VIPER was designed to eliminate these issues. The VIPER joints have the additional advantage of being demountable, meaning they can be taken apart and reused with no impact on performance.</p>
1031
1032 <p>Prestemon notes that the cable’s innovative architecture directly impacts real-world challenges in operating fusion reactors of the future. “In an actual commercial fusion-energy-producing facility, intense heat and radiation deep inside the reactor will require routine component replacements,” he says. “Being able to take these joints apart and put them back together is a significant step towards making fusion a cost-effective proposition.”</p>
1033
1034 <p>The 12 VIPER cables that Hartwig’s team built, running between one and 12 meters in length, were evaluated with bending tests, thousands of sudden “on-off” mechanical cycles, multiple cryogenic thermal cycles, and dozens of quench-like events to simulate the kind of punishing conditions encountered in the magnets of a fusion device. The group successfully completed four multi-week test campaigns in four months at the SULTAN facility, a leading center for superconducting cable evaluation operated by <a href="https://www.epfl.ch/research/domains/swiss-plasma-center/">Swiss Plasma Center</a>, affiliated with Ecole Polytechnique Fédérale de Lausanne in Switzerland.</p>
1035
1036 <p>“This unprecedented rate of HTS cable testing at SULTAN shows the speed that technology can be advanced by an outstanding team with the mindset to go fast, the willingness to take risks, and the resources to execute,” says Hartwig. It is a sentiment that serves as the foundation of the SPARC project.</p>
1037
1038 <p>The SPARC team continues to improve VIPER cable and is moving on to the next project milestone in mid-2021: “We’ll be building a multi-ton model coil that will be similar to the size of a full-scale magnet for SPARC,” says Sorbom. These research activities will continue to advance the foundational magnet technologies for SPARC and enable the demonstration of net energy from fusion, a key achievement that signals fusion is a viable energy technology. “That will be a watershed moment for fusion energy,” says Hartwig.</p>
1039
1040 <p>Funding for this research was provided by CFS.</p>
1041 </content:encoded>
1042 <media:content url="https://news.mit.edu/sites/default/files/styles/news_article__cover_image__original/public/images/202010/MIT-viper-cable-rendering_0.jpg?itok=6PJEGO5Y" medium="image" type="image/jpeg" width="390" height="260">
1043 <media:description type="plain">The assembly designed for the third SULTAN cable test features two 3-meter VIPER HTS cables in parallel and connected with a copper joint at the bottom; cryogenic helium and electrical current are injected at the top.</media:description>
1044 <media:credit>Photo: Jose Estrada/PSFC</media:credit>
1045 </media:content>
1046 </item>
1047 <item>
1048 <title>India’s culture of coping with cancer</title>
1049 <link>https://news.mit.edu/2020/enduring-cancer-book-Banerjee-1007</link>
1050 <description>Dwaipayan Banerjee’s new book examines the psychological and social terrain of living with cancer in a country where the disease has long been downplayed.</description>
1051 <pubDate>Wed, 07 Oct 2020 17:00:00 -0400</pubDate>
1052 <guid isPermaLink="true">https://news.mit.edu/2020/enduring-cancer-book-Banerjee-1007</guid>
1053 <dc:creator>Peter Dizikes | MIT News Office</dc:creator>
1054 <content:encoded><p>When Dwaipayan Banerjee began studying the lives of poor cancer patients in and around Delhi, India, he noticed something distinctive: Virtually none of them used the word “cancer” itself. One elderly man Banerjee met got upset at seeing a medical van with the words “caring for cancer” on the side; the man insisted he was actually suffering from “oncology.”</p>
1055
1056 <p>Banerjee also learned, from a medical resident at a hospital, to think of these patients as experiencing “shak,” a Hindi word implying doubt, skepticism, and suspicion. For a patient, a diagnosis would create fears about not only physical well-being, but also the social stakes of the disease.</p>
1057
1058 <p>“The commonplace use of the word ‘shak’ by families and patients indicated not only the doubt that a lump or growth could be a tumor, but also revealed longer and deeper misgivings,” Banerjee says. “These doubts and suspicions were often about a lack of faith in the public health system that had failed them before, and skepticism about finding support from doctors and kin.”</p>
1059
1060 <p>For many people, having cancer creates “a sense of being unmoored from prior certainties about oneself and one’s place in the world,” writes Banerjee, an associate professor in the MIT Program in Science, Technology, and Society, in a new book exploring the world of Indian cancer patients.</p>
1061
1062 <p>In the book, “Enduring Cancer: Life, Death, and Diagnosis in Delhi,” published by Duke University Press, Banerjee delves into the psychological and social worlds of generally low-income cancer patients. The work illuminates the culture that has emerged around cancer in India — such as the tendency to avoid naming the disease — and aims to spur thought about how other cultures cope with cancer.</p>
1063
1064 <p>“It’s not as if people aren’t very careful about how they talk about cancer in the U.S. or anywhere else,” Banerjee says. “It’s still something that evinces a lot of stigma, in the same way that other diseases strongly associated with death do. People are uncomfortable around death and the possibility of dying, anywhere in the world.”</p>
1065
1066 <p><strong>A more visible problem</strong></p>
1067
1068 <p>Banerjee conducted much of the research for “Enduring Cancer” as anthropological fieldwork, studying patients and medical care as an inside observer with two organizations: CanSupport, an NGO dedicated to helping lower-income urban cancer patients, and the All India Institute of the Medical Sciences (AIIMS), a prestigious hospital in Delhi.</p>
1069
1070 <p>The services offered by CanSupport and AIIMS add a layer of care beyond medical procedures, helping patients cope psychologically with cancer or deal with lingering physical pain they experience.</p>
1071
1072 <p>“I grew up in Delhi, and I had never heard of doctors and organizations doing this kind of work,” Banerjee says. “That’s because there were none, at least until these last two decades. That says something about the rise of cancer as a visible problem, as well as the need for palliative care. Very distinctively, at AIIMS, cancer pain is treated as a biomedical condition deserving of attention by the country’s best anesthesiologists, rather than a symptom that can be ignored and left for nonspecialists. The United States has much to learn from this approach.”</p>
1073
1074 <p>To an extent, Banerjee observes, the presence of cancer had long been ignored in the global south by public health experts and policymakers, who associated the disease more with industrialized societies and even claimed that vegetarian diets (where present in India) reduced incidence of the disease. And yet, he notes, even British doctors in India in the 19th century were sending reports home about treating cancer patients, frustrated by the lack of attention paid to the disease by the colonial government.</p>
1075
1076 <p>“The disease has been in India as long as it has been elsewhere in the world,” Banerjee says. “It’s one of the big myths I try to unravel in the book.”</p>
1077
1078 <p>Banerjee also notes how cancer treatment at AIIMS has expanded to include an emphasis on pain management as a part of patient care: “The commitment from them to [address] pain, as highly trained specialists and not just leaving it to public health workers, is really remarkable. Pain is not a mere symptom; it’s part of the disease.”</p>
1079
1080 <p>During his fieldwork, Banerjee closely observed how a cancer diagnosis reverberates around affected families, as well as marriages, both revealing and reshaping relationship dynamics. Many times, he observed, there are delicate decisions about how much information is distributed among families.</p>
1081
1082 <p>“There is this way of distributing knowledge across families rather than putting all the burden on an individual,” Banerjee says. “Doctors are as well-aware of this as anybody else.”</p>
1083
1084 <p>But Banjerjee also found gender differences at work in families, adding tensions to the whole process of family support.</p>
1085
1086 <p>“There are imbalances within families — who gets to be told and who doesn’t — that reveal who has power within the family,” Banerjee says. “It’s often women who are not told and it’s often the male kin who make these decisions. Cancer doesn’t just easily map itself onto what hierarchies and imbalances exist, it changes them and puts pressure on what already exists.”</p>
1087
1088 <p><strong>By the book</strong></p>
1089
1090 <p>Even as patients avoid using the word “cancer” and may sometimes seem unclear about the nature of their illnesses, Banerjee believes, they are well-aware of their diagnoses.</p>
1091
1092 <p>“Of course they know,” he says. “They know through different nonverbal cues, from repeated hospital visits, and extrapolate from the treatments they receive. They know how to read between the lines. The decision not to talk about it, or to carefully negotiate speech around it, is more often than not a way of demonstrating care and concern, at a different register than explicit talk. It is a way through which they weave this terrifying disease into their everyday worlds.”</p>
1093
1094 <p>But as Banerjee discusses in multiple chapters, there are some explicit discussions of life as a cancer patient in Indian culture, including books and films. Banerjee has mixed views about this material. Like elsewhere in the world, most mass-market books about cancer in India are self-help volumes that provide encouragement, but may also burden people by making them feel unduly responsible for their own wellness.</p>
1095
1096 <p>“These books urge patients to be strong in the face of the disease and to be a survivor and to transcend the pain by the sheer force of personal will,” Banerjee says. “I would be wrong if I said a self-help book does not offer a degree of identification and comfort. It absolutely does. There’s a reason they’ve been popular. But this structure can end up distracting from the more political aspects of the disease.”</p>
1097
1098 <p>Consider some of the book’s titles: ‘The Joy of Cancer,” “To Cancer with Love,” “My Date with Cancer,” and “Cancer Made Me,” which Banerjee says sends a clear message to patients: “Don’t let them know you’ve lost your hair or are experiencing so much pain. Learn to love the disease and let it teach you about not just surviving, but becoming a better person. The assumption here is that the habits of the patient’s life before the disease contributed to its occurrence. For women, this can be especially dangerous, as they are often accused of transitioning too quickly to a “modern” working lifestyle and not taking care of their own health.</p>
1099
1100 <p>“Of course, I see the value of encouraging patients to be resilient,” Banerjee says. “But in the end, I think the books’ intent is to make everyone around patients feel comfortable; not the patients themselves. Instead of ‘loving’ or ‘accepting’ cancer, there is a case to be made for being justifiably upset and angry —&nbsp;angry at the lack of political will to address the environmental containments and inequalities of medicine&nbsp;that have become part of public health systems all over the world.”</p>
1101
1102 <p>“Enduring Cancer” has received praise from other scholars in the field. Vincanne Adams, a professor of medical anthropology at the University of California at San Francisco, calls the book a “wonderful ethnography,” in which Banerjee “shows how cancer in India exists across many relationships, aspirations, frustrations, gendered battles, caregiving gestures, medical sciences, and familial trials.” Harrison Solomon, a professor of cultural anthropology and global health at Duke University, describes the book as “a landmark for thinking about survival and endurance in medical anthropology, science studies, public health, and South Asian studies.”&nbsp;</p>
1103
1104 <p>For his part, Banerjee says he wants to help readers consider how cancer is experienced by people, and what everyone can do to help.</p>
1105
1106 <p>“It’s not a battle that your psychological state got you into, and it’s not a battle that your psychological state alone solves,” Banerjee says. “Increasingly, there is a new kind of activism around cancer that I find very heartening. In the best of circumstances, it’s a difficult disease for a patient to present a cheerful face about. The onus is on the rest of us to make sure treatments are available, and the patient is supported.”</p>
1107 </content:encoded>
1108 <media:content url="https://news.mit.edu/sites/default/files/styles/news_article__cover_image__original/public/images/202010/MIT-EnduringCancer-01-Press_1.jpg?itok=uADiMYXP" medium="image" type="image/jpeg" width="390" height="260">
1109 <media:description type="plain">MIT professor Dwaipayan Banerjee’s new book, “Enduring Cancer,” takes a close look at the efforts of low-income residents of Delhi, India, as they battle cancer and try to find social support in moments of crisis.</media:description>
1110 <media:credit>Photo: Jon Sachs/MIT SHASS Communications </media:credit>
1111 </media:content>
1112 </item>
1113 <item>
1114 <title>Stressed on the job? An AI teammate may know how to help</title>
1115 <link>https://news.mit.edu/2020/stressed-job-ai-teammate-may-know-how-help-1026</link>
1116 <description>Researchers are working toward intelligent machines that can sense cognitive fatigue and suggest interventions to help a human improve performance.</description>
1117 <pubDate>Mon, 26 Oct 2020 12:55:00 -0400</pubDate>
1118 <guid isPermaLink="true">https://news.mit.edu/2020/stressed-job-ai-teammate-may-know-how-help-1026</guid>
1119 <dc:creator>Kylie Foy | MIT Lincoln Laboratory</dc:creator>
1120 <content:encoded><p>Humans have been teaming up with machines throughout history to achieve goals, be it by using simple machines to move materials or complex machines to travel in space. But advances in artificial intelligence today bring possibilities for even more sophisticated teamwork — true human-machine teams that cooperate to solve complex problems.</p>
1121
1122 <p>Much of the development of these human-machine teams focuses on the machine, tackling the technology challenges of training AI algorithms to perform their role in a mission effectively. But less focus, MIT Lincoln Laboratory researchers say, has been given to the human side of the team. What if the machine works perfectly, but the human is struggling?</p>
1123
1124 <p>"In the area of human-machine teaming, we often think about the technology — for example, how do we monitor it, understand it, make sure it's working right. But teamwork is a two-way street, and these considerations aren't happening both ways. What we're doing is looking at the flip side, where the machine is monitoring and enhancing the other side — the human," says Michael Pietrucha, a tactical systems specialist at the laboratory.&nbsp;</p>
1125
1126 <p>Pietrucha is among a team of laboratory researchers that aims to develop AI systems that can sense when a person's cognitive fatigue is interfering with their performance. The system would then suggest interventions, or even take action in dire scenarios, to help the individual recover or to prevent harm.&nbsp;</p>
1127
1128 <p>"Throughout history, we see human error leading to mishaps, missed opportunities, and sometimes disastrous consequences," says Megan Blackwell, former deputy lead of internally funded biological science and technology research at the laboratory. "Today, neuromonitoring is becoming more specific and portable. We envision using technology to monitor for fatigue or cognitive overload. Is this person attending to too much? Will they run out of gas, so to speak? If you can monitor the human, you could intervene before something bad happens."</p>
1129
1130 <p>This vision has its roots in decades-long research at the laboratory in using technology to "read" a person's cognitive or emotional state. By collecting biometric data — such as video and audio recordings of a person speaking — and processing these data with advanced AI algorithms, researchers have uncovered biomarkers of various psychological and neurobehavioral conditions. These biomarkers have been used to train models that can accurately estimate the level of a person's depression, for example.</p>
1131
1132 <p>In this work, the team will apply their biomarker research to AI that can analyze an individual's cognitive state, encapsulating how fatigued, stressed, or overloaded a person is feeling. The system will use biomarkers derived from physiological data such as vocal and facial recordings, heart rate, EEG and optical indications of brain activity, and eye movement to gain these insights.</p>
1133
1134 <p>The first step will be to build a cognitive model of an individual. "The cognitive model will integrate the physiological inputs and monitor the inputs to see how they change as a person performs particular fatiguing tasks," says Thomas Quatieri, who leads several neurobehavioral biomarker research efforts at the laboratory. "Through this process, the system can establish patterns of activity and learn a person's baseline cognitive state involving basic task-related functions needed to avoid injury or undesirable outcomes, such as auditory and visual attention and response time."</p>
1135
1136 <p>Once this individualized baseline is established, the system can start to recognize deviations from normal and predict if those deviations will lead to mistakes or poor performance.</p>
1137
1138 <p>"Building a model is hard. You know you got it right when it predicts performance," says William Streilein, principal staff in the Lincoln Lab's Homeland Protection and Air Traffic Control Division. "We've done well if the system can identify a deviation, and then actually predict that the deviation is going to interfere with the person's performance on a task. Humans are complex; we compensate naturally to stress or fatigue. What's important is building a system that can predict when that deviation won't be compensated for, and to only intervene then."</p>
1139
1140 <p>The possibilities for interventions are wide-ranging. On one end of the spectrum are minor adjustments a human can make to restore performance: drink coffee, change the lighting, get fresh air. Other interventions could suggest a shift change or transfer of a task to a machine or other teammate. Another possibility is using transcranial direct current stimulation, a&nbsp;<a href="https://www.frontiersin.org/articles/10.3389/fnhum.2019.00114/full">performance-restoring&nbsp;technique</a>&nbsp;that uses electrodes to stimulate parts of the brain and has been show to be&nbsp;<a href="https://pubmed.ncbi.nlm.nih.gov/28851554/">more effective than caffeine</a>&nbsp;in countering fatigue, with fewer side effects.</p>
1141
1142 <p>On the other end of the spectrum, the machine might take actions necessary to ensure the survival of the human team member when the human is incapable of doing so. For example, an AI teammate could make the "ejection decision" for a fighter pilot who has lost consciousness or the physical ability to eject themselves. Pietrucha, a retired colonel in the U.S. Air Force who has had many flight hours as a fighter/attack aviator, sees the promise of such a system that "goes beyond the mere analysis of flight parameters and includes analysis of the cognitive state of the aircrew, intervening only when the aircrew can't or wont," he says.&nbsp;</p>
1143
1144 <p>Determining the most helpful intervention, and its effectiveness, depends on a number of factors related to the task at hand, dosage of the intervention, and even a user's demographic background. "There's a lot of work to be done still in understanding the effects of different interventions and validating their safety," Streilein says. "Eventually, we want to introduce personalized cognitive interventions and assess their effectiveness on mission performance."</p>
1145
1146 <p>Beyond its use in combat aviation, the technology could benefit other demanding or dangerous jobs, such as those related to air traffic control, combat operations, disaster response, or emergency medicine. "There are scenarios where combat medics are vastly outnumbered, are in taxing situations, and are as every bit as tired as everyone else. Having this kind of over-the-shoulder help, something to help monitor their mental status and fatigue, could help prevent medical errors or even alert others to their level of fatigue," Blackwell says.</p>
1147
1148 <p>Today, the team is pursuing sponsorship to help develop the technology further. The coming year will be focused on collecting data to train their algorithms. The first subjects will be intelligence analysts, outfitted with sensors as they play a serious game that simulates the demands of their job. "Intelligence analysts are often overwhelmed by data and could benefit from this type of system," Streilein says. "The fact that they usually do their job in a 'normal' room environment, on a computer, allows us to easily instrument them to collect physiological data and start training."</p>
1149
1150 <p>"We'll be working on a basis set of capabilities in the near term," Quatieri says, "but an ultimate goal would be to leverage those capabilities so that, while the system is still individualized, it could be a more turnkey capability that could be deployed widely, similar to how Siri, for example, is universal but adapts quickly to an individual." In the long view, the team sees the promise of a universal background model that could represent anyone and be adapted for a specific use.&nbsp;</p>
1151
1152 <p>Such a capability may be key to advancing human-machine teams of the future. As AI progresses to achieve more human-like capabilities, while being immune from the human condition of mental stress, it's possible that humans may present the greatest risk to mission success. An AI teammate may know just how to lift their partner up.</p>
1153 </content:encoded>
1154 <media:content url="https://news.mit.edu/sites/default/files/styles/news_article__cover_image__original/public/images/202010/AI-Teaming.png?itok=iTmJ-rPd" medium="image" type="image/jpeg" width="390" height="260">
1155 <media:description type="plain">Lincoln Laboratory researchers aim to develop artificial intelligence "teammates" that can sense cognitive fatigue and suggest actions to help a human recover. </media:description>
1156 <media:credit>Image: Bryan Mastergeorge</media:credit>
1157 </media:content>
1158 </item>
1159 <item>
1160 <title>A step toward a universal flu vaccine</title>
1161 <link>https://news.mit.edu/2020/universal-flu-vaccine-1007</link>
1162 <description>With computer models and lab experiments, researchers are working on a strategy for vaccines that could protect against any influenza virus.</description>
1163 <pubDate>Wed, 07 Oct 2020 11:00:00 -0400</pubDate>
1164 <guid isPermaLink="true">https://news.mit.edu/2020/universal-flu-vaccine-1007</guid>
1165 <dc:creator>Anne Trafton | MIT News Office</dc:creator>
1166 <content:encoded><p>Each year, the flu vaccine has to be redesigned to account for mutations that the virus accumulates, and even then, the vaccine is often not fully protective for everyone.</p>
1167
1168 <p>Researchers at MIT and the Ragon Institute of MIT, MGH, and Harvard are now working on strategies for designing a universal flu vaccine that could work against any flu strain. In a study appearing today, they describe a vaccine that triggers an immune response against an influenza protein segment that rarely mutates but is normally not targeted by the immune system.</p>
1169
1170 <p>The vaccine consists of nanoparticles coated with flu proteins that train the immune system to create the desired antibodies. In studies of mice with humanized immune systems, the researchers showed that their vaccine can elicit an antibody response targeting that elusive protein segment, raising the possibility that the vaccine could be effective against any flu strain.</p>
1171
1172 <p>“The reason we’re excited about this work is that it is a small step toward developing a flu shot that you just take once, or a few times, and the resulting antibody response is likely to protect against seasonal flu strains and pandemic strains as well,” says Arup K. Chakraborty, the Robert T. Haslam Professor in Chemical Engineering and professor of physics and chemistry at MIT, and a member of MIT’s Institute for Medical Engineering and Science and the Ragon Institute of MGH, MIT, and Harvard.</p>
1173
1174 <p>Chakraborty and Daniel Lingwood, an assistant professor at Harvard Medical School and a group leader at the Ragon Institute, are the senior authors of the study, which appears today in <em>Cell Systems</em>. MIT research scientist Assaf Amitai is the lead author of the paper.</p>
1175
1176 <p><strong>Targeting flu</strong></p>
1177
1178 <p>Most flu vaccines consist of inactivated flu viruses. These viruses are coated with a protein called hemagglutinin (HA), which helps them bind to host cells. After vaccination, the immune system generates squadrons of antibodies that target the HA protein. These antibodies almost always bind to the head of the HA protein, which is the part of the protein that mutates the most rapidly. Parts of the HA stem, on the other hand, very rarely mutate.</p>
1179
1180 <p>“We don't understand the complete picture yet, but for many reasons, the immune system is intrinsically not good at seeing the conserved parts of these proteins, which if effectively targeted would elicit an antibody response that would neutralize multiple influenza types,” Lingwood says.</p>
1181
1182 <p>In their new study, the researchers set out to study why the immune system ends up targeting the HA head rather than the stem, and to find ways to refocus the immune system’s attention on the stem. Such a vaccine could elicit antibodies known as “broadly neutralizing antibodies,” which would respond to any flu strain. In principle, this kind of vaccine could end the arms race between vaccine designers and rapidly mutating flu viruses.</p>
1183
1184 <p>One factor that was already known to contribute to antibody preference for the HA head is that HA proteins are densely clustered on the surface of the virus, so it’s difficult for antibodies to access the stem region. The head region is much more accessible.</p>
1185
1186 <p>The researchers developed a computational model that helped them to further explore the “immunodominance” of the protein’s head region. “We hypothesized that the surface geometry of the virus could be key to its ability to survive by protecting its vulnerable parts from antibodies,” Amitai says.</p>
1187
1188 <p>The researchers explored the effects of geometry on immunodominance using a technique called molecular dynamics simulation. They further modeled a process called antibody affinity maturation. This process, which occurs after B cells encounter a virus (or a vaccine), determines which antibodies will predominate during the immune response.</p>
1189
1190 <p>Each B cell has on its surface proteins called B cell receptors, which bind to different foreign proteins. Once a particular B cell receptor binds strongly to the HA protein, that B cell becomes activated and starts to multiply rapidly. This process introduces new mutations into the B cell receptors, some of which bind more strongly. These better binders tend to survive, while the weaker binders die. At the end of this process, which takes one or two weeks, there is a population of B cells that is very good at binding strongly to the HA protein. These B cells secrete antibodies that bind to the HA protein.</p>
1191
