Before being awarded the Koç University Rahmi M. Koç Medal of Science in her native Turkey, Bilge Yildiz was nervous. But it wasn’t standing in front of an audience of hundreds that stressed the Breene M. Kerr Professor in the departments of Nuclear Science and Engineering and of Materials Science and Engineering (DMSE). It wasn’t having to do interviews with journalists. Rather, it was discussing her research in Turkish.
“Two weeks before the award ceremony, I was learning Turkish again,” says Yildiz. Her language of scholarship is English; she came to the United States more than two decades ago, first to do her PhD in nuclear science at MIT, then postdoctoral work in electrochemistry and research at Argonne National Laboratory in Illinois. Even her bachelor’s studies in nuclear science at Hacettepe University in Turkey, where many courses are taught in English, were not in her mother tongue. “I did not learn these materials in Turkish. I never talked about them in Turkish. And all the technical terms especially — I had no idea.”
But Yildiz isn’t known for shying away from new things.
Trained as a nuclear scientist, using artificial intelligence algorithms to safely operate nuclear power plants, she ventured into electrochemistry, studying the chemical reactions that use or make electricity. Her lab today works on a huge array of projects, all centered on the movement of charged atoms in materials. They include human-intelligence-inspired computer processors; fuel cells, which convert hydrogen and oxygen into electricity; and electrolysis, which uses electricity to cause chemical reactions — for example, to produce hydrogen and other useful industrial chemicals. She recently contributed to a NASA project to turn the carbon dioxide in the Martian atmosphere into oxygen, charting one course toward human habitation on Mars.
Named an American Physical Society Fellow in 2021 and a Royal Society of Chemistry Fellow in 2022, Yildiz has been recognized the world over, but never in the country of her birth — until December, when Koç University made her the seventh recipient of the annual award. The Rahmi M. Koç Medal of Science recognizes scientists of Turkish origin younger than 50 who have made outstanding contributions to their fields. It’s given to people from various disciplines, from biological and physical sciences and engineering to social sciences.
Yildiz is the second MIT scholar awarded the medal. Institute Professor Daron Acemoglu, in the Department of Economics, was recognized in 2017 for contributions to labor and political economics and macroeconomics.
“A great honor”
In introducing Yildiz at the award ceremony at the Rahmi M. Koç Museum in Istanbul, Koç University professor Umran Inan spoke of the need for visionaries who are comfortable working across disciplines.
“She is a very young and extraordinarily successful researcher and scientist who can combine different disciplines and outputs to obtain novel and impactful results,” Inan said.
For Yildiz, being acknowledged in her home country was a “great honor.” And in Turkey, whose economy has been devastated by runaway inflation and a collapsing currency, recognition of the importance of scientific research is a good sign.
“It's an opportunity for people in Turkey to get to know about my work and my story,” she said. “I’m also hoping that this motivates young students, that it shows them what’s possible no matter where you come from.”
Yildiz is from Tire, a rural district about 60 miles southeast of Izmir. The region is known for its agriculture and dairy, especially milk and yogurt. After the award ceremony, which was covered by scores of journalists asking for interviews and snapping photos, Yildiz joked that she’d become “as famous as Tire dairy products.”
As happy as Yildiz is for the recognition her work has gotten, she stressed that it’s the science that’s important.
“I don’t want to leave the message that for young people to feel they’re successful they have to win awards,” she said. “Scientists don’t do their science to win awards. We do it to satisfy our curiosity. It gives us energy, and we contribute to society this way. That should be the priority. I feel lucky and grateful every day for the opportunities that enabled me to become a scientist and work with brilliant and motivated students, young scientists, and colleagues at MIT.”
A devoted scientist, mentor
Yildiz’s colleagues and students described a hardworking researcher and teacher committed to relaying her passion for science to others.
“I certainly recognized that she had something rather special in her ability to really take and master both the experimental techniques needed and the theoretical aspects,” said John Kilner, a former professor and now senior research investigator at Imperial College London, who worked with Yildiz on surfaces of oxides in fuel cells, in her early days in academia. “She attracted a lot of very bright young people, and that makes a big difference to your research group to get these people and to foster them as well.”
Kilner spoke in a video about Yildiz jointly produced by Koç University and MIT Video Productions.
Yang Shao-Horn, the JR East Professor of Engineering in the Department of Mechanical Engineering and DMSE, says Yildiz is an accomplished advisor who looks out for the interests of her students.
“She’s a deeply caring mentor who looks at individuals and supports them, asking tough questions and seeking opportunities to make them grow professionally and scientifically,” says Shao-Horn, who advised Yildiz on her postdoctoral work. “This is really an important part of being a professor. We do research and we conduct research projects and we write papers — and these are all tools, mechanisms to train people and empower people and bring the best out of people. And I think Bilge is an excellent role model in doing so.”
Miranda Schwacke, a graduate student in Yildiz’s research group, pointed to her advisor’s commitment to using science to build a better, cleaner society.
“One of the main ways she inspires me is that she really cares about not just making batteries or fuel cells or whatever we work on better,” Schwacke says, “but also really understanding what's going on and exactly how we're making things better, which I think is a better approach, long term.”
Shao-Horn noted that Yildiz, a mother of two, is also an inspiration for women in academia seeking a balance between work and home life.
“Bilge is an incredible role model for many female and other scientists who want to be able to do both — to have an amazing career but also have a happy family,” Shao-Horn says. “And that requires really hard work and discipline.”
Health care has always been ripe for innovation. Whether it’s increasing safety in operating rooms, developing systems to reduce patient wait times, or improving drug delivery, there are endless opportunities to improve the efficacy and efficiency of health care. The Covid-19 pandemic made the need for these solutions all the more pressing.
“There were a number of startups from MIT that addressed problems related to the pandemic,” says George Whitfield, entrepreneur in residence at the Martin Trust Center for MIT Entrepreneurship. “One company, Biobot Analytics, developed a technology to monitor disease spread by looking at wastewater in sewers. In a case of unbelievable serendipity, they developed this right as Covid was starting to spread.”
Another startup inspired by the Covid-19 pandemic, Teal Bio, developed a comfortable, reusable, and transparent respirator that can be worn by health care professionals on long shifts. The company has identified a number of benefits to their design, including lower costs, decreased waste, and an improved ability to identify emotions. Teal Bio was co-founded by Department of Mechanical Engineering (MechE) Leaders for Global Operations alumnus Jason Troutner MBA ’19, SM ’19 and Giovanni Traverso, assistant professor of mechanical engineering at MIT.
Traverso is no stranger to startups. He has co-founded seven of them. An MD-PhD, Traverso is both an assistant professor at MIT and a physician at Brigham and Women’s Hospital. His companies range in size from one employee to 140 employees. With the exception of Teal Bio, the thread that connects his companies is gastroenterology.
“These companies are launching systems that make it easier for patients to receive medication one way or another, particularly through the GI tract,” says Traverso.
One of the companies that Traverso co-founded, Lyndra Therapeutics, hopes to revolutionize how patients take medications. They have developed an oral drug-delivery platform called LYNX, which consistently delivers one, two, or four weeks of medication in one capsule that releases the medication over a specific time period. The capsule dissolves in the stomach and a star-shaped drug delivery system emerges.
The arms of the “star” are made of a polymer that holds the medication and are connected to a central core through degradable linkers. Once the dosing period is complete, the linkers disintegrate, the arms separate, and the entire system safely moves from the stomach into the small intestines, where it passes through the gastrointestinal tract. The platform is being studied with a variety of drugs, including an oral memantine for Alzheimer’s disease.
“Many patients need a loved one or caretaker to help them take oral medication daily, so giving them the ability to take a pill once a week or once a month would positively affect adherence and be hugely impactful on their quality of life,” says Traverso.
Lyndra has raised $240 million to date. One of the therapies they developed to deliver drugs used to treat schizophrenia has advanced to phase-two clinical trials.
Clinical trials are one example of the unique hurdles that medtech startups like Lyndra face on the path to commercialization. Bodies like the U.S. Food and Drug Administration (FDA) and the National Institute for Occupational Safety and Health require strict regulations that need to be met before any medical device, drug, or health care platform can be sold to end users.
“Having an understanding of the regulatory, manufacturing, and business challenges that need to be met to launch a successful product is really crucial. It speaks to the resources that are required to actually be able to execute on these regulations,” adds Traverso. In his first year on MIT’s faculty, Traverso introduced a new class, 2.S988 (Translational Engineering), which aims to introduce these critical elements to students.
Ellen Roche, associate professor of mechanical engineering, is currently trying to determine the regulatory needs for her own startup. In May, she won the grand prize at the inaugural MIT Future Founders Initiative Prize Competition for her pitch.
Roche has developed a minimally invasive technology that occludes the left atrial appendage in patients with atrial fibrillation. The technology, which she developed alongside Professor Jennifer Lewis at Harvard University, decreases the likelihood that blood clots will dislodge, thereby preventing stroke.
“The Future Founders program was invaluable for refining the vision for our company and identifying the correct regulatory and commercialization path to move forward,” says Roche. “Creating a pitch deck forced us to really think through aspects such as our beachhead market, our clinical target population, our funding, and IP [intellectual property] strategy, all the while having access to a network of experts.”
In September, Roche and her team also won the Lab Central Ignite Golden Ticket to support startup founders from traditionally underrepresented groups in the biotech industry.
Both Traverso and Roche have served as instructors for mechanical engineering class 2.75 (Medical Device Design), alongside Professor Alexander Slocum and Nevan Hanumara. The class culminates in a project in which students work with clinicians from Boston-area hospitals and representatives from industry on designing medical devices that address a particular problem. Throughout the class, regulatory experts introduce students to the unique challenges of starting a company or launching a product in the health-care space.
One former student of 2.75, Adam Sachs ’13, co-founded the startup Vicarious Surgical. The company has developed a robotic system that enables minimally invasive surgery. A camera and two robotic instruments enter the abdomen via an incision smaller than the size of a dime. The surgeon can then operate with 360-degree visibility inside a patient’s body.
“Course 2.75 gave me a deep understanding of the entire medical device design process, which was incredibly valuable when we founded Vicarious Surgical. It helped me understand the needs of a user, showed me how to deliver on a product, and allowed me to dip my toes into the process of developing a device from start to finish — much of which I still reference as the company grows and we continue to develop our system,” says Sachs.
Vicarious Surgical, which is based in Waltham, Massachusetts, and currently has just over 200 full-time employees, is in the development process. They have received positive feedback from surgeons regarding their Beta 2 prototypes. After securing the appropriate approvals from the FDA, Sachs and his team plan to bring their product to market for use in hernia and other general surgery procedures.
Traverso sees mechanical engineers, like himself, Roche, and Sachs, as being particularly well-suited to launch medtech startups.
“A huge part of our program is hands-on experience, which we introduce and nurture through many of our course offerings. I think that’s so valuable when you’re developing a device that will be engaging with another human being,” he says.
Student presentations tackled themes of identity, nation-building, racism, multiculturalism, and more, as reflected in the rich traditions of Brazilian music at “The Beat of Brazil” last month at the Lewis Music Library. The presentations were by students of Portuguese enrolled in class 21G.821 (The Beat of Brazil: Portuguese Language Through Brazilian Society), taught by Nilma Dominique.
Three professional musicians were invited to perform as part of the event: Anna Borges and Bill Ward (from the duo Receita de Samba), along with Grammy Award-winning drummer Rafael Barata. After each student spoke about the historical and cultural context for a particular Brazilian style of music, the musicians performed selections in the discussed styles.
Theo St. Francis (an MIT senior in aeronautics and astronautics) explained Brazil’s history of enslaving Africans to work in the sugar fields. He said, “With the workers came their traditions in food, crafts, religion, and mythology, and especially music and dance. … This forced mixing of cultures rendered a brilliant mix of rhythms and sounds among the three primary influences — African, European, and Native Brazilian.” He explained that the music genre choro is imbued with “influences from lundu dances from Angola, the maxixe or Brazilian tango (itself a mix of African rhythms with European polka dance of that time), as well as flute rhythms from European musicians.”
Samba is not really just one genre, but “is better thought of as the backbone of most genres in Brazil,” explained Alessandre Santos, an MIT senior in mathematics. Samba has strong roots in African music, and contains within it many sub-genres. But throughout Brazil’s history, music has been a site for conflicting forces. Modern samba has been reclaimed by Afro-Brazilians as a kind of resistance to racist oppression. Santos explained that samba-canção is a very poetic variant of samba, characterized by soft melodies and slow rhythms, and two examples of this style were performed.
Laura Leal de Souza, a junior at Wellesley College majoring in Latin American Studies, made a remote presentation over a large monitor. She explained that the samba-exaltação nomenclature first appeared in 1939. She explained, “Composers began to write lyrics that worshiped Brazil and the government.” This was used “to create a sense of ‘Brazilianness’ in Brazilian citizens in order to facilitate the dictatorship later implemented by [then-president] Vargas.” She also described the emergence of samba-enredo in the same period, which became strongly identified with government-backed “samba schools,” becoming “the main rhythm of the Brazilian carnival, characterized by its strong percussion and themes that portray specific elements of Brazilian culture.” Leal de Souza also presented on the protest music of the 1960s later in the program.
The role of television and radio was discussed by Ygor Moura, an MIT junior majoring in chemistry, who explained how that media helped to propagate and essentialize aspects of Brazilian culture. The U.S. government’s “Good Neighbor Policy” of the 1930s sought to advance U.S. interests in Central and South America through trade and other means. He pointed to the promotion of Broadway actress and film star Carmen Miranda as the “origin of most of the visual stereotypes Americans have about Brazilians.” Moura also discussed the role of Brazilian music festivals, which became sites for protest during the Brazilian dictatorship of 1964-85.
One of the best-known Brazilian songs, “The Girl from Ipanema,” is an example of bossa nova (Portuguese for “new wave”), explained Dasha Castillo, an MIT senior in computation and cognition. Castillo explained that this music moved away from samba’s larger group ensembles, toward arrangements typically focus on a “lone singer with a guitar, or a singer with another accompanist on another instrument like a piano.” The best-known artists associated with bossa nova are Antonio Carlos Jobim, Vinícius de Moraes, and João Gilberto.
The upbeat music interludes inspired some audience members to get up and dance during the event. One highlight of the evening was a performance by three MIT students (Allessandre Santos, Ygor Moura, and Dasha Castillo) of the well-known Brazilian song “Águas de Março,” known in English as “Waters of March,” accompanied by drummer Rafael Barata.
