MIT Research News' Journal
[Most Recent Entries]
[Calendar View]
Tuesday, October 18th, 2016
| Time |
Event |
| 12:00a |
Making a splash in health care economics When Heidi Williams was in the fifth grade, she wanted to become an industrial organizational psychologist. That could have been an interesting career. Later, when Williams was in college, she wanted to become a cryptologist, and even completed an internship at the National Security Agency. That also could have been an interesting career. But Williams did not stick with either one.
“You have to be honest with yourself about what you’re most excited about,” says Williams. “As a researcher, you want to find a set of questions that you really want to answer. That’s not an automatic ticket to success in life, but I think it’s a necessary precondition for being motivated to get up in the morning and being excited to work.”
Instead, Williams is now the Class of 1957 Career Development Associate Professor in MIT’s Department of Economics, and what motivates her to get up in the morning is a set of questions about innovation, medical research, and health care. Do gene patents restrict or enhance medical advances? What is the effect of patent law on cancer research? To what extent does the use of medical technology drive health care cost growth?
Since she was a graduate student, Williams has kept a list of research questions that intrigue her, and she has developed a distinctive method of trying to answer them. For each new study, she essentially builds an all-new data set from the ground up, linking scientific records about medical research with economic and financial records.
“That linkage is what I find most exciting,” says Williams. “MIT is wonderful because in addition to [finding] an amazing home in economics, I can talk with and learn from a lot of people on the science side.”
Working this way takes time, and many junior faculty, under pressure to publish frequent research papers, might shy away from this approach. But Williams’ efforts have paid off, and in a short time she has had a significant impact on her field. In 2015, Williams was awarded a MacArthur fellowship — the so-called “genius grant” — to further her research. And this spring, she was granted tenure at MIT, just five years after joining the Institute as an assistant professor.
Mentors make the difference
Williams grew up in Williston, North Dakota, a town near the Montana border that had about 12,000 people in it when she lived there. (It has roughly doubled in size since then due to the oil-drilling boom.)
“It’s small, but it is relatively a big city for North Dakota,” Williams offers.
Her mother, who was a social worker, and her father, who was an eye doctor, were “just amazing, always very supportive of [the idea that] I should do whatever I want to do,” Williams adds.
Williams’ education took a decisive turn when, as a high school student, she submitted an entry to a college-level competition for cryptology papers. After some time passed and the results were not announced, Williams contacted the contest’s director, Brian Winkel, now an emeritus math professor at the United States Military Academy at West Point.
“I would never do that today,” she says. But although she didn’t know it, Williams had been judged winner of the contest, and her paper — suggesting improvements to a method of decoding text sent by Enigma machines, the German message-sending devices that the British cracked during World War II — would be published in the journal Cryptologia. Winkel, far from being miffed, took an interest in Williams’ future. He suggested she apply to Dartmouth and consider working with a math professor there named Dorothy Wallace.
As it happens, Williams was considering applying to Dartmouth anyway, because, apparently weary of the bright lights of Williston, “I already had my heart set on moving to a small town for college,” she says.
Williams was accepted to Dartmouth, attended, and as a math major, worked with Wallace. As is so often the case with students, direct mentoring made a big difference to Williams.
“She was, as promised, just a fantastic advisor,” Williams says, about Wallace. “My life would have been really different if Brian and Dorothy hadn’t gone out of their way to mentor me at that stage.”
Williams did seriously consider entering the field of cryptography and calls her NSA internship “extremely interesting work.” But she was looking for some way to apply her quantitative background more directly to social issues. She received a master’s degree in economics from Oxford University and entered Harvard University’s PhD program in economics.
The genome wars and more
At Harvard, working with advisors David Cutler, Lawrence Katz, Michael Kremer, and MIT’s health care economist Amy Finkelstein, Williams began carefully researching the project that became her PhD thesis and first big published paper, about intellectual property rights and the human genome. The private firm Celera had famously raced the public-sector Human Genome Project to finish decoding a complete copy of a human genome, claiming intellectual property rights on some of its sequenced genes along the way.
Williams documented evidence that Celera’s intellectual property rights reduced subsequent scientific research and product development by 20 to 30 percent, compared to the sequenced genes the Human Genome Project placed in the public domain. However, Williams also thinks the issue is “a lot more nuanced” than that one statistic might imply, since Celera did not have outright patents on its sequenced genes and instead developed an alternative, contract-based form of intellectual property rights when the firm wasn’t granted gene patents.
“This … alternative form of intellectual property that Celera used seems to have discouraged scientific researchers,” Williams notes, adding that her results from this study together with a subsequent research paper suggest that both Celera and society might “have been better off if we’d just granted Celera its patents in the first place.”