1192 <p>“As time goes on, after infection, the antibodies get better and better at targeting this particular antigen,” Chakraborty says.</p>
1193
1194 <p>The researchers’ computer simulations of this process revealed that when a typical flu vaccine is given, B cell receptors that bind strongly to the HA stem are at a competitive disadvantage during the maturation process, because they can’t reach their targets as easily as B cell receptors that bind strongly to the HA head.</p>
1195
1196 <p>The researchers also used their computer model to simulate this maturation process with a nanoparticle vaccine developed at the National Institutes of Health, which is now in a phase 1 clinical trial. This particle carries HA stem proteins spaced out at lower density. The model showed that this arrangement makes the proteins more accessible to antibodies, which are Y-shaped, allowing the antibodies to grab onto the proteins with both arms. The simulations revealed that those stem-targeting antibodies predominated at the end of the maturation process.</p>
1197
1198 <p><strong>Refocused immunity</strong></p>
1199
1200 <p>The researchers also used their computational model to predict the outcome of several possible vaccination strategies. One strategy that appears promising is to immunize with an HA stem from a virus that is similar to, but not the same as, strains that the recipient has previously been exposed to. In 2009, many people around the world were either infected with or vaccinated against a novel H1N1 strain. The modeling led the researchers to hypothesize that if they vaccinated with nanoparticles displaying HA-like proteins from a strain that is different from the 2009 version, it should elicit the kind of broadly neutralizing antibodies that may confer universal immunity.</p>
1201
1202 <p>Using mice with human immune cells, the researchers tested this strategy, first immunizing them against the 2009 H1N1 strain, followed by a nanoparticle vaccine carrying the HA stem protein from a different H1N1 strain. They found that this approach was much more successful at eliciting broadly neutralizing antibodies than any of the other strategies that they tested.</p>
1203
1204 <p>“We discovered that this particular event in our immune history can actually be harnessed with this particular nanoparticle to refocus the immune system’s attention on one of these so-called universal vaccine targets,” Lingwood says. “When there's a refocusing event, that means we can swing the antibody response against that target, which under other conditions is simply not seen. We have shown in previous studies that when you're able to elicit this kind of response, it’s protective against flu strains that mimic pandemic threats.”</p>
1205
1206 <p>The research was funded by the National Institutes of Health, the Harvard University Milton Award, the Gilead Research Scholars Program, and the National Science Foundation Research Fellowship Program.</p>
1207 </content:encoded>
1208 <media:content url="https://news.mit.edu/sites/default/files/styles/news_article__cover_image__original/public/images/202010/MIT-Flu-Vaccine-01-PRESS.jpg?itok=6f9TDzBz" medium="image" type="image/jpeg" width="390" height="260">
1209 <media:description type="plain">Researchers are working on strategies for designing a universal flu vaccine that could work against any flu strain.</media:description>
1210 <media:credit>Image: courtesy of the researchers, edited by MIT News</media:credit>
1211 </media:content>
1212 </item>
1213 <item>
1214 <title>Learning by doing, remotely</title>
1215 <link>https://news.mit.edu/2020/learning-by-doing-remotely-1019</link>
1216 <description>Despite the disruption caused by the pandemic, MIT students have carved out meaningful hands-on experiences. </description>
1217 <pubDate>Mon, 19 Oct 2020 14:05:00 -0400</pubDate>
1218 <guid isPermaLink="true">https://news.mit.edu/2020/learning-by-doing-remotely-1019</guid>
1219 <dc:creator>Elizabeth Durant | Office of the Vice Chancellor</dc:creator>
1220 <content:encoded><p>Experiential learning is alive and well at MIT — even when it’s remote.</p>
1221
1222 <p>Just ask Julian Zulueta, a sophomore in biological engineering. Last May, he spotted an intriguing social impact internship opportunity in the PKG Public Service Center newsletter: The CDC Foundation, a Congressionally-chartered nonprofit created to support the Centers for Disease Control and Prevention (CDC), was seeking remote students to assist with the Covid-19 response.</p>
1223
1224 <p>He applied — one of 60 candidates for two spots — and got the position. As a member of the CDC Foundation’s Workforce Strike Team, Zulueta interfaced with state and local health departments, with a particular focus on the Midwest region. Drawing on his introductory Python experience at MIT, he analyzed requisition data and created visualizations to detect trends in resources and in the effectiveness of medical interventions. He also studied correlations in universities’ response to Covid-19 and helped establish new professional growth policies within the CDC Foundation.</p>
1225
1226 <p>The internship was eye-opening, and it stoked his interest in exploring public health careers further. “To my surprise, I realized that public health was more than just the opinions of doctors and nurses. Rather, it extends to incorporate ideas related to public policy design and statistics, which can favor majority groups and lead to disparities in health outcomes,” Zulueta says.</p>
1227
1228 <p>Zulueta’s experience is heartening to Kate Trimble, senior associate dean and director of the Office of Experiential Learning (OEL). “When the pandemic first hit, we were very concerned that students were going to miss out on the hands-on experiences that are so critical to their personal and professional development,” she says.</p>
1229
1230 <p>The PKG Center newsletter that changed Zulueta’s trajectory was the result of concerted efforts by OEL and other campus partners to help students whose summer plans had fallen through or were up in the air — efforts that seem to have paid off. While data from industry partners are not available, statistics from the PKG Center and the Undergraduate Research Opportunities Program (UROP) show a marked uptick in participation this summer. A total of 213 students worked remotely in intensive social impact programs through the PKG Center (compared to 136 in 2019), and 1,523 students participated in UROP (compared to 975 last year).</p>
1231
1232 <p>And, even better, those efforts served as a roadmap for rethinking similar experiences for the mostly-remote fall academic semester. “It was really inspiring to see the MIT community spring into action to adapt in-person UROPs and internships — and even global experiences through MISTI — to a remote format,” says Trimble. “In retrospect, we shouldn’t have been surprised; the ‘magic’ of MIT lies in hands-on learning, and everyone here excels at problem-solving.”</p>
1233
1234 <p><strong>Retooling the best laid plans</strong></p>
1235
1236 <p>Some students were able to extend or reconfigure in-person opportunities into remote versions. When sophomore Sherry Nyeo realized that she would not be able to intern for the summer at a biotech company in Israel, she applied to continue a UROP she started in February at the Whitehead Institute for Biomedical Research, working on RNA secondary structure. Nyeo, who is majoring in electrical engineering and computer science and biology, remotely analyzed the lab’s data and ran the data pipeline.</p>
1237
1238 <p>“I do appreciate that I got a firsthand experience of what goes behind research, and I had a lot of opportunities to present papers to my lab during journal club,” Nyeo says. Her computational data analysis, along with figures she generated, have been incorporated into a paper on the SARS CoV-2 genome, on which she is listed as a co-author.</p>
1239
1240 <p>Marisa Gaetz, a 2020 MIT graduate who is staying on to pursue a PhD in mathematics, managed to tweak her ESG-PKG Fellowship for The Educational Justice Institute (TEJI) at MIT, a nonprofit that leverages education and technology to address mass incarceration.</p>
1241
1242 <p>Before the pandemic, she was planning to help facilitate a summer program for Boston-area youth who have been drawn into the criminal justice system. Instead, she adapted the in-person elements of the program into engaging online activities that encourage discussions about wrongdoing, ethical dilemmas, and moral worth. In addition, Gaetz researched interactive boards with Zoom capabilities and secured funding to install the technology in correctional facilities in Suffolk County, Massachusetts, and Maine, ensuring that TEJI can continue to offer its <a href="https://www.youtube.com/watch?time_continue=357&amp;v=6Xnm_TYGjC0&amp;feature=emb_logo">signature classes</a> in these facilities.</p>
1243
1244 <p>“Even though many experiences have to be remote right now, Covid has also exposed a lot of different needs, and so there’s a lot of new opportunities to do impactful work as well,” Gaetz says.</p>
1245
1246 <p>Sophomore Catherine Lu is one of many participants in MIT International Science and Technology Initiatives (MISTI) who were able to convert their global education experiences into remote versions. Originally, the civil and environmental major was slated to travel to Tulum, Mexico, to help restore a lighthouse into a coral education center. Instead, Lu designed and created a virtual reality experience of the lighthouse, which allows users to visualize and interact with the physical space and, by extension, promotes awareness of coral restoration efforts.</p>
1247
1248 <p>“Through this virtual reality world, we are able to expand even more on the idea of coral education, since much of the audience we’re targeting are people who might not live near coral or are not able to physically travel there,” Lu says.</p>
1249
1250 <p><strong>Seizing a singular opportunity</strong></p>
1251
1252 <p>For juniors Carlos Mercado-Lara and Evan Gwozdz, shifting gears to a remote summer opened up unique opportunities they could not have anticipated.</p>
1253
1254 <p>Once Mercado-Lara found out that his MISTI-France program was cancelled, he channeled his time and efforts into <a href="https://sciteens.org">SciTeens</a>, as a PKG Center IDEAS grantee. SciTeens, a nonprofit he co-founded in high school, connects underrepresented high school students from underserved communities with online mentors in STEM. This summer, Mercado-Lara and the SciTeens team collaborated with an organization in Zimbabwe to mentor local teens working on science projects.</p>
1255
1256 <p>“That was great, because it was our first time working internationally with students and establishing trust with another organization,” says Mercado-Lara, who is majoring in biological engineering. The experience also helped him shift his focus for the future. “If I had done an internship, it would have helped my career and allowed me to explore a career path, but over the summer I was able to realign some of the things that I want to continue doing for the next few years while I’m at MIT, and hopefully grad school.”</p>
1257
1258 <p>Gwozdz ditched his plans to find an internship in March, since many of them were being cancelled. He reached out to faculty doing interesting research in his major, chemical engineering, and landed a remote UROP in the Zack Smith lab, which investigates polymer membranes for gas separations.</p>
1259
1260 <p>Since he couldn’t physically be in the lab, he focused on learning about molecular simulations, using software to model experimental polymer systems. “Simulations are heavily used in the field, but they haven’t been explored thoroughly in the Smith lab,” Gwozdz explains. By delving into simulations, he created a niche for himself and has become a valuable member of the team. “With this remote project, I think I was able to contribute as much to them as I received myself,” he says.</p>
1261
1262 <p><strong>Expanding ELOs to all undergraduates</strong></p>
1263
1264 <p>The experiences that Zulueta, Nyeo, Gaetz, Lu, Mercado-Lara, and Gwozdz had are indicative of the diversity and range of opportunities available to students, Trimble says.</p>
1265
1266 <p>“The all-remote summer allowed some students to think outside the box and explore amazing experiences — like social impact internships or work in their own communities — that they might not have considered before. And many students used their positions to make a difference on urgent issues like the pandemic, climate change, and racial justice.”</p>
1267
1268 <p>At the same time, Trimble notes, OEL and other offices learned a great deal over the spring and summer about how to support virtual experiential learning. “We’re putting all of that into practice this academic year,” she says.</p>
1269
1270 <p>To that end, MIT is guaranteeing all undergraduates a paid experiential learning opportunity (ELO) this year. Students who are on campus or at home can earn up to $1,900 while working in a wide variety of remote or on-campus ELOs. The <a href="https://elo.mit.edu/">OEL’s new website</a> serves as a guidepost, with resources organized into six tracks: research; public service and social impact; innovation and entrepreneurship; global opportunities; teaching and learning; and opportunities for first-year students.</p>
1271 </content:encoded>
1272 <media:content url="https://news.mit.edu/sites/default/files/styles/news_article__cover_image__original/public/images/202010/MIT-Student-Remote-01-Press-Option.jpg?itok=5Rt658CL" medium="image" type="image/jpeg" width="390" height="260">
1273 <media:description type="plain">Clockwise from top left: MIT students Catherine Lu, Carlos Mercado-Lara, Sherry Nyeo, Julian Zulueta, Evan Gwozdz, and Marisa Gaetz.</media:description>
1274 </media:content>
1275 </item>
1276 <item>
1277 <title>Turning diamond into metal</title>
1278 <link>https://news.mit.edu/2020/diamond-metal-conductor-1005</link>
1279 <description>Normally an insulator, diamond becomes a metallic conductor when subjected to large strain in a new theoretical model.</description>
1280 <pubDate>Mon, 05 Oct 2020 15:00:00 -0400</pubDate>
1281 <guid isPermaLink="true">https://news.mit.edu/2020/diamond-metal-conductor-1005</guid>
1282 <dc:creator>David L. Chandler | MIT News Office</dc:creator>
1283 <content:encoded><p>Long known as the hardest of all natural materials, diamonds are also exceptional thermal conductors and electrical insulators. Now, researchers have discovered a way to tweak tiny needles of diamond in a controlled way to transform their electronic properties, dialing them from insulating, through semiconducting, all the way to highly conductive, or metallic. This can be induced dynamically and reversed at will, with no degradation of the diamond material.</p>
1284
1285 <p>The research, though still at an early proof-of-concept stage, may open up a wide array of potential applications, including new kinds of broadband solar cells, highly efficient LEDs and power electronics, and new optical devices or quantum sensors, the researchers say.</p>
1286
1287 <p>Their findings, which are based on simulations, calculations, and previous experimental results, are reported this week in the <em>Proceedings of the National Academy of Sciences. </em>The<em> </em>paper is by MIT Professor Ju Li and graduate student Zhe Shi; Principal Research Scientist Ming Dao; Professor Subra Suresh, who is president of Nanyang Technological University in Singapore as well as former dean of engineering and Vannevar Bush Professor Emeritus at MIT; and Evgenii Tsymbalov and Alexander Shapeev at the Skolkovo Institute of Science and Technology in Moscow.</p>
1288
1289 <p>The team used a combination of quantum mechanical calculations, analyses of mechanical deformation, and machine learning to demonstrate that the phenomenon, long theorized as a possibility, really can occur in nanosized diamond.</p>
1290
1291 <p>The concept of straining a semiconductor material such as silicon to improve its performance found applications in the microelectronics industry more than two decades ago. However, that approach entailed small strains on the order of about 1 percent. Li and his collaborators have spent years developing the concept of elastic strain engineering. This is based on the ability to cause significant changes in the electrical, optical, thermal, and other properties of materials simply by deforming them — putting them under moderate to large mechanical strain, enough to alter the geometric arrangement of atoms in the material’s crystal lattice, but without disrupting that lattice.</p>
1292
1293 <p>In a major advance in 2018, a team led by Suresh, Dao, and Yang Lu from the City University of Hong Kong showed that tiny needles of diamond, just a few hundred nanometers across, could be <a href="https://news.mit.edu/2018/bend-stretch-diamond-ultrafine-needles-0419">bent without fracture</a> at room temperature to large strains. They were able to repeatedly bend these nanoneedles to tensile strain as much as 10 percent; the needles can then return intact to their original shape.</p>
1294
1295 <p>Key to this work is a property known as <a href="https://news.mit.edu/2010/explained-bandgap-0723">bandgap</a>, which essentially determines how readily electrons can move through a material. This property is thus key to the material’s electrical conductivity. Diamond normally has a very wide bandgap of 5.6 electron volts, meaning that it is a strong electrical insulator that electrons do not move through readily. In their latest simulations, the researchers show that diamond’s bandgap can be gradually, continuously, and reversibly changed, providing a wide range of electrical properties, from insulator through semiconductor to metal.</p>
1296
1297 <p>“We found that it’s possible to reduce the bandgap from 5.6 electron volts all the way to zero,” Li says. “The point of this is that if you can change continuously from 5.6 to 0 electron volts, then you cover all the range of bandgaps. Through strain engineering, you can make diamond have the bandgap of silicon, which is most widely used as a semiconductor, or gallium nitride, which is used for LEDs. You can even have it become an infrared detector or detect a whole range of light all the way from the infrared to the ultraviolet part of the spectrum.”</p>
1298
1299 <p>“The ability to engineer and design electrical conductivity in diamond without changing its chemical composition and stability offers unprecedented flexibility to custom-design its functions,” says Suresh. “The methods demonstrated in this work could be applied to a broad range of other semiconductor materials of technological interest in mechanical, microelectronics, biomedical, energy and photonics applications, through strain engineering.”</p>
1300
1301 <p>So, for example, a single tiny piece of diamond, bent so that it has a gradient of strain across it, could become a solar cell capable of capturing all frequencies of light on a single device — something that currently can only be achieved through tandem devices that couple different kinds of solar cell materials together in layers to combine their different absorption bands. These might someday be used as broad-spectrum photodetectors for industrial or scientific applications.</p>
1302
1303 <p>One constraint, which required not only the right amount of strain but also the right orientation of the diamond’s crystalline lattice, was to prevent the strain from causing the atomic configuration to cross a tipping point and turning into graphite, the soft material used in pencils.</p>
1304
1305 <p>The process can also make diamond into two types of semiconductors, either “direct” or “indirect” bandgap semiconductors, depending on the intended application. For solar cells, for example, direct bandgaps provide a much more efficient collection of energy from light, allowing them to be much thinner than materials such as silicon, whose indirect bandgap requires a much longer pathway to collect a photon’s energy.</p>
1306
1307 <p>The process could be relevant for a wide variety of potential applications, Li suggests, such as for highly sensitive quantum-based detectors that use defects and dopant atoms in a diamond. “Using strain, we can control the emission and absorption levels of these point defects,” he says, allowing novel ways of controlling their electronic and nuclear quantum states.</p>
1308
1309 <p>But given the great variety of conditions made possible by the different dimensions of strain variations, Li says, “if we have a particular application in mind, then we could optimize toward that application target. And what is nice about the elastic straining approach is that it is dynamic,” so that it can be continuously varied in real time as needed.</p>
1310
1311 <p>This early-stage proof-of-concept work is not yet at the point where they can begin to design practical devices, the researchers say, but with the ongoing research they expect that practical applications could be possible, partly because of promising work being done around the world on the growth of homogeneous diamond materials.</p>
1312
1313 <p>The work was supported by the U.S. Office of Naval Research.</p>
1314 </content:encoded>
1315 <media:content url="https://news.mit.edu/sites/default/files/styles/news_article__cover_image__original/public/images/202010/MIT-Metallic-Diamond-01-Press.jpg?itok=A1hYK-fG" medium="image" type="image/jpeg" width="390" height="260">
1316 <media:description type="plain">Researchers have discovered a way to transform the electronic properties of nanoscale needles of diamond.</media:description>
1317 <media:credit>Image: MIT News</media:credit>
1318 </media:content>
1319 </item>
1320 <item>
1321 <title>Fotini Christia named director of the Sociotechnical Systems Research Center</title>
1322 <link>https://news.mit.edu/2020/fotini-christia-named-director-sociotechnical-systems-research-center-1014</link>
1323 <description>Political science professor will spearhead the Institute’s interdisciplinary center that studies high-impact, complex societal challenges.</description>
1324 <pubDate>Wed, 14 Oct 2020 15:40:00 -0400</pubDate>
1325 <guid isPermaLink="true">https://news.mit.edu/2020/fotini-christia-named-director-sociotechnical-systems-research-center-1014</guid>
1326 <dc:creator>Terri Park | MIT Schwarzman College of Computing</dc:creator>
1327 <content:encoded><p>Professor <a href="http://fotini.mit.edu/" target="_blank">Fotini Christia</a> has been named the director of the Sociotechnical Systems Research Center (SSRC) at MIT.</p>
1328
1329 <p>A professor in the Department of Political Science, Christia stepped into her new role with SSRC on Oct. 1. The interdisciplinary center, part of the Institute for Data, Systems, and Society in the MIT Stephen A. Schwarzman College of Computing, focuses on the study of high-impact, complex societal challenges that shape our world.</p>
1330
1331 <p>Christia succeeds Ali Jadbabaie, the JR East Professor of Engineering, who has led SSRC since 2016. Jadbabaie recently stepped down to become the new head of the Department of Civil and Environmental Engineering.</p>
1332
1333 <p>“Fotini’s breadth as a social scientist, on-the-ground approach, use of data science and computational techniques, and application of novel methods to understand how societies are being shaped in diverse areas, made her a natural fit to lead SSRC into the next chapter,” says Daniel Huttenlocher, dean of the MIT Schwarzman College of Computing. “I’m delighted to welcome her and look forward to collaborating on behalf of the college and the Institute.”</p>
1334
1335 <p>Christia’s research interests deal with the political economy of conflict and&nbsp;development in the Muslim world, for which she has done extensive experimental and survey-related fieldwork in Afghanistan, Bosnia-Herzegovina, Iraq, and Yemen. She is presently using cellphone and social media data in ongoing research on refugee return in Syria, and on gender-based violence in Egypt during Covid-19.</p>
1336
1337 <p>She is the author of “Alliance Formation in Civil War” (Cambridge University Press, 2012), which rationalizes that warring group alliances are not eternally divided along ethnic or religious lines, but rather are dynamic, formed for more instrumental reasons that often reflect shifts in the balance of power. Her book was awarded the Luebbert Award for Best Book in Comparative Politics, the Lepgold Prize for Best Book in International Relations, and a Distinguished Book Award from the International Studies Association.&nbsp;</p>
1338
1339 <p>Her research has also appeared in<em> Science</em>, <em>Review of Economic Studies, </em><em>IEEE Transactions on Network Science and Engineering,<em> </em></em>and<em><em> American Political Science Review, </em></em>among other journals, and her opinion pieces have been published in <em>Foreign Affairs</em>, <em>The New York Times</em>, and <em>The</em> <em>Washington Post</em>.<em> </em>She has been awarded an inaugural Andrew Carnegie fellowship and a Harvard Academy fellowship.</p>
1340
1341 <p>A native of Greece, where she grew up in the port city of Salonika, Christia moved to the United States to attend college at Columbia University. She graduated magna cum laude in 2001 with a joint BA in economics–operations research and an MA in international affairs. She joined the MIT faculty in 2008 after receiving her PhD in public policy from Harvard University.</p>
1342 </content:encoded>
1343 <media:content url="https://news.mit.edu/sites/default/files/styles/news_article__cover_image__original/public/images/202010/fotini-christia-mit-a1_0.jpg?itok=w3VCk8jA" medium="image" type="image/jpeg" width="390" height="260">
1344 <media:description type="plain">Fotini Chrisita, MIT professor of political science, has been named director of the Sociotechnical Systems Research Center.</media:description>
1345 <media:credit>Photo: MIT Political Science</media:credit>
1346 </media:content>
1347 </item>
1348 <item>
1349 <title>Nanoparticles can turn off genes in bone marrow cells</title>
1350 <link>https://news.mit.edu/2020/nanoparticles-bone-marrow-rnai-1005</link>
1351 <description>Using these new particles, researchers could develop treatments for heart disease and other conditions.</description>
1352 <pubDate>Mon, 05 Oct 2020 11:00:00 -0400</pubDate>
1353 <guid isPermaLink="true">https://news.mit.edu/2020/nanoparticles-bone-marrow-rnai-1005</guid>
1354 <dc:creator>Anne Trafton | MIT News Office</dc:creator>
1355 <content:encoded><p>Using specialized nanoparticles, MIT engineers have developed a way to turn off specific genes in cells of the bone marrow, which play an important role in producing blood cells. These particles could be tailored to help treat heart disease or to boost the yield of stem cells in patients who need stem cell transplants, the researchers say.</p>
1356
1357 <p>This type of genetic therapy, known as RNA interference, is usually difficult to target to organs other than the liver, where nanoparticles would tend to accumulate. The MIT researchers were able to modify their particles in such a way that they would accumulate in the cells found in the bone marrow.</p>
1358
1359 <p>“If we can get these particles to hit other organs of interest, there could be a broader range of disease applications to explore, and one that we were really interested in this paper was the bone marrow. The bone marrow is a site for hematopoiesis of blood cells, and these give rise to a whole lineage of cells that contribute to various types of diseases,” says Michael Mitchell, a former MIT postdoc and one of the lead authors of the study.</p>
1360
1361 <p>In a study of mice, the researchers showed that they could use this approach to improve recovery after a heart attack by inhibiting the release of bone marrow blood cells that promote inflammation and contribute to heart disease.</p>
1362