Nilma Dominique has offered several classes over the years that teach Portuguese language skills through the vehicle of either film or music. “Language and culture go hand-in-hand,” she said. In her class 21G.821, Dominique guides students in examining Brazilian music genres with within a historical context, and “analyzing cultural production from a transnational perspective ... Throughout the course, there was a strong emphasis on developing critical thinking, often centering discussions on how Brazilian musical production reflects questions of identity, social class, race, inequality, and politics.”
Nilma Dominique explained that the Lewis Music library was not just the venue for the event. The staff and student workers at Lewis played a key role in publicizing the event, and event setup and staffing. She added, “This was actually the second time I had the pleasure of working in partnership with the Lewis Library to put on a program of Brazilian music."
After the event, Avery Boddie, the Rosalind Denny Lewis Music Library department head, explained that the library was involved as a continuation of the tradition of “offering engaging programming and outreach to the MIT community through workshops, lectures, and in this case, concerts.” He also pointed out that the library has a “vast collection on music from most regions of the world, including Brazilian music, so any opportunity to support research and education in diverse genres of music across different cultures is something that we value strongly in our department. And who doesn’t enjoy a little samba?”
Funding for the program was from the Council for the Arts at MIT, the Kelly Douglas Fund through the MIT School of Humanities, Arts, and Social Sciences, and the MIT Brazilian Student Association.
Some attendees provided post-event comments. Abby Mrvos said, “The Beat of Brazil show was an absolute treat. The performances were a joy to watch, and the introductions by the students really gave color and context to the experience. I would love to see more events like this in the future!”
Sam Heath agreed, saying: “Listening to a captivating live performance of a selection of Brazilian music along with its historical context, my mind took a journey through time in Brazil, a much-needed escape after a long semester.”
If a scientist wanted to forecast ocean currents to understand how pollution travels after an oil spill, she could use a common approach that looks at currents traveling between 10 and 200 kilometers. Or, she could choose a newer model that also includes shorter currents. This might be more accurate, but it could also require learning new software or running new computational experiments. How to know if it will be worth the time, cost, and effort to use the new method?
A new approach developed by MIT researchers could help data scientists answer this question, whether they are looking at statistics on ocean currents, violent crime, children’s reading ability, or any number of other types of datasets.
The team created a new measure, known as the “c-value,” that helps users choose between techniques based on the chance that a new method is more accurate for a specific dataset. This measure answers the question “is it likely that the new method is more accurate for this data than the common approach?”
Traditionally, statisticians compare methods by averaging a method’s accuracy across all possible datasets. But just because a new method is better for all datasets on average doesn’t mean it will actually provide a better estimate using one particular dataset. Averages are not application-specific.
So, researchers from MIT and elsewhere created the c-value, which is a dataset-specific tool. A high c-value means it is unlikely a new method will be less accurate than the original method on a specific data problem.
In their proof-of-concept paper, the researchers describe and evaluate the c-value using real-world data analysis problems: modeling ocean currents, estimating violent crime in neighborhoods, and approximating student reading ability at schools. They show how the c-value could help statisticians and data analysts achieve more accurate results by indicating when to use alternative estimation methods they otherwise might have ignored.
“What we are trying to do with this particular work is come up with something that is data specific. The classical notion of risk is really natural for someone developing a new method. That person wants their method to work well for all of their users on average. But a user of a method wants something that will work on their individual problem. We’ve shown that the c-value is a very practical proof-of-concept in that direction,” says senior author Tamara Broderick, an associate professor in the Department of Electrical Engineering and Computer Science (EECS) and a member of the Laboratory for Information and Decision Systems and the Institute for Data, Systems, and Society.
She’s joined on the paper by Brian Trippe PhD ’22, a former graduate student in Broderick’s group who is now a postdoc at Columbia University; and Sameer Deshpande ’13, a former postdoc in Broderick’s group who is now an assistant professor at the University of Wisconsin at Madison. An accepted version of the paper is posted online in the Journal of the American Statistical Association.
The c-value is designed to help with data problems in which researchers seek to estimate an unknown parameter using a dataset, such as estimating average student reading ability from a dataset of assessment results and student survey responses. A researcher has two estimation methods and must decide which to use for this particular problem.
The better estimation method is the one that results in less “loss,” which means the estimate will be closer to the ground truth. Consider again the forecasting of ocean currents: Perhaps being off by a few meters per hour isn’t so bad, but being off by many kilometers per hour makes the estimate useless. The ground truth is unknown, though; the scientist is trying to estimate it. Therefore, one can never actually compute the loss of an estimate for their specific data. That’s what makes comparing estimates challenging. The c-value helps a scientist navigate this challenge.
The c-value equation uses a specific dataset to compute the estimate with each method, and then once more to compute the c-value between the methods. If the c-value is large, it is unlikely that the alternative method is going to be worse and yield less accurate estimates than the original method.
“In our case, we are assuming that you conservatively want to stay with the default estimator, and you only want to go to the new estimator if you feel very confident about it. With a high c-value, it’s likely that the new estimate is more accurate. If you get a low c-value, you can’t say anything conclusive. You might have actually done better, but you just don’t know,” Broderick explains.
Probing the theory
The researchers put that theory to the test by evaluating three real-world data analysis problems.
For one, they used the c-value to help determine which approach is best for modeling ocean currents, a problem Trippe has been tackling. Accurate models are important for predicting the dispersion of contaminants, like pollution from an oil spill. The team found that estimating ocean currents using multiple scales, one larger and one smaller, likely yields higher accuracy than using only larger scale measurements.
“Oceans researchers are studying this, and the c-value can provide some statistical ‘oomph’ to support modeling the smaller scale,” Broderick says.
In another example, the researchers sought to predict violent crime in census tracts in Philadelphia, an application Deshpande has been studying. Using the c-value, they found that one could get better estimates about violent crime rates by incorporating information about census-tract-level nonviolent crime into the analysis. They also used the c-value to show that additionally leveraging violent crime data from neighboring census tracts in the analysis isn’t likely to provide further accuracy improvements.
“That doesn’t mean there isn’t an improvement, that just means that we don’t feel confident saying that you will get it,” she says.
Now that they have proven the c-value in theory and shown how it could be used to tackle real-world data problems, the researchers want to expand the measure to more types of data and a wider set of model classes.
The ultimate goal is to create a measure that is general enough for many more data analysis problems, and while there is still a lot of work to do to realize that objective, Broderick says this is an important and exciting first step in the right direction.
This research was supported, in part, by an Advanced Research Projects Agency-Energy grant, a National Science Foundation CAREER Award, the Office of Naval Research, and the Wisconsin Alumni Research Foundation.
The Federal Laboratory Consortium (FLC) has awarded 2023 Excellence in Technology Transfer Awards at the national level to two MIT Lincoln Laboratory software products developed to improve security: Keylime and the Forensic Video Exploitation and Analysis (FOVEA) tool suite. Keylime increases the security and privacy of data and services in the cloud, while FOVEA expedites the process of reviewing and extracting useful information from existing surveillance videos. These technologies both previously won FLC Northeast regional awards for Excellence in Technology Transfer, as well as R&D 100 Awards.
"Lincoln Laboratory is honored to receive these two national FLC awards, which demonstrate the capacity of government-nonprofit-industry partnerships to enhance our national security while simultaneously driving new economic growth," says Louis Bellaire, acting chief technology ventures officer at the laboratory. "These awards are particularly meaningful because they show Lincoln Laboratory teams at their best, developing transformative R&D [research and development] and transferring these results to achieve the strongest benefits for the nation."
A nationwide network of more than 300 government laboratories, agencies, and research centers, FLC helps facilitate the transfer of technologies out of research labs and into the marketplace. Ultimately, the goal of FLC — organized in 1974 and formally chartered by the Federal Technology Transfer Act of 1986 — is to “increase the impact of federal laboratories’ technology transfer for the benefit of the U.S. economy, society, and national security.” Each year, FLC confers awards to commend outstanding technology transfer efforts of employees of FLC member labs and their partners from industry, academia, nonprofit, or state and local government. The Excellence in Technology Transfer Award recognizes exemplary work in transferring federally developed technology.
Keylime: Enabling trust in the cloud
Cloud computing services are an increasingly convenient way for organizations to store, process, and disseminate data and information. These services allow organizations to rent computing resources from a cloud provider, who handles the management and security of those rented machines. Although cloud providers claim that the machines are secure, customers have no way to verify this security. As a result, organizations with sensitive data, such as U.S. government agencies and financial institutions, are reluctant to reap the benefits of flexibility and low cost that commercial cloud providers offer.
Keylime is an open-source software that enables customers with sensitive data to continuously verify the security of cloud machines, and edge and internet-of-things (IoT) devices. To enact its constant security checks, Keylime leverages a piece of hardware called a trusted platform module (TPM). The TPM generates a hash (a string of characters representing data) that will change significantly if data are tampered with. Keylime was designed to make TPMs compatible with cloud technology and reacts to a TPM hash change within seconds to shut down a compromised machine. Keylime also enables users to securely bootstrap secrets (in other words, upload cryptographic keys, passwords, and certificates into the rented machines) without divulging these secrets to the cloud provider.
Lincoln Laboratory transitioned Keylime to the public via an open-source license and distribution strategy that involved a series of partnerships. In 2015, after completing a prototype of Keylime, laboratory researchers Charles Munson and Nabil Schear collaborated with Boston University and Northeastern University to implement it as a core security component in the Mass Open Cloud (MOC) alliance, a public cloud service supporting thousands of researchers in the state. That experience led the team to work with Red Hat (under a pilot program funded by the U.S. Department of Homeland Security) to mature the technology in the open-source community.
Through the efforts of the Red Hat partnership, Keylime was accepted into the Linux Foundation’s highly selective Cloud Native Computing Foundation as a Sandbox project technology in 2019, a significant step in establishing the technology's prestige. More than 50 open-source developers are now contributing to Keylime from around the world, and large organizations, including IBM, are deploying the technology to their cloud machines. Most recently, Red Hat released Keylime into its Enterprise Linux 9.1 operating system.
"We are proud that the Keylime team, our partners, and open-source developers have been recognized for their hard work and dedication with this national FLC award. We look forward to maintaining and building impactful collaborations, and helping the Keylime open-source community continue to grow," says Munson.
The team members recognized with the FLC award are Munson and Schear (creators of Keylime at Lincoln Laboratory); Orran Krieger (MOC and Boston University); Luke Hinds and Michael Peters (Red Hat); Gheorghe Almasi (IBM); and Dan Dardani (formerly of the MIT Technology Licensing Office).
FOVEA: Accelerating video surveillance review
While significant investments have improved camera coverage and video quality, the burden on video operators to analyze and obtain meaningful insights from surveillance footage — still a largely manual process — has greatly increased. The large-scale closed-circuit television systems patrolling public and commercial spaces can comprise hundreds or thousands of cameras, making daily investigation tasks burdensome. Examples of these tasks include searching for events of interest, investigating abandoned objects, and piecing together people's activity from multiple cameras. As with any investigation, time is of the essence in apprehending persons of interest before they have inflicted widespread harm.
FOVEA dramatically reduces the time required for such forensic video analysis. With FOVEA, security personnel can review hours of video in minutes and perform complex investigations in hours rather than days, translating to faster reaction times to in-progress events and a stronger overall security posture. No pre-analysis video curation or proprietary server equipment are required; the add-on suite of video analytic capabilities can be applied to any video stream in an on-demand fashion and support both routine investigations and unforeseen or catastrophic circumstances such as terrorist threats. This suite includes capabilities for jump back, which automatically rewinds video to critical times and detects general scene changes; video summarization, which condenses all motion activity from long raw video into a short visual summary; multicamera navigation and path reconstruction, which tracks activity over place and time and camera to camera in chronological order; and on-demand person search, which scans neighboring cameras for persons of similar appearance.
Lincoln Laboratory began developing FOVEA under sponsorship from the U.S. Department of Homeland Security to address the critical needs of security operators in mass transit security centers. Through an entrepreneurial training program based on the National Science Foundation's Innovation Corps, Lincoln Laboratory conducted a broad set of customer interviews, which ultimately led to Doradus Labs licensing FOVEA. The Colorado-based software development and technical support small business offered FOVEA to two of their casino customers and is now introducing the technology to their customers in the educational and transportation industries.
The laboratory team members recognized with the FLC award are Marianne DeAngelus and Jason Thornton (technology invention and primary contact with Doradus); Natalya Luciw, Diane Staheli, Sanjeev Mohindra, and (formerly) Tyler Shube (customer discovery); Ronald Duarte, Zach Elko, Brett Levasseur (software design and technology demonstrations); Jesslyn Alekseyev, Heather Griffin, and Kimberlee Chang and (formerly) Christine Russ, Aaron Yahr, and Marc Valliant (algorithm and software development); Dan Dardani (formerly of the MIT Technology Licensing Office) and Louis Bellaire (licensing); and Drinalda Kume, Jayme Selinger, and Zach Sweet (contracting services).
“It is wonderful to see the software team’s efforts recognized with this award,” says DeAngelus. “I am grateful for the many friendly people across Lincoln Laboratory and MIT who made this transition happen — especially the licensing, contracts, and communications offices.”
The FLC 2023 award winners will be recognized on March 29 at an awards reception and ceremony during the FLC National Meeting.
Treating cancer with combinations of drugs can be more effective than using a single drug. However, figuring out the optimal combination of drugs, and making sure that all of the drugs reach the right place, can be challenging.
To help address those challenges, MIT chemists have designed a bottlebrush-shaped nanoparticle that can be loaded with multiple drugs, in ratios that can be easily controlled. Using these particles, the researchers were able to calculate and then deliver the optimal ratio of three cancer drugs used to treat multiple myeloma.
“There’s a lot of interest in finding synergistic combination therapies for cancer, meaning that they leverage some underlying mechanism of the cancer cell that allows them to kill more effectively, but oftentimes we don’t know what that right ratio will be,” says Jeremiah Johnson, an MIT professor of chemistry and one of the senior authors of the study.
In a study of mice, the researchers showed that nanoparticles carrying three drugs in the synergistic ratio they identified shrank tumors much more than when the three drugs were given at the same ratio but untethered to a particle. This nanoparticle platform could potentially be deployed to deliver drug combinations against a variety of cancers, the researchers say.
Irene Ghobrial, a professor of medicine at Harvard Medical School and Dana-Farber Cancer Institute, and P. Peter Ghoroghchian, president of Ceptur Therapeutics and a former MIT Koch Institute Clinical Investigator, are also senior authors of the paper, which appears today in Nature Nanotechnology. Alexandre Detappe, an assistant professor at the Strasbourg Europe Cancer Institute, and Hung Nguyen PhD ’19 are the paper’s lead authors.
Using nanoparticles to deliver cancer drugs allows the drugs to accumulate at the tumor site and reduces toxic side effects because the particles protect the drugs from being released prematurely. However, only a handful of nanoparticle drug formulations have received FDA approval to treat cancer, and only one of these particles carries more than one drug.
For several years, Johnson’s lab has been working on polymer nanoparticles designed to carry multiple drugs. In the new study, the research team focused on a bottlebrush-shaped particle. To make the particles, drug molecules are inactivated by binding to polymer building blocks and then mixed together in a specific ratio for polymerization. This forms chains that extend from a central backbone, giving the molecule a bottlebrush-like structure with inactivated drugs — prodrugs — along the bottlebrush backbone. Cleavage of the linker that holds the drug to the backbone release the active agent.
“If we want to make a bottlebrush that has two drugs or three drugs or any number of drugs in it, we simply need to synthesize those different drug conjugated monomers, mix them together, and polymerize them. The resulting bottlebrushes have exactly the same size and shape as the bottlebrush that only has one drug, but now they have a distribution of two, three, or however many drugs you want within them,” Johnson says.
In this study, the researchers first tested particles carrying just one drug: bortezomib, which is used to treat multiple myeloma, a cancer that affects a type of B cells known as plasma cells. Bortezomib is a proteasome inhibitor, a type of drug that prevents cancer cells from breaking down the excess proteins they produce. Accumulation of these proteins eventually causes the tumor cells to die.
When bortezomib is given on its own, the drug tends accumulate in red blood cells, which have high proteasome concentrations. However, when the researchers gave their bottlebrush prodrug version of the drug to mice, they found that the particles accumulated primarily in plasma cells because the bottlebrush structure protects the drug from being released right away, allowing it to circulate long enough to reach its target.
Using the bottlebrush particles, the researchers were also able to analyze many different drug combinations to evaluate which were the most effective.
Currently, researchers test potential drug combinations by exposing cancer cells in a lab dish to different concentrations of multiple drugs, but those results often don’t translate to patients because each drug is distributed and absorbed differently inside the human body.
“If you inject three drugs into the body, the likelihood that the correct ratio of those drugs will arrive at the cancer cell at the same time can be very low. The drugs have different properties that cause them to go to different places, and that hinders the translation of these identified synergistic drug ratios quite immensely,” Johnson says.
However, delivering all three drugs together in one particle could potentially overcome that obstacle and make it easier to deliver synergistic ratios. Because of the ease of creating bottlebrush particles with varying concentrations of drugs, the researchers were able to compare particles carrying different ratios of bortezomib and two other drugs used to treat multiple myeloma: an immunostimulatory drug called pomalidomide, and dexamethasone, an anti-inflammatory drug.
Exposing these particles to cancer cells in a lab dish revealed combinations that were synergistic, but these combinations were different from the synergistic ratios that had been identified using drugs not bound to the bottlebrush.
“What that tells us is that whenever you are trying to develop a synergistic drug combination that you ultimately plan to administer in a nanoparticle, you should measure synergy in the context of the nanoparticle,” Johnson says. “If you measure it for the drugs alone, and then try to make a nanoparticle with that ratio, you can’t guarantee it will be as effective.”
In tests in two mouse models of multiple myeloma, the researchers found that three-drug bottlebrushes with a synergistic ratio significantly inhibited tumor growth compared to the free drugs given at the same ratio and to mixtures of three different single-drug bottlebrushes. They also discovered that their bortezomib-only bottlebrushes were very effective at slowing tumor growth when given in higher doses. Although it is approved for blood cancers such as multiple myeloma, bortezomib has never been approved for solid tumors due to its limited therapeutic window and bioavailability.
“We were happy to see that the bortezomib bottlebrush prodrug on its own was an excellent drug, displaying improved efficacy and safety compared to bortezomib, and that has led us to pursue trying to bring this molecule to the clinic as a next-generation proteasome inhibitor,” Johnson says. “It has completely different properties than bortezomib and gives you the ability to have a wider therapeutic index to treat cancers that bortezomib has not been used in before.”
Johnson, Nguyen, and Yivan Jiang PhD ’19 have founded a company called Window Therapeutics, which is working on further developing these particles for testing in clinical trials. The company also hopes to explore other drug combinations that could be used against other types of cancer.
Johnson’s lab is also working on using these particles to deliver therapeutic antibodies along with drugs, as well as combining them with larger particles that could deliver messenger RNA along with drug molecules. “The versatility of this platform gives us endless opportunities to create new combinations,” he says.
The research was funded, in part, by the U.S. National Institutes of Health, the Leukemia and Lymphoma Society, the U.S. National Science Foundation, The Deshpande Center for Technological Innovation, and the Koch Institute Support (core) Grant from the National Cancer Institute.
Growing up in Idaho, Catherine Ji found herself with a lot of time to write.
“Idaho is a great environment for writing because it’s isolated and there’s a bunch of nature,” says Ji. “I wrote so much poetry — a lot of really messy poetry. I just loved it so much. It really defined my childhood.”
Now a senior majoring in physics and mathematics, Ji finds time to write despite a heavy class load and a variety of other activities. She has taken a number of literature and poetry classes during her undergraduate career, and has been recognized for her work as a winner of the Ilona Karmel Writing Prize in poetry and essay categories.
Beyond writing, Ji has done research in math, physics, and economics, sung in an a cappella group, co-chaired advocacy groups and math mentorship initiatives, and taught as a TA, volunteer mentor, and MIT-Italy GTL instructor.
Still, Ji wishes she could more thoroughly explore MIT’s offerings. “There are just too many cool things to do here, and never enough time,” she says.
In the town where she grew up, which was 40 minutes outside of Idaho’s rapidly growing capital, tensions between the left-leaning city and its more conservative surroundings were often on display. Ji says experiencing that contributed to her ability to adapt in a variety of environments.
“I think Idaho has also taught me a lot of skills that I’m thankful for. Because of the intense politics, you get good at adapting and actively coexisting,” she says. “How do I interact with people who have totally different worldviews? It’s one thing when such discussions on connectivity are theoretical or far-removed and there are structures in place to easily disengage. It’s another when you’re talking about most of your family, neighbors, friends, and teachers.”
The wide range of opinions and beliefs taught her how to advocate for herself and others.
“On some level, I just had to decide, ‘This is what I believe about all these things, and stand by it,’” Ji says.
She carried these skills with her to college. For example, as co-chair of the Council for Math Majors, an advocacy group focused on improving the undergraduate experience in MIT’s math department, she has worked on diversity, equity, and inclusion efforts for the department in collaboration with faculty and staff.
At MIT, “I had to learn a completely new toolbox of language and interacting with others, as well as institutions,” Ji says. “A big question is how to self-advocate in an institution that supports DEI but is super decentralized. It’s ‘mens et manus’ [MIT’s motto of “mind and hand”] in that we have to be the change we want to see; however, higher connectivity on all levels would make life easier.”
These lessons also helped her find her place in the Logarhythms, MIT’s oldest a cappella group that was historically all-male until accepting its first non-male member in 2018. Ji was the second.
“As with any institution that experiences this type of change, it’s a process. At first, I felt very complicated about being in the group, but I’m extremely happy that I stuck with it,” she says.
Ji didn’t only stick with it; in the fall semester, she ran rehearsals and organized concerts as co-director. She also served as president and is now an historian going into her final undergraduate semester.
Pursuing interdisciplinary interests
Whether exploring poetry, scientific research, or challenging historical norms to make student life at MIT more inclusive, Ji is deliberate about doing things her own way.
Navigating MIT with a wide array of interests, Ji jokingly describes her college experience to be “like gradient descent,” referring to the way a machine-learning algorithm tries different paths to a solution and “descends” each time as it gets closer to the correct answer. “I ask myself: What do I enjoy most in the moment? And then I move in that direction,” she says.
Ji’s experiences led her to biophysics, where she has studied the mechanics of polymers with Professor Jörn Dunkel and the dynamics of starfish embryo crystals with Associate Professor Nikta Fakhri. “The same math can model anything from the glug-glug of a draining bottle to solar flares to epidemics,” she explains. “All of these systems in nature are connected through math, which is really cool.”
Considering her roles as a scientist, poet, and community member, she says: “These realms of my life are closely intertwined, and this is the type of life that I want.”
“In a way, we all try to make meaning and joy through connectivity, from mathematicians to writers,” she says. “This is a lesson felt perhaps most strongly during a pandemic. At the risk of being overly theoretical, I see literature, making a cappella music, advocacy, mentorship, and physical applied math research as all contributing to the same purpose of prescribing meaning.”
For her final undergraduate semester, Ji has many things to look forward to. There are exciting classes to take like 18.212 (Algebraic Combinatorics), a potential tour with the Logs in Tokyo, and an upcoming poetry publication in Electric Lit, a digital literary magazine. In between, she’s taking time to breathe and enjoy the people who have made her time at MIT so special.
“I think MIT students strongly believe that putting your all into every area of life, from being a good housemate to a jazz ensemble player to a fire-spinner, naturally leads to better innovators or ‘makers.’ They do their best. And I really admire that.”
Following graduation, Ji plans to pursue a PhD in physics or applied math. In parallel, she wants to find avenues to advise science policy as well as lead advocacy and outreach efforts. And she hopes that her publication in Electric Lit is one of many.
Mary Louise Morrissey, whose career at MIT spanned 45 years, including her service as director of the Information and Special Events Center, passed away peacefully on Jan. 17 at the age of 95.
Morrissey joined the MIT community in 1950, working in the Registrar’s Office. At the time, all student transcripts were handwritten in India ink, and it was perhaps there — if not during her student days in Catholic schools — that she developed her precise and elegant script. Another responsibility of the registrar’s staff at the time was to notarize students’ draft-board and other documents and — during the Cold War era — faculty signatures on loyalty oaths. One student at the time, Paul Gray ’54, SM ’55, ScD ’60, recalled that he was not the only one who found multiple reasons to have things notarized by the charming but also somewhat intimidating Morrissey. A few decades later, she organized his inauguration as the 14th president of MIT.
Over the years, her responsibilities grew, as the Information Center became part of the greater President’s Office. The center coordinated the logistics for faculty-sponsored conferences and major Institute events. It oversaw the student-guided campus tours for prospective students and greeted visitors from throughout the country and overseas. In time, the center developed an array of services to facilitate the appointment and support of scholars and researchers from all over the world — what is today the International Scholars Office.
Morrissey was perhaps best known at MIT as the impresario of countless Institute events, including new building dedications, milestone anniversaries, presidential inaugurations, memorial services, and the annual commencement exercises. David Ferriero, recently retired archivist of the United States, worked with Morrissey on events when he was a librarian at MIT: “Mary Morrissey brought style, grace, and magic to commencement and inauguration ceremonies, new building dedications, and the quality of Institute life for students, faculty, and staff.”
Among the many innovations she orchestrated was moving the Commencement exercises in 1979 from Rockwell Cage (where they had been held since 1927) outdoors to Killian Court — despite skepticism and predictions of disaster from many quarters. Her main faculty partner in this endeavor was Professor Emeritus Gerald L. Wilson, former dean of the School of Engineering and chair of the Commencement Committee at the time, who became a good friend, as did so many who worked with her. “Mary had very high standards for everything she did on behalf of the Institute,” Wilson says. “She did so with a seriousness of purpose that never blurred her wonderful sense of humor. For many, her throaty laugh belied that stern look when 'Miss Mary' was not amused. She was a great friend to many. She certainly was one of the special ones that made being in the MIT community not just a vocation, but a wonderful experience.”
An exemplar of the iron fist in a velvet glove, she was able to persuade people from all corners of the Institute — faculty members, carpenters, students, administrators, custodians, and trustees — to share her vision of how best to mark significant occasions in the life of the Institute, and come together to make it reality. In recognition of her many contributions to MIT, she was made an honorary member of the Alumni Association in 1995, the same year in which she retired.
Through it all, she was a gifted mentor to individuals from many quarters of the Institute, but none more so than to her immediate staff.
Gayle Gallagher, recently retired as executive director of MIT Institute Events, says, "Mary was one of my dearest friends for more than 40 years. We shared an abundance of laughs, a few tears, and great affection and respect always. In the early years of my career, I most valued her role as mentor for nearly 15 years. Mary was politically astute, a tremendous judge of character, and a fearless and imaginative events creator. The lessons I learned were innumerable and she was my most stalwart champion. She gave me confidence I would never have known without her nurturing and guiding hand. I loved her dearly.”
Members of the MIT engineering faculty receive many awards in recognition of their scholarship, service, and overall excellence. The School of Engineering periodically recognizes their achievements by highlighting the honors, prizes, and medals won by faculty working in our academic departments, labs, and centers.
In 1941, the National Academy of Sciences appointed a committee to assess the use of gas turbine engines — which use heat released during fuel combustion to produce thrust for propulsion — in aviation. The group of luminaries concluded that due to the temperature limitations of existing materials, gas turbines did not have much of a future in propelling airplanes.
However, “Unknown to the committee, the first jet engine was already successfully run in Germany in 1940: the Junkers Jumo,” says Professor Zoltán Spakovszky, director of the MIT Gas Turbine Laboratory (GTL) and the T.A. Wilson Professor in Aeronautics and Astronautics. Although the committee had correctly identified the temperature limitations, “the German engineers and designers redefined the problem and introduced turbine cooling,” he explains.
The Junkers Jumo, the world’s first turbojet engine in production, was put in operation during World War II, while separately, Sir Frank Whittle had been leading progress on the development of the turbojet engine in Great Britain. With the United States falling behind Germany and Britain in developing turbojet engines, Professor Jerome C. Hunsaker had the vision of establishing a laboratory dedicated to gas turbine propulsion at MIT. Hunsaker, an aviation pioneer in his own right and member of the National Advisory Committee of Aeronautics, gathered funds and support from six U.S. industries and the U.S. Navy to get started.
On Oct. 7, 1947, the GTL, led by Professor Edward Story Taylor as its founding director, officially launched with all major U.S. aviation and aircraft companies of that time in attendance at the opening ceremony. Over the course of 75 years, the GTL, now housed in the Department of Aeronautics and Astronautics, has been at the cutting edge of applied research. It continues to do so by delivering “new perspectives on integration of propulsion systems with new aircraft concepts and high-impact collaborative projects cutting across disciplines,” Spakovszky says.