1363 <p>Marvin Krohn-Grimberghe, a cardiologist at the Freiburg University Heart Center in Germany, and Maximilian Schloss, a research fellow at Massachusetts General Hospital, are also lead authors of the paper, which appears today in <em>Nature Biomedical Engineering</em>. The paper’s senior authors are Daniel Anderson, a professor of chemical engineering at MIT and a member of MIT’s Koch Institute for Integrative Cancer Research and Institute for Medical Engineering and Science, and Matthias Nahrendorf, a professor of radiology at MGH.</p>
1364
1365 <p><strong>Targeting the bone marrow</strong></p>
1366
1367 <p>RNA interference is a strategy that could potentially be used to treat a variety of diseases by delivering short strands of RNA that block specific genes from being turned on in a cell. So far, the biggest obstacle to this kind of therapy has been the difficulty in delivering it to the right part of the body. When injected into the bloodstream, nanoparticles carrying RNA tend to accumulate in the liver, which some biotech companies have taken advantage of to develop new experimental treatments for liver disease.</p>
1368
1369 <p>Anderson’s lab, working with MIT Institute Professor Robert Langer, who is also an author of the new study, has previously developed a type of polymer nanoparticles that can <a href="https://news.mit.edu/2018/new-materials-improve-delivery-therapeutic-messenger-rna-0716">deliver RNA</a> to organs other than the liver. The particles are coated with lipids that help stabilize them, and they can target organs such as the lungs, heart, and spleen, depending on the particles’ composition and molecular weight.</p>
1370
1371 <p>“RNA nanoparticles are currently FDA-approved as a liver-targeted therapy but hold promise for many diseases, ranging from Covid-19 vaccines to drugs that can permanently repair disease genes,” Anderson says. “We believe that engineering nanoparticles to deliver RNA to different types of cells and organs in the body is key to reaching the broadest potential of genetic therapy.”</p>
1372
1373 <p>In the new study, the researchers set out to adapt the particles so that they could reach the bone marrow. The bone marrow contains stem cells that produce many different types of blood cells, through a process called hematopoiesis. Stimulating this process could enhance the yield of hematopoietic stem cells for stem cell transplantation, while repressing it could have beneficial effects on patients with heart disease or other diseases.</p>
1374
1375 <p>“If we could develop technologies that could control cellular activity in bone marrow and the hematopoietic stem cell niche, it could be transformative for disease applications,” says Mitchell, who is now an assistant professor of bioengineering at the University of Pennsylvania.</p>
1376
1377 <p>The researchers began with the particles they had previously used to target the lungs and created variants that had different arrangements of a surface coating called polyethylene glycol (PEG). They tested 15 of these particles and found one that was able to avoid being caught in the liver or the lungs, and that could effectively accumulate in endothelial cells of the bone marrow. They also showed that RNA carried by this particle could reduce the expression of a target gene by up to 80 percent.</p>
1378
1379 <p>The researchers tested this approach with two genes that they believed could be beneficial to knock down. The first, SDF1, is a molecule that normally prevents hematopoietic stem cells from leaving the bone marrow. Turning off this gene could achieve the same effect as the drugs that doctors often use to induce hematopoietic stem cell release in patients who need to undergo radiation treatments for blood cancers. These stem cells are later transplanted to repopulate the patient’s blood cells.</p>
1380
1381 <p>“If you have a way to knock down SDF1, you can cause the release of these hematopoietic stem cells, which could be very important for a transplantation so you can harvest more from the patient,” Mitchell says.</p>
1382
1383 <p>The researchers showed that when they used their nanoparticles to knock down SDF1, they could boost the release of hematopoietic stem cells fivefold, which is comparable to the levels achieved by the drugs that are now used to enhance stem cell release. They also showed that these cells could successfully differentiate into new blood cells when transplanted into another mouse.</p>
1384
1385 <p>“We are very excited about the latest results,” says Langer, who is also the David H. Koch Institute Professor at MIT. “Previously we have developed high-throughput synthesis and screening approaches to target the liver and blood vessel cells, and now in this study, the bone marrow. We hope this will lead to new treatments for diseases of the bone marrow like multiple myeloma and other illnesses.”</p>
1386
1387 <p><strong>Combatting heart disease</strong></p>
1388
1389 <p>The second gene that the researchers targeted for knockdown is called MCP1, a molecule that plays a key role in heart disease. When MCP1 is released by bone marrow cells after a heart attack, it stimulates a flood of immune cells to leave the bone marrow and travel to the heart, where they promote inflammation and can lead to further heart damage.</p>
1390
1391 <p>In a study of mice, the researchers found that delivering RNA that targets MCP1 reduced the number of immune cells that went to the heart after a heart attack. Mice that received this treatment also showed improved healing of heart tissue following a heart attack.</p>
1392
1393 <p>“We now know that immune cells play such a key role in the progression of heart attack and heart failure,” Mitchell says. “If we could develop therapeutic strategies to stop immune cells that originate from bone marrow from getting into the heart, it could be a new means of treating heart attack. This is one of the first demonstrations of a nucleic-acid-based approach of doing this.”</p>
1394
1395 <p>At his lab at the University of Pennsylvania, Mitchell is now working on new nanotechnologies that target bone marrow and immune cells for treating other diseases, especially blood cancers such as multiple myeloma.</p>
1396
1397 <p>The research was funded in part by the National Institutes of Health, the European Union’s Horizon 2020 research and innovation program, the MGH Research Scholar Program, a Burroughs Wellcome Fund Career Award at the Scientific Interface, a Koch-Prostate Cancer Foundation Award in Nanotherapeutics, the Koch Institute Marble Center for Cancer Nanomedicine, and the Koch Institute Support (core) Grant from the National Cancer Institute.</p>
1398 </content:encoded>
1399 <media:content url="https://news.mit.edu/sites/default/files/styles/news_article__cover_image__original/public/images/202010/MIT-Bone-Marrow-01-PRESS.jpg?itok=OS8PEbMh" medium="image" type="image/jpeg" width="390" height="260">
1400 <media:description type="plain">MIT researchers have shown they can deliver RNAi nanoparticles to the bone marrow, influencing their function. At top right, the bone marrow is not yet treated with particles that turn off a gene called SDF1. At bottom right, the number of neutrophils (blue) decreases, indicating that they have been released from bone marrow after treatment. At left, treatment with a control nanoparticle does not affect the number of neutrophils before and after treatment.</media:description>
1401 <media:credit>Image courtesy of the researchers</media:credit>
1402 </media:content>
1403 </item>
1404 <item>
1405 <title>Neuroscientists discover a molecular mechanism that allows memories to form</title>
1406 <link>https://news.mit.edu/2020/engram-memories-form-1005</link>
1407 <description>Modifications to chromosomes in “engram” neurons control the encoding and retrieval of memories.</description>
1408 <pubDate>Mon, 05 Oct 2020 11:00:00 -0400</pubDate>
1409 <guid isPermaLink="true">https://news.mit.edu/2020/engram-memories-form-1005</guid>
1410 <dc:creator>Anne Trafton | MIT News Office</dc:creator>
1411 <content:encoded><p>When the brain forms a memory of a new experience, neurons called engram cells encode the details of the memory and are later reactivated whenever we recall it. A new MIT study reveals that this process is controlled by large-scale remodeling of cells’ chromatin.</p>
1412
1413 <p>This remodeling, which allows specific genes involved in storing memories to become more active, takes place in multiple stages spread out over several days. Changes to the density and arrangement of chromatin, a highly compressed structure consisting of DNA and proteins called histones, can control how active specific genes are within a given cell.</p>
1414
1415 <p>“This paper is the first to really reveal this very mysterious process of how different waves of genes become activated, and what is the epigenetic mechanism underlying these different waves of gene expression,” says Li-Huei Tsai, the director of MIT’s Picower Institute for Learning and Memory and the senior author of the study.</p>
1416
1417 <p>Asaf Marco, an MIT postdoc, is the lead author of the paper, which appears today in <em>Nature Neuroscience</em>.</p>
1418
1419 <p><strong>Epigenomic control</strong></p>
1420
1421 <p>Engram cells are found in the hippocampus as well as other parts of the brain. Many recent studies have shown that these cells form networks that are associated with particular memories, and these networks are activated when that memory is recalled. However, the molecular mechanisms underlying the encoding and retrieval of these memories are not well-understood.</p>
1422
1423 <p>Neuroscientists know that in the very first stage of memory formation, genes known as immediate early genes are turned on in engram cells, but these genes soon return to normal activity levels. The MIT team wanted to explore what happens later in the process to coordinate the long-term storage of memories.</p>
1424
1425 <p>“The formation and preservation of memory is a very delicate and coordinated event that spreads over hours and days, and might be even months — we don’t know for sure,” Marco says. “During this process, there are a few waves of gene expression and protein synthesis that make the connections between the neurons stronger and faster.”</p>
1426
1427 <p>Tsai and Marco hypothesized that these waves could be controlled by epigenomic modifications, which are chemical alterations of chromatin that control whether a particular gene is accessible or not. <a href="http://news.mit.edu/2012/alzheimers-hdac2-inhibitors-0301">Previous studies</a> from Tsai’s lab have shown that when enzymes that make chromatin inaccessible are too active, they can interfere with the ability to form new memories.</p>
1428
1429 <p>To study epigenomic changes that occur in individual engram cells over time, the researchers used genetically engineered mice in which they can permanently tag engram cells in the hippocampus with a fluorescent protein when a memory is formed. These mice received a mild foot shock that they learned to associate with the cage in which they received the shock. When this memory forms, the hippocampal cells encoding the memory begin to produce a yellow fluorescent protein marker.</p>
1430
1431 <p>“Then we can track those neurons forever, and we can sort them out and ask what happens to them one hour after the foot shock, what happens five days after, and what happens when those neurons get reactivated during memory recall,” Marco says.</p>
1432
1433 <p>At the very first stage, right after a memory is formed, the researchers found that many regions of DNA undergo chromatin modifications. In these regions, the chromatin becomes looser, allowing the DNA to become more accessible. To the researchers’ surprise, nearly all of these regions were in stretches of DNA where no genes are found. These regions contain noncoding sequences called enhancers, which interact with genes to help turn them on. The researchers also found that in this early stage, the chromatin modifications did not have any effect on gene expression.</p>
1434
1435 <p>The researchers then analyzed engram cells five days after memory formation. They found that as memories were consolidated, or strengthened, over those five days, the 3D structure of the chromatin surrounding the enhancers changed, bringing the enhancers closer to their target genes. This still doesn’t turn on those genes, but it primes them to be expressed when the memory is recalled.</p>
1436
1437 <p>Next, the researchers placed some of the mice back into the chamber where they received the foot shock, reactivating the fearful memory. In engram cells from those mice, the researchers found that the primed enhancers interacted frequently with their target genes, leading to a surge in the expression of those genes.</p>
1438
1439 <p>Many of the genes turned on during memory recall are involved in promoting protein synthesis at the synapses, helping neurons strengthen their connections with other neurons. The researchers also found that the neurons’ dendrites — branched extensions that receive input from other neurons — developed more spines, offering further evidence that their connections were further strengthened.</p>
1440
1441 <p><strong>Primed for expression</strong></p>
1442
1443 <p>The study is the first to show that memory formation is driven by epigenomically priming enhancers to stimulate gene expression when a memory is recalled, Marco says.</p>
1444
1445 <p>“This is the first work that shows on the molecular level how the epigenome can be primed to gain accessibility. First, you make the enhancers more accessible, but the accessibility on its own is not sufficient. You need those regions to physically interact with the genes, which is the second phase,” he says. “We are now realizing that the 3D genome architecture plays a very significant role in orchestrating gene expression.”</p>
1446
1447 <p>The researchers did not explore how long these epigenomic modifications last, but Marco says he believes they may remain for weeks or even months. He now hopes to study how the chromatin of engram cells is affected by Alzheimer’s disease. <a href="http://news.mit.edu/2007/picower-team-reverses-alzheimers-symptoms-mice">Previous work</a> from Tsai’s lab has shown that treating a mouse model of Alzheimer’s with an HDAC inhibitor, a drug that helps to reopen inaccessible chromatin, can help to restore lost memories.</p>
1448
1449 <p>The research was funded by the JBP Foundation and the Alzheimer’s Association.</p>
1450 </content:encoded>
1451 <media:content url="https://news.mit.edu/sites/default/files/styles/news_article__cover_image__original/public/images/202010/MIT-Engram-Mod-01-Press.jpg?itok=x88BNGz3" medium="image" type="image/jpeg" width="390" height="260">
1452 <media:description type="plain">A new MIT study reveals that encoding memories in engram cells is controlled by large-scale remodeling of the proteins and DNA that make up cells’ chromatin. In this image of the brain, the hippocampus is the large yellow structure near the top. Green indicates neurons that were activated in memory formation; red shows the neurons that were activated in memory recall; blue shows the DNA of the cells; and yellow shows neurons that were activated in both memory formation and recall, and are thus considered to be the engram neurons.</media:description>
1453 <media:credit>Courtesy of the researchers </media:credit>
1454 </media:content>
1455 </item>
1456 <item>
1457 <title>How we make moral decisions</title>
1458 <link>https://news.mit.edu/2020/moral-decisions-universalization-1002</link>
1459 <description>In some situations, asking “what if everyone did that?” is a common strategy for judging whether an action is right or wrong.</description>
1460 <pubDate>Fri, 02 Oct 2020 16:22:00 -0400</pubDate>
1461 <guid isPermaLink="true">https://news.mit.edu/2020/moral-decisions-universalization-1002</guid>
1462 <dc:creator>Anne Trafton | MIT News Office</dc:creator>
1463 <content:encoded><p>Imagine that one day you’re riding the train and decide to hop the turnstile to avoid paying the fare. It probably won’t have a big impact on the financial well-being of your local transportation system. But now ask yourself, “What if everyone did that?” The outcome is much different — the system would likely go bankrupt and no one would be able to ride the train anymore.</p>
1464
1465 <p>Moral philosophers have long believed this type of reasoning, known as universalization, is the best way to make moral decisions. But do ordinary people spontaneously use this kind of moral judgment in their everyday lives?</p>
1466
1467 <p>In a study of several hundred people, MIT and Harvard University researchers have confirmed that people do use this strategy in particular situations called “threshold problems.” These are social dilemmas in which harm can occur if everyone, or a large number of people, performs a certain action. The authors devised a mathematical model that quantitatively predicts the judgments they are likely to make. They also showed, for the first time, that children as young as 4 years old can use this type of reasoning to judge right and wrong.</p>
1468
1469 <p>“This mechanism seems to be a way that we spontaneously can figure out what are the kinds of actions that I can do that are sustainable in my community,” says Sydney Levine, a postdoc at MIT and Harvard and the lead author of the study.</p>
1470
1471 <p>Other authors of the study are Max Kleiman-Weiner, a postdoc at MIT and Harvard; Laura Schulz, an MIT professor of cognitive science; Joshua Tenenbaum, a professor of computational cognitive science at MIT and a member of MIT’s Center for Brains, Minds, and Machines and Computer Science and Artificial Intelligence Laboratory (CSAIL); and Fiery Cushman, an assistant professor of psychology at Harvard. The paper is appearing this week in the <em>Proceedings of the National Academy of Sciences</em>.</p>
1472
1473 <p><strong>Judging morality</strong></p>
1474
1475 <p>The concept of universalization has been included in philosophical theories since at least the 1700s. Universalization is one of several strategies that philosophers believe people use to make moral judgments, along with outcome-based reasoning and rule-based reasoning. However, there have been few psychological studies of universalization, and many questions remain regarding how often this strategy is used, and under what circumstances.</p>
1476
1477 <p>To explore those questions, the MIT/Harvard team asked participants in their study to evaluate the morality of actions taken in situations where harm could occur if too many people perform the action. In one hypothetical scenario, John, a fisherman, is trying to decide whether to start using a new, more efficient fishing hook that will allow him to catch more fish. However, if every fisherman in his village decided to use the new hook, there would soon be no fish left in the lake.</p>
1478
1479 <p>The researchers found that many subjects did use universalization to evaluate John’s actions, and that their judgments depended on a variety of factors, including&nbsp;the number of people who were interested in using the new hook and the number of people using it that would trigger a harmful outcome.</p>
1480
1481 <p>To tease out the impact of those factors, the researchers created several versions of the scenario. In one, no one else in the village was interested in using the new hook, and in that scenario, most participants deemed it acceptable for John to use it. However, if others in the village were interested but chose not to use it, then John’s decision to use it was judged to be morally wrong.</p>
1482
1483 <p>The researchers also found that they could use their data to create a mathematical model that explains how people take different factors into account, such as the number of people who want to do the action and the number of people doing it that would cause harm. The model accurately predicts how people’s judgments change when these factors change.</p>
1484
1485 <p>In their last set of studies, the researchers created scenarios that they used to test judgments made by children between the ages of 4 and 11. One story featured a child who wanted to take a rock from a path in a park for his rock collection. Children were asked to judge if that was OK, under two different circumstances: In one, only one child wanted a rock, and in the other, many other children also wanted to take rocks for their collections.</p>
1486
1487 <p>The researchers found that most of the children deemed it wrong to take a rock if everyone wanted to, but permissible if there was only one child who wanted to do it. However, the children were not able to specifically explain why they had made those judgments.</p>
1488
1489 <p>“What's interesting about this is we discovered that if you set up this carefully controlled contrast, the kids seem to be using this computation, even though they can't articulate it,” Levine says. “They can't introspect on their cognition and know what they're doing and why, but they seem to be deploying the mechanism anyway.”</p>
1490
1491 <p>In future studies, the researchers hope to explore how and when the ability to use this type of reasoning develops in children.</p>
1492
1493 <p><strong>Collective action</strong></p>
1494
1495 <p>In the real world, there are many instances where universalization could be a good strategy for making decisions, but it’s not necessary because rules are already in place governing those situations.</p>
1496
1497 <p>“There are a lot of collective action problems in our world that can be solved with universalization, but they're already solved with governmental regulation,” Levine says. “We don't rely on people to have to do that kind of reasoning, we just make it illegal to ride the bus without paying.”</p>
1498
1499 <p>However, universalization can still be useful in situations that arise suddenly, before any government regulations or guidelines have been put in place. For example, at the beginning of the Covid-19 pandemic, before many local governments began requiring masks in public places, people contemplating wearing masks might have asked themselves what would happen if everyone decided not to wear one.</p>
1500
1501 <p>The researchers now hope to explore the reasons why people sometimes don’t seem to use universalization in cases where it could be applicable, such as combating climate change. One possible explanation is that people don’t have enough information about the potential harm that can result from certain actions, Levine says.</p>
1502
1503 <p>The research was funded by the John Templeton Foundation, the Templeton World Charity Foundation, and the Center for Brains, Minds, and Machines.</p>
1504 </content:encoded>
1505 <media:content url="https://news.mit.edu/sites/default/files/styles/news_article__cover_image__original/public/images/202010/MIT-Moral-Cognition-01-Press.jpg?itok=9B-c24Yf" medium="image" type="image/jpeg" width="390" height="260">
1506 <media:description type="plain">Researchers at MIT and Harvard have shown that people use a type of reasoning known as universalization to help them make moral decisions in certain types of situations. This strategy is most applicable in social dilemmas called “threshold problems,” in which harm can occur if everyone, or a large number of people, perform a certain action.</media:description>
1507 <media:credit>Image: iStock illustration edited by MIT News</media:credit>
1508 </media:content>
1509 </item>
1510 <item>
1511 <title>The real promise of synthetic data </title>
1512 <link>https://news.mit.edu/2020/real-promise-synthetic-data-1016</link>
1513 <description>MIT researchers release the Synthetic Data Vault, a set of open-source tools meant to expand data access without compromising privacy.</description>
1514 <pubDate>Fri, 16 Oct 2020 12:10:00 -0400</pubDate>
1515 <guid isPermaLink="true">https://news.mit.edu/2020/real-promise-synthetic-data-1016</guid>
1516 <dc:creator>Laboratory for Information and Decision Systems</dc:creator>
1517 <content:encoded><p>Each year, the world generates more data than the previous year. In 2020 alone, <a href="https://www.idc.com/getdoc.jsp?containerId=prUS46286020">an estimated 59 zettabytes of data</a> will be “created, captured, copied, and consumed,” according to the International Data Corporation — enough to fill about a trillion 64-gigabyte hard drives.</p>
1518
1519 <p>But just because data are proliferating doesn't mean everyone can actually use them. Companies and institutions, rightfully concerned with their users' privacy, often restrict access to datasets — sometimes within their own teams. And now that the Covid-19 pandemic has shut down labs and offices, preventing people from visiting centralized data stores, sharing information safely is even more difficult.</p>
1520
1521 <p>Without access to data, it's hard to make tools that actually work. Enter synthetic data: artificial information developers and engineers can use as a stand-in for real data.</p>
1522
1523 <p>Synthetic data is a bit like diet soda. To be effective, it has to resemble the “real thing” in certain ways. Diet soda should look, taste, and fizz like regular soda. Similarly, a synthetic dataset must have the same mathematical and statistical properties as the real-world dataset it's standing in for. “It looks like it, and has formatting like it,” says Kalyan Veeramachaneni, principal investigator of the Data to AI (DAI) Lab and a principal research scientist in MIT’s Laboratory for Information and Decision Systems. If it's run through a model, or used to build or test an application, it performs like that real-world data would.</p>
1524
1525 <p>But — just as diet soda should have fewer calories than the regular variety — a synthetic dataset must also differ from a real one in crucial aspects. If it's based on a real dataset, for example, it shouldn't contain or even hint at any of the information from that dataset.</p>
1526
1527 <p>Threading this needle is tricky. After years of work, Veeramachaneni and his collaborators recently unveiled a set of open-source data generation tools — a one-stop shop where users can get as much data as they need for their projects, in formats from tables to time series. They call it the <a href="https://sdv.dev/" target="_blank">Synthetic Data Vault</a>.</p>
1528
1529 <p><strong>Maximizing access while maintaining privacy</strong></p>
1530
1531 <p>Veeramachaneni and his team first tried to create synthetic data in 2013. They had been tasked with analyzing a large amount of information from the online learning program edX, and wanted to bring in some MIT students to help. The data were sensitive, and couldn't be shared with these new hires, so the team decided to create artificial data that the students could work with instead — figuring that “once they wrote the processing software, we could use it on the real data,” Veeramachaneni says.</p>
1532
1533 <p>This is a common scenario. Imagine you're a software developer contracted by a hospital. You've been asked to build a dashboard that lets patients access their test results, prescriptions, and other health information. But you aren't allowed to see any real patient data, because it's private.</p>
1534
1535 <p>Most developers in this situation will make “a very simplistic version" of the data they need, and do their best, says Carles Sala, a researcher in the DAI lab. But when the dashboard goes live, there's a good chance that “everything crashes,” he says, “because there are some edge cases they weren't taking into account.”</p>
1536
1537 <p>High-quality synthetic data — as complex as what it's meant to replace — would help to solve this problem. Companies and institutions could share it freely, allowing teams to work more collaboratively and efficiently. Developers could even carry it around on their laptops, knowing they weren't putting any sensitive information at risk.</p>
1538
1539 <p><strong>Perfecting the formula — and handling constraints</strong></p>
1540
1541 <p>Back in 2013, Veeramachaneni's team gave themselves two weeks to create a data pool they could use for that edX project. The timeline “seemed really reasonable,” Veeramachaneni says. “But we failed completely.” They soon realized that if they built a series of synthetic data generators, they could make the process quicker for everyone else.</p>