To describe the work of laboratory, Professor Edward Greitzer, a former GTL director and the H.N. Slater Professor in Aeronautics and Astronautics, quotes former prime minister of Singapore Lee Kuan Yew, who spoke of not “perfecting the known,” but rather reaching for the unknown. “That’s what we have always tried to do at the GTL,” Greitzer says. “We do our best to think strategically about things we could do that would not only be intellectually interesting but would also have an impact.”
The GTL “is still going very strong, tackling new and different challenges,” Spakovszky says. “Today, we’re not only working on the propulsion system, jet engines, and power plants, we’re also working on integrating jet engines into aircraft and on forward-looking challenges like electrification of aviation.”
In the early years, projects focused on one discipline and addressed one specific problem, Greitzer points out, but today’s GTL works on “problems with larger scope and scale, cutting across disciplines and sometimes organizations.” For example, a project working on a conceptual design of a fuel-efficient aircraft led to a test in a large wind tunnel, at a NASA facility.
Equally important, Spakovszky adds, is the lab’s focus on industry. True to its roots, the GTL continues to work on “projects that don’t just go into theses and sit on the shelf; they actually move the needle and start with real applications in industry,” he says. Super-high-pressure ratio compressors for carbon sequestration and ultrashort aeroengine inlets to reduce fuel burn are examples of the many different industry-focused projects that the GTL has worked on.
Fostering excellence, passion, and collaboration
Over the past three-quarters of a century, close to 500 students have called the GTL their academic home. In addition to being steeped in academic rigor, students came away with technical communication skills, says Borislav “Bobby” Sirakov SM ’01, PhD ’04. The ability, “developed at the GTL, to summarize and explain a complex topic in simple words has served me well in my career,” he states.
Andras Kiss ’13, SM ’15, PhD ’21, worked at the GTL from his sophomore year in Course 16 until he completed his doctoral degree in aerospace engineering. “The first thing that Zolti or Ed would say when you wrote a report or made a presentation was 'answer the Heilmeier questions [a series of questions addressing risks, costs and more] in plain language,’ Kiss laughs, “It was all about distilling your work into very approachable, clear language so you know exactly what you’re trying to do. Otherwise it’s very easy to hide behind detail.”
Kiss has many fond memories of the GTL, including the time he spent designing the electrical and fuel systems for a turbofan engine and having it work smoothly after 18 months of effort. “It was a real thrill, seeing the engine start up for the first time,” he remembers.
Phil Mullan SM ’59, ME ’62, ScD ’64, who majored in mechanical engineering at MIT while working in the GTL, loved the academic rigor. “The lab environment was very invigorating for me because the other research assistants were really bright people,” Mullan says. “They came from different backgrounds and had lots of good ideas to share and were always willing to help.” He remembers looking forward to the midmorning and midafternoon coffee breaks in the library.
According to Spakovszky, ideas that pushed the boundaries of so-called conventional wisdom have been an important differentiator of the GTL. Two research initiatives in this regard have been Micro-Engines, shirt-button-sized gas turbine engines for portable power made using computer chip manufacturing, and the Silent Aircraft Initiative, focused on the conceptual design of an aircraft whose noise would be imperceptible outside airport boundaries.
This approach was also evident when approaching challenges earlier in the lab's history, like finding the original drive system for the De Laval wind tunnel and air system. Not to be confused with MIT AeroAstro’s Wright Brothers Wind Tunnel, the reconfigurable De Laval wind tunnel is located within Building 31 and provides air to various test facilities. “The logistical challenge was getting a motor to run the compressor,” Spakovszky says. “It turns out that the USS Halibut, a Gato-class submarine, had run ashore and was decommissioned in New Hampshire in 1945. Eddie Taylor bought the motor drive system out of that submarine and put it here in 1947. We operated that equipment until we renovated a few years ago (in 2017) and now have a new electric motor to drive the De Laval air system.”
According to Greitzer, there have been pleasant technological surprises along the way since he joined the GTL (from Pratt and Whitney) in 1977. One of these was the Silent Aircraft Initiative. “My expectation was that we’d have a trade-off of performance — fuel burn for noise,” Greitzer says. “But we found that if you think about opening up the design of the aircraft … you don’t have to make those compromises and you can get both less noise and improved fuel burn performance.”
Celebrating a roaring future
In his 1947 welcome speech inaugurating the lab, Taylor said: “It hardly seems necessary to stress the growing importance of the gas turbine as a prime mover.” In the speech he also referred to the GTL as a “a new laboratory specifically designed for research in problems encountered in gas turbines.”
On Oct. 7, 2022, 75 years later to the day, Spakovszky addressed a room full of more than 140 alumni, industry members, and academic luminaries who came together from all over the world to return to campus and celebrate the historic milestone for the GTL. Mullan — with his grandson, an engineer with Pratt and Whitney, in tow — Sirakov, and Kiss were among the laboratory alumni in attendance.
“The challenges are different now compared to 75 years ago, but the way we do research and the way we collaborate has not changed. Today, we’re looking at electrifying aviation and working with new fuels like hydrogen,” Spakovszky says. “The bottom line is that our name has not changed, we’re still the Gas Turbine Lab, but we’re doing more than gas turbines, and addressing different aspects of the field.”
The lab’s invigorating environment and a passion for gas turbine technology were on full display at the celebrations where attendees were delighted to catch up with old friends and mentors and go down memory lane while touring the GTL’s renovated facilities to learn more about the latest research and even view a Junkers Jumo 004 engine on display, an emblem of the field’s history embedded in the present.
While 2022 marked an important milestone in the history of the GTL, Sirakov believes that the lab will always be at the forefront of advancements.
“I was very happy to see so many new test rigs and experimental projects going on,” he says. “I am proud of the long history of the lab, the long list of contributions to the field, and the powerful beginning with all the aerospace leaders attending the [launch]. It’s remarkable that after so many years the MIT GTL lab is still very relevant to the fields of aeronautics, space, and automotive [research] and to all of the new and exciting horizons like electrification and clean energy.”
Greitzer agrees. “The feeling that came through at the 75th anniversary celebration is that the Gas Turbine Lab is a special place, it’s distinctive and it’s different,” Greitzer says. “We continue on our voyage of discovery to learn the unknown.”
Measuring activity in the human brain remains one of the greatest challenges in science and medicine. Despite recent technological advances in areas such as imaging and nanoscience, researchers still struggle to accurately detect cognition. Currently, functional magnetic resonance imaging (MRI) is used to measure brain activity, but this method requires the patient to lie still in a large, noisy, and expensive apparatus. A portable and noninvasive method is needed to illuminate how the brain functions within a more natural setting while performing daily life activities.
In 2013, the National Institutes of Health launched an initiative to encourage more research into neuroscience by funding projects in key areas of the field. One such project is led by the Massachusetts General Hospital (MGH) Athinoula A. Martinos Center for Biomedical Imaging, in collaboration with MIT Lincoln Laboratory and Boston University, to develop a high-performance brain imaging method that can monitor cerebral blood flow with more accuracy than ever before. The brain regulates blood flow differently depending on what mental and physical tasks a person is doing. Accurately mapping cerebral blood flow with a portable system would give researchers insight into cognition.
"This new method is called time-domain diffuse correlation spectroscopy (TD-DCS) and it works by transmitting laser light to and from the brain using fiber optics," says Jonathan Richardson, a research team member from Lincoln Laboratory’s Advanced Imager Technology Group. The method will be integrated into a system resembling a cap that has 64 transmission and 192 receive points that are organized into groups called optodes, spaced 1 centimeter apart to cover nearly the entire scalp. "The light diffuses from the transmitter of each optode, bounces off of hemoglobin in red blood cells, and returns to several of the surrounding receivers."
Blood cells are constantly moving, and the faster they move, the more rapidly the intensity of the returning light signal will fluctuate. Researchers can use the rate of that fluctuation to measure blood flow velocity.
Early on in the program, the team worked to optimize the wavelength of light being used for the pulses. Tissue and blood absorb and scatter light differently at different wavelengths. These effects can swallow a light signal such that nothing bounces back to the receivers. Through modeling and measurements, they determined that a 1,064-nanometer laser could safely deliver almost 11 times more photons and reach a 25 percent deeper region of the brain than the shorter wavelengths that are used currently. In addition, a 1,064-nanometer laser is readily produced by commercial pulsed fiber laser technology.
To make the receivers sensitive to faint light signals returning from deep in the brain, the team used a custom detector technology, developed at Lincoln Laboratory, called Geiger-mode avalanche photodiodes (GmAPDs).
"The GmAPD is a device that can give a fast electrical pulse in response to a single photon," says researcher Brian Aull. "We can detect that pulse and measure its time of occurrence digitally, which makes the detector exquisitely sensitive. We need that because most of the light scatters off in random directions and only a fraction of it scatters in the right direction to reach the detector."
Twenty years in the making, these GmAPDs have been involved in many critical programs at Lincoln Laboratory. This project is the first medical application of GmAPDs, which are coupled with a novel readout integrated circuit (ROIC) that was designed specifically for this use.
"After many years of development and demonstration for astronomy and national security applications, we are pleased to see our detector technology make an impact in medicine," says Erik Duerr, leader of the Advanced Imager Technology Group.
The GmAPD technology also addresses the issue of irrelevant returning light signals — in particular, those that bounce off cells in the scalp rather than blood in the brain — that can confound results.
"They are gated," says Aull, “meaning they can be turned on only during selected time intervals.” Photons bouncing off of the scalp will return to the optodes more quickly than those coming from deeper in the brain. "By using a delayed turn-on, the system can ignore these early photons."
So far, the team successfully demonstrated TD-DCS at 1,064 nanometers in human subjects using individual commercial detectors. They are now focused on implementing and testing the ROIC and GmAPD integrated detector. In 2024, they plan to transition the system to the MGH team, who will then integrate it with their laser system.
"This technology has immediate clinical relevance to the diagnosis and tracking of traumatic brain injuries and can monitor brain perfusion during field-forward trauma care," says Richardson. "In the longer term, we hope this technology can assist in treatment of psychiatric conditions such as post-traumatic stress disorder, depression, and suicidality among soldiers as well."
Members of the Department of Mathematics community — including faculty, students, and alumni — were recognized for their achievements at the recent 2023 Joint Mathematics Meetings in Boston.
Professor Tom Mrowka and his Harvard University collaborator Peter Kronheimer received the 2023 Leroy P. Steele Prize for Seminal Contribution to Research, awarded by the American Mathematical Society (AMS), for their joint paper “Gauge Theory for Embedded Surfaces.”
The AMS’ 2023 Joseph L. Doob Prize was awarded to Professor Bjorn Poonen for his 2017 book “Rational Points on Varieties,” in the series “Graduate Studies in Mathematics.” The citation called his book “an essential reference for anybody who wishes to apply the tools and techniques of modern algebraic geometry to the venerable area of Diophantine equations.”
Professor Scott Sheffield and former MIT postdoc and instructor Jason P. Miller, now at the University of Cambridge, have been awarded the AMS’ 2023 Leonard Eisenbud Prize in Mathematics and Physics. They earned this award “for their monumental series of papers on Liouville Quantum Gravity.”
CLE Moore instructor Jia Shi received the Association for Women in Mathematics’ Dissertation Prize for her thesis that “proves major results on two separate topics in fluid mechanics, a hard classical field.”
The association also honored two MIT seniors who were nominees for the Alice T. Schafer Prize for excellence in mathematics by an undergraduate woman: Anqi Li was the 2023 runner-up, and Ilani Axelrod-Freed earned an honorable mention.
Letong Carina Hong ’22, currently at Oxford University as a Rhodes Scholar for China, received the 2023 AMS-MAA-SIAM Frank and Brennie Morgan Prize for Outstanding Research in Mathematics by an Undergraduate Student, for proving a number of results and solving conjectures in combinatorics, number theory, and probability.
The Ruth Lyttle Satter Prize in Mathematics went to Rutgers University Associate Professor Nataša Šešum PhD ’04 and Panagiota Daskalopoulos of Columbia University “for groundbreaking work in the study of ancient solutions to geometric evolution equations.”
J-PAL North America has announced that Matthew “Matt” Notowidigdo ’03, MEng ’04, PhD ’10, professor of economics at the University of Chicago Booth School of Business, is joining Amy Finkelstein as co-scientific director of the organization, replacing Lawrence “Larry” Katz.
Katz is stepping down after nearly 10 years of supporting the growth and development of J-PAL North America, having worked closely with Finkelstein to launch the regional office of J-PAL in 2013. He will continue his role as a co-chair of J-PAL North America’s Worker Prosperity Initiative and as an active affiliated researcher.
“Over the past decade, J-PAL North America has had a major impact in expanding the field’s capacity to conduct randomized evaluations on key policy issues in the region,” says Katz. “I am excited to pass the baton to new leadership at the scientific director level. Matt is an excellent choice to join Amy in guiding the organization in its mission over the next decade.”
“We truly would not be where we are today without Larry’s efforts, including helping to launch the office, advising the launch of multiple research initiatives, reviewing over a hundred proposals as a review board member, expanding our research network, supporting high-quality evidence synthesis and policy outreach, and contributing to major fundraising successes,” says Finkelstein. “I now look forward to collaborating with Matt in providing scientific direction to the next phase of J-PAL North America’s work and driving forward a new generation of evidence.”
In the co-scientific director role, Notowidigdo, alongside Finkelstein, will support and guide J-PAL North America in developing rigorous research on economic mobility and advise as the organization builds a focus on diversity, equity, and inclusion and develops a research agenda for racial equity.
“Matt holds a deep commitment to our mission, a history of public service at J-PAL, excitement about randomized evaluations, and many ideas around how to strengthen our diversity, equity, and inclusion work,” says Vincent Quan, co-executive director of J-PAL North America. “His thought leadership in these areas will be extremely valuable to the growth of our organization.”
A J-PAL affiliated researcher since 2015, Notowidigdo’s work spans labor market issues, social protection policies, and health interventions. He has experience conducting a number of randomized evaluations and working effectively with both nonprofit and government partners. Highlights of his evaluations in partnership with J-PAL North America include an influential study on increasing SNAP take-up in Pennsylvania and an ongoing study on managed care organizations with the South Carolina Department of Health and Human Services.
Notowidigdo also serves as a co-chair for the Worker Prosperity Initiative and has been an active reviewer of proposals, a participant at J-PAL North America events, and a reviewer for the Invited Researcher Search Committee. He has provided guidance on J-PAL North America’s strategies to build a diverse economics pipeline, informed by his direct experience in serving as a mentor for both the Russell Sage Foundation’s Pipeline Grants Competition and for junior economists with the Committee on the Status of Women in the Economics Profession.