1542
1543 <p>In 2016, the team completed an algorithm that accurately captures correlations between the different fields in a real dataset — think a patient's age, blood pressure, and heart rate — and creates a synthetic dataset that preserves those relationships, without any identifying information. When data scientists were asked to solve problems using this synthetic data, their solutions were as effective as those made with real data 70 percent of the time. The team <a href="http://news.mit.edu/2017/artificial-data-give-same-results-as-real-data-0303">presented this research</a> at the 2016 IEEE International Conference on Data Science and Advanced Analytics.</p>
1544
1545 <p>For the next go-around, the team reached deep into the machine learning toolbox. In 2019, PhD student Lei Xu presented his new algorithm, <a href="https://papers.nips.cc/paper/8953-modeling-tabular-data-using-conditional-gan.pdf">CTGAN</a>, at the 33rd Conference on Neural Information Processing Systems in Vancouver. CTGAN (for "conditional tabular generative adversarial networks) uses GANs to build and perfect synthetic data tables. GANs are pairs of neural networks that “play against each other,” Xu says. The first network, called a generator, creates something — in this case, a row of synthetic data — and the second, called the discriminator, tries to tell if it's real or not.</p>
1546
1547 <p>“Eventually, the generator can generate perfect [data], and the discriminator cannot tell the difference,” says Xu. GANs are more often used in artificial image generation, but they work well for synthetic data, too: CTGAN outperformed classic synthetic data creation techniques in 85 percent of the cases tested in Xu's study.</p>
1548
1549 <p>Statistical similarity is crucial. But depending on what they represent, datasets also come with their own vital context and constraints, which must be preserved in synthetic data. DAI lab researcher Sala gives the example of a hotel ledger: a guest always checks out after he or she checks in. The dates in a synthetic hotel reservation dataset must follow this rule, too: “They need to be in the right order,” he says.</p>
1550
1551 <p>Large datasets may contain a number of different relationships like this, each strictly defined. “Models cannot learn the constraints, because those are very context-dependent,” says Veeramachaneni. So the team recently finalized an interface that allows people to tell a synthetic data generator where those bounds are. “The data is generated within those constraints,” Veeramachaneni says.</p>
1552
1553 <p>Such precise data could aid companies and organizations in many different sectors. One example is banking, where increased digitization, along with new data privacy rules, have “triggered a growing interest in ways to generate synthetic data,” says Wim Blommaert, a team leader at ING financial services. Current solutions, like data-masking, often destroy valuable information that banks could otherwise use to make decisions, he said. A tool like SDV has the potential to sidestep the sensitive aspects of data while preserving these important constraints and relationships.</p>
1554
1555 <p><strong>One vault to rule them all</strong></p>
1556
1557 <p>The Synthetic Data Vault combines everything the group has built so far into “a whole ecosystem,” says Veeramachaneni. The idea is that stakeholders — from students to professional software developers — can come to the vault and get what they need, whether that's a large table, a small amount of time-series data, or a mix of many different data types.</p>
1558
1559 <p>The vault is open-source and expandable. “There are a whole lot of different areas where we are realizing synthetic data can be used as well,” says Sala. For example, if a particular group is underrepresented in a sample dataset, synthetic data can be used to fill in those gaps — a sensitive endeavor that requires a lot of finesse. Or companies might also want to use synthetic data to plan for scenarios they haven't yet experienced, like a huge bump in user traffic.</p>
1560
1561 <p>As use cases continue to come up, more tools will be developed and added to the vault, Veeramachaneni says. It may occupy the team for another seven years at least, but they are ready: “We're just touching the tip of the iceberg.”</p>
1562 </content:encoded>
1563 <media:content url="https://news.mit.edu/sites/default/files/styles/news_article__cover_image__original/public/images/202010/LIDS-Synthetic-Data-Vault_0.png?itok=JgdJqhlU" medium="image" type="image/jpeg" width="390" height="260">
1564 <media:description type="plain">After years of work, MIT's Kalyan Veeramachaneni and his collaborators recently unveiled a set of open-source data generation tools — a one-stop shop where users can get as much data as they need for their projects, in formats from tables to time series. They call it the Synthetic Data Vault.</media:description>
1565 <media:credit>Image: Arash Akhgari</media:credit>
1566 </media:content>
1567 </item>
1568 <item>
1569 <title>Finding patterns in the noise</title>
1570 <link>https://news.mit.edu/2020/finding-patterns-noise-sean-liu-1014</link>
1571 <description>Using novel computational approaches, graduate student Sean Liu develops better tools for analyzing data.</description>
1572 <pubDate>Wed, 14 Oct 2020 15:15:00 -0400</pubDate>
1573 <guid isPermaLink="true">https://news.mit.edu/2020/finding-patterns-noise-sean-liu-1014</guid>
1574 <dc:creator>Leda Zimmerman | Department of Political Science</dc:creator>
1575 <content:encoded><p>When social scientists administer surveys and questionnaires, they cannot always count on the scrupulous cooperation of their respondents: It’s human nature to get distracted when faced with a form. So how can researchers sort through what may be unreliable data to identify statistically significant answers to their questions? That’s where <a href="https://polisci.mit.edu/people/sean-shiyao-liu">Shiyao "Sean" Liu</a> comes in.</p>
1576
1577 <p>“I have designed a tool for reducing measurement errors when respondents don’t pay serious attention to online questions,” says Liu, a sixth-year PhD candidate in political science. Through statistical methods he has devised, Liu can detect and eliminate random-seeming answers that make for a noisy dataset. “With cleaner data, it’s easier to discover patterns and generate meaningful results.”</p>
1578
1579 <p>Liu’s work on this computational tool earned the Best Graduate Student Poster Award from the Society of Political Methodology in 2019. It is one thrust of his dissertation research, which focuses on optimizing social science survey methods and data analysis.</p>
1580
1581 <p><strong>Fast start</strong></p>
1582
1583 <p>After arriving at MIT in 2015 from Peking University with undergraduate degrees both in statistics and in philosophy, politics, and economics, Liu found himself in demand. During his first summer in Cambridge, Massachusetts, he was engaged by Lily Tsai, faculty director of the MIT Governance Lab, as both a field researcher and methodologist.</p>
1584
1585 <p>“She was writing about public support for authoritarian regimes, and investigating the role played by retributive justice,” says Liu. “The idea is that when authoritarian leaders punish their lower-level officials for wrongdoing, public support rises for the leaders because people perceive that the regime is pursuing justice.”</p>
1586
1587 <p>Liu visited China twice, and with MIT colleagues and local Chinese collaborators, helped conduct 1,600 face-to-face interviews.</p>
1588
1589 <p>Some political science theories suggest that punishing corruption improves the image of authoritarian leaders because it makes them appear more competent. But with computational help from Liu, the research team learned that authoritarian image-building wasn’t just about making the trains run on time.</p>
1590
1591 <p>“By leveraging new surveying methods, I was able to show that increased support for top leadership also flowed from people’s belief that leaders were behaving in a moral way — that they knew the difference between right and wrong,” he says.</p>
1592
1593 <p>Liu, who co-authored a forthcoming paper with Tsai on the popularity of anti-corruption punishment, believes this research provides a useful prism for examining governments today.</p>
1594
1595 <p>“When the Soviet Union collapsed in 1989, people thought democracy would be the only surviving political system in the world, but it wasn’t the case,” he says. “Regimes emerged that frequently used anti-corruption campaigns as a way of building up their popularity among the people.” Think Duterte in the Philippines and Bolsonaro in Brazil, says Liu. People are eager to prop up even the most non-democratic dictators, if they perceive them as pursuing justice.</p>
1596
1597 <p><strong>Statistics for social sciences</strong></p>
1598
1599 <p>Liu was born and grew up in Shanghai, the son of a businessman and a warehouse worker. He quickly discovered an affinity for history and political science, with a special fascination for the workings of different political systems and ideologies.</p>
1600
1601 <p>After Liu’s stellar results in China’s high-pressure college entrance exam (he scored at the very top of Shanghai’s cohort of 30,000 high schoolers), he received a full scholarship to Peking University. There he plunged into graduate-level courses on the social sciences and statistics. One of his undergraduate theses looked at whether mass shootings in different congressional districts in the United States had the effect of reducing votes for the representatives from those districts. He collected and analyzed 10 years of historical data, geocoding the sites of mass shooting events.</p>
1602
1603 <p>By the end of college, says Liu, “I realized I wanted a career where I could basically use statistics to solve problems about society, in either politics or economics.” After serving in business consulting and research positions, he felt certain that his future lay in academics, and that a graduate degree in political science, at MIT in particular, was the right path.</p>
1604
1605 <p>Liu’s decision to study abroad coincided with a more open time in China. “My generation saw the country playing more according to international rules,” he says. “Demand was building from the government and Chinese business to give Chinese students international experience so we could help them communicate better with the rest of the world, to minimize confusion and misunderstanding.” This, says Liu, helped a large cohort of Chinese students to transform themselves.</p>
1606
1607 <p><strong>Cleaning up messy data</strong></p>
1608
1609 <p>At MIT, Liu was drawn to the <a href="https://polisci.mit.edu/research/projects/political-methodology-lab-pml">Political Methodology Lab</a> and the work of its director, Associate Professor Teppei Yamamoto. Under Yamamoto’s mentorship, Liu began to zero in on ways he might advance quantitative methods for political science research.</p>
1610
1611 <p>Liu turned toward the problem of clearing up statistical noise in social science studies: “I was doing online surveys myself, and realized that some people were spending 30 seconds, and others, 30 minutes,” he recalls. Liu came to adopt terms for these distinct types of respondents: “fast-forwarders” are those who click through with little thought about the content, and the “wanderers” are those who take a longer time to answer questions because they are simultaneously looking at their mail or meandering through YouTube. “Both the fast-forwarders and wanderers don’t pay attention, and their answers lead to great randomness in the dataset.”</p>
1612
1613 <p>Liu embarked upon developing a statistical program that can identify both types of respondents in a dataset. “It’s impossible to say precisely that one person is a wanderer and another is a fast-forwarder,” says Liu. “But my program uses probability to show that particular respondents are likely to be one of these types, and to kick them out of the dataset.”</p>
1614
1615 <p>Liu believes this tool could prove useful to many social scientists whose research deploys online surveys. “We could save researchers who are struggling to find a pattern in the noise of their data,” he says.</p>
1616
1617 <p><strong>Forward through the pandemic</strong></p>
1618
1619 <p>Since departing Cambridge in March due to the Covid-19 pandemic, Liu has been preparing drafts of his working papers for publication, while residing in his family home in Shanghai. His physical absence from MIT has posed a range of challenges: “I miss face-to-face check-ins with my advisors and colleagues, which inspired me, gave me new ideas about research, and created a sense of community,” he says. He also pines for the regular dinners, karaoke parties, and hikes with friends. The tense state of affairs between the United States and China, restricting travel, is “making Chinese students' lives difficult, including mine,” says Liu.</p>
1620
1621 <p>But there are consolations to being home, including ample support from friends and family, along with home-cooked meals. “The pandemic is an obstacle for everyone, and we’re all trying to overcome it in our own ways,” he says. “For me, it means pushing forward on my dissertation, and exploring faculty and postdoc positions all around the world.”</p>
1622 </content:encoded>
1623 <media:content url="https://news.mit.edu/sites/default/files/styles/news_article__cover_image__original/public/images/202010/Sean-Liu.jpg?itok=nxqAPqQy" medium="image" type="image/jpeg" width="390" height="260">
1624 <media:description type="plain">Through statistical methods he has devised, political science doctoral candidate Shiyao "Sean" Liu can detect and eliminate random-seeming answers that make for a noisy data set.</media:description>
1625 <media:credit>Photo: Cindy Zhao</media:credit>
1626 </media:content>
1627 </item>
1628 <item>
1629 <title>Antarctic sea ice may not cap carbon emissions as much as previously thought</title>
1630 <link>https://news.mit.edu/2020/antarctic-ice-carbon-emissions-1001</link>
1631 <description>Study suggests sea ice blocks the flow of carbon both into and out of the ocean, in roughly equal measure.</description>
1632 <pubDate>Thu, 01 Oct 2020 00:00:00 -0400</pubDate>
1633 <guid isPermaLink="true">https://news.mit.edu/2020/antarctic-ice-carbon-emissions-1001</guid>
1634 <dc:creator>Jennifer Chu | MIT News Office</dc:creator>
1635 <content:encoded><p>The Southern Ocean surrounding Antarctica is a region where many of the world’s carbon-rich deep waters can rise back up to the surface. Scientists have thought that the vast swaths of sea ice around Antarctica can act as a lid for upwelling carbon, preventing the gas from breaking through the ocean’s surface and returning to the atmosphere.</p>
1636
1637 <p>However, researchers at MIT have now identified a counteracting effect that suggests Antarctic sea ice may not be as powerful a control on the global carbon cycle as scientists had suspected.</p>
1638
1639 <p>In a study published in the August issue of the journal <em>Global Biogeochemical Cycles</em>, the team has found that indeed, sea ice in the Southern Ocean can act as a physical barrier for upwelling carbon. But it can also act as a shade, blocking sunlight from reaching the surface ocean. Sunlight is essential for phytosynthesis, the process by which phytoplankton and other ocean microbes take up carbon from the atmosphere to grow.</p>
1640
1641 <p>The researchers found that when sea ice blocks sunlight, biological activity — and the amount of carbon that microbes can sequester from the atmosphere — decreases significantly. And surprisingly, this shading effect is almost equal and opposite to that of sea ice’s capping effect. Taken together, both effects essentially cancel each other out.&nbsp;</p>
1642
1643 <p>“In terms of future climate change, the expected loss of sea ice around Antarctica may therefore not increase the carbon concentration in the atmosphere,” says lead author Mukund Gupta, who carried out the research as a graduate student in MIT’s Department of Earth, Atmospheric and Planetary Sciences (EAPS).</p>
1644
1645 <p>He emphasizes that sea ice does have other effects on the global climate, foremost through its albedo, or ability to reflect solar radiation.</p>
1646
1647 <p>“When the Earth warms up, it loses sea ice and absorbs more of this solar radiation, so in that sense, the loss of sea ice can accelerate climate change,” Gupta says. “What we can say here is, sea ice changes may not have such a strong effect on carbon outgassing around Antarctica through this capping and shading effect.”</p>
1648
1649 <p>Gupta’s coauthors are EAPS Professor Michael “Mick” Follows, and EAPS research scientist Jonathan Lauderdale.</p>
1650
1651 <p><strong>The role of ice</strong></p>
1652
1653 <p>Each winter, wide swaths of the Southern Ocean freeze over, forming vast sheets of sea ice that extend out from Antarctica for millions of square miles. The role of Antarctic sea ice in regulating the climate and the carbon cycle has been much debated, though the prevailing theory has been that sea ice can act as a lid to keep carbon in the ocean from escaping to the atmosphere.</p>
1654
1655 <p>“This theory is mostly thought of in the context of ice ages, when the Earth was much colder and the atmospheric carbon was lower,” Gupta says. “One of the theories explaining this low carbon concentration argues that because it was colder, a thick sea ice cover extended further into the ocean, blocking carbon exchanges with the atmosphere and effectively trapping it in the deep ocean.”</p>
1656
1657 <p>Gupta and his colleagues wondered whether an effect other than capping may also be in play. In general, the researchers have sought to understand how various features and processes in the ocean interact with ocean biology such as phytoplankton. They assumed that there might be less biological activity as a result of sea ice blocking microbes’ vital sunlight — but how strong would this shading effect be?</p>
1658
1659 <p><strong>Equal and opposite</strong></p>
1660
1661 <p>To answer that question, the researchers used the MITgcm, a global circulation model that simulates the many physical, chemical, and biological processes involved in the circulation of the atmosphere and ocean. With MITgcm, they simulated a vertical slice of the ocean spanning 3,000 kilometers wide and about 4,000 meters deep, and with conditions similar to today’s Southern Ocean. They then ran the model multiple times, each time with a different concentration of sea ice.</p>
1662
1663 <p>“At 100 percent concentration, there are no leaks in the ice, and it’s really compacted together, versus very low concentrations representing loose and sparse ice floes moving around,” Gupta explains.</p>
1664
1665 <p>They set each simulation to one of three scenarios: one where only the capping effect is active, and sea ice is only influencing the carbon cycle by preventing carbon from leaking back out to the atmosphere; another where only the shading effect is active, and sea ice is only blocking sunlight from penetrating the ocean; and the last in which both capping and shading effects are in play.</p>
1666
1667 <p>For every simulation, the researchers observed how the conditions they set affected the overall carbon flux, or amount of carbon that escaped from the ocean to the atmosphere.</p>
1668
1669 <p>They found that capping and shading had opposite effects on the carbon cycle, reducing the amount of carbon to the atmosphere in the former case and increasing it in the latter, by equal amounts. In the scenarios where both effects were considered, one canceled the other out almost entirely, across a wide range of sea ice concentrations, leading to no significant change in the carbon flux. Only when sea ice was at its highest concentration did capping have the edge, with a decrease in carbon escaping to the atmosphere.</p>
1670
1671 <p>The results suggest that Antarctic sea ice may effectively trap carbon in the ocean, but only when that ice cover is very expansive and thick. Otherwise, it seems that sea ice’s shading effect on the underlying organisms may counteract its capping effect.</p>
1672
1673 <p>“If one just considered the physics and the pure capping, or carbon barrier idea, that would be an incomplete way of thinking about it,” Gupta says. “This shows that we need to understand more of the biology under sea ice and how it underlies this effect.”</p>
1674
1675 <p>This research was supported in part by the U.S. National Science Foundation.</p>
1676 </content:encoded>
1677 <media:content url="https://news.mit.edu/sites/default/files/styles/news_article__cover_image__original/public/images/202009/MIT-Ice-Shade-01.jpg?itok=NzI0GQnU" medium="image" type="image/jpeg" width="390" height="260">
1678 <media:description type="plain">The prevailing theory has been that sea ice can act as a lid to keep carbon in the ocean from escaping back to the atmosphere. However, researchers at MIT have now identified a counteracting effect that suggests Antarctic sea ice may not be as powerful a control on the global carbon cycle as scientists had suspected.</media:description>
1679 </media:content>
1680 </item>
1681 <item>
1682 <title>Validating the physics behind the new MIT-designed fusion experiment</title>
1683 <link>https://news.mit.edu/2020/physics-fusion-studies-0929</link>
1684 <description>Seven studies describe progress thus far and challenges ahead for a revolutionary zero-emissions power source.</description>
1685 <pubDate>Tue, 29 Sep 2020 09:00:00 -0400</pubDate>
1686 <guid isPermaLink="true">https://news.mit.edu/2020/physics-fusion-studies-0929</guid>
1687 <dc:creator>David L. Chandler | MIT News Office</dc:creator>
1688 <content:encoded><p>Two and a half years ago, MIT entered into a research agreement with startup company Commonwealth Fusion Systems to develop a next-generation fusion research experiment, called SPARC, as a precursor to a practical, emissions-free power plant.</p>
1689
1690 <p>Now, after many months of intensive research and engineering work, the researchers charged with defining and refining the physics behind the ambitious tokamak design have published a series of papers summarizing the progress they have made and outlining the key research questions SPARC will enable.</p>
1691
1692 <p>Overall, says Martin Greenwald, deputy director of MIT’s Plasma Science and Fusion Center and one of the project’s lead scientists, the work is progressing smoothly and on track. This series of papers provides a high level of confidence in the plasma physics and the performance predictions for SPARC, he says. No unexpected impediments or surprises have shown up, and the remaining challenges appear to be manageable. This sets a solid basis for the device’s operation once constructed, according to Greenwald.</p>
1693
1694 <p>Greenwald wrote the introduction for a set of seven research papers authored by 47 researchers from 12 institutions and published today in a special issue of the <em>Journal of Plasma Physics</em>. Together, the papers outline the theoretical and empirical physics basis for the new fusion system, which the consortium expects to start building next year.</p>
1695
1696 <p>SPARC is planned to be the first experimental device ever to achieve a “burning plasma” — that is, a self-sustaining fusion reaction in which different isotopes of the element hydrogen fuse together to form helium, without the need for any further input of energy. Studying the behavior of this burning plasma — something never before seen on Earth in a controlled fashion — is seen as crucial information for developing the next step, a working prototype of a practical, power-generating power plant.</p>
1697
1698 <p>Such fusion power plants might significantly reduce greenhouse gas emissions from the power-generation sector, one of the major sources of these emissions globally. The MIT and CFS project is one of the largest privately funded research and development projects ever undertaken in the fusion field.</p>
1699
1700 <p>"The MIT group is pursuing a very compelling approach to fusion energy." says Chris Hegna, a professor of engineering physics at the University of Wisconsin at Madison, who was not connected to this work. "They realized the emergence of high-temperature superconducting technology enables a high magnetic field approach to producing net energy gain from a magnetic confinement system. This work is a potential game-changer for the international fusion program."</p>
1701
1702 <p>The SPARC design, though about twice the size as MIT’s now-retired <a href="https://news.mit.edu/2016/alcator-c-mod-tokamak-nuclear-fusion-world-record-1014">Alcator C-Mod</a> experiment and similar to several other research fusion machines currently in operation, would be far more powerful, achieving fusion performance comparable to that expected in the much larger ITER tokamak being built in France by an international consortium. The high power in a small size is made possible by advances in superconducting magnets that allow for a much stronger magnetic field to confine the hot plasma.</p>
1703
1704 <p>The SPARC project was launched in early 2018, and work on its first stage, the development of the superconducting magnets that would allow smaller fusion systems to be built, has been proceeding apace. The new set of papers represents the first time that the underlying physics basis for the SPARC machine has been outlined in detail in peer-reviewed publications. The seven papers explore the specific areas of the physics that had to be further refined, and that still require ongoing research to pin down the final elements of the machine design and the operating procedures and tests that will be involved as work progresses toward the power plant.</p>
1705
1706 <p>The papers also describe the use of calculations and simulation tools for the design of SPARC, which have been tested against many experiments around the world. The authors used cutting-edge simulations, run on powerful supercomputers, that have been developed to aid the design of ITER. The large multi-institutional team of researchers represented in the new set of papers aimed to bring the best consensus tools to the SPARC machine design to increase confidence it will achieve its mission.</p>
1707
1708 <p>The analysis done so far shows that the planned fusion energy output of the SPARC tokamak should be able to meet the design specifications with a comfortable margin to spare. It is designed to achieve a Q factor — a key parameter denoting the efficiency of a fusion plasma — of at least 2, essentially meaning that twice as much fusion energy is produced as the amount of energy pumped in to generate the reaction. That would be the first time a fusion plasma of any kind has produced more energy than it consumed.</p>
1709
1710 <p>The calculations at this point show that SPARC could actually achieve a Q ratio of 10 or more, according to the new papers. While Greenwald cautions that the team wants to be careful not to overpromise, and much work remains, the results so far indicate that the project will at least achieve its goals, and specifically will meet its key objective of producing a burning plasma, wherein the self-heating dominates the energy balance.</p>
1711
1712 <p>Limitations imposed by the Covid-19 pandemic slowed progress a bit, but not much, he says, and the researchers are back in the labs under new operating guidelines.</p>
1713
1714 <p>Overall, “we’re still aiming for a start of construction in roughly June of ’21,” Greenwald says. “The physics effort is well-integrated with the engineering design. What we’re trying to do is put the project on the firmest possible physics basis, so that we’re confident about how it’s going to perform, and then to provide guidance and answer questions for the engineering design as it proceeds.”</p>
1715
1716 <p>Many of the fine details are still being worked out on the machine design, covering the best ways of getting energy and fuel into the device, getting the power out, dealing with any sudden thermal or power transients, and how and where to measure key parameters in order to monitor the machine’s operation.</p>
1717
1718 <p>So far, there have been only minor changes to the overall design. The diameter of the tokamak has been increased by about 12 percent, but little else has changed, Greenwald says. “There’s always the question of a little more of this, a little less of that, and there’s lots of things that weigh into that, engineering issues, mechanical stresses, thermal stresses, and there’s also the physics — how do you affect the performance of the machine?”</p>
1719
1720 <p>The publication of this special issue of the journal, he says, “represents a summary, a snapshot of the physics basis as it stands today.” Though members of the team have discussed many aspects of it at physics meetings, “this is our first opportunity to tell our story, get it reviewed, get the stamp of approval, and put it out into the community.”</p>
1721
1722 <p>Greenwald says there is still much to be learned about the physics of burning plasmas, and once this machine is up and running, key information can be gained that will help pave the way to commercial, power-producing fusion devices, whose fuel — the hydrogen isotopes deuterium and tritium — can be made available in virtually limitless supplies.</p>
1723
1724 <p>The details of the burning plasma “are really novel and important,” he says. “The big mountain we have to get over is to understand this self-heated state of a plasma.”</p>
1725
1726 <p>"The analysis presented in these papers will provide the world-wide fusion community with an opportunity to better understand the physics basis of the SPARC device and gauge for itself the remaining challenges that need to be resolved," says George Tynan, professor of mechanical and aerospace engineering at the University of California at San Diego, who was not connected to this work. "Their publication marks an important milestone on the road to the study of burning plasmas and the first demonstration of net energy production from controlled fusion, and I applaud the authors for putting this work out for all to see."</p>
1727
1728 <p>Overall, Greenwald says, the work that has gone into the analysis presented in this package of papers “helps to validate our confidence that we will achieve the mission. We haven’t run into anything where we say, ‘oh, this is predicting that we won’t get to where we want.” In short, he says, “one of the conclusions is that things are still looking on-track. We believe it’s going to work.”</p>
1729 </content:encoded>
1730 <media:content url="https://news.mit.edu/sites/default/files/styles/news_article__cover_image__original/public/images/202009/MIT-Fusion-Progress-01-PRESS.jpg?itok=Lq2CRpL3" medium="image" type="image/jpeg" width="390" height="260">
1731 <media:description type="plain">This image shows a cutaway rendering of SPARC, a compact, high-field, DT burning tokamak, currently under design by a team from the Massachusetts Institute of Technology and Commonwealth Fusion Systems. Its mission is to create and confine a plasma that produces net fusion energy.</media:description>
1732 <media:credit>Image: CFS/MIT-PSFC — CAD Rendering by T. Henderson</media:credit>
1733 </media:content>
1734 </item>
1735 <item>
1736 <title>Less scatterbrained scatterplots</title>
1737 <link>https://news.mit.edu/2020/less-scatterbrained-scatterplots-1007</link>
1738 <description>Large datasets are difficult to depict as scatterplots — but that may change with a new CSAIL project for creating interactive visualizations.</description>
1739 <pubDate>Wed, 07 Oct 2020 14:30:00 -0400</pubDate>
1740 <guid isPermaLink="true">https://news.mit.edu/2020/less-scatterbrained-scatterplots-1007</guid>
1741 <dc:creator>Adam Conner-Simons | MIT CSAIL</dc:creator>
1742 <content:encoded><p>Scatterplots. You might not know them by name, but if you spend more than 10 minutes online you’ll find them everywhere. They’re popular in <a href="https://www.nytimes.com/2020/01/02/learning/whats-going-on-in-this-graph-internet-privacy-policies.html"><u>news articles</u></a>, in <a href="https://towardsdatascience.com/everything-you-need-to-know-about-scatter-plots-for-data-visualisation-924144c0bc5"><u>the data science community</u></a>, and, perhaps most crucially, for <a href="https://www.mentalfloss.com/article/64631/9-scatterplots-good-laugh"><u>internet memes about the digestive quality of pancakes</u></a>.&nbsp;</p>
1743
1744 <p>By depicting data as a mass of points across two axes, scatterplots are effective in visualizing trends, correlations, and anomalies. But using them for large datasets often leads to overlapping dots that make them <a href="https://www.data-to-viz.com/caveat/overplotting.html"><u>more or less unreadable</u></a>.</p>
1745
1746 <p>Researchers from MIT’s <a href="https://www.csail.mit.edu/">Computer Science and Artificial Intelligence Laboratory</a> (CSAIL) say they’ve solved this with a new open-source system that makes it possible to create interactive scatterplots based on large-scale datasets that have upwards of billions of distinct data points.&nbsp;</p>
1747 <p>Called “Kyrix-S,” the system has an interface that allows users to pan, zoom, and jump around a scatterplot as if they were looking at directions on Google Maps. Whereas other systems developed for large datasets often focus on very specific applications, Kyrix-S is generalizable enough to work for a wide range of visualization styles, including heat maps, pie charts, and radar-style graphics. (The team showed that the system allows users to create visualizations with 800 percent less code compared to a similar state-of-the-art authoring system.)</p>
1748
1749 <p>Users can produce a scatterplot by just writing a few dozen lines of JSON, a human-readable text format.</p>
1750
1751 <p>Lead developer Wenbo Tao, a PhD student at MIT CSAIL, gives the&nbsp;example of <a href="https://www.nytimes.com/interactive/2014/03/15/business/higher-income-longer-lives.html"><u>a static New York Times scatterplot</u></a>&nbsp;that he says would improve by being made interactive via a system like Kyrix-S.</p>
1752 <p>“In these scatterplots, you are able to see overall trends and outliers, but the overplotting and the static nature of the plot limit the user's ability to interact with the chart,” says Tao.&nbsp;</p>
1753
1754 <p>In contrast, Kyrix-S can produce a version (below) that puts data in several zoom levels, enabling interaction with each county. To avoid overplotting, Kyrix-S’ scatterplot also shows only the most important examples, like the most populous counties.</p>
1755
1756 <p>Kyrix-S is currently being used by <a href="https://www.vldb.org/pvldb/vol12/p1954-rezig.pdf"><u>Data Civilizer 2.0</u></a>, a data integration platform developed at MIT. An earlier version was also employed to help Massachusetts General Hospital analyze a massive brain activity dataset (EEG) that clocks in at 30 terabytes — the equivalent of more than 50,000 hours of digital music. (The goal of that study was to train a model that predicts seizures, given a series of 2-second EEG segments.)</p>
1757
1758 <p>Moving forward, the researchers will be adapting Kyrix-S to work as part of a graphical user interface. They also plan to add functionality so that the system can handle data that are&nbsp;being continuously updated.</p>
1759
1760 <p>Tao wrote a paper about Kyrix-S alongside MIT Adjunct Professor Mike Stonebraker, researchers Xinli Hou and Adam Sah, Leilani Battle SM '13, PhD '17, and Professor Remco Chang of Tufts University. It will be presented virtually at IEEE’s VIS data visualization conference Oct. 25.</p>
1761 </content:encoded>
1762 <media:content url="https://news.mit.edu/sites/default/files/styles/news_article__cover_image__original/public/images/202009/Kyrix-S-scatterplot.png?itok=a6peC4Jh" medium="image" type="image/jpeg" width="390" height="260">
1763 <media:description type="plain">The Kyrix-S system developed at MIT CSAIL makes it easier to automatically create visualizations that users can interact with dynamically.</media:description>
1764 <media:credit>Image courtesy of the researchers.</media:credit>
1765 </media:content>
1766 </item>
1767 <item>
1768 <title>3 Questions: Anat Biletzki on the Human Rights and Technology Fellowship Program</title>
1769 <link>https://news.mit.edu/2020/3-questions-anat-biletzki-human-rights-technology-fellowship-program-1013</link>
1770 <description>Designed for students, the program explores on a grand scale how technology can aid or hinder human rights.</description>
1771 <pubDate>Tue, 13 Oct 2020 12:40:00 -0400</pubDate>
1772 <guid isPermaLink="true">https://news.mit.edu/2020/3-questions-anat-biletzki-human-rights-technology-fellowship-program-1013</guid>
1773 <dc:creator>Center for International Studies</dc:creator>
1774 <content:encoded><p><em>MIT Center for International Studies (CIS) research affiliate Anat Biletzki<strong> </strong>is the&nbsp;Albert Schweitzer Professor of Philosophy at Quinnipiac University. From 2001 to 2006, she served as chair of&nbsp;B’Tselem ― the Israeli Information Center for Human Rights in the Occupied Territories, and in 2015 was honored as a nominee of the 1,000 Women for the Nobel Peace Prize initiative. Her most recent book is&nbsp;“Philosophy of Human Rights: A Systematic Introduction”&nbsp;(Routledge, 2019).</em></p>
1775
1776 <p><em>Biletzki is a founding co-director of the CIS&nbsp;<a href="https://cis.mit.edu/programs/human-rights-technology">Human Rights and Technology Fellowship Program</a>. The program offers research fellowships to MIT students with the intent of producing new knowledge about the relationship of human rights and technology. The fellowship program is open to both undergraduate and graduate students and invites proposals for its 2020-21 cohort of fellows through Oct. 26. She speaks here on the fellowship program.</em></p>
1777
1778 <p><strong>Q: </strong>Can you tell us about the work that the Human Rights and Technology Program does and what it offers to MIT students?&nbsp;</p>
1779
1780 <p><strong>A: </strong>The program is invested in teaching human rights, but teaching in a very deep sense of the word “teaching.” It is not about classes. It is about actually getting students to engage with human rights. The “work” is getting students to think of their own projects, which can be completed in a semester or a year, that link human rights with technology.</p>
1781
1782 <p>When thinking of a human rights program at CIS, John Tirman [executive director and a principal research scientist at CIS], Richard Samuels [director of CIS and Ford International Professor of Political Science], and I determined that the program should focus on MIT’s strength in technology. Our vision, then, was to add the human rights component and thus explore on a grand scale how technology either aids or hinders human rights.</p>
1783
1784 <p>Tirman and I co-direct the program, and each fall we send out a request for proposals to MIT students to apply for the annual fellowship. The program started in 2018 with its first cohort of students.&nbsp;Our 2019-20 awardees involved seven projects and 10 students. The projects are amazingly diverse and come from both undergraduate and graduate students across the Institute.&nbsp;</p>
1785
1786 <p>One student was working on the use of technology in monitoring migrants at the U.S.-Mexico border.&nbsp;Another student was working in Micronesia, looking at Facebook Groups and the issue of labor exploitation of migrant workers. One evolving group project began with looking at how social media promotes activism for workers' rights. And we have other students working on questions of indigenous knowledge, indigenous culture and indigenous groups, and how access to their own resources is supported or harmed by technology.</p>
1787
1788 <p><strong>Q: </strong>What sort of adaptations has the program and its current fellows made due to the Covid-19 pandemic?</p>
1789
1790 <p><strong>A: </strong>On a very mundane, technical level, the part of the program that is predicated on meetings among fellows and personal reports and encounters all went online. There is nothing exclusive here <em>―</em> merely the easy success of Zoom meetings with all our fellows who continued to deliberate together about their ongoing projects.</p>
1791
1792 <p>More significant was the fact that a few projects involved travel abroad (to Vietnam, Mexico, India) while others demanded face-to-face contact with interviewees who were not accessible via internet, or internet contact with interviewees whose technological resources became scarce. These projects underwent certain transformations <em>―</em> some couldn’t travel, some were “stuck” in their alternative space, some meetings couldn’t take place <em>―</em> and the fellows managed to admirably convert their methodology and even their project-goals to something doable in Covid-19 times.</p>
1793
1794 <p>Clearly, the 2019-20 cohort had to make their changes and “deliver the goods” on an ad hoc basis. Looking forward to 2020-21 and beyond, and as strange as it may sound, I see great potential for the program in our current Covid-19 predicament. Projects that are articulated now cannot ignore the limitations and demands that this new context will require of students, especially when thinking about their objectives and tools, since both are now impacted in the matter of technology. Just as relevant, to my mind, are the effects of Covid-19 on questions of human rights. A whole new slew of human rights abuses, now so much with us, are precisely matters of human rights and technology. In working with new fellows, we will need to adapt to our new physical, geographical, and conceptual frameworks.</p>
1795
1796 <p><strong>Q: </strong>How do you hope the program will continue to grow in the future?&nbsp;</p>
1797
1798 <p><strong>A: </strong>Of course, we’d like it to grow in the very mundane sense of having more people. For example, if we could have 20 projects a year, we’d have a more vibrant program. What we’re noticing now is how the projects are enriching one another and how the group as a whole is working together. If it’s a bigger group with more projects, it widens the horizons of what we can do.</p>
1799
1800 <p>On a less concrete level, I want the program to be asking deeper questions about whether technology is good or bad for human rights, and grappling with how we deal with the encroachment of technology.&nbsp;In that sense, I see this program as being a great contribution in the way&nbsp;human rights is perceived and done all&nbsp;around the world, not just at MIT.</p>
1801
1802 <p>What we are seeing, and what gives me great hope that the program will thrive, is that the students involved are getting more and more excited. I have been incredibly amazed at the speed and depth with which they do their work. Within two weeks of our first meeting, they are human rights “experts.” They read, they investigate, they absorb everything they hear.</p>
1803 </content:encoded>
1804 <media:content url="https://news.mit.edu/sites/default/files/styles/news_article__cover_image__original/public/images/202010/anat-biletzki-mit-a1.jpg?itok=KXzAIca3" medium="image" type="image/jpeg" width="390" height="260">
1805 <media:description type="plain">"What we are seeing, and what gives me great hope that the program will thrive, is that the students involved are getting more and more excited," says Anat Biletzki. I have been incredibly amazed at the speed and depth with which they do their work. Within two weeks of our first meeting, they are human rights ‘experts.’ They read, they investigate, they absorb everything they hear."</media:description>
1806 <media:credit>Photo courtesy Anat Biletzki.</media:credit>
1807 </media:content>
1808 </item>
1809 <item>
1810 <title>Scientists uncover new clues about Parkinson’s disease</title>
1811 <link>https://news.mit.edu/2020/scientists-uncover-new-clues-about-parkinsons-disease-1013</link>
1812 <description>Tool developed at MIT simultaneously measures chemical and electrical brain signals, revealing unexpectedly complex relationship between brain signals.</description>
1813 <pubDate>Tue, 13 Oct 2020 11:50:00 -0400</pubDate>
1814 <guid isPermaLink="true">https://news.mit.edu/2020/scientists-uncover-new-clues-about-parkinsons-disease-1013</guid>
1815 <dc:creator>Jennifer Michalowski | McGovern Institute for Brain Research</dc:creator>
1816 <content:encoded><p>As the brain processes information, electrical charges zip through its circuits and neurotransmitters pass molecular messages from cell to cell. Both forms of communication are vital, but because they are usually studied separately, little is known about how they work together to control our actions, regulate mood, and perform the other functions of a healthy brain.</p>
1817
1818 <p>Neuroscientists in <a href="https://mcgovern.mit.edu/profile/ann-graybiel/">Ann Graybiel</a>’s laboratory at MIT’s McGovern Institute for Brain Research are taking a closer look at the relationship between these electrical and chemical signals. “Considering electrical signals side by side with chemical signals is really important to understand how the brain works,” says <a href="http://web.mit.edu/schwerdt/www/index.html">Helen Schwerdt</a>, a postdoc in Graybiel’s lab. Understanding that relationship is also crucial for developing better ways to diagnose and treat nervous system disorders and mental illness, she says, noting that the drugs used to treat these conditions typically aim to modulate the brain’s chemical signaling, yet studies of brain activity are more likely to focus on electrical signals, which are easier to measure.</p>
1819
1820 <p>Schwerdt and colleagues in Graybiel’s lab have developed new tools so that chemical and electrical signals can, for the first time, be measured simultaneously in the brains of primates. In a <a href="https://advances.sciencemag.org/content/6/39/eabb9226/tab-article-info">study</a> published Sept. 25 in <em>Science Advances</em>, they used those tools to reveal an unexpectedly complex relationship between two types of signals that are disrupted in patients with <a href="https://mcgovern.mit.edu/research-areas/parkinsons-disease/">Parkinson’s disease</a> — dopamine signaling and coordinated waves of electrical activity known as beta-band oscillations.</p>
1821
1822 <p><strong>Complicated relationship</strong></p>
1823
1824 <p>Graybiel’s team focused its attention on beta-band activity and dopamine signaling because studies of patients with Parkinson’s disease had suggested a straightforward inverse relationship between the two. The tremors, slowness of movement, and other symptoms associated with the disease develop and progress as the brain’s production of the neurotransmitter dopamine declines, and, at the same time, beta-band oscillations surge to abnormal levels. Beta-band oscillations are normally observed in parts of the brain that control movement when a person is paying attention or planning to move. It’s not clear what they do or why they are disrupted in patients with Parkinson’s disease. But because patients’ symptoms tend to be worst when beta activity is high — and because beta activity can be measured in real time with sensors placed on the scalp or with a deep-brain stimulation device that has been implanted for treatment, researchers have been hopeful that it might be useful for monitoring the disease’s progression and patients’ response to treatment. In fact, clinical trials are already underway to explore the effectiveness of modulating deep-brain stimulation treatment based on beta activity.</p>
1825
1826 <p>When Schwerdt and colleagues examined these two types of signals in the brains of rhesus macaques, they discovered that the relationship between beta activity and dopamine is more complicated than previously thought. Their new tools allowed them to simultaneously monitor both signals with extraordinary precision, targeting specific parts of the striatum — a region deep within the brain involved in controlling movement, where dopamine is particularly abundant — and taking measurements on the millisecond time scale to capture neurons’ rapid-fire communications.</p>
1827
1828 <p>They took these measurements as the monkeys performed a simple task, directing their gaze in a particular direction in anticipation of a reward. This allowed the researchers to track chemical and electrical signaling during the active, motivated movement of the animals’ eyes. They found that beta activity did increase as dopamine signaling declined — but only in certain parts of the striatum and during certain tasks. The reward value of a task, an animal’s past experiences,&nbsp;and the particular movement the animal performed all impacted the relationship between the two types of signals.</p>
1829
1830 <p>“What we expected is there in the overall view, but if we just look at a different level of resolution, all of a sudden the rules don’t hold,” says Graybiel, who is an MIT Institute Professor. “It doesn’t destroy the likelihood that one would want to have a treatment related to this presumed opposite relationship, but it does say there’s something more here that we haven’t known about.”</p>
1831
1832 <p>The researchers say it’s important to investigate this more nuanced relationship between dopamine signaling and beta activity, and that understanding it more deeply might lead to better treatments for patients with Parkinson’s disease and related disorders. While they plan to continue to examine how the two types of signals relate to one another across different parts of the brain and under different behavioral conditions, they hope that other teams will also take advantage of the tools they have developed. “As these methods in neuroscience become more and more precise and dazzling in their power, we’re bound to discover new things,” says Graybiel.</p>
1833
1834 <p>Joining Graybiel and Schwerdt on the study are Michael Cima, the David H. Koch Professor of Engineering in the Department of Materials Science and Engineering and a member of MIT’s Koch Institute for Integrative Cancer Research; Robert Langer, the David H. Koch Institute Professor and a member of the Koch Institute; Robert Desimone, director of the McGovern Institute; MIT research scientists Ken Amemori, Dan Gibson, Narcisse Bichot, Satoko Amemori; and Graybiel lab technicians Lauren Stanwicks and Tomoko Yoshida.</p>
1835
1836 <p>This study was supported by the National Institute of Biomedical Imaging and Bioengineering, the National Institute of Neurological Disorders and Stroke, the U.S. Army Research Office, the Saks Kavanaugh Foundation, the National Science Foundation, Kristin R. Pressman and Jessica J. Pourian ’13 Fund, and Robert Buxton.</p>
1837 </content:encoded>
1838 <media:content url="https://news.mit.edu/sites/default/files/styles/news_article__cover_image__original/public/images/202009/helen-schwerdt.jpg?itok=wgT33nI_" medium="image" type="image/jpeg" width="390" height="260">
1839 <media:description type="plain">Helen Schwerdt, a postdoc in Ann Graybiel's lab, builds ultrathin probes that target brain microstructures with pinpoint accuracy. </media:description>
1840 <media:credit>Photo: Michael D. Spencer</media:credit>
1841 </media:content>
1842 </item>
1843 <item>
1844 <title>Anticipating heart failure with machine learning</title>
1845 <link>https://news.mit.edu/2020/anticipating-heart-failure-machine-learning-1001</link>
1846 <description>Many health issues are tied to excess fluid in the lungs. A new algorithm can detect the severity by looking at a single X-ray.</description>
1847 <pubDate>Thu, 01 Oct 2020 14:00:00 -0400</pubDate>
1848 <guid isPermaLink="true">https://news.mit.edu/2020/anticipating-heart-failure-machine-learning-1001</guid>
1849 <dc:creator>Adam Conner-Simons | MIT CSAIL</dc:creator>
1850 <content:encoded><p>Every year, roughly one out of eight U.S. deaths is caused at least in part <a href="https://wonder.cdc.gov/ucd-icd10.html">by heart failure</a>. One of acute heart failure’s most common warning signs is <a href="https://pubmed.ncbi.nlm.nih.gov/20354029/">excess fluid in the lungs</a>, a condition known as “pulmonary edema.”&nbsp;</p>
1851
1852 <p>A patient’s exact level of excess fluid often dictates the doctor’s course of action, but&nbsp;making such determinations is difficult and requires clinicians to rely on subtle features in X-rays that sometimes lead to inconsistent diagnoses and treatment plans.</p>
1853
1854 <p>To better handle that kind of nuance, a group led by researchers at MIT’s Computer Science and Artificial Intelligence Lab (CSAIL) has developed a machine learning model that can look at an X-ray to quantify how severe the edema is, on a four-level scale ranging from 0 (healthy) to 3 (very, very bad). The system determined the right level more than half of the time, and correctly diagnosed level 3 cases 90 percent of the time.</p>
1855
1856 <p>Working with <a href="https://www.bidmc.org/">Beth Israel Deaconess Medical Center</a> (BIDMC) and <a href="https://www.philips.com/a-w/research/home">Philips</a>, the team plans to integrate the model into BIDMC’s emergency-room workflow this fall.</p>
1857
1858 <p>“This project is meant to augment doctors’ workflow by providing additional information that can be used to inform their diagnoses as well as enable retrospective analyses,” says PhD student Ruizhi Liao, who was the co-lead author of a related paper with fellow PhD student Geeticka Chauhan and MIT professors Polina Golland and Peter Szolovits.&nbsp;</p>
1859
1860 <p>The team says that better edema diagnosis would help doctors manage not only acute heart issues, but other conditions like sepsis and kidney failure that are strongly associated with edema.&nbsp;</p>
1861
1862 <p>As part of a separate journal article, Liao and colleagues also took <a href="https://news.mit.edu/2019/mimic-chest-x-ray-database-0201">an existing public dataset of X-ray images</a> and <a href="https://github.com/RayRuizhiLiao/regex_pulmonary_edema">developed new annotations</a> of severity labels that were agreed upon by a team of four radiologists. Liao’s hope is that these consensus labels can serve as a universal standard to benchmark future machine learning development.</p>
1863
1864 <p>An important aspect of the system is that it was trained not just on more than 300,000 X-ray images, but also on the corresponding text of reports about the X-rays that were written by radiologists. The team was pleasantly surprised that their system found such success using these reports, most of which didn’t have labels explaining the exact severity level of the edema.</p>
1865
1866 <p>“By learning the association between images and their corresponding reports, the method has the potential for a new way of automatic report generation from the detection of image-driven findings,<strong>” </strong>says Tanveer Syeda-Mahmood, a <a href="https://arxiv.org/abs/2007.13831">researcher</a> not involved in the project who serves as chief scientist for IBM’s <a href="https://researcher.watson.ibm.com/researcher/view_group.php?id=4384">Medical Sieve Radiology Grand Challenge</a>. “Of course, further experiments would have to be done for this to be broadly applicable to other findings and their fine-grained descriptors.”</p>
1867
1868 <p>Chauhan’s efforts focused on helping the system make sense of the text of the reports, which could often be as short as a sentence or two. Different radiologists write with varying tones and use a range of terminology, so the researchers had to develop a set of linguistic rules and substitutions to ensure that data could be analyzed consistently across reports. This was in addition to the technical challenge of designing a model that can jointly train the image and text representations in a meaningful manner.</p>
1869
1870 <p>“Our model can turn both images and text into compact numerical abstractions from which an interpretation can be derived,” says Chauhan. “We trained it to minimize the difference between the representations of the X-ray images and the text of the radiology reports, using the reports to improve the image interpretation.”</p>