“I've had the pleasure of working with Matt on numerous research projects during his time as a J-PAL affiliate,” says Laura Feeney, co-executive director of J-PAL North America. “In addition to the experience he brings to his new role, I am excited for the mentorship he will bring to our staff, his enthusiasm for the work, and his commitment to supporting economics scholars from all backgrounds."
Notowidigdo holds a BS in economics, a BS in computer engineering, an MEng in computer science, and a PhD in economics from MIT. In addition to his work with J-PAL, he has served as a co-editor of the American Economic Journal: Economic Policy, a research associate at the National Bureau of Economics Research, and an associate editor at the Quarterly Journal of Economics.
In science and technology, there has been a long and steady drive toward improving the accuracy of measurements of all kinds, along with parallel efforts to enhance the resolution of images. An accompanying goal is to reduce the uncertainty in the estimates that can be made, and the inferences drawn, from the data (visual or otherwise) that have been collected. Yet uncertainty can never be wholly eliminated. And since we have to live with it, at least to some extent, there is much to be gained by quantifying the uncertainty as precisely as possible.
Expressed in other terms, we’d like to know just how uncertain our uncertainty is.
That issue was taken up in a new study, led by Swami Sankaranarayanan, a postdoc at MIT’s Computer Science and Artificial Intelligence Laboratory (CSAIL), and his co-authors — Anastasios Angelopoulos and Stephen Bates of the University of California at Berkeley; Yaniv Romano of Technion, the Israel Institute of Technology; and Phillip Isola, an associate professor of electrical engineering and computer science at MIT. These researchers succeeded not only in obtaining accurate measures of uncertainty, they also found a way to display uncertainty in a manner the average person could grasp.
Their paper, which was presented in December at the Neural Information Processing Systems Conference in New Orleans, relates to computer vision — a field of artificial intelligence that involves training computers to glean information from digital images. The focus of this research is on images that are partially smudged or corrupted (due to missing pixels), as well as on methods — computer algorithms, in particular — that are designed to uncover the part of the signal that is marred or otherwise concealed. An algorithm of this sort, Sankaranarayanan explains, “takes the blurred image as the input and gives you a clean image as the output” — a process that typically occurs in a couple of steps.
First, there is an encoder, a kind of neural network specifically trained by the researchers for the task of de-blurring fuzzy images. The encoder takes a distorted image and, from that, creates an abstract (or “latent”) representation of a clean image in a form — consisting of a list of numbers — that is intelligible to a computer but would not make sense to most humans. The next step is a decoder, of which there are a couple of types, that are again usually neural networks. Sankaranarayanan and his colleagues worked with a kind of decoder called a “generative” model. In particular, they used an off-the-shelf version called StyleGAN, which takes the numbers from the encoded representation (of a cat, for instance) as its input and then constructs a complete, cleaned-up image (of that particular cat). So the entire process, including the encoding and decoding stages, yields a crisp picture from an originally muddied rendering.
But how much faith can someone place in the accuracy of the resultant image? And, as addressed in the December 2022 paper, what is the best way to represent the uncertainty in that image? The standard approach is to create a “saliency map,” which ascribes a probability value — somewhere between 0 and 1 — to indicate the confidence the model has in the correctness of every pixel, taken one at a time. This strategy has a drawback, according to Sankaranarayanan, “because the prediction is performed independently for each pixel. But meaningful objects occur within groups of pixels, not within an individual pixel,” he adds, which is why he and his colleagues are proposing an entirely different way of assessing uncertainty.
Their approach is centered around the “semantic attributes” of an image — groups of pixels that, when taken together, have meaning, making up a human face, for example, or a dog, or some other recognizable thing. The objective, Sankaranarayanan maintains, “is to estimate uncertainty in a way that relates to the groupings of pixels that humans can readily interpret.”
Whereas the standard method might yield a single image, constituting the “best guess” as to what the true picture should be, the uncertainty in that representation is normally hard to discern. The new paper argues that for use in the real world, uncertainty should be presented in a way that holds meaning for people who are not experts in machine learning. Rather than producing a single image, the authors have devised a procedure for generating a range of images — each of which might be correct. Moreover, they can set precise bounds on the range, or interval, and provide a probabilistic guarantee that the true depiction lies somewhere within that range. A narrower range can be provided if the user is comfortable with, say, 90 percent certitude, and a narrower range still if more risk is acceptable.
The authors believe their paper puts forth the first algorithm, designed for a generative model, which can establish uncertainty intervals that relate to meaningful (semantically-interpretable) features of an image and come with “a formal statistical guarantee.” While that is an important milestone, Sankaranarayanan considers it merely a step toward “the ultimate goal. So far, we have been able to do this for simple things, like restoring images of human faces or animals, but we want to extend this approach into more critical domains, such as medical imaging, where our ‘statistical guarantee’ could be especially important.”
Suppose that the film, or radiograph, of a chest X-ray is blurred, he adds, “and you want to reconstruct the image. If you are given a range of images, you want to know that the true image is contained within that range, so you are not missing anything critical” — information that might reveal whether or not a patient has lung cancer or pneumonia. In fact, Sankaranarayanan and his colleagues have already begun working with a radiologist to see if their algorithm for predicting pneumonia could be useful in a clinical setting.
Their work may also have relevance in the law enforcement field, he says. “The picture from a surveillance camera may be blurry, and you want to enhance that. Models for doing that already exist, but it is not easy to gauge the uncertainty. And you don’t want to make a mistake in a life-or-death situation.” The tools that he and his colleagues are developing could help identify a guilty person and help exonerate an innocent one as well.
Much of what we do and many of the things happening in the world around us are shrouded in uncertainty, Sankaranarayanan notes. Therefore, gaining a firmer grasp of that uncertainty could help us in countless ways. For one thing, it can tell us more about exactly what it is we do not know.
Angelopoulos was supported by the National Science Foundation. Bates was supported by the Foundations of Data Science Institute and the Simons Institute. Romano was supported by the Israel Science Foundation and by a Career Advancement Fellowship from Technion. Sankaranarayanan's and Isola’s research for this project was sponsored by the U.S. Air Force Research Laboratory and the U.S. Air Force Artificial Intelligence Accelerator and was accomplished under Cooperative Agreement Number FA8750-19-2- 1000. MIT SuperCloud and the Lincoln Laboratory Supercomputing Center also provided computing resources that contributed to the results reported in this work.
Since its founding in 1998, Vecna Technologies has developed a number of ways to help hospitals care for patients. The company has produced intake systems to respond to Covid-19 patient surges, prediction systems to manage health complications in maternity wards, and telepresence robots that have allowed sick people to stay connected with friends and loved ones.
The differences among those products have also led to a number of transformations and spinoffs, including material handling company Vecna Robotics and the health care nonprofit VecnaCares. Vecna Technologies co-founders Deborah Noel Theobald ’95 and Daniel Theobald ’95, SM ’98 say each of those pivots has been driven by a desire to build a robotics company that makes a positive impact on the world.
“We knew we wanted to do robotics and do something good in the world,” Deborah says of the team’s mindset. “We founded Vecna thinking, ‘How can these new web technologies influence and improve health care?’ That’s the arc MIT set me on and something I’ve been excited to pursue ever since.”
“A fun ride”
As a child, Deborah Theobald wanted to be an astronaut. The desire led her to MIT, which had one of the few aerospace engineering programs for undergraduates. She got interested in the health care industry while studying the health effects of long-term space exploration with Professor Dava Newman, the Apollo Program Professor of Astronautics at MIT who is now also the director of the Media Lab.
Deborah also met Daniel Theobald at MIT. Daniel had been building robots since he was a child and was majoring in mechanical engineering.
The two began thinking about starting a company, and Daniel even applied to the MIT $100K Entrepreneurship Competition (then the $10K) with a rough idea for a robotics company.
For their master’s degrees, Deborah went to the University of Maryland to continue studying health effects in space, while Daniel stayed at MIT, working on several robotics projects. When Daniel graduated in 1998, Vecna was born.
From day one, the company had a policy of paying employees to spend 10 percent of their work week doing community service.
“We found that our focus on giving back benefited the business in so many ways that it was absolutely, unambiguously the right thing to do,” Daniel says. “For one, it was a self-filtering mechanism. People joined Vecna who believed in giving back and wanted to be part of something that was socially responsible. And we found those are also the people that make amazing employees.”
The founders got their first big break with a government contract to build a health care portal that allowed patients, managers, and providers to communicate and share documents. The contract also provided flexibility for the founders to explore other avenues for the business.
The pair went on to earn a number of government grants for one-off projects, some of which blossomed into successful commercial products. Another grant tasked them with building models to help hospitals predict and manage hospital acquired infections (HAIs), which kill tens of thousands of people in the U.S. each year. The resulting tool ended up being deployed in about 100 hospitals.
“At the time, people were using spreadsheets to pull in data from different systems … and trying to comprehend what kind of infection it was,” Deborah says, noting that doctors usually start infected patients on general antibiotics before they can classify the disease. “Our tool allowed them to pull that information together faster, reducing their stay in hospitals — and all the trauma and pain that goes with that — by weeks.”
The company’s next product was a patient registration system that used kiosks to streamline patient intake at hospitals. During the Covid-19 pandemic, Vecna turned the platform into a text-based check-in service for clinics. The service is being used by thousands of hospitals today.
Subsequent mobile versions of that system have been used to deliver medication, allow doctors to hold virtual consultations, and even help immunocompromised students to attend school virtually and avoid isolation.
Vecna’s emphasis on community service led the team to explore ways to apply the company’s technologies in low-resource settings, leading to the creation of the company’s nonprofit arm, VecnaCares.
In 2014, VecnaCares brought their VGo mobile robot to Liberia and Sierra Leone to help with the Ebola response, allowing doctors to see patients without going through a time-consuming decontamination process. The company’s patient intake software was also used to register and manage patients with Ebola and other diseases.
VecnaCares has since partnered with groups including the International Rescue Committee, the International Committee of the Red Cross, International Medical Corp, and the Special Olympics for a variety of projects. It’s also honed its algorithms to help low-resource hospitals manage staff shortages in maternity wards, helping nurses focus their attention on the babies and mothers most at risk of complications.
“One of the places we’re deployed has 10,000 births a year, so at any one time there may be 40 women laboring in that hospital, which has one operating room for all C-sections,” Deborah explains. “Our tool can intake women, do an assessment, and notify clinicians if someone’s high risk and needs checking-in on. It leads to better outcomes and helps manage some of the complications that have led to a high rate of infant and maternal mortality in these areas.”
After years of robot development and commercialization, the founders decided their robots may be better suited for warehouses than health care. In 2017, Daniel spun out Vecna Robotics to focus exclusively on robotics for industrial settings like manufacturing, logistics, and order fulfillment.
“We’ve sort of done four different growths and exits,” Deborah explains. “It’s been a fun ride.”
Continuing to innovate
As it nears the 25th anniversary of its founding, Vecna Technologies is far from finished. Its leaders believe the firm’s products and expertise can play a significant role in the burgeoning home health care and extended care industries, helping patients stay out of hospitals while remaining safe.
“As we look at the aging population, that burden of care is really going to fall on family members as well as [health care organizations],” Deborah says. “I’d love to be able to provide better tools for them to care for loved ones, which is often unpaid and unrecognized.”
Later this year, the company will release an inexpensive home care robot that can move autonomously or by remote control to help care for people struggling with diseases like Alzheimer’s. The robots will be part of Vecna’s “Be There Network” that health care providers can use to provide care for large numbers of patients despite staff shortages.
“Now you can see and hear and feel like you’re actually there to more seamlessly interact with the environment,” Deborah says. “We see that as the wave of the future now that people have begun to embrace telepresence. There are so many uses for this robot. People keep coming up with more ideas as they catch the vision.”
No matter what the future holds for Vecna — whose motto is “Better technology, better world” — the founders say the company will continue exploring new applications where its technologies could make a real difference in people’s lives.
Visitors to MIT’s AgeLab in the Center for Transportation and Logistics are greeted silently by a shiny mannequin in a jumpsuit and chunky red goggles, standing a little ominously in a glass-walled studio. While the mannequin itself cuts a striking appearance, it’s the accessories under the jumpsuit that are the real attraction: a collection of weights and bungie cords, some unwieldy gloves, and a pair of Crocs with blocks of foam glued to the bottom of them — as well as the red goggles.
Taken together, these items make up AGNES, which stands for the Age Gain Now Empathy System. AGNES is an empathy and research tool designed by the MIT AgeLab to simulate for the wearer some of what it may feel like to live in one’s early 80s with a few chronic health conditions. The weights approximate muscle loss, the bungies the reduction of range of motion and flexibility that can affect the joints with age. The foam-platform Crocs simulate the erosion of balance, and the heavy, awkward gloves evoke the loss of tactile sensation. Finally, the red goggles simulate a range of impairments to vision, from impaired acuity to diabetic retinopathy.
“The development of AGNES has been a collaborative and iterative effort by MIT researchers and students over time,” says Joe Coughlin, director of the MIT AgeLab. “It began with a neck brace and elastic bands that we used to better understand the challenges of automobile ingress and egress for older users. Today, we use AGNES to give researchers and students a taste of the friction, frustration, and fatigue that older adults often experience.” AGNES is also a key instructional tool in Coughlin’s course 11.547/SCM 287 (Global Aging & The Built Environment).
Putting on AGNES approximates the effect of aging a person in a moment, an experience that is startling, if not overwhelming, for many people. But the suit is not just for shock value. It is used to help designers, engineers, executives, and helping professionals understand a little better the physical and social world as a version of their future self, so that they can design better products and services for older users. AGNES has been used globally to inform the design of public transportation systems, retail environments, medical devices, and product packaging.
Since her debut in 2006, aside from unnerving and instructing MIT students, research sponsors, and visitors, AGNES has ventured into show business, making an appearance with the popular YouTubers “The Try Guys” and playing a prominent role in the PBS documentary “Fast Forward.”
Most recently, AGNES appears in a new documentary series titled “Limitless with Chris Hemsworth,” which is produced by National Geographic and streaming on Disney+. The series is directed by Darren Aronofsky (“Noah,” “Black Swan”) and stars actor Chris Hemsworth.
Most of the six-episode series features Hemsworth braving extreme situations like freezing temperatures, high altitudes, and days of fasting, which serve as avenues to understand — and overcome — the limits of the human body. But the final episode of the documentary takes a surprising turn. Instead of attempting to surmount yet another challenge, Hemsworth is tasked with accepting the unavoidable limits of human potential: the fact that he will get older, that he will physically decline, and that he will die.