1871
1872 <p>On top of that, the team’s system was also able to “explain” itself, by showing which parts of the reports and areas of X-ray images correspond to the model prediction. Chauhan is hopeful that future work in this area will provide more detailed lower-level image-text correlations, so that clinicians can build a taxonomy of images, reports, disease labels and relevant correlated regions.&nbsp;</p>
1873
1874 <p>“These correlations will be valuable for improving search through a large database of X-ray images and reports, to make retrospective analysis even more effective,” Chauhan says.</p>
1875
1876 <p>Chauhan, Golland, Liao and Szolovits co-wrote the paper with MIT Assistant Professor Jacob Andreas, Professor William Wells of Brigham and Women’s Hospital, Xin Wang of Philips, and Seth Berkowitz and Steven Horng of BIDMC. The paper will be presented Oct. 5 (virtually) at the International Conference on Medical Image Computing and Computer-Assisted Intervention (MICCAI).&nbsp;</p>
1877
1878 <p>The work was supported in part by the MIT Deshpande Center for Technological Innovation, the MIT Lincoln Lab, the National Institutes of Health, Philips, Takeda, and the Wistron Corporation.</p>
1879 </content:encoded>
1880 <media:content url="https://news.mit.edu/sites/default/files/styles/news_article__cover_image__original/public/images/202009/Pulmonary-Edema-xray.png?itok=oDUWzDcy" medium="image" type="image/jpeg" width="390" height="260">
1881 <media:description type="plain">It can be difficult for clinicians to distinguish between different severity levels of pulmonary edema (excess fluid in the lungs). A new algorithm aims to make that distinction automatically by looking at a single X-ray image.</media:description>
1882 <media:credit>Image: James Heilman/Creative Commons</media:credit>
1883 </media:content>
1884 </item>
1885 <item>
1886 <title>J-PAL North America launches MIT Roybal Center for Translational Research to Improve Health Care for the Aging</title>
1887 <link>https://news.mit.edu/2020/j-pal-north-america-launches-mit-roybal-center-translational-research-1005</link>
1888 <description>Center will work with affiliated researchers to test low-cost, high-impact behavioral interventions to improve health-care delivery and health outcomes for aging adults in the United States.</description>
1889 <pubDate>Mon, 05 Oct 2020 15:30:00 -0400</pubDate>
1890 <guid isPermaLink="true">https://news.mit.edu/2020/j-pal-north-america-launches-mit-roybal-center-translational-research-1005</guid>
1891 <dc:creator>J-PAL North America</dc:creator>
1892 <content:encoded><p>With support from the National Institute on Aging, J-PAL North America, a research center in the MIT Department of Economics, recently launched the <a href="https://www.povertyactionlab.org/initiative/mit-roybal-center-translational-research-improve-health-care-aging">MIT Roybal Center for Translational Research to Improve Health Care for the Aging</a>. The center will support randomized evaluations of low-cost, high-impact behavioral interventions to improve health-care delivery and health outcomes for older adults in the United States.</p>
1893
1894 <p>According to the <a href="https://www.ncoa.org/news/resources-for-reporters/get-the-facts/economic-security-facts/">National Council on Aging</a>, more than 25 million Americans over the age of 60 are economically insecure, living at below 250 percent of the federal poverty line ($2,452 per month for one person). In addition to experiencing high rates of poverty, the aging population interacts with the health-care system more frequently and experiences high rates of chronic disease and other illnesses.</p>
1895
1896 <p>However, these challenges are not insurmountable and can be effectively addressed through the right policies and reforms. Health-care delivery interventions, for example, have the potential to dramatically improve health outcomes and reduce the cost of health care for the aging population, and especially for aging individuals experiencing poverty. Yet, more research is needed to identify effective interventions that can help improve the health of this population.&nbsp;</p>
1897
1898 <p>To fill this knowledge gap, the MIT Roybal Center will support pilot randomized evaluations of low-cost, high-impact behavioral interventions aimed at improving health-care delivery and health outcomes for aging individuals. Past Roybal-funded pilots have examined interventions such as <a href="https://chibe.upenn.edu/project/structuring-incentives-to-attract-and-retain-wellness-program-participants/">novel incentive structures to attract and retain wellness program participants</a>, and <a href="https://academic.oup.com/innovateage/article/2/1/igy007/4962182">behavioral strategies to increase physical activity</a>. J-PAL’s network of affiliated researchers will be able to apply for up to $50,000 of funding through the MIT Roybal Center.</p>
1899
1900 <p>“Over the last seven years, J-PAL North America’s health sector has focused on building an evidence base around policy interventions that make health care more effective, efficient, and equitable. Becoming a Roybal center will allow J-PAL North America to expand this evidence base to have a greater focus on aging individuals,” says Amy Finkelstein, the John and Jennie S. MacDonald Professor of Economics and co-chair of the MIT Roybal Center. “This population — one that is more likely to interact with the health system and experience poverty — stands to really benefit from the types of low-cost behavioral interventions that the center’s research will focus on.”</p>
1901
1902 <p>As co-chairs, Finkelstein and Marcella Alsan, professor of public policy at Harvard Kennedy School, will help build the MIT Roybal Center’s research agenda and will oversee the proposal review process. Additionally, an advisory committee of six members, including three physicians, will review study proposals and provide strategic guidance. The MIT Roybal Center aims to leverage the interdisciplinary expertise of the advisory committee to ensure that the research funded by the center is implementable, policy-relevant, and meeting the needs of the aging population.</p>
1903
1904 <p>“The center will be anchored by an interdisciplinary advisory committee with representatives from the health care and policy communities. Supporting interdisciplinary research is a priority for the center and we look forward to forging new partnerships that cut across disciplines to identify the most effective policy solutions,” Alsan says.</p>
1905
1906 <p>One of 15 Roybal centers in&nbsp;the United States, the MIT Roybal Center was made possible with support from a five-year P30 center grant from the National Institute of Aging. Those interested in learning more about the MIT Roybal Center are encouraged to contact center Manager <a href="https://www.povertyactionlab.org/person/reuter">Hannah Reuter</a>.</p>
1907 </content:encoded>
1908 <media:content url="https://news.mit.edu/sites/default/files/styles/news_article__cover_image__original/public/images/202009/old-people-exercise.jpg?itok=FthRwGAw" medium="image" type="image/jpeg" width="390" height="260">
1909 <media:description type="plain">Over 25 million Americans over the age of 60 are economically insecure and face rising housing and health-care costs, lack of transportation, and reduced savings.</media:description>
1910 </media:content>
1911 </item>
1912 <item>
1913 <title>New Semilab SE-2000 Spectroscopic Ellipsometer at MIT.nano</title>
1914 <link>https://news.mit.edu/2020/semilab-se-2000-spectroscopic-ellipsometer-mitnano-0925</link>
1915 <description>Instrument can provide information on various properties of thin films via a combination of optical characterization and modeling.</description>
1916 <pubDate>Fri, 25 Sep 2020 13:45:00 -0400</pubDate>
1917 <guid isPermaLink="true">https://news.mit.edu/2020/semilab-se-2000-spectroscopic-ellipsometer-mitnano-0925</guid>
1918 <dc:creator>MIT.nano</dc:creator>
1919 <content:encoded><p>In nanofabrication, it is important to characterize and understand the physical, mechanical, and chemical properties of the various thin-film materials used to build devices. Ellipsometry is a particularly valuable technique for non-destructive optical characterization of a wide variety of thin-film materials — such as semiconductors, dielectrics, polymers, and metals — as well as complex multilayered structures. By measuring the change in the light polarization upon interaction with the matter, information such as thickness, refractive index, dielectric permittivity, composition, and doping can be revealed.</p>
1920
1921 <p>To expand thin-film measurement capabilities available to its researchers, MIT.nano recently acquired a Semilab SE-2000 Spectroscopic Ellipsometer. The instrument has been installed and qualified in the third-floor clean room in Building 12, and is now available for training and use.</p>
1922
1923 <p>The Semilab ellipsometer will augment the MIT.nano clean room metrology capabilities by addressing the need for a broad spectral range (from 245 nm–1,700 nm), autofocus, automatic goniometer for incidence angle adjustment (20-90 degrees), and large-area (300 mm) automatic stage with mapping capabilities, according to Anna Osherov, assistant director of user services for Characterization.nano.&nbsp;</p>
1924
1925 <p>“We’re delighted to add this tool to the growing set of instruments and equipment available to MIT.nano users,” she says. “We’re developing a powerful set of ex-situ and in-situ capabilities for characterization and metrology in our facilities, and ellipsometry is yet another fundamental technique that will advance research for many users.”</p>
1926
1927 <p>With Microspot optics and an optical camera for region-of-interest visualization, the instrument is able to uniformly assess samples ranging from pieces up to 300 mm wafers. The SE-2000 replaces the capabilities that had been provided by the decommissioned Tencor UV1280, which was located in the second-floor Integrated Circuits Laboratory in Building 39.</p>
1928
1929 <p><a href="https://semilab.com/">Semilab</a> designs and manufactures metrology equipment for the characterization of semiconductor and photovoltaic materials, and for the R&amp;D and manufacturing processes of semiconductor devices, flat-panel displays, and solar cells.</p>
1930
1931 <p>For more information about MIT.nano’s tools and instruments, visit <a href="https://nanousers.mit.edu/">nanousers.mit.edu</a>.</p>
1932 </content:encoded>
1933 <media:content url="https://news.mit.edu/sites/default/files/styles/news_article__cover_image__original/public/images/202010/mit-nano-semilab-00.jpg?itok=U-3UTt6o" medium="image" type="image/jpeg" width="390" height="260">
1934 <media:description type="plain">MIT’s new Semilab SE-2000 Spectroscopic Ellipsometer is located in MIT.nano’s third-floor clean room and is available for use by trained researchers.</media:description>
1935 <media:credit>Photo courtesy of MIT.nano.</media:credit>
1936 </media:content>
1937 </item>
1938 <item>
1939 <title>Cancer researchers collaborate, target DNA damage repair pathways for cancer therapy</title>
1940 <link>https://news.mit.edu/2020/cancer-researchers-collaborate-target-pathways-for-cancer-therapies-1001</link>
1941 <description>MIT researchers find blocking the expression of the genes XPA and MK2 enhances the tumor-shrinking effects of platinum-based chemotherapies in p53-mutated cancers.</description>
1942 <pubDate>Thu, 01 Oct 2020 15:00:00 -0400</pubDate>
1943 <guid isPermaLink="true">https://news.mit.edu/2020/cancer-researchers-collaborate-target-pathways-for-cancer-therapies-1001</guid>
1944 <dc:creator>Koch Institute</dc:creator>
1945 <content:encoded><p>Cancer therapies that target specific molecular defects arising from mutations in tumor cells are currently the focus of much anticancer drug development. However, due to the absence of good targets and to the genetic variation in tumors, platinum-based chemotherapies are still the mainstay in the treatment of many cancers, including those that have a mutated version of the tumor suppressor gene p53. P53 is mutated in a majority of cancers, which enables tumor cells to develop resistance to platinum-based chemotherapies. But these defects can still be exploited to selectively target tumor cells by targeting a second gene to take down the tumor cell, leveraging a phenomenon known as synthetic lethality.</p>
1946
1947 <p>Focused on understanding and targeting cell signaling in cancer, the laboratory of Michael Yaffe, the David H. Koch Professor Science and director of the MIT Center for Precision Cancer Medicine, seeks to identify pathways that are synthetic lethal with each other, and to develop therapeutic strategies that capitalize on that relationship. His group has <a href="https://news.mit.edu/2013/biologists-id-new-cancer-weakness-1114">already identified MK2</a> as a key signaling pathway in cancer and a partner to p53 in a synthetic lethal combination.</p>
1948
1949 <p>Now, working with a team of fellow researchers at MIT’s Koch Institute for Integrative Cancer Research, Yaffe’s lab added a new target, the gene XPA, to the combination. Appearing in <em>Nature Communications</em>, the work demonstrates the potential of “augmented synthetic lethality,” where depletion of a third gene product enhances a combination of targets already known to show synthetic lethality. Their work not only demonstrates the effectiveness of teaming up cancer targets, but also of the collaborative teamwork for which the Koch Institute is known.</p>
1950
1951 <p>P53 serves two functions: first, to give cells time to repair DNA damage by pausing cell division, and second, to induce cell death if DNA damage is too severe. Platinum-based chemotherapies work by inducing enough DNA damage to initiate the cell’s self-destruct mechanism. In their previous work, the Yaffe lab found that when cancer cells lose p53, they can re-wire their signaling circuitry to recruit MK2 as a backup pathway. However, MK2 only restores the ability to orchestrate DNA damage repair, but not to initiate cell death.</p>
1952
1953 <p>The Yaffe group reasoned that targeting MK2, which is only recruited when p53 function is absent, would be a unique way to create a synthetic lethality that specifically kills p53-defective tumors, by blocking their ability to coordinate DNA repair after chemotherapy. Indeed, the Yaffe Lab was able to show in pre-clinical models of non-small cell lung cancer tumors with mutations in p53, that silencing MK2 in combination with chemotherapy treatment caused the tumors to shrink significantly.</p>
1954
1955 <p>Although promising, MK2 has proven difficult to drug. Attempts to create target-specific, clinically viable small-molecule MK2 inhibitors have so far been unsuccessful. Researchers led by co-lead author Yi Wen Kong, then a postdoc in the Yaffe lab, have been exploring the use of RNA interference (siRNA) to stop expression of the MK2 gene, but siRNA’s tendency to degrade rapidly in the body presents new challenges.</p>
1956
1957 <p>Enter the potential of nanomaterials, and a team of nanotechnology experts in the laboratory of Paula Hammond, the David H. Koch Professor of Engineering, head of the MIT Department of Chemical Engineering, and the Yaffe group’s upstairs neighbor. There, Kong found a willing collaborator in then-postdoc Erik Dreaden, whose team had developed a delivery vehicle known as a nanoplex to protect siRNA until it gets to a cancer cell. In studies of non-small cell lung cancer models where mice were given the MK2-targeting nanocomplexes and standard chemotherapy, the combination clearly enhanced tumor cell response to chemotherapy. However, the overall increase in survival was significant, but relatively modest.</p>
1958
1959 <p>Meanwhile, Kong had identified XPA, a key protein involved in another DNA repair pathway called NER, as a potential addition to the MK2-p53 synthetic lethal combination. As with MK2, efforts to target XPA using traditional small-molecule drugs have not yet proven successful, and RNA interference emerged as the team’s tool of choice. The flexible and highly controllable nature of the Hammond group’s nanomaterials assembly technologies allowed Dreaden to incorporate siRNAs against both XPA and MK2 into the nanocomplexes.</p>
1960
1961 <p>Kong and Dreaden tested these dual-targeted nanocomplexes against established tumors in an immunocompetent, aggressive lung cancer model developed in collaboration between the laboratories of professor of biology Michael Hemann and Koch Institute Director Tyler Jacks. They let the tumors grow even larger before treatment than they had in their previous study, thus raising the bar for therapeutic intervention.</p>
1962
1963 <p>Tumors in mice treated with the dual-targeted nanocomplexes and chemotherapy were reduced by up to 20-fold over chemotherapy alone, and similarly improved over single-target nanocomplexes and chemotherapy. Mice treated with this regimen survived three times longer than with chemotherapy alone, and much longer than mice receiving nanocomplexes targeting MK2 or XPA alone.</p>
1964
1965 <p>Overall, these data demonstrate that identification and therapeutic targeting of augmented synthetic lethal relationships — in this case between p53, MK2 and XPA — can produce a safe and highly effective cancer therapy by re-wiring multiple DNA damage response pathways, the systemic inhibition of which may otherwise be toxic.</p>
1966
1967 <p>The nanocomplexes are modular and can be adapted to carry other siRNA combinations or for use against other cancers in which this augmented synthetic lethality combination is relevant. Beyond application in lung cancer, the researchers — including Kong, who is now a research scientist at the Koch Institute, and Dreaden, who is now an assistant professor at Georgia Tech and Emory School of Medicine — are working to test this strategy for use against ovarian and other cancers.</p>
1968
1969 <p>Additional collaborations and contributions were made to this project by the laboratories of Koch Institute members Stephen Lippard and Omer Yilmaz, the Eisen and Chang Career Development Professor.&nbsp;</p>
1970
1971 <p>This work was supported in part by a Mazumdar-Shaw International Oncology Fellowship, a postdoctoral fellowship from the S. Leslie Misrock (1949) Frontier Fund for Cancer Nanotechnology, and by the Charles and Marjorie Holloway Foundation, the Ovarian Cancer Research Foundation, and the Breast Cancer Alliance.</p>
1972 </content:encoded>
1973 <media:content url="https://news.mit.edu/sites/default/files/styles/news_article__cover_image__original/public/images/202009/MIT-Koch-Institute-Augmented-Synthetic-Lethality_0.jpg?itok=fEJH4yuj" medium="image" type="image/jpeg" width="390" height="260">
1974 <media:description type="plain">Nanocomplexes could be used to treat ovarian tumors like these by delivering combinations of siRNAs that target genes with synthetic lethal relationships. </media:description>
1975 <media:credit>Image courtesy of Erik C. Dreaden, Yi Wen Kong, Michael Yaffe, and Paula T. Hammond</media:credit>
1976 </media:content>
1977 </item>
1978 <item>
1979 <title>MIT researchers highlight the impacts of logjams in river restoration projects</title>
1980 <link>https://news.mit.edu/2020/mit-research-impacts-logjams-river-restoration-projects-0925</link>
1981 <description>New analysis can help scientists and engineers better assess flood risk and habitat creation associated with restoration efforts.</description>
1982 <pubDate>Fri, 25 Sep 2020 13:30:00 -0400</pubDate>
1983 <guid isPermaLink="true">https://news.mit.edu/2020/mit-research-impacts-logjams-river-restoration-projects-0925</guid>
1984 <dc:creator>Department of Civil and Environmental Engineering</dc:creator>
1985 <content:encoded><p>Researchers at MIT have modeled how engineered and natural wood jams change river water levels, enabling an assessment of the trade-offs in flood risk and habitat creation for river restoration projects.</p>
1986
1987 <p>In a recent <a href="https://agupubs.onlinelibrary.wiley.com/doi/10.1029/2020GL089346" target="_blank">paper</a> published in <em>Geophysical Research Letters</em>, researchers <a href="https://www.cardiff.ac.uk/people/view/2416119-follett-elizabeth" target="_blank">Elizabeth Follett</a> ’09 PhD ’16, postdoc Isabella Schalko, and Donald and Martha Harleman Professor of Civil and Environmental Engineering Heidi Nepf detail their analysis of 584 experiments measuring the backwater rise induced by model logjams in an experimental flume. Schalko ran these experiments, with the hope of filling gaps of the previously understudied physical processes to better explain just how water flow is impacted by large, densely packed groups of logs and to better inform current and future flood risk as well as river restoration projects.</p>
1988
1989 <p>“We’ve been missing a way to describe the physical mechanisms by which large groups of wood pieces affect the river water level,”&nbsp;says Follett, who is the lead author on the paper and a Royal Academy of Engineering Research Fellow at Cardiff University. “Our work allows researchers to characterize structural properties of wood jams from field measurements, by measuring the river water level up- and downstream of the jam and applying our new model.”</p>
1990
1991 <p>The team hopes that the structural metrics will be useful for a wide range of scientists and engineers. The paper has also had an unintended benefit: bridging gaps between research groups.</p>
1992
1993 <p>“What I like most about the paper is that it brings together two research communities; those who look more at in-stream wood, and those more interested in canopy shear flows,” says Schalko.</p>
1994
1995 <p>The findings could have significant implications for government or non-profit organizations engaging in restoration projects. According to the researchers, there is growing interest all over the world in river restoration projects; up until now, it was understood that adding wood to rivers was good for restoration because wood increases flow heterogeneity by increasing water depth. Despite the growing popularity of wood as a solution, the physical processes have not been studied in depth and are not always accounted for in flood prediction models.</p>
1996
1997 <p>“Flood risk and river restoration projects have attracted recent investments, but up to now it has been difficult to include the effect of wood in flood models to improve the design and assessment of these projects,” says Follett. “This is a first step in the direction of being able to theoretically describe how wood alters the flow conditions in a river.” When joined with existing information, the new data on wood jams should better inform flood risk and river restoration efforts in the future.</p>
1998 </content:encoded>
1999 <media:content url="https://news.mit.edu/sites/default/files/styles/news_article__cover_image__original/public/images/202009/isabella-schalko-logjam.jpg?itok=Ie2Q5Knl" medium="image" type="image/jpeg" width="390" height="260">
2000 <media:description type="plain">Researchers Elizabeth Follett and Isabella Schalko, and Professor Heidi Nepf, detail their analysis of 584 experiments measuring the backwater rise induced by model logjams in an experimental flume. </media:description>
2001 </media:content>
2002 </item>
2003 <item>
2004 <title>The Committee on Animal Care solicits feedback</title>
2005 <link>https://news.mit.edu/2020/committee-animal-care-solicits-feedback-1002</link>
2006 <description/>
2007 <pubDate>Fri, 02 Oct 2020 12:30:00 -0400</pubDate>
2008 <guid isPermaLink="true">https://news.mit.edu/2020/committee-animal-care-solicits-feedback-1002</guid>
2009 <dc:creator>MIT News Office</dc:creator>
2010 <content:encoded><p>The Committee on Animal Care (CAC) and the MIT vice president for research welcome any information which would aid our efforts to assure the humane care of research animals used at MIT and the Whitehead Institute for Biomedical Research.<br />
2011 &nbsp;<br />
2012 Established to ensure that MIT researchers working with animals comply with federal, state, local and institutional regulations on animal care, the CAC inspects animals, animal facilities, and laboratories, and reviews all research and teaching exercises that involve animals before experiments are performed.<br />
2013 &nbsp;<br />
2014 If you have concerns about animal welfare, please contact the Committee on Animal Care by calling 617-324-6892, or send your concern in writing to the CAC Office (Room 16-408) to <a href="cacpo@mit.edu">cacpo@mit.edu</a>. The issue will be forwarded to the chair of the CAC and the attending veterinarian. You may also contact any of the following:</p>
2015
2016 <ul>
2017 <li>Vice president for research: 617-253-3206, mtz@mit.edu</li>
2018 <li>Director of the Division of Comparative Medicine or attending veterinarian: 617-253-1735, jgfox@mit.edu</li>
2019 <li>CAC chair: 617-285-5156, hheller@mit.edu</li>
2020 </ul>
2021
2022 <p>All concerns about animal welfare will remain confidential; the identity of individuals who contact the CAC with concerns will be treated as confidential, and individuals will be protected against reprisal and discrimination consistent with MIT policies. The Committee on Animal Care will report its findings and actions to correct the issue to the vice president for research, the director of comparative medicine, the individual who reported the concern (if not reported anonymously), and oversight agencies as applicable.</p>
2023 </content:encoded>
2024 <media:content url="https://news.mit.edu/sites/default/files/styles/news_article__cover_image__original/public/images/202009/MIT-seal_0.jpg?itok=PqV0Z9JV" medium="image" type="image/jpeg" width="390" height="260">
2025 </media:content>
2026 </item>
2027 <item>
2028 <title>Astronomers discover an Earth-sized “pi planet” with a 3.14-day orbit</title>
2029 <link>https://news.mit.edu/2020/earth-sized-pi-planet-0921</link>
2030 <description>The rocky world, with its baking-hot surface, is likely not habitable.</description>
2031 <pubDate>Mon, 21 Sep 2020 00:00:00 -0400</pubDate>
2032 <guid isPermaLink="true">https://news.mit.edu/2020/earth-sized-pi-planet-0921</guid>
2033 <dc:creator>Jennifer Chu | MIT News Office</dc:creator>
2034 <content:encoded><p>In a delightful alignment of astronomy and mathematics, scientists at MIT and elsewhere have discovered a “pi Earth” — an Earth-sized planet that zips around its star every 3.14 days, in an orbit reminiscent of the universal mathematics constant.</p>
2035
2036 <p>The researchers discovered signals of the planet in data taken in 2017 by the NASA Kepler Space Telescope’s K2 mission. By zeroing in on the system earlier this year with SPECULOOS, a network of ground-based telescopes, the team confirmed that the signals were of a planet orbiting its star. And indeed, the planet appears to still be circling its star today, with a pi-like period, every 3.14 days.</p>
2037
2038 <p>“The planet moves like clockwork,” says Prajwal Niraula, a graduate student in MIT’s Department of Earth, Atmospheric and Planetary Sciences (EAPS), who is the lead author of a paper published today in the <em>Astronomical Journal</em>, titled: “π Earth: a 3.14-day Earth-sized Planet from K2’s Kitchen Served Warm by the SPECULOOS Team.”</p>
2039
2040 <p>“Everyone needs a bit of fun these days,” says co-author Julien de Wit, of both the paper title and the discovery of the pi planet itself.</p>
2041
2042 <p><strong>Planet extraction</strong></p>
2043
2044 <p>The new planet is labeled K2-315b; it’s the 315th planetary system discovered within K2 data — just one system shy of an even more serendipitous place on the list.</p>
2045
2046 <p>The researchers estimate that K2-315b has a radius of 0.95 that of Earth’s, making it just about Earth-sized. It orbits a cool, low-mass star that is about one-fifth the size of the sun. The planet circles its star every 3.14 days, at a blistering 81 kilometers per second, or about 181,000 miles per hour.</p>