The producers of “Limitless” go to surprising lengths to immerse Hemsworth in the world of older age. An entire retirement community, complete with residents and aides, is constructed for Hemsworth to live in. An ID card with an aged version of his face is printed for him to carry on a lanyard. And Hemsworth is outfitted in a custom version of AGNES so that he can experience how his body might change when he gets older.
Early in the episode, Hemsworth complains about wearing AGNES — “this suit sucks, by the way” — and attempts to overcome its limitations by brute effort. He loses a game of ping pong and exhausts himself in an aerobics class. But by the second day of wearing the suit, he realizes that there is no way for him to defy the limits that AGNES brings. Instead, he begins to learn how to adapt to them and finally to accept them, and to allow himself to depend more on other people.
“We often talk about AGNES as an empathy tool, but the suit’s appearances in popular media also suggest its power as a tool for storytelling about increased longevity,” says Taylor Patskanick, a researcher at the MIT AgeLab who is involved in research and organizing workshops using AGNES. The striking appearance of the suit, and the vivid responses that its wearers have to the experience, can prompt surprise, discussion, and reflection in an audience, leading to new forms of understanding.
“I like to say that the story is the oldest technology that we have. Stories engage us, instruct us, and give us a sense of what’s possible,” says Coughlin. “AGNES is not necessarily destiny, but it does provide users and observers with insights into what their older self might be. It’s an opportunity for us to imagine our future today so that we might have a better life tomorrow.”
Imagine living and sharing your passions with hundreds of MIT students while experiencing the fun and singular energy of living in an on-campus residence. Sound fun? Welcome to the Residential Scholars Program.
The program is managed by the Office of Residential Education in the Division of Student Life (DSL), which is committed to developing welcoming, safe, and inclusive living and learning communities. “Programs like the Residential Scholars foster intellectual, physical, spiritual, and personal development by connecting students and community members who bring new perspectives on life, art, and careers to campus through diverse and enriching experiential learning opportunities in the unique and exciting culture that is MIT,” says Judy Robinson, senior associate dean for residential education and executive director for DSL strategic initiatives.
There are four residential scholars at MIT — one in New Vassar, one in International House, and two in Simmons Hall. They hail from all over the world and bring unique experiences that aid the development of community and individual student growth within their residence halls.
Andrea Bolnick from Johannesburg, South Africa, saw the opportunity to become a residential scholar in New Vassar House as a way to immerse herself in MIT and interact with students. She is also the managing director of Ikhayalami Development Services and visiting scholar at the Leventhal Center for Applied Urbanism.
This semester, Bolnick plans to show the sci-fi film “District 9” about apartheid in South Africa and bring in someone to speak about emerging financial trends.
She and her 5-year-old daughter enjoy meeting students, especially at dinnertime. “I think the students are amazing. Most of them play sports, are active, and are sociable and appear light-hearted — this is quite extraordinary considering that they are also so smart,” says Bolnick.
Jeff Behrens grew up in Framingham, Massachusetts, and is the CEO of LabShares Newton, a biotech incubator. Like Bolnick, he finds that working with MIT students is the best perk of being a residential scholar.
“Getting to know some of the students and renewing our optimism in the leaders and innovators of the future has been the best part of the role. I’ve enjoyed setting up small, intimate events where you can talk in more detail to a few students. It’s well worth it!” says Behrens.
Behrens held a resume review session and also invited MIT alums Dan Nussbaum ’85, SM ’88, PhD ’93, a former member of the MIT Blackjack team, and Warren Katz ’86, a software entrepreneur, to speak on separate occasions.
From Sao Paulo, Brazil, Hannah Arcuschin Machado is an MIT SPURS fellow who applied for the role so she could become immersed in MIT's community and exchange knowledge and experience with students.
Machado quickly engaged New House residents by offering Portuguese-speaking brunches and dinners. She also invited residents to watch World Cup matches together.
“This spring, I will offer a workshop to fix the abandoned bikes in the New House Bike Park and transform them into shared bikes for the collective use of everyone — since you can bike almost everywhere in Cambridge. There is a New House undergrad resident that is excited by this idea and plans to create a GPS system to locate the bikes and manage their use,” she says.
Charles Evavold and his partner Isabella Fraschilla are both from Georgia and work in the Cambridge area. Evavold runs a research lab as a principal investigator and fellow of the Ragon Institute of MGH, MIT, and Harvard. Fraschilla is a postdoc at MIT studying cancer biology at the Koch Institute for Integrative Cancer Research.
As the residential scholars in Simmons Hall, they both have enjoyed getting to know MIT students on a personal level and are grateful to be immersed in the MIT community.
“Watching the transition from new social groups to established friend groups over the course of the semester has been wonderful,” says Evavold. Fraschilla adds, “Through a resume workshop and informal conversations, I have found many students have humanities and artistic talents beyond my assumption of stellar STEM skills.”
The two have joined study breaks and cake decorating and plant potting events. They hope to host a movie night and an ice-skating event this semester.
All of the scholars agree that meeting and reaching out to students early in the semester is the key to success for future residential scholars. The group has also found it helpful to bounce ideas off of each other and meet for lunch or coffee on occasion. They will continue to cultivate and facilitate opportunities for learning that promote creative thinking, leadership, citizenship, inclusion, and a commitment to lifelong learning and, of course, fun.
Email the Residential Scholars Program to learn more or to apply when positions become available.
The name Sybil has its origins in the oracles of Ancient Greece, also known as sibyls: feminine figures who were relied upon to relay divine knowledge of the unseen and the omnipotent past, present, and future. Now, the name has been excavated from antiquity and bestowed on an artificial intelligence tool for lung cancer risk assessment being developed by researchers at MIT's Abdul Latif Jameel Clinic for Machine Learning in Health, Mass General Cancer Center (MGCC), and Chang Gung Memorial Hospital (CGMH).
Lung cancer is the No. 1 deadliest cancer in the world, resulting in 1.7 million deaths worldwide in 2020, killing more people than the next three deadliest cancers combined.
"It’s the biggest cancer killer because it’s relatively common and relatively hard to treat, especially once it has reached an advanced stage,” says Florian Fintelmann, MGCC thoracic interventional radiologist and co-author on the new work. “In this case, it’s important to know that if you detect lung cancer early, the long-term outcome is significantly better. Your five-year survival rate is closer to 70 percent, whereas if you detect it when it’s advanced, the five-year survival rate is just short of 10 percent.”
Although there has been a surge in new therapies introduced to combat lung cancer in recent years, the majority of patients with lung cancer still succumb to the disease. Low-dose computed tomography (LDCT) scans of the lung are currently the most common way patients are screened for lung cancer with the hope of finding it in the earliest stages, when it can still be surgically removed. Sybil takes the screening a step further, analyzing the LDCT image data without the assistance of a radiologist to predict the risk of a patient developing a future lung cancer within six years.
In their new paper published in the Journal of Clinical Oncology, Jameel Clinic, MGCC, and CGMH researchers demonstrated that Sybil obtained C-indices of 0.75, 0.81, and 0.80 over the course of six years from diverse sets of lung LDCT scans taken from the National Lung Cancer Screening Trial (NLST), Mass General Hospital (MGH), and CGMH, respectively — models achieving a C-index score over 0.7 are considered good and over 0.8 is considered strong. The ROC-AUCs for one-year prediction using Sybil scored even higher, ranging from 0.86 to 0.94, with 1.00 being the highest score possible.
Despite its success, the 3D nature of lung CT scans made Sybil a challenge to build. Co-author Peter Mikhael, an MIT PhD student in electrical engineering and computer science, and affiliate of Jameel Clinic and the MIT Computer Science and Artificial Intelligence Laboratory (CSAIL), likened the process to “trying to find a needle in a haystack.” The imaging data used to train Sybil was largely absent of any signs of cancer because early-stage lung cancer occupies small portions of the lung — just a fraction of the hundreds of thousands of pixels making up each CT scan. Denser portions of lung tissue are known as lung nodules, and while they have the potential to be cancerous, most are not, and can occur from healed infections or airborne irritants.
To ensure that Sybil would be able to accurately assess cancer risk, Fintelmann and his team labeled hundreds of CT scans with visible cancerous tumors that would be used to train Sybil before testing the model on CT scans without discernible signs of cancer.
MIT electrical engineering and computer science PhD student Jeremy Wohlwend, co-author of the paper and Jameel Clinic and CSAIL affiliate, was surprised by how highly Sybil scored despite the lack of any visible cancer. “We found that while we [as humans] couldn’t quite see where the cancer was, the model could still have some predictive power as to which lung would eventually develop cancer,” he recalls. “Knowing [Sybil] was able to highlight which side was the most likely side was really interesting to us.”
Co-author Lecia V. Sequist, a medical oncologist, lung cancer expert, and director of the Center for Innovation in Early Cancer Detection at MGH, says the results the team achieved with Sybil are important “because lung cancer screening is not being deployed to its fullest potential in the U.S. or globally, and Sybil may be able to help us bridge this gap.”
Lung cancer screening programs are underdeveloped in regions of the United States hardest hit by lung cancer due to a variety of factors. These range from stigma against smokers to political and policy landscape factors like Medicaid expansion, which varies from state to state.
Moreover, many patients diagnosed with lung cancer today have either never smoked or are former smokers who quit over 15 ago — traits that make both groups ineligible for lung cancer CT screening in the United States.
“Our training data consisted only of smokers because this was a necessary criterion for enrolling in the NLST,” Mikhael says. “In Taiwan, they screen nonsmokers, so our validation data is expected to contain people who didn’t smoke, and it was exciting to see Sybil generalize well to that population.”
“An exciting next step in the research will be testing Sybil prospectively on people at risk for lung cancer who have not smoked or who quit decades ago,” says Sequist. “I treat such patients every day in my lung cancer clinic and it’s understandably hard for them to reconcile that they would not have been candidates to undergo screening. Perhaps that will change in the future.”
There is a growing population of patients with lung cancer who are categorized as nonsmokers. Women nonsmokers are more likely to be diagnosed with lung cancer than men who are nonsmokers. Globally, over 50 percent of women diagnosed with lung cancer are nonsmokers, compared to 15 to 20 percent of men.
MIT Professor Regina Barzilay, a paper co-author and the Jameel Clinic AI faculty lead, who is also a member of the Koch Institute for Integrative Cancer Research, credits MIT and MGH’s joint efforts on Sybil to Sylvia, the sister to a close friend of Barzilay and one of Sequist’s patients. "Sylvia was young, healthy and athletic — she never smoked,” Barzilay recalls. “When she started coughing, neither her doctors nor her family initially suspected that the cause could be lung cancer. When Sylvia was finally diagnosed and met Dr. Sequist, the disease was too advanced to revert its course. When mourning Sylvia's death, we couldn't stop thinking how many other patients have similar trajectories.”
This work was supported by the Bridge Project, a partnership between the Koch Institute at MIT and the Dana-Farber/Harvard Cancer Center; the MIT Jameel Clinic; Quanta Computer; Stand Up To Cancer; the MGH Center for Innovation in Early Cancer Detection; the Bralower and Landry Families; Upstage Lung Cancer; and the Eric and Wendy Schmidt Center at the Broad Institute of MIT and Harvard. The Cancer Center of Linkou CGMH under Chang Gung Medical Foundation provided assistance with data collection and R. Yang, J. Song and their team (Quanta Computer Inc.) provided technical and computing support for analyzing the CGMH dataset. The authors thank the National Cancer Institute for access to NCI’s data collected by the National Lung Screening Trial, as well as patients who participated in the trial.
People at MIT know “mens et manus,” or “mind and hand,” as the school motto. But it’s also a good framework for early childhood education. Kids often learn best when they’re allowed to explore the environment around them, building models of the world by picking things up and moving them around.
Back in 2014, that insight was the inspiration for a Media Lab project that designed new learning environments. Four years later, that project became the inspiration for a startup called Learning Beautiful.
Learning Beautiful makes tactile materials to inspire hands-on learning for kids between the ages of 3 and 9. The materials, which are designed to explain simple concepts in computer science, promote child-driven, physical learning that aligns with the Montessori method of education.
“For young children, being able to build and then experience with their hands is so important,” Learning Beautiful founder Kim Smith Claudel SM ’17 says. “I don’t think I need to do much convincing about the importance of limiting screen time for kids. I focus more on the positive things we can give to children, and I think giving them these sensorial, tactile materials is a developmentally enriching opportunity.”
The company’s materials include things like binary cards and pixel boards made from sustainably sourced wood, cork, and canvas. To date, Learning Beautiful has sold over 2,000 materials to schools and libraries and trained about 500 teachers to guide learning activities.
Smith Claudel believes the concepts illuminated by the materials are a great primer to more advanced computer science education later in life.
“If we think about how we scaffold learning for subjects like reading and writing and math, we have all these things in place to build a strong foundation in early childhood to help progression in these subjects,” Smith Claudel says. “But there really wasn’t something that did the same thing for computer science.”
From project to product
In 2013, Smith Claudel began collaborating with Sepandar Kamvar, who was a professor of media arts and sciences at MIT and the director of the Social Computing group at the Media Lab. After Smith Claudel worked on a show Kamvar was organizing in Sweden, he asked her to join his lab as a research scientist.
“His vision was to bring together a lot of different people,” Smith Claudel recalls. “My background was art and design, and we had architects, computer scientists, videographers, biologists, educators, and philosophers.”
The diverse team soon began exploring alternative approaches to education, partially inspired by Kamvar’s own struggles to find a good prekindergarten school for his child. Their ideas coalesced into the first of what the team called “Wildflower Schools” described as open-source learning environments inspired by the century-old Montessori learning method that emphasizes self-directed learning activities based on children’s natural interests.
The schools served as test beds for experiments in teaching and learning, with the project advertised as “blurring the boundaries between home-schooling and institutional schooling, between scientists and teachers, between schools and the neighborhoods around them.”
“I worked in the school for a year doing art projects with the kids, and that was my crash course in Montessori education,” Smith Claudel says.
The first school sparked interest in the Cambridge community, so the group opened more. Each one featured aspects of the research going on in Kamvar’s lab, including small-scale agriculture projects and experiments with different learning materials — even some of the teachers were members of the lab.
“The idea was to test different things with the community and cultivate this research within the school,” Smith Claudel says. “It became a link to what we were doing in the Media Lab.”
Smith Claudel became enamored with some of the materials being used in the classrooms and intrigued by the research showing young children learn more effectively by physically interacting with their environment. She officially enrolled in the Media Lab as a graduate student in 2015.
After hearing frustration from MIT computer scientists that too many educational materials were screen-based and focused solely on coding, Smith Claudel and others in her lab worked with them to build materials that demonstrated different computational concepts.