2047
2048 <p>While its mass is yet to be determined, scientists suspect that K2-315b is terrestrial, like the Earth. But the pi planet is likely not habitable, as its tight orbit brings the planet close enough to its star to heat its surface up to 450 kelvins, or around 350 degrees Fahrenheit — perfect, as it turns out, for baking actual pie.</p>
2049
2050 <p>“This would be too hot to be habitable in the common understanding of the phrase,” says Niraula, who adds that the excitement around this particular planet, aside from its associations with the mathematical constant pi, is that it may prove a promising candidate for studying the characteristics of its atmosphere.</p>
2051
2052 <p>“We now know we can mine and extract planets from archival data, and hopefully there will be no planets left behind, especially these really important ones that have a high impact,” says de Wit, who is an assistant professor in EAPS, and a member of MIT’s Kavli Institute for Astrophysics and Space Research.</p>
2053
2054 <p>Niraula and de Wit’s MIT co-authors include Benjamin Rackham and Artem Burdanov, along with a team of international collaborators.</p>
2055
2056 <p><strong>Dips in the data</strong></p>
2057
2058 <p>The researchers are members of SPECULOOS, an acronym for The Search for habitable Planets EClipsing ULtra-cOOl Stars, and named for a network of four 1-meter telescopes in Chile’s Atacama Desert, which scan the sky across the southern hemisphere. Most recently, the network added a fifth telescope, which is the first to be located in the northern hemisphere, named Artemis — a project that was spearheaded by researchers at MIT.</p>
2059
2060 <p>The SPECULOOS telescopes are designed to search for Earth-like planets around nearby, ultracool dwarfs — small, dim stars that offer astronomers a better chance of spotting an orbiting planet and characterizing its atmosphere, as these stars lack the glare of much larger, brighter stars.</p>
2061
2062 <p>“These ultracool dwarfs are scattered all across the sky,” Burdanov says. “Targeted ground-based surveys like SPECULOOS are helpful because we can look at these ultracool dwarfs one by one.”</p>
2063
2064 <p>In particular, astronomers look at individual stars for signs of transits, or periodic dips in a star’s light, that signal a possible planet crossing in front of the star, and briefly blocking its light.</p>
2065
2066 <p>Earlier this year, Niraula came upon a cool dwarf, slightly warmer than the commonly accepted threshold for an ultracool dwarf, in data collected by the K2 campaign — the Kepler Space Telescope’s second observing mission, which monitored slivers of the sky as the spacecraft orbited around the sun.</p>
2067
2068 <p>Over several months in 2017, the Kepler telescope observed a part of the sky that included the cool dwarf, labeled in the K2 data as EPIC 249631677. Niraula combed through this period and found around 20 dips in the light of this star, that seemed to repeat every 3.14 days.</p>
2069
2070 <p>The team analyzed the signals, testing different potential astrophysical scenarios for their origin, and confirmed that the signals were likely of a transiting planet, and not a product of some other phenomena such as a binary system of two spiraling stars.</p>
2071
2072 <p>The researchers then planned to get a closer look at the star and its orbiting planet with SPECULOOS. But first, they had to identify a window of time when they would be sure to catch a transit.</p>
2073
2074 <p>“Nailing down the best night to follow up from the ground is a little bit tricky,” says Rackham, who developed a forecasting algorithm to predict when a transit might next occur. “Even when you see this 3.14 day signal in the K2 data, there’s an uncertainty to that, which adds up with every orbit.”</p>
2075
2076 <p>With Rackham’s forecasting algorithm, the group narrowed in on several nights in February 2020 during which they were likely to see the planet crossing in front of its star. They then pointed SPECULOOS’ telescopes in the direction of the star and were able to see three clear transits: two with the network’s Southern Hemisphere telescopes, and the third from Artemis, in the Northern Hemisphere.</p>
2077
2078 <p>The researchers say the new pi planet may be a promising candidate to follow up with the James Webb Space Telescope (JWST), to see details of the planet’s atmosphere. For now, the team is looking through other datasets, such as from NASA’s TESS mission, and are also directly observing the skies with Artemis and the rest of the SPECULOOS network, for signs of Earthlike planets.</p>
2079
2080 <p>“There will be more interesting planets in the future, just in time for JWST, a telescope designed to probe the atmosphere of these alien worlds,” says Niraula. “With better algorithms, hopefully one day, we can look for smaller planets, even as small as Mars.”</p>
2081
2082 <p>This research was supported in part by the Heising-Simons Foundation, and the European Research Council.</p>
2083 </content:encoded>
2084 <media:content url="https://news.mit.edu/sites/default/files/styles/news_article__cover_image__original/public/images/202009/MIT-Pi-Planet-01-PRESS_1.jpg?itok=Q0pN-hSf" medium="image" type="image/jpeg" width="390" height="260">
2085 <media:description type="plain">Scientists at MIT and elsewhere have discovered an Earth-sized planet that zips around its star every 3.14 days. </media:description>
2086 <media:credit>Image credit: NASA Ames/JPL-Caltech/T. Pyle, Christine Daniloff, MIT</media:credit>
2087 </media:content>
2088 </item>
2089 <item>
2090 <title>The wobbling shadow of the M87* black hole</title>
2091 <link>https://news.mit.edu/2020/wobbling-shadow-m87-black-hole-0923</link>
2092 <description>Analysis of Event Horizon Telescope observations from 2009 to 2017 reveals turbulent evolution of the M87* black hole image.</description>
2093 <pubDate>Wed, 23 Sep 2020 12:00:00 -0400</pubDate>
2094 <guid isPermaLink="true">https://news.mit.edu/2020/wobbling-shadow-m87-black-hole-0923</guid>
2095 <dc:creator>Nancy Wolfe Kotary | MIT Haystack Observatory</dc:creator>
2096 <content:encoded><p>In 2019, the Event Horizon Telescope (EHT) Collaboration, including a team of <a href="http://www.haystack.mit.edu" target="_blank">MIT Haystack Observatory</a> scientists, delivered <a href="https://eventhorizontelescope.org/press-release-april-10-2019-astronomers-capture-first-image-black-hole">the first image of a black hole</a>, revealing M87* — the supermassive object in the center of the M87 galaxy. The EHT team has used the lessons learned last year to analyze the archival data sets from 2009 to 2013, some of which were not published before. The analysis reveals the behavior of the black hole image across multiple years, indicating persistence of the crescent-like shadow feature, but also variation of its orientation — the crescent appears to be wobbling. The full results appear today in <em>The Astrophysical Journal</em> in an article titled, “<a href="https://iopscience.iop.org/article/10.3847/1538-4357/abac0d" target="_blank">Monitoring the Morphology of M87* in 2009–2017 with the Event Horizon Telescope</a>.”</p>
2097
2098 <p>The EHT is a global array of telescopes, performing synchronized observations using the technique of very long baseline interferometry. Together they form a virtual Earth-sized radio dish, providing a uniquely high image resolution. In 2009–13, M87* was observed by early-EHT prototype arrays, with telescopes located at three geographical sites from 2009 to 2012 and four sites in 2013. In 2017, the EHT reached maturity with telescopes located at five distinct geographical sites across the globe.</p>
2099
2100 <p>Datasets for this research were fully correlated at MIT Haystack Observatory. The 2009–2013 observations consist of less data than the ones performed in 2017, making it impossible to create an image. But the EHT team was able to use statistical modeling to look at changes in the appearance of M87* over time. In the modeling approach, the data are compared to a family of geometric templates, in this case rings of non-uniform brightness. A statistical framework is then employed to determine if the data are consistent with such models and to find the best-fitting model parameters.</p>
2101
2102 <p>“This is a beautiful example of creative data analysis. Extracting important new astrophysical understanding&nbsp;and squeezing new insight out of previous observations is an imaginative example of how scientists can maximally use the information content of such painstakingly collected data,” says Colin Lonsdale, director of MIT Haystack Observatory and chair of the EHT Collaboration Board. “The behavior of this event horizon scale structure over a period of years allows important additional constraints to be placed on the properties of this fascinating object.”</p>
2103
2104 <p>Expanding the analysis to the 2009–2017 observations, EHT scientists have shown that M87* adheres to theoretical expectations. The black hole’s shadow diameter has remained consistent with the prediction of Einstein’s theory of general relativity for a black hole of 6.5 billion solar masses.</p>
2105
2106 <p>“In this study, we show that the general morphology, or presence of an asymmetric ring, most likely persists on timescales of several years,” says Kazu Akiyama, research scientist at MIT Haystack Observatory and a participant in the project. “The consistency throughout multiple observational epochs gives us more confidence than ever about the nature of M87* and the origin of the shadow.”</p>
2107
2108 <p>Although the crescent diameter remained consistent, the EHT team found that the data were hiding a surprise: The ring is wobbling, and that means big news for scientists. For the first time, they can get a glimpse of the dynamical structure of the accretion flow so close to the black hole’s event horizon, in extreme gravity conditions. Studying this region holds the key to understanding phenomena such as relativistic jet launching, and will allow scientists to formulate new tests of the theory of general relativity.</p>
2109
2110 <p>The gas falling onto a black hole heats up to billions of degrees, ionizes, and becomes turbulent in the presence of magnetic fields. “Because the flow of matter is turbulent, the crescent appears to wobble with time,” says Maciek Wielgus of the Harvard and Smithsonian Center for Astrophysics, who is a Black Hole Initiative fellow, and lead author of the paper. “Actually, we see quite a lot of variation there, and not all theoretical models of accretion allow for so much wobbling. What it means is that we can start ruling out some of the models based on the observed source dynamics.”</p>
2111
2112 <p>“MIT Haystack Observatory was instrumental in organizing these early observations, correlating the massive amounts of data returned on large numbers of hard drives, and reducing the data,” says Vincent Fish, research scientist at Haystack Observatory. “While we were able to place important constraints on the size and nature of the emission in M87* at the time, the images made from the much better 2017 array data provided critical context for fully understanding what the earlier data were trying to tell us.”</p>
2113
2114 <p>Haystack scientist Geoff Crew adds, “After working on EHT technology for a decade, I’m gratified that M87* has been making equally good use of its time.”</p>
2115 </content:encoded>
2116 <media:content url="https://news.mit.edu/sites/default/files/styles/news_article__cover_image__original/public/images/202009/m87-black-hole-snapshots.png?itok=YmLbKIYd" medium="image" type="image/jpeg" width="390" height="260">
2117 <media:description type="plain">Snapshots of the M87* black hole obtained through imaging/geometric modeling, and the EHT array of telescopes in 2009 to 2017. The diameter of all rings is similar, but the location of the bright side varies.</media:description>
2118 <media:credit>Image courtesy of M. Wielgus, D. Pesce, and the EHT Collaboration.</media:credit>
2119 </media:content>
2120 </item>
2121 <item>
2122 <title>Engineers produce a fisheye lens that’s completely flat</title>
2123 <link>https://news.mit.edu/2020/flat-fisheye-lens-0918</link>
2124 <description>The single piece of glass produces crisp panoramic images.</description>
2125 <pubDate>Fri, 18 Sep 2020 00:00:00 -0400</pubDate>
2126 <guid isPermaLink="true">https://news.mit.edu/2020/flat-fisheye-lens-0918</guid>
2127 <dc:creator>Jennifer Chu | MIT News Office</dc:creator>
2128 <content:encoded><p>To capture panoramic views in a single shot, photographers typically use fisheye lenses — ultra-wide-angle lenses made from multiple pieces of curved glass, which distort incoming light to produce wide, bubble-like images. Their spherical, multipiece design makes fisheye lenses inherently bulky and often costly to produce.</p>
2129
2130 <p>Now engineers at MIT and the University of Massachusetts at Lowell have designed a wide-angle lens that is completely flat. It is the first flat fisheye lens to produce crisp, 180-degree panoramic images. The design is a type of “metalens,” a wafer-thin material patterned with microscopic features that work together to manipulate light in a specific way.</p>
2131
2132 <p>In this case, the new fisheye lens consists of a single flat, millimeter-thin piece of glass covered on one side with tiny structures that precisely scatter incoming light to produce panoramic images, just as a conventional curved, multielement fisheye lens assembly would. The lens works in the infrared part of the spectrum, but the researchers say it could be modified to capture images using visible light as well.</p>
2133
2134 <p>The new design could potentially be adapted for a range of applications, with thin, ultra-wide-angle lenses built directly into smartphones and laptops, rather than physically attached as bulky add-ons. The low-profile lenses might also be integrated into medical imaging devices such as endoscopes, as well as in virtual reality glasses, wearable electronics, and other computer vision devices.</p>
2135
2136 <p>“This design comes as somewhat of a surprise, because some have thought it would be impossible to make a metalens with an ultra-wide-field view,” says Juejun Hu, associate professor in MIT’s Department of Materials Science and Engineering. “The fact that this can actually realize fisheye images is completely outside expectation.</p>
2137
2138 <p>This isn’t just light-bending — it’s mind-bending.”</p>
2139
2140 <p>Hu and his colleagues have published their results today in the journal <em>Nano Letters</em>. Hu’s MIT coauthors are Mikhail Shalaginov, Fan Yang, Peter Su, Dominika Lyzwa, Anuradha Agarwal, and Tian Gu, along with Sensong An and Hualiang Zhang of UMass Lowell.</p>
2141 <p><strong>Design on the back side</strong></p>
2142
2143 <p>Metalenses, while still largely at an experimental stage, have the potential to significantly reshape the field of optics. Previously, scientists have designed metalenses that produce high-resolution and relatively wide-angle images of up to 60 degrees. To expand the field of view further would traditionally require additional optical components to correct for aberrations, or blurriness — a workaround that would add bulk to a metalens design.</p>
2144
2145 <p>Hu and his colleagues instead came up with a simple design that does not require additional components and keeps a minimum element count. Their new metalens is a single transparent piece made from calcium fluoride with a thin film of lead telluride deposited on one side. The team then used lithographic techniques to carve a pattern of optical structures into the film.</p>
2146
2147 <p>Each structure, or “meta-atom,” as the team refers to them, is shaped into one of several nanoscale geometries, such as a rectangular or a bone-shaped configuration, that refracts light in a specific way. For instance, light may take longer to scatter, or propagate off one shape versus another — a phenomenon known as phase delay.</p>
2148
2149 <p>In conventional fisheye lenses, the curvature of the glass naturally creates a distribution of phase delays that ultimately produces a panoramic image. The team determined the corresponding pattern of meta-atoms and carved this pattern into the back side of the flat glass.</p>
2150
2151 <p>‘We’ve designed the back side structures in such a way that each part can produce a perfect focus,” Hu says.</p>
2152
2153 <p>On the front side, the team placed an optical aperture, or opening for light.</p>
2154
2155 <p>“When light comes in through this aperture, it will refract at the first surface of the glass, and then will get angularly dispersed,” Shalaginov explains. “The light will then hit different parts of the backside, from different and yet continuous angles. As long as you design the back side properly, you can be sure to achieve high-quality imaging across the entire panoramic view.”</p>
2156
2157 <p><strong>Across the panorama</strong></p>
2158
2159 <p>In one demonstration, the new lens is tuned to operate in the mid-infrared region of the spectrum. The team used the imaging setup equipped with the metalens to snap pictures of a striped target. They then compared the quality of pictures taken at various angles across the scene, and found the new lens produced images of the stripes that were crisp and clear, even at the edges of the camera’s view, spanning nearly 180 degrees.</p>
2160
2161 <p>“It shows we can achieve perfect imaging performance across almost the whole 180-degree view, using our methods,” Gu says.</p>
2162
2163 <p>In another study, the team designed the metalens to operate at a near-infrared wavelength using amorphous silicon nanoposts as the meta-atoms. They plugged the metalens into a simulation used to test imaging instruments. Next, they fed the simulation a scene of Paris, composed of black and white images stitched together to make a panoramic view. They then ran the simulation to see what kind of image the new lens would produce.</p>
2164
2165 <p>“The key question was, does the lens cover the entire field of view? And we see that it captures everything across the panorama,” Gu says. “You can see buildings and people, and the resolution is very good, regardless of whether you’re looking at the center or the edges.”</p>
2166
2167 <p>The team says the new lens can be adapted to other wavelengths of light. To make a similar flat fisheye lens for visible light, for instance, Hu says the optical features may have to be made smaller than they are now, to better refract that particular range of wavelengths. The lens material would also have to change. But the general architecture that the team has designed would remain the same.</p>
2168
2169 <p>The researchers are exploring applications for their new lens, not just as compact fisheye cameras, but also as panoramic projectors, as well as depth sensors built directly into smartphones, laptops, and wearable devices.</p>
2170
2171 <p>“Currently, all 3D sensors have a limited field of view, which is why when you put your face away from your smartphone, it won’t recognize you,” Gu says. “What we have here is a new 3D sensor that enables panoramic depth profiling, which could be useful for consumer electronic devices.”</p>
2172
2173 <p>This research was funded in part by DARPA under the EXTREME Program.</p>
2174 </content:encoded>
2175 <media:content url="https://news.mit.edu/sites/default/files/styles/news_article__cover_image__original/public/images/202009/MIT-Fisheye-Lens-01.jpg?itok=L4TWOyCg" medium="image" type="image/jpeg" width="390" height="260">
2176 <media:description type="plain">3D artistic illustration of the wide-field-of-view metalens capturing a 180° panorama of MIT’s Killian Court and producing a high-resolution monochromatic flat image.”</media:description>
2177 <media:credit>Image: Mikhail Shalaginov, Tian Gu, Christine Daniloff, Felice Hankel, Juejun Hu</media:credit>
2178 </media:content>
2179 </item>
2180 <item>
2181 <title>Helping robots avoid collisions</title>
2182 <link>https://news.mit.edu/2020/realtime-robots-motion-0917</link>
2183 <description>Realtime Robotics has created a controller that helps robots safely move around on the fly.</description>
2184 <pubDate>Thu, 17 Sep 2020 15:00:00 -0400</pubDate>
2185 <guid isPermaLink="true">https://news.mit.edu/2020/realtime-robots-motion-0917</guid>
2186 <dc:creator>Zach Winn | MIT News Office</dc:creator>
2187 <content:encoded><p>George Konidaris still remembers his disheartening introduction to robotics.</p>
2188
2189 <p>“When you’re a young student and you want to program a robot, the first thing that hits you is this immense disappointment at how much you can’t do with that robot,” he says.</p>
2190
2191 <p>Most new roboticists want to program their robots to solve interesting, complex tasks — but it turns out that just moving them through space without colliding with objects is more difficult than it sounds.</p>
2192
2193 <p>Fortunately, Konidaris is hopeful that future roboticists will have a more exciting start in the field. That’s because roughly four years ago, he co-founded Realtime Robotics, a startup that’s solving the “motion planning problem” for robots.</p>
2194
2195 <p>The company has invented a solution that gives robots the ability to quickly adjust their path to avoid objects as they move to a target. The Realtime controller is a box that can be connected to a variety of robots and deployed in dynamic environments.</p>
2196
2197 <p>“Our box simply runs the robot according to the customer’s program,” explains Konidaris, who currently serves as Realtime’s chief roboticist. “It takes care of the movement, the speed of the robot, detecting obstacles, collision detection. All [our customers] need to say is, ‘I want this robot to move here.’”</p>
2198
2199 <p>Realtime’s key enabling technology is a unique circuit design that, when combined with proprietary software, has the effect of a plug-in motor cortex for robots. In addition to helping to fulfill the expectations of starry-eyed roboticists, the technology also represents a fundamental advance toward robots that can work effectively in changing environments.</p>
2200
2201 <p><strong>Helping robots get around</strong></p>
2202
2203 <p>Konidaris was not the first person to get discouraged about the motion planning problem in robotics. Researchers in the field have been working on it for 40 years. During a four-year postdoc at MIT, Konidaris worked with School of Engineering Professor in Teaching Excellence Tomas Lozano-Perez, a pioneer in the field who was publishing papers on motion planning before Konidaris was born.</p>
2204
2205 <p>Humans take collision avoidance for granted. Konidaris points out that the simple act of grabbing a beer from the fridge actually requires a series of tasks such as opening the fridge, positioning your body to reach in, avoiding other objects in the fridge, and deciding where to grab the beer can.</p>
2206
2207 <p>“You actually need to compute more than one plan,” Konidaris says. “You might need to compute hundreds of plans to get the action you want. … It’s weird how the simplest things humans do hundreds of times a day actually require immense computation.”</p>
2208
2209 <p>In robotics, the motion planning problem revolves around the computational power required to carry out frequent tests as robots move through space. At each stage of a planned path, the tests help determine if various tiny movements will make the robot collide with objects around it. Such tests have inspired researchers to think up ever more complicated algorithms in recent years, but Konidaris believes that’s the wrong approach.</p>
2210
2211 <p>“People were trying to make algorithms smarter and more complex, but usually that’s a sign that you’re going down the wrong path,” Konidaris says. “It’s actually not that common that super technically sophisticated techniques solve problems like that.”</p>
2212
2213 <p>Konidaris left MIT in 2014 to join the faculty at Duke University, but he continued to collaborate with researchers at MIT’s Computer Science and Artificial Intelligence Laboratory (CSAIL). Duke is also where Konidaris met Realtime co-founders Sean Murray, Dan Sorin, and Will Floyd-Jones. In 2015, the co-founders collaborated to make a new type of computer chip with circuits specifically designed to perform the frequent collision tests required to move a robot safely through space. The custom circuits could perform operations in parallel to more efficiently test short motion collisions.</p>
2214
2215 <p>“When I left MIT for Duke, one thing bugging me was this motion planning thing should really be solved by now,” Konidaris says. “It really did come directly out of a lot of experiences at MIT. I wouldn’t have been able to write a single paper on motion planning before I got to MIT.”</p>
2216
2217 <p>The researchers founded Realtime in 2016 and quickly brought on robotics industry veteran Peter Howard SM&nbsp;’87, who currently serves as Realtime’s CEO and is also considered a co-founder.</p>
2218
2219 <p>“I wanted to start the company in Boston because I knew MIT and lot of robotics work was happening there,” says Konidaris, who moved to Brown University in 2016. “Boston is a hub for robotics. There’s a ton of local talent, and I think a lot of that is because MIT is here — PhDs from MIT became faculty at local schools, and those people started robotics programs. That network effect is very strong.”</p>
2220
2221 <p><strong>Removing robot restraints</strong></p>
2222
2223 <p>Today the majority of Realtime’s customers are in the automotive, manufacturing, and logistics industries. The robots using Realtime’s solution are doing everything from spot welding to making inspections to picking items from bins.</p>
2224
2225 <p>After customers purchase Realtime’s control box, they load in a file describing the configuration of the robot’s work cell, information about the robot such as its end-of-arm tool, and the task the robot is completing. Realtime can also help optimally place the robot and its accompanying sensors around a work area. Konidaris says Realtime can shorten the process of deploying robots from an average of 15 weeks to one week.</p>
2226
2227 <p>Once the robot is up and running, Realtime’s box controls its movement, giving it instant collision-avoidance capabilities.</p>
2228
2229 <p>“You can use it for any robot,” Konidaris says. “You tell it where it needs to go and we’ll handle the rest.”</p>
2230
2231 <p>Realtime is part of MIT’s Industrial Liaison Program (ILP), which helps companies make connections with larger industrial partners, and it recently joined ILP’s STEX25 startup accelerator.</p>
2232
2233 <p>With a few large rollouts planned for the coming months, the Realtime team’s excitement is driven by the belief that solving a problem as fundamental as motion planning unlocks a slew of new applications for the robotics field.</p>
2234
2235 <p>“What I find most exciting about Realtime is that we are a true technology company,” says Konidaris. “The vast majority of startups are aimed at finding a new application for existing technology; often, there’s no real pushing of the technical boundaries with a new app or website, or even a new robotics ‘vertical.’ But we really did invent something new, and that edge and that energy is what drives us. All of that feels very MIT to me.”</p>
2236 </content:encoded>
2237 <media:content url="https://news.mit.edu/sites/default/files/styles/news_article__cover_image__original/public/images/202009/MIT-RealtimeRobotics-01-press.jpg?itok=NbRHRNPO" medium="image" type="image/jpeg" width="390" height="260">
2238 <media:description type="plain">The startup Realtime Robotics is helping robots solve the motion planning problem by giving them collision avoidance capabilities. Here, a robot avoids a researcher’s waving hand.</media:description>
2239 <media:credit>Courtesy of Realtime Robotics</media:credit>
2240 </media:content>
2241 </item>
2242 <item>
2243 <title>Rapid test for Covid-19 shows improved sensitivity</title>