“The children are very helpful because it either works or it doesn’t work,” Smith Claudel says. “Feedback from teachers is also helpful because either they understand it or they don’t, and if they don’t then we’ve failed.”
Smith Claudel went through the MIT DesignX accelerator run through the School of Architecture, where they started hearing from people who wanted copies of their research materials for their classrooms and libraries.
“DesignX shifted the whole paradigm of how I thought about the research, and turned it into ‘How can we take this solid foundation and spin it into a business?’” Smith Claudel says.
As Smith Claudel neared graduation in 2017, she got her first order for materials from the Chicago Public Library, which had seen her work develop at the Media Lab. She still remembers juggling finishing her master’s work with building each of those early sets by hand in MIT’s makerspaces, using CNC machines and spending hours sanding, painting, and gluing.
The company’s first series of materials includes pixel boards that demonstrate how computers represent images through 1s and 0s and a “binary tree” that introduces the concept of data structures as the child connects the branches and builds the tree.
“With the binary tree, a 2- or 3-year-old might start playing using what we call sensorial exploration,” Smith Claudel says. “What they’re doing is experimenting and discovering through a physical process. They’re starting to see things fit together. They’re starting to build something, getting a sense of balance. They’re also noticing the pieces are different shapes, different colors, so they’re building these models. They’re learning from that whole process.”
Learning Beautiful also provides support and educational materials for teachers.
“We learned early on you can’t just hand someone new materials and expect them to be comfortable with an unfamiliar subject area, so we created children’s books, a full curriculum, lesson plans, and then training,” Smith Claudel says.
When school shut down during the pandemic, the team developed instructions for at-home learning activities and offered them for free to parents and teachers. The slowdown also gave them time to plan their next series of materials, which will be released over the next year.
“A pause can be a healthy thing,” Smith Claudel says. “Especially in the beginning [of the pandemic], our attitude was what could we make that would be helpful right now?”
Helping everyone learn beautiful
Lately the company has been focusing on scaling its teacher training efforts, including by building a virtual training program.
Last fall, after partnering with a school district in Iowa, Learning Beautiful hosted a training workshop with 250 teachers, giving them each their own set of materials to bring back to their classrooms.
Smith Claudel also believes her materials can help a broader set of children than computer-based learning programs. Learning Beautiful has even begun conversations with schools in other countries that don’t have access to electricity.
“I think accessibility is really important on a few different levels,” Smith Claudel says. “We all learn differently, so to provide a variety of different kinds of learning opportunities is crucial. We use sound and touch in our materials, and we’ve had early conversations about working with blind children, because the materials are not solely dependent on sight.”
Learning Beautiful’s next products will expand beyond computer science to encourage ecological thinking, helping children understand environmental systems around them and their schools.
As the company’s sales grow, it’s developed a program where proceeds from sales to one community can help fund donations to communities with fewer resources.
“Hands-on learning is effective for all of us,” Smith Claudel says. “For children, most of their brain development is happening between zero and three, so physical interaction is so rich — understanding spatial relationships, how to hold things, how to use their body, how to take inputs from the world and process them in their minds. That’s what MIT’s ‘mind and hand’ motto is about: this connection between the physical experiences and what we’re building in our mind.”
In patients with Huntington’s disease, neurons in a part of the brain called the striatum are among the hardest-hit. Degeneration of these neurons contributes to patients’ loss of motor control, which is one of the major hallmarks of the disease.
Neuroscientists at MIT have now shown that two distinct cell populations in the striatum are affected differently by Huntington’s disease. They believe that neurodegeneration of one of these populations leads to motor impairments, while damage to the other population, located in structures called striosomes, may account for the mood disorders that are often seen in the early stages of the disease.
“As many as 10 years ahead of the motor diagnosis, Huntington’s patients can experience mood disorders, and one possibility is that the striosomes might be involved in these,” says Ann Graybiel, an MIT Institute Professor, a member of MIT’s McGovern Institute for Brain Research, and one of the senior authors of the study.
Using single-cell RNA sequencing to analyze the genes expressed in mouse models of Huntington’s disease and postmortem brain samples from Huntington’s patients, the researchers found that cells of the striosomes and another structure, the matrix, begin to lose their distinguishing features as the disease progresses. The researchers hope that their mapping of the striatum and how it is affected by Huntington’s could help lead to new treatments that target specific cells within the brain.
This kind of analysis could also shed light on other brain disorders that affect the striatum, such as Parkinson’s disease and autism spectrum disorder, the researchers say.
Myriam Heiman, an associate professor in MIT’s Department of Brain and Cognitive Sciences and a member of The Picower Institute for Learning and Memory, and Manolis Kellis, a professor of computer science in MIT’s Computer Science and Artificial Intelligence Laboratory (CSAIL) and a member of the Broad Institute of MIT and Harvard, are also senior authors of the study. Ayano Matsushima, a McGovern Institute research scientist, and Sergio Sebastian Pineda, an MIT graduate student, are the lead authors of the paper, which appears in Nature Communications.
Huntington’s disease leads to degeneration of brain structures called the basal ganglia, which are responsible for control of movement and also play roles in other behaviors, as well as emotions. For many years, Graybiel has been studying the striatum, a part of the basal ganglia that is involved in making decisions that require evaluating the outcomes of a particular action.
Many years ago, Graybiel discovered that the striatum is divided into striosomes, which are clusters of neurons, and the matrix, which surrounds the striosomes. She has also shown that striosomes are necessary for making decisions that require an anxiety-provoking cost-benefit analysis.
In a 2007 study, Richard Faull of the University of Auckland discovered that in postmortem brain tissue from Huntington’s patients, the striosomes showed a great deal of degeneration. Faull also found that while those patients were alive, many of them had shown signs of mood disorders such as depression before their motor symptoms developed.
To further explore the connections between the striatum and the mood and motor effects of Huntington’s, Graybiel teamed up with Kellis and Heiman to study the gene expression patterns of striosomal and matrix cells. To do that, the researchers used single-cell RNA sequencing to analyze human brain samples and brain tissue from two mouse models of Huntington’s disease.
Within the striatum, neurons can be classified as either D1 or D2 neurons. D1 neurons are involved in the “go” pathway, which initiates an action, and D2 neurons are part of the “no-go” pathway, which suppresses an action. D1 and D2 neurons can both be found within either the striosomes and the matrix.
The analysis of RNA expression in each of these types of cells revealed that striosomal neurons are harder hit by Huntington’s than matrix neurons. Furthermore, within the striosomes, D2 neurons are more vulnerable than D1.
The researchers also found that these four major cell types begin to lose their identifying molecular identities and become more difficult to distinguish from one another in Huntington’s disease. “Overall, the distinction between striosomes and matrix becomes really blurry,” Graybiel says.
The findings suggest that damage to the striosomes, which are known to be involved in regulating mood, may be responsible for the mood disorders that strike Huntington’s patients in the early stages of the disease. Later on, degeneration of the matrix neurons likely contributes to the decline of motor function, the researchers say.
In future work, the researchers hope to explore how degeneration or abnormal gene expression in the striosomes may contribute to other brain disorders.
Previous research has shown that overactivity of striosomes can lead to the development of repetitive behaviors such as those seen in autism, obsessive compulsive disorder, and Tourette’s syndrome. In this study, at least one of the genes that the researchers discovered was overexpressed in the striosomes of Huntington’s brains is also linked to autism.
Additionally, many striosome neurons project to the part of the brain that is most affected by Parkinson’s disease (the substantia nigra, which produces most of the brain’s dopamine).
“There are many, many disorders that probably involve the striatum, and now, partly through transcriptomics, we’re working to understand how all of this could fit together,” Graybiel says.
The research was funded by the Saks Kavanaugh Foundation, the CHDI Foundation, the National Institutes of Health, the Nancy Lurie Marks Family Foundation, the Simons Foundation, The JPB Foundation, the Kristin R. Pressman and Jessica J. Pourian ’13 Fund, and Robert Buxton.
If you mention Leslie Regan’s name to any alum of MIT’s mechanical engineering graduate program, their face will break into a smile. For nearly five decades, Regan’s kind, caring presence was a mainstay for thousands of mechanical engineering students. Now, after 47 years, Regan can reflect back on an impactful journey as she begins her retirement.
Regan joined MIT’s staff in September 1974. She started as an administrative assistant supporting three faculty members, including Professor David Wormley. It was in that role that Regan first discovered her love of working with students.
After Wormley became department head in 1982, Regan joined him in the Department of Mechanical Engineering (MechE) headquarters. She spent four years working alongside Wormley, learning the ins and outs of the Institute’s policies and procedures. In 1986, there was an opening for the academic administrator in the graduate office and Regan seized an opportunity to get back to where she felt she belonged: working with students.
“My heart was with the students. To me it was never really a job, it was a lifetime mission. I really loved seeing them grow and helping them if they had issues,” says Regan. As academic administrator, Regan helped countless graduate students navigate the often-tumultuous experiences of securing funding, passing qualifying exams, and finishing their theses. For many students, especially international students, she was a safe harbor far from home.
One such student was Shangzhi Wu SM ’81, MBA ’84, PhD ’85. Wu was the first student from mainland China accepted into MIT after the Cultural Revolution. The week he was set to leave for Cambridge, Wu and his wife quickly got married in Beijing since there was so much uncertainty about their visas. The wedding took place in a civil office without a ceremony or reception. Two days later, Wu boarded a flight to Boston.
Over the next two years, Wu developed a close friendship with Regan. When his wife finally joined him in Cambridge, Regan organized the wedding party they never had. “When Leslie learned that we never had an official wedding ceremony, she organized a party with many friends at her house,” says Wu. “We really appreciated Leslie’s friendship and kindness to organize that in her house.”
From Regan’s perspective, she gained just as much from her relationships with international students as they did from her.
“I just feel like we can make an impact no matter where we are. People would always ask me ‘Do you travel?’ And I say that I do. I travel all over the country and all over the world through the eyes of our students,” Regan says. “There’s no place I haven’t been. They come into MIT and I get to know their culture.”
As the years went on, Leslie became an institution within MechE. She ushered the graduate program from a time when master’s and doctoral theses were written on typewriters and copied on mimeographs, through to the digital age and the pandemic, which saw students defending their PhD theses on Zoom.
Over the years, her office on the first floor of MIT’s Building 3 became a shrine to the many students she cared for, helped, and mentored over the years. The walls were decorated with mementos sent from every corner of the world, and even beyond.
On one of his trips to the International Space Station, alumnus and former NASA astronaut Mike Massimino SM ’88, PhD ’92 had the opportunity to bring an item to space as a gift when he returned to Earth. He chose a T-shirt with MechE’s name on it in honor of the department’s — and by extension Leslie’s — impact on his life.
“Leslie made every one of us feel like family. I wanted to fly something for the department specifically, and Leslie had a lot to do with that because the department took care of me. They educated me and gave me these great opportunities,” says Massimino. “A lot of people come in and out of your life to make these things possible, and Leslie was one of them.”
After decades of keeping the graduate office running, Regan is adjusting to the slower pace of retirement. She has been cleaning and organizing boxes and boxes of keepsakes, including dozens of cherished letters and emails from alumni. She also is very active in her church community and keeps busy with volunteering.
When she reflects on her career at MIT, Regan only has positive things to say about each student she worked with.
“I never felt like this was a job. I felt this is really the place I should be and I loved every minute of it,” says Regan. “Some people would say to me, ‘There must be one student you didn’t like.’ And I would say never, not one. Sure, their needs are different, but I cared for every single student who went through our program.”
Sulafa Zidani is an assistant professor in the Comparative Media Studies Program whose work focuses on digital culture: the social, political, and cultural dynamics in which technology operates and the role it plays in transnational power. She is working on her first book, which focuses on multilinguistic memes and centers the creators of these memes. By looking into the lives and work of these global meme makers, the book tells the story of globalization in the digital era as it is expressed through untranslatability, feelings, and humor.
Zidani spoke with SHASS Communications about her research and her experience at MIT.
Q: The intersections of the technological and the creative are central to your work. How does the interdisciplinary nature of an MIT education lend itself to this kind of humanistic and technological thinking?
A: This is my second year at MIT, and in that time I have I seen how aligned MIT’s educational approach here is with my work. Studying technology and culture from a critical transnational perspective, I believe it is crucial to be part of conversations that cross between disciplines and departments, conversations where the designers, engineers, and entrepreneurs of tomorrow are learning critical perspectives and confronting ethical dilemmas. MIT is the place where these conversations are taking place. I have students in fields ranging from business to engineering, biology, to astrophysics. Together, we examine some of the urgent questions our society is facing: What characterizes online cultures today? How did we get to this place where misinformation and racism spread widely online? What makes some online spaces more connective or more divisive?
One of the aspects I appreciate the most in MIT education is the push for thinking about solutions. When confronting difficult questions, it is understandable that many of us get stuck in the challenges of the present. Yet, MIT students demonstrate in my class that they have a forward-looking approach that continually returns to questions like: How do we make things better? How do we create better content or invent better technology without replicating existing problems? In this way, MIT is a great place to enhance prolific thinking around technology and culture.
Q: Online civic engagement is a central part of so many people’s political experience and exposure — and you examine this engagement on a transnational scale! How do you approach online fieldwork, especially engagement on issues that rely on local nuances and humor, remixed with transnational culture? How are various kinds of power at play in those exchanges?
A: In my work on transnational online civic engagement, context is key. Oftentimes in research, when we scale up to large datasets or transnational case studies, we compromise on a deep and intimate knowledge of the data. In my work, I maintain a global scale while also centering knowledge of the context that the data is stemming from. I research internet content in languages that I speak, from places in which I have lived, and cultures which I know through my heritage knowledge, lived experience, and my education.
Context helps us understand research data better. For example, in my paper on mashup and remix culture in the Middle East, I examine memes and videos in Arabic. My understanding of the language and culture helps me identify what hides between the lines, that might be based on the accent being used, terms that are specific to a region, to a generation, or to a subculture, and the political backdrop that online content might be conversing with. This is especially important in humor, which relies on unstated aspects to provoke laughter.
Another reason that context is central in my work is because symbols we use in one place might not hold the same meaning in a different context. One example which demonstrates this difference is the image of Pepe the Frog, which I discuss in an interview with journalist Nancy Guan. In the context of the U.S., the image of Pepe the Frog is mostly used in misogynist and antisemitic alt-right memes. However, in Hong Kong, the face of Pepe the Frog shows up in memes, graffiti, and protest signs as a representation of pro-democracy activists.
Researching everyday communication is fascinating. To truly understand power in these kinds of mundane-yet-creative forms of content, especially to understand the nuance around them, we must spend time getting to know the history that led to them and the events and culture occurring around them.
Q: You’ve written on decolonizing syllabi in media, communications, and cultural studies. What does that process of inclusive pedagogy look like in the classroom? What has been your experience of bringing that pedagogy to MIT?
A: My approach to inclusive pedagogy is centered around embracing differences, which I interpret as inviting our differences into the classroom rather than pushing them out in favor of consensus. In the first few weeks of class, as everyone is getting to know one another better, I pay special attention to the knowledge and experience that students already have. I then guide students to connect concepts to their existing knowledge, be that their life experience or knowledge they acquired in other classes. I’ve found that this method enriches our class discussions and leads to a deeper understanding of the course material.
In terms of pedagogy, working with MIT students has been an intellectual delight. I am regularly amazed at the variety of skills students bring to the class and their eagerness to engage in discussion. Students add in perspectives based on their interests, their majors and minors, and their desired career paths. They do this by raising questions that concern them, like “How do we build a better social media environment?” and by sharing their experiences being part of social movements or fan cultures. Since I aim to bring a global and critical perspective into my classes, MIT’s diverse student body means that students can also add some contextual knowledge or draw our attention to important relevant events taking place in other places around the world.
Many media studies courses, especially foundational and introductory courses, have traditionally favored perspectives that center what we call the “Western” world, especially scholarship produced by white European and North American men. Such syllabi present this type of knowledge as the canon, which then puts knowledge produced by women, Indigenous people, Black people, and other people of color — both in and outside of “the West” — as less important. Many academics have tried to address this by adding one week in their syllabus with readings from underrepresented perspectives, but I think this type of solution can cement the view of these perspectives as marginal. I have written more about this in my article in Media, Culture & Society where I suggest actual strategies for creating more inclusive syllabi and classrooms.
While one syllabus or one class cannot alone rid us of the shadows of colonialism that we have inherited in higher education, I believe that centering our students is a great place to start.
Pushing a shovel through snow, planting an umbrella on the beach, wading through a ball pit, and driving over gravel all have one thing in common: They all are exercises in intrusion, with an intruding object exerting some force to move through a soft and granular material.
Predicting what it takes to push through sand, gravel, or other soft media can help engineers drive a rover over Martian soil, anchor a ship in rough seas, and walk a robot through sand and mud. But modeling the forces involved in such processes is a huge computational challenge that often takes days to weeks to solve.
Now, engineers at MIT and Georgia Tech have found a faster and simpler way to model intrusion through any soft, flowable material. Their new method quickly maps the forces it would take to push, wiggle, and drill an object through granular material in real-time. The method can apply to objects and grains of any size and shape, and does not require complex computational tools as other methods do.
“We now have a formula that can be very useful in settings where you have to check through lots of options as fast as possible,” says Ken Kamrin, professor of mechanical engineering at MIT.
“This is especially useful for applications such as real-time path-planning for vehicles traveling through vast deserts and other off-road terrains, that cannot wait for existing slower simulation methods to decide their path,” adds Shashank Agarwal SM ’19, PhD ’22.
Kamrin and Agarwal detail their new method in a study appearing this week in the journal Proceedings of the National Academy of Sciences. The study also includes Daniel I. Goldman, professor of physics at Georgia Tech.
A fluid connection
In order to know how much to push on an object to move it through sand, one could go grain by grain, using discrete element modeling, or DEM — an approach that systematically calculates each individual grain’s motion in response to a given force. DEM is precise but slow, and it can take weeks to fully solve a practical problem involving just a handful of sand. As a faster alternative, scientists can develop continuum models, which simulate granular behavior in generalized chunks, or grain groupings. This more simplified approach can still generate a detailed picture of how grains flow, in a way that can shave a weeks-long problem down to days or even hours.
“We wanted to see if we could do even better than that and cut that process down to seconds,” Agarwal says.
The team looked to previous work by Goldman. In 2014, he was studying how animals and robots move through dry, granular material such as sand and soil. In looking for ways to quantitatively describe their movements, he found he could do so with a quick relationship that was originally meant to describe fluid swimmers.
The formulation, Resistive Force Theory (RFT), works by considering an object’s surface as a collection of small plates. (Imagine representing a sphere as a soccer ball.) As an object moves through a fluid, each plate experiences a force, and RFT claims that the force on each plate depends only on its local orientation and movement. The equation takes all this into account, along with the fluid’s individual characteristics, to ultimately describe how the object as a whole moves through a fluid.
Surprisingly, Goldman found this simple approach was also accurate when applied to granular intrusion. Specifically, it predicted the forces lizards and snakes exert to slither through sand, as well as how small, legged robots walk over soil. The question, Kamrin says, was why?
“It was this weird mystery why this theory, which was originally derived for moving through viscous fluid, would even work at all in granular media, which has completely different flow behavior,” he says.
Kamrin took a closer look at the math and found a connection between RFT and a continuum model he had derived to describe granular flow. In other words, the physics checked out, and RFT could indeed be an accurate way to predict granular flow, in a simpler and faster way than conventional models. But there was one big limitation: The approach was mainly workable for two-dimensional problems.
To model intrusion using RFT, one needs to know what will happen if one moves a plate every which way possible — a task that is manageable in two dimensions, but not in three. The team then needed some shortcut to simplify 3D’s complexity.
In their new study, the researchers adapted RFT to 3D by adding an extra ingredient to the equation. That ingredient is a plate’s twist angle, measuring how plate orientation changes as the entire object is rotated. When they incorporated this extra angle, in addition to a plate’s tilt and direction of motion, the team had enough information to define the force acting on the plate as it moves through a material in 3D. Importantly, by exploiting the connection to continuum modeling, the resulting 3D-RFT is generalizable, and can be easily recalibrated to apply to many dry granular media on Earth, and even on other planetary bodies.
The researchers demonstrated the new method using a variety of three-dimensional objects, from simple cylinders and cubes to more complex bunny- and monkey-shaped geometries. They first tiled the objects, representing them each as a collection of hundreds to thousands of tiny plates. Then they applied the tweaked RFT formula to each individual plate and calculated the forces that would be needed over time to drill each plate, and ultimately the entire object, down through a bed of sand.
“For more wacky objects, like the bunny, you can imagine having to consistently shift your loads to keep drilling it straight down,” Kamrin says. “And our method can even predict those little wiggles, and the distribution of force all around the bunny, in less than a minute.”
The new approach provides a fast and accurate way to model granular intrusion, which can be applied to a host of practical problems, from driving a rover through Martian soil, to characterizing the movement of animals through sand, and even predicting what it would take to uproot a tree.
“Can I predict how hard it is to uproot natural plants? You might want to know, is this storm going to knock over this tree?” Kamrin says. “Here is a way to get an answer fast.”
This research was supported, in part, by the Army Research Office, the U.S. Army DEVCOM Ground Vehicle Systems Center, and NASA.
Taking what they learned conceptually about artificial intelligence and machine learning (ML) this year, students from across the Greater Boston area had the opportunity to apply their new skills to real-world industry projects as part of an experiential learning opportunity offered through Break Through Tech AI at MIT.
Hosted by the MIT Schwarzman College of Computing, Break Through Tech AI is a pilot program that aims to bridge the talent gap for women and underrepresented genders in computing fields by providing skills-based training, industry-relevant portfolios, and mentoring to undergraduate students in regional metropolitan areas in order to position them more competitively for careers in data science, machine learning, and artificial intelligence.
“Programs like Break Through Tech AI gives us opportunities to connect with other students and other institutions, and allows us to bring MIT’s values of diversity, equity, and inclusion to the learning and application in the spaces that we hold,” says Alana Anderson, assistant dean of diversity, equity, and inclusion for the MIT Schwarzman College of Computing.
The inaugural cohort of 33 undergraduates from 18 Greater Boston-area schools, including Salem State University, Smith College, and Brandeis University, began the free, 18-month program last summer with an eight-week, online skills-based course to learn the basics of AI and machine learning. Students then split into small groups in the fall to collaborate on six machine learning challenge projects presented to them by MathWorks, MIT-IBM Watson AI Lab, and Replicate. The students dedicated five hours or more each week to meet with their teams, teaching assistants, and project advisors, including convening once a month at MIT, while juggling their regular academic course load with other daily activities and responsibilities.
The challenges gave the undergraduates the chance to help contribute to actual projects that industry organizations are working on and to put their machine learning skills to the test. Members from each organization also served as project advisors, providing encouragement and guidance to the teams throughout.
“Students are gaining industry experience by working closely with their project advisors,” says Aude Oliva, director of strategic industry engagement at the MIT Schwarzman College of Computing and the MIT director of the MIT-IBM Watson AI Lab. “These projects will be an add-on to their machine learning portfolio that they can share as a work example when they’re ready to apply for a job in AI.”
Over the course of 15 weeks, teams delved into large-scale, real-world datasets to train, test, and evaluate machine learning models in a variety of contexts.
In December, the students celebrated the fruits of their labor at a showcase event held at MIT in which the six teams gave final presentations on their AI projects. The projects not only allowed the students to build up their AI and machine learning experience, it helped to “improve their knowledge base and skills in presenting their work to both technical and nontechnical audiences,” Oliva says.
For a project on traffic data analysis, students got trained on MATLAB, a programming and numeric computing platform developed by MathWorks, to create a model that enables decision-making in autonomous driving by predicting future vehicle trajectories. “It’s important to realize that AI is not that intelligent. It’s only as smart as you make it and that’s exactly what we tried to do,” said Brandeis University student Srishti Nautiyal as she introduced her team’s project to the audience. With companies already making autonomous vehicles from planes to trucks a reality, Nautiyal, a physics and mathematics major, shared that her team was also highly motivated to consider the ethical issues of the technology in their model for the safety of passengers, drivers, and pedestrians.
Using census data to train a model can be tricky because they are often messy and full of holes. In a project on algorithmic fairness for the MIT-IBM Watson AI Lab, the hardest task for the team was having to clean up mountains of unorganized data in a way where they could still gain insights from them. The project — which aimed to create demonstration of fairness applied on a real dataset to evaluate and compare effectiveness of different fairness interventions and fair metric learning techniques — could eventually serve as an educational resource for data scientists interested in learning about fairness in AI and using it in their work, as well as to promote the practice of evaluating the ethical implications of machine learning models in industry.
Other challenge projects included an ML-assisted whiteboard for nontechnical people to interact with ready-made machine learning models, and a sign language recognition model to help disabled people communicate with others. A team that worked on a visual language app set out to include over 50 languages in their model to increase access for the millions of people that are visually impaired throughout the world. According to the team, similar apps on the market currently only offer up to 23 languages.
Throughout the semester, students persisted and demonstrated grit in order to cross the finish line on their projects. With the final presentations marking the conclusion of the fall semester, students will return to MIT in the spring to continue their Break Through Tech AI journey to tackle another round of AI projects. This time, the students will work with Google on new machine learning challenges that will enable them to hone their AI skills even further with an eye toward launching a successful career in AI.
After decades of fundamental scientific and drug discovery research, Alzheimer’s disease has remained inscrutable and incurable, with a bare minimum of therapeutic progress. But in a new review article in Nature Neuroscience, MIT scientists write that by employing the new research capability of single-cell profiling, the field has rapidly achieved long-sought insights with strong potential for both explaining Alzheimer’s disease and doing something meaningful about it. By analyzing this new evidence, for instance, the authors show that the disease’s disruptions converge on five main areas of cellular function, or “pathways,” in each of five major brain cell types.
Single-cell profiling technologies produce comprehensive measurements of genetic activity in individual cells, such as levels of RNA, which is transcribed from DNA, so that the cell’s functions and roles in the biology of the brain, and the pathology of disease, can be assessed. Single-cell profiling technologies go beyond genome sequencing, which catalogs the DNA present in most every cell of a person, by revealing how each cell is uniquely making use of that common set of instructions. In studying Alzheimer’s disease, scientists have been using single-cell profiling to see how various brain cells, such as distinct types of neurons and microglia and astrocytes, act differently in disease compared to how they behave in a healthy brain.
In the article, MIT Department of Brain and Cognitive Sciences doctoral student Mitch Murdock and Picower Professor Li-Huei Tsai, director of MIT’s Picower Institute for Learning and Memory and Aging Brain Initiative, write that while the findings of single-cell profiling studies confirm that the disease’s terrible effects are complex and far-reaching, there appear to also be five pathways that become perturbed in each of five major cell types. Investigating these pathways, they write, could produce valuable biomarkers of disease and yield meaningful targets for therapeutic intervention:
For each of these pathways in neurons, microglia, astrocytes, oligodendrocytes, and oligodendrocyte precursor cells, Tsai and Murdock identify specific differences in gene regulation, found in single-cell studies, that significantly occur in brains of Alzheimer’s patients or mouse models compared to healthy control samples.
For example, Tsai and Murdock highlight more than a dozen genes all intimately involved in lipid processing whose expression is altered in various ways in various cells in the brain’s prefrontal cortex. For another example they show that all five cell types show impairments in DNA repair, albeit by changed expression of different genes in each.
“By identifying vulnerable cell types and the molecular programs that give rise to them, therapeutic interventions might reverse aberrant cellular trajectories,” Murdock and Tsai write in Nature Neuroscience. “While many transcriptional alterations are cell-type specific, these changes ultimately might converge on shared signaling pathways across cell types that might represent targets for new therapeutic strategies.”
To be sure, the authors note, there is still plenty of work to be done, both in refining and improving on single-cell techniques and also exploiting newer related opportunities. The paper notes a number of issues that must be carefully considered in producing valid single-cell profiling results, including where cells are sampled in the brain for sequencing, from whom, and in what condition. Moreover, it’s not always straightforward to show how changes in gene expression necessarily affect biology and it’s even harder to know whether any particular intervention, for instance to target altered inflammation pathways, will prove safe and effective as a therapy.
Future directions, meanwhile could include making greater use of “spatial transcriptomics,” which measures gene transcription in cells where they are situated within the brain, rather than removing them for analysis. Studies should be expanded to incorporate more human samples so that varying disease and demographic differences can be fully accounted for. Datasets should be shared and integrated, the authors write, and better comparisons between human and mouse samples are necessary to better understand how well, or not, they overlap.
“Single-cell profiling facilitates a nuanced portrait of the diverse cellular processes perturbed in the AD brain,” Tsai and Murdock conclude. “These varied molecular programs help explain the divergence between healthy aging and cognitive decline, and highlight cell-type-specific molecular programs involved in AD. Core signaling modules are disrupted across multiple cell types, and manipulating disrupted cellular states will pave the way for new therapeutic opportunities.”