2244 <link>https://news.mit.edu/2020/rapid-covid-19-test-0917</link>
2245 <description>A CRISPR-based test developed at MIT and the Broad Institute can detect nearly as many cases as the standard Covid-19 diagnostic.</description>
2246 <pubDate>Thu, 17 Sep 2020 11:00:00 -0400</pubDate>
2247 <guid isPermaLink="true">https://news.mit.edu/2020/rapid-covid-19-test-0917</guid>
2248 <dc:creator>Anne Trafton | MIT News Office</dc:creator>
2249 <content:encoded><p>Since the start of the Covid-19 pandemic, researchers at MIT and the Broad Institute of MIT and Harvard, along with their collaborators at the University of Washington, Fred Hutchinson Cancer Research Center, Brigham and Women's Hospital, and the Ragon Institute, have been working on a CRISPR-based diagnostic for Covid-19 that can produce results in 30 minutes to an hour, with similar accuracy as the standard PCR diagnostics now used.</p>
2250
2251 <p>The new test, known as STOPCovid, is still in the research stage but, in principle, could be made cheaply enough that people could test themselves every day. In a study appearing today in the <em>New England Journal of Medicine</em>, the researchers showed that on a set of patient samples, their test detected 93 percent of the positive cases as determined by PCR tests for Covid-19.</p>
2252
2253 <p>“We need rapid testing to become part of the fabric of this situation so that people can test themselves every day, which will slow down outbreak,” says Omar Abudayyeh, an MIT McGovern Fellow working on the diagnostic.</p>
2254
2255 <p>Abudayyah is one of the senior authors of the study, along with Jonathan Gootenberg, a McGovern Fellow, and Feng Zhang, a core member of the Broad Institute, investigator at the MIT McGovern Institute and Howard Hughes Medical Institute, and the James and Patricia Poitras ’63 Professor of Neuroscience at MIT. The first authors of the paper are MIT biological engineering graduate students Julia Joung and Alim Ladha in the Zhang lab.</p>
2256 <p><strong>A streamlined test</strong></p>
2257
2258 <p>Zhang’s laboratory began collaborating with the Abudayyeh and Gootenberg laboratory to work on the Covid-19 diagnostic soon after the SARS-CoV-2 outbreak began. They focused on making an assay, called STOPCovid, that was simple to carry out and did not require any specialized laboratory equipment. Such a test, they hoped, would be amenable to future use in point-of-care settings, such as doctors’ offices, pharmacies, nursing homes, and schools.&nbsp;</p>
2259
2260 <p>“We developed STOPCovid so that everything could be done in a single step,” Joung says. “A single step means the test can be potentially performed by nonexperts outside of laboratory settings.”</p>
2261
2262 <p>In the new version of STOPCovid reported today, the researchers incorporated a process to concentrate the viral genetic material in a patient sample by adding magnetic beads that attract RNA, eliminating the need for expensive purification kits that are time-intensive and can be in short supply due to high demand.&nbsp;This concentration step boosted the test’s sensitivity so that it now approaches that of PCR.</p>
2263
2264 <p>“Once we got the viral genomes onto the beads, we found that that could get us to very high levels of sensitivity,” Gootenberg says.</p>
2265
2266 <p>Working with collaborators Keith Jerome at Fred Hutchinson Cancer Research Center and Alex Greninger at the University of Washington, the researchers tested STOPCovid on 402 patient samples — 202 positive and 200 negative — and found that the new test detected 93 percent of the positive cases as determined by the standard CDC PCR test.</p>
2267
2268 <p>“Seeing STOPCovid working on actual patient samples was really gratifying,” Ladha says.</p>
2269
2270 <p>They also showed, working with Ann Woolley and Deb Hung at Brigham and Women’s Hospital, that the STOPCovid test works on samples taken using the less invasive anterior nares swab. They are now testing it with saliva samples, which could make at-home tests even easier to perform. The researchers are continuing to develop the test with the hope of delivering it to end users to help fight the COVID-19 pandemic.</p>
2271
2272 <p>“The goal is to make this test easy to use and sensitive, so that we can tell whether or not someone is carrying the virus as early as possible,” Zhang says.</p>
2273
2274 <p>The research was funded by the National Institutes of Health, the Swiss National Science Foundation, the Patrick J. McGovern Foundation, the McGovern Institute for Brain Research, the Massachusetts Consortium on Pathogen Readiness Evergrande Covid-19 Response Fund, the Mathers Foundation, the Howard Hughes Medical Institute, the Open Philanthropy Project, J. and P. Poitras, and R. Metcalfe.</p>
2275 </content:encoded>
2276 <media:content url="https://news.mit.edu/sites/default/files/styles/news_article__cover_image__original/public/images/202009/MIT-Covid-Test-01-press.jpg?itok=DArEWVKk" medium="image" type="image/jpeg" width="390" height="260">
2277 <media:description type="plain">This scanning electron microscope image shows SARS-CoV-2 (round gold objects) emerging from the surface of cells cultured in the lab.</media:description>
2278 <media:credit>Credit: NIAID</media:credit>
2279 </media:content>
2280 </item>
2281 <item>
2282 <title>Live imaging method brings structure to mapping brain function</title>
2283 <link>https://news.mit.edu/2020/live-imaging-method-brings-structure-mapping-brain-function-0924</link>
2284 <description>Scientists distinguish brain regions based on what they do, but now have a new way to overlay information about how they are built.</description>
2285 <pubDate>Thu, 24 Sep 2020 14:00:00 -0400</pubDate>
2286 <guid isPermaLink="true">https://news.mit.edu/2020/live-imaging-method-brings-structure-mapping-brain-function-0924</guid>
2287 <dc:creator>David Orenstein | Picower Institute for Learning and Memory</dc:creator>
2288 <content:encoded><p>To understand the massive capabilities and complexities of the brain, neuroscientists segment it into regions based on what they appear to do — such as processing what we sense, or how to move. What’s been lacking, however, is an ability to tie those functional maps precisely and consistently to matching distinctions of physical structure, especially in live animals while they are performing the functions of interest. In a new study, MIT researchers demonstrate a new way to do that, providing an unprecedented pairing of functional mapping in live mice with distinguishing structural information for each region all the way through the cortex into deeper tissue below.</p>
2289
2290 <p>“Our study shows for the first time that structural and functional coupling of visual areas in the mouse brain can be detected at sub-cellular resolution in vivo,” wrote the authors based in the lab of Mriganka Sur, the Newton Professor of Neuroscience in The Picower Institute for Learning and Memory and the Department of Brain and Cognitive Sciences at MIT.</p>
2291
2292 <p>The technique could give scientists more precise ways to distinguish the borders and contents of regions they wish to study and could help them better understand the way that structural distinctions develop within individuals in different functional regions over time. Sur’s lab, for instance, is intensely interested in understanding the especially complex development of vision. Humans have 35 or so distinct functional regions that contribute to processing vision, Sur notes, and even mice have 10.</p>
2293
2294 <p>“There is something profound in the way that vision is represented and created in mammalian brains,” Sur says. “Where do these areas come from, what do they mean and what do they do? It has not been easy to understand how they differ. The critical thing is to precisely map or match the functional representation of each area with its anatomical uniqueness.”</p>
2295
2296 <p>To develop tools to help answer those questions, postdoc Murat Yildirim led the study published in <em>Biomedical Optics Express. </em>In it he describes how the research team combined a method of charting functional areas — retinotopic mapping — with deep structural information measured by a technology he has <a href="https://picower.mit.edu/news/scope-advance-reveals-first-look-through-all-cortical-layers-awake-brain">helped to pioneer</a> — third-harmonic generation (THG) three-photon microscopy.</p>
2297
2298 <p>In retinotopic mapping, researchers can identify functional regions by engineering neurons to flash when they become electrically active (and show changes in calcium) in response to a particular stimulation. For example, scientists could show a mouse a pattern moving across a screen and mark where neurons light up, with each area showing a characteristic location and pattern of response.</p>
2299
2300 <p>Three-photon microscopy can finely resolve individual cells and their smaller substructures as deep as a millimeter or more — enough to see all the way through the cortex. THG, meanwhile, adds the capability to finely resolve both blood vessels and the fibers of a material called myelin that wrap the long, tendrilous axons of many neurons. THG does not require adding any labeling dyes or chemicals.</p>
2301 <p>Crucially, THG yields an important optical measure called effective attenuation length (EAL), which is a measure of how much the light is absorbed or scattered as it moves through the tissue. In the study, Yildirim and co-authors show that EAL specifically depends on each region’s unique architecture of cells, blood vessels, and myelin. They measured EAL in each of six visual functional regions and showed that the EAL significantly differed among neighboring visual areas, providing a structural signature of sorts for each functional area. Their measurements were so precise, in fact, that they could show how EAL varied within functional regions, being most unique toward the middle and blending closer to the values of neighboring regions out toward the borders.</p>
2302
2303 <p>In other words, by combining the retinotopic mapping with THG three-photon microscopy, Yildirim said, scientists can identify distinct regions by both their function and structure while continuing to work with animals in live experiments. This can produce more accurate and faster results than making observations during behavior and then dissecting tissue in hopes of relocating those same exact positions in preserved brain sections later.</p>
2304
2305 <p>“We would like to combine the strength of retinotopic mapping with three-photon imaging to get more structural information,” Yildirim says. “Otherwise there may be some discrepancies when you do the live imaging of brain activity but then take the tissue out, stain it and try to find the same region.”</p>
2306
2307 <p>Especially as three-photon microscopy gains wider adoption and imaging speeds improve — right now, imaging a millimeter-deep column of cortex takes about 15 minutes, the authors acknowledge — the team expects that its new method could be used not only for studies of the visual system, but also in regions all around the cortex. Moreover, it may help characterize disease states as well as healthy brain structure and function.</p>
2308
2309 <p>“This advance should enable similar studies of structural and functional coupling in other sensory and non-sensory cortical areas in the brains of mice and other animal models,” they wrote. “We believe that the structural and functional correlation in visual areas that we describe for the first time points to crucial developmental mechanisms that set up these areas, thus our work would lead to a better fundamental understanding of brain development, and of disorders such as Alzheimer’s, stroke, and aging.”</p>
2310
2311 <p>In addition to Yildirim and Sur, the paper’s other authors are Ming Hu, Nhat Le, Hiroki Sugihara, and Peter So.</p>
2312
2313 <p>The National Institutes of Health, the National Science Foundation, The JPB Foundation and the Massachusetts Life Sciences Initiative provided funding for the study.</p>
2314 </content:encoded>
2315 <media:content url="https://news.mit.edu/sites/default/files/styles/news_article__cover_image__original/public/images/202009/mouse-brain-vessels-myelin.png?itok=t0uWdHeS" medium="image" type="image/jpeg" width="390" height="260">
2316 <media:description type="plain">A distinct thicket of vessels and myelin fibers (which wrap around the long extensions of many neurons) are evident in each color-coded visual processing region in the cortex of a mouse. The columns are formed by stacking images taken at 5-micron increments through a millimeter of depth in each of the regions.</media:description>
2317 <media:credit>Image: Murat Yildirim/Sur Lab</media:credit>
2318 </media:content>
2319 </item>
2320 <item>
2321 <title>Making tuberculosis more susceptible to antibiotics</title>
2322 <link>https://news.mit.edu/2020/tuberculosis-antibiotics-galactin-0916</link>
2323 <description>Shortening carbohydrates in the bacterial cell wall makes them more vulnerable to certain drugs.</description>
2324 <pubDate>Wed, 16 Sep 2020 14:00:00 -0400</pubDate>
2325 <guid isPermaLink="true">https://news.mit.edu/2020/tuberculosis-antibiotics-galactin-0916</guid>
2326 <dc:creator>Anne Trafton | MIT News Office</dc:creator>
2327 <content:encoded><p>Every living cell is coated with a distinctive array of carbohydrates, which serves as a unique cellular “ID” and helps to manage the cell’s interactions with other cells.</p>
2328
2329 <p>MIT chemists have now discovered that changing the length of these carbohydrates can dramatically affect their function. In a study of mycobacteria, the type of bacteria that cause tuberculosis and other diseases, they found that shortening the length of a carbohydrate called galactan impairs some cell functions and makes the cells much more susceptible to certain antibiotics.</p>
2330
2331 <p>The findings suggest that drugs that interfere with galactan synthesis could be used along with existing antibiotics to create more effective treatments, says Laura Kiessling, the Novartis Professor of Chemistry at MIT and the senior author of the study.</p>
2332
2333 <p>“There are a lot of TB strains that are resistant to the current set of antibiotics,” Kiessling says. “TB kills over a million people every year and is the number one infectious disease killer.”</p>
2334
2335 <p>Former MIT graduate student Alexander Justen is the lead author of the paper, which appears today in <em>Science Advances</em>.</p>
2336
2337 <p><strong>The long and short of it</strong></p>
2338
2339 <p>Galactan, a polysaccharide, is a component of the cell wall of mycobacteria, but little is known about its function. Until now, its only known role was to form links between molecules called peptidoglycans, which make up most of the bacterial cell wall, and other sugars and lipids. However, the version of galactan found in mycobacteria is much longer than it needs to be to perform this linker function.</p>
2340
2341 <p>“What was so strange is that the galactan is about 30 sugar molecules long, but the branch points for the other sugars that it links to are at eight, 10, and 12. So, why is the cell expending so much energy to make galactan longer than 12 units?” Kiessling says.</p>
2342
2343 <p>That question led Kiessling and her research group to investigate what might happen if galactan were shorter. A team led by Justen genetically engineered a type of mycobacteria called <em>Mycobacterium smegmatis</em> (which is related to <em>Mycobacterium tuberculosis</em> but is not harmful to humans) so that their galactan chains would contain only 12 sugar molecules.</p>
2344
2345 <p>As a result of this shortening, cells lost their usual shape and developed “blebs,” or bulges from their cell membranes. Shortening galactan also shrank the size of a compartment called the periplasm, a space that is found between a bacterial cell’s inner and outer cell membranes. This compartment is involved in absorbing nutrients from the cell’s environment.</p>
2346
2347 <p>Truncating galactan also made the cells more susceptible to certain antibiotics — specifically, antibiotics that are hydrophobic. Mycobacteria cell walls are relatively impermeable to hydrophobic antibiotics, but the shortened galactan molecules make the cells more permeable, so these drugs can get inside more easily.</p>
2348
2349 <p>“This suggests that drugs that would lead to these truncated chains could be valuable in combination with hydrophobic antibiotics,” Kiessling says. “I think it validates this part of the cell as a good target.”</p>
2350
2351 <p>Her lab is currently working on developing drugs that could block galactan synthesis, which is not targeted by any existing TB drugs. Patients with TB are usually given drug combinations that have to be taken for six months, and some strains have developed resistance to the existing drugs.</p>
2352
2353 <p><strong>Unexpected roles</strong></p>
2354
2355 <p>Kiessling’s lab is also studying the question of why it is useful for bacteria to alter the length of their carbohydrate molecules. One hypothesis is that it helps them to shield themselves from the immune system, she says. Some studies have shown that a dense coating of longer carbohydrate chains could help to achieve a stealth effect by preventing host immune cells from interacting with proteins on the bacterial cell surface.</p>
2356
2357 <p>If that hypothesis is confirmed, then drugs that interfere with the length of galactan or other carbohydrates might also help the immune system fight off bacterial infection, Kiessling says. This could be useful for treating not only tuberculosis but also other diseases caused by mycobacteria, such as chronic obstructive pulmonary disease (COPD) and leprosy. Other strains of mycobacteria (known as “flesh-eating bacteria”) cause a potentially deadly infection called necrotizing fasciitis. All of these mycobacteria have galactan in their cell walls, and there are no good vaccines against any of them.</p>
2358
2359 <p>Although the research may end up helping scientists to develop better drugs, Kiessling first became interested in this topic as a basic science question.</p>
2360
2361 <p>“The reason I like this paper is because while it does have implications for treating tuberculosis, it also shows a fundamentally new role for carbohydrates, which I love. People are finding that they can have unexpected roles, and this is another unexpected result,” she says.</p>
2362
2363 <p>The research was funded by the National Institute of Allergy and Infectious Disease and the National Institutes of Health Common Fund.</p>
2364 </content:encoded>
2365 <media:content url="https://news.mit.edu/sites/default/files/styles/news_article__cover_image__original/public/images/202009/MIT-Antibiotic-Susceptibility-01-press_0.jpg?itok=xfLrAD9z" medium="image" type="image/jpeg" width="390" height="260">
2366 <media:description type="plain">MIT chemists have discovered that changing the length of the carbohydrate galactan can dramatically affect its function. In a study of mycobacteria, the type of bacteria that cause tuberculosis and other diseases, they found that shortening galactan impairs some cell functions and makes the cells much more susceptible to certain antibiotics. This artistic impression shows the carbohydrate galactan sliced in half in the foreground, while the background shows an image of tuberculosis bacteria. </media:description>
2367 <media:credit>Image: Jose-Luis Olivares, MIT. Tuberculosis image courtesy of NIAID</media:credit>
2368 </media:content>
2369 </item>
2370 <item>
2371 <title>Provably exact artificial intelligence for nuclear and particle physics</title>
2372 <link>https://news.mit.edu/2020/provably-exact-artificial-intelligence-nuclear-particle-physics-0924</link>
2373 <description>MIT-led team uses AI and machine learning to explore fundamental forces.</description>
2374 <pubDate>Thu, 24 Sep 2020 13:50:00 -0400</pubDate>
2375 <guid isPermaLink="true">https://news.mit.edu/2020/provably-exact-artificial-intelligence-nuclear-particle-physics-0924</guid>
2376 <dc:creator>Sandi Miller | Department of Physics</dc:creator>
2377 <content:encoded><p>The Standard Model&nbsp;of particle physics describes all the known elementary particles and three of the four&nbsp;fundamental forces governing the universe; everything except gravity. These three forces — electromagnetic, strong,&nbsp;and weak — govern&nbsp;how&nbsp;particles are formed, how they interact, and how the particles decay.</p>
2378
2379 <p>Studying particle and&nbsp;nuclear physics within this framework, however, is difficult, and relies on large-scale numerical studies.&nbsp;For example,&nbsp;many aspects of the strong force require numerically simulating the dynamics at the scale of 1/10th to 1/100th the&nbsp;size of a proton to answer fundamental questions about the properties of protons, neutrons, and nuclei.</p>
2380
2381 <p>“Ultimately,&nbsp;we are computationally limited in the study of proton and nuclear structure using lattice field theory,” says assistant professor of physics <a href="https://web.mit.edu/physics/people/faculty/shanahan_phiala.html">Phiala Shanahan</a>. “There are a lot of interesting problems that we know how to address in principle, but&nbsp;we just don’t have enough compute, even though we run on the largest supercomputers in the world.”</p>
2382
2383 <p>To push past these limitations, Shanahan leads a group that combines theoretical physics with machine learning&nbsp;models. In their paper “<a href="https://doi.org/10.1103/PhysRevLett.125.121601" target="_blank">Equivariant flow-based sampling for lattice gauge theory</a>,” published this month in <em>Physical&nbsp;Review Letters</em>, they show how incorporating the symmetries of physics theories into&nbsp;machine learning and artificial intelligence architectures can provide much faster algorithms for theoretical physics.&nbsp;</p>
2384
2385 <p>“We are using machine learning not to analyze large amounts of data, but to accelerate first-principles theory in a&nbsp;way which doesn’t compromise the rigor of the approach,” Shanahan says. “This particular work demonstrated that&nbsp;we can build machine learning architectures with some of the symmetries of the Standard Model of particle and&nbsp;nuclear physics built in, and accelerate the sampling problem we are targeting by orders of magnitude.”&nbsp;</p>
2386
2387 <p>Shanahan launched the project with MIT graduate student <a href="https://www.csail.mit.edu/person/gurtej-kanwar">Gurtej Kanwar</a> and with Michael Albergo, who is now&nbsp;at NYU. The project expanded to include Center for Theoretical Physics postdocs Daniel Hackett and Denis Boyda,&nbsp;NYU Professor Kyle Cranmer, and physics-savvy machine-learning scientists at Google Deep Mind, Sébastien Racanière and Danilo Jimenez Rezende.</p>
2388
2389 <p>This month’s paper is one in a series aimed at enabling studies&nbsp;in theoretical physics that are currently computationally intractable. “Our aim is to&nbsp;develop new algorithms for a key component of numerical calculations in theoretical physics,” says Kanwar. “These calculations&nbsp;inform us about the inner workings of the Standard Model of particle physics, our most fundamental theory of matter.&nbsp;Such calculations are of vital importance to compare against results from particle physics experiments, such as the&nbsp;Large Hadron Collider at CERN, both to constrain the model more precisely and to discover where the model breaks&nbsp;down and must be extended to something even more fundamental.”</p>
2390
2391 <p>The only known systematically controllable method of studying the Standard Model of particle physics in the&nbsp;nonperturbative regime is based on a sampling of snapshots of quantum fluctuations in the vacuum. By measuring&nbsp;properties of these fluctuations, one can infer properties of the particles and collisions of interest.</p>
2392
2393 <p>This technique&nbsp;comes with challenges, Kanwar explains. “This sampling is expensive, and we are looking to use physics-inspired machine learning techniques to draw samples far more efficiently,” he says. “Machine learning has already made great&nbsp;strides on generating images, including, for example, recent work by <a href="https://thispersondoesnotexist.com/">NVIDIA to generate images of faces</a> 'dreamed&nbsp;up' by neural networks. Thinking of these snapshots of the vacuum as images,&nbsp;we think it's quite natural to turn to similar methods for our problem.”</p>
2394
2395 <p>Adds Shanahan, “In our approach to sampling these quantum snapshots, we optimize a model that takes us from a space that is&nbsp;easy to sample to the target space: given a trained model, sampling is then efficient since you just need to take&nbsp;independent samples in the easy-to-sample space, and transform them via the learned model.”</p>
2396
2397 <p>In particular, the group has introduced a framework for building machine-learning models that exactly respect a&nbsp;class of symmetries, called "gauge symmetries," crucial for studying high-energy physics.</p>
2398
2399 <p>As a proof of principle, Shanahan and colleagues used their framework to train machine-learning models to simulate&nbsp;a theory in two dimensions, resulting in orders-of-magnitude efficiency gains over state-of-the-art techniques and&nbsp;more precise predictions from the theory. This paves the way for significantly accelerated research into the&nbsp;fundamental forces of nature using physics-informed machine learning.</p>
2400
2401 <p>The group’s first few papers as a collaboration discussed applying the machine-learning technique to a simple lattice&nbsp;field theory, and developed this class of approaches on compact, connected manifolds which describe the more&nbsp;complicated field theories of the Standard Model. Now they are working to scale the techniques to state-of-the-art&nbsp;calculations.</p>
2402
2403 <p>“I think we have shown over the past year that there is a lot of promise in combining physics knowledge&nbsp;with machine learning techniques,” says Kanwar. “We are actively thinking about how to tackle the remaining&nbsp;barriers in the way of performing full-scale simulations using our approach. I hope to see the first application of&nbsp;these methods to calculations at scale in the next couple of years. If we are able to overcome the last few obstacles,&nbsp;this promises to extend what we can do with limited resources, and I dream of performing calculations soon that give&nbsp;us novel insights into what lies beyond our best understanding of physics today.”<br />
2404 <br />
2405 This idea of physics-informed machine learning is also known by the team as “ab-initio AI,” a key theme of the&nbsp;recently launched MIT-based National Science Foundation <a href="http://www.iaifi.org/">Institute for Artificial Intelligence and Fundamental Interactions</a> (IAIFI), where Shanahan is research coordinator for physics theory.</p>
2406
2407 <p>Led by the <a href="http://web.mit.edu/lns/">Laboratory for Nuclear&nbsp;Science</a>, the IAIFI is comprised of both&nbsp;physics and AI researchers&nbsp;at MIT and Harvard, Northeastern, and Tufts universities.</p>
2408
2409 <p>“Our&nbsp;collaboration is a great example of the spirit of IAIFI, with a team with diverse backgrounds coming together to&nbsp;advance AI and physics simultaneously” says Shanahan. As well as research like Shanahan’s targeting physics&nbsp;theory, IAIFI researchers are also working to use AI to enhance the scientific potential of various facilities, including&nbsp;the Large Hadron Collider and the Laser Interferometer Gravity Wave Observatory, and to advance AI&nbsp;itself.&nbsp;</p>
2410 </content:encoded>
2411 <media:content url="https://news.mit.edu/sites/default/files/styles/news_article__cover_image__original/public/images/202009/ShanahanKanwarAIMLPhysics_0.jpg?itok=FfkVB-pw" medium="image" type="image/jpeg" width="390" height="260">
2412 <media:description type="plain">Physics team leader Phiala Shanahan and graduate student Tej Kanwar</media:description>
2413 </media:content>
2414 </item>
2415
2416 </channel>
2417 </rss>