Williams joined MIT in 2011 and has since published additional large-scale findings on medical innovation and health care. In a 2015 paper, Williams and two co-authors published evidence that pharmaceutical firms “underinvest” in drugs to fight early-stage cancers, partly because clinical trials for those drugs take longer and cost more to develop than the clinical trials for drugs treating late-stage cancers — and leave firms with less time to control patented drug therapies on the market.
“It’s amazing. Patents are incredibly controversial, and we really don’t have rigorous studies to point to that quantify the potential benefits of patents in terms of spurred research investments, or whether those benefits justify the costs that patents impose on society,” Williams says. “And I’m excited to continue to work on trying to generate empirical evidence to inform that question.”
Some of her other research tracks the economics of medical care, including breast cancer therapies and infant mortality prevention. And later this fall, Williams, Finkelstein, and another co-author will also publish a paper on a big-time topic in health care research: whether the regional variation in U.S. health care spending comes from entrenched medical practices or from the varying degrees of sickness in the local populations. (The short answer: a bit of both.)
Williams is quick to credit her MIT faculty colleagues — including Finkelstein, David Autor, James Poterba, and Scott Stern — with helping her make the transition to faculty life at the Institute. She calls them “phenomenal role models,” adding, “I have been so incredibly appreciative of their support.”
As a rising star, Williams has plenty of options in choosing her future research directions. Intriguingly, she is considering undertaking studies on the effects of patents in multiple industries, not just in health care. But whatever intellectual direction Williams moves in, her method will likely remain the same: careful construction of unique data sets on big topics.
“It is usually the case that not every part of a paper is perfect in terms of your ability to learn something from it,” Williams says. “But in the long run, making large investments in constructing new data, and working on a body of several papers in a given area, seems like a useful approach. I’ve been very lucky at MIT to feel like I have the time I’ve needed need to work on the best possible versions of the research projects that I’m most excited to pursue.” | | 8:00a |
A new player in appetite control MIT neuroscientists have discovered that brain cells called glial cells play a critical role in controlling appetite and feeding behavior. In a study of mice, the researchers found that activating these cells stimulates overeating, and that when the cells are suppressed, appetite is also suppressed.
The findings could offer scientists a new target for developing drugs against obesity and other appetite-related disorders, the researchers say. The study is also the latest in recent years to implicate glial cells in important brain functions. Until about 10 years ago, glial cells were believed to play more of a supporting role for neurons.
“In the last few years, abnormal glial cell activities have been strongly implicated in neurodegenerative disorders. There is more and more evidence to point to the importance of glial cells in modulating neuronal function and in mediating brain disorders,” says Guoping Feng, the James W. and Patricia Poitras Professor of Neuroscience. Feng is also a member of MIT’s McGovern Institute for Brain Research and the Stanley Center for Psychiatric Research at the Broad Institute.
Feng is one of the senior authors of the study, which appears in the Oct. 18 edition of the journal eLife. The other senior author is Weiping Han, head of the Laboratory of Metabolic Medicine at the Singapore Bioimaging Consortium in Singapore. Naiyan Chen, a postdoc at the Singapore Bioimaging Consortium and the McGovern Institute, is the lead author.
Turning on appetite
It has long been known that the hypothalamus, an almond-sized structure located deep within the brain, controls appetite as well as energy expenditure, body temperature, and circadian rhythms including sleep cycles. While performing studies on glial cells in other parts of the brain, Chen noticed that the hypothalamus also appeared to have a lot of glial cell activity.
“I was very curious at that point what glial cells would be doing in the hypothalamus, since glial cells have been shown in other brain areas to have an influence on regulation of neuronal function,” she says.
Within the hypothalamus, scientists have identified two key groups of neurons that regulate appetite, known as AgRP neurons and POMC neurons. AgRP neurons stimulate feeding, while POMC neurons suppress appetite.
Until recently it has been difficult to study the role of glial cells in controlling appetite or any other brain function, because scientists haven’t developed many techniques for silencing or stimulating these cells, as they have for neurons. Glial cells, which make up about half of the cells in the brain, have many supporting roles, including cushioning neurons and helping them form connections with one another.
In this study, the research team used a new technique developed at the University of North Carolina to study a type of glial cell known as an astrocyte. Using this strategy, researchers can engineer specific cells to produce a surface receptor that binds to a chemical compound known as CNO, a derivative of clozapine. Then, when CNO is given, it activates the glial cells.
The MIT team found that turning on astrocyte activity with just a single dose of CNO had a significant effect on feeding behavior.
“When we gave the compound that specifically activated the receptors, we saw a robust increase in feeding,” Chen says. “Mice are not known to eat very much in the daytime, but when we gave drugs to these animals that express a particular receptor, they were eating a lot.”
The researchers also found that in the short term (three days), the mice did not gain extra weight, even though they were eating more.
“This raises the possibility that glial cells may also be modulating neurons that control energy expenditures, to compensate for the increased food intake,” Chen says. “They might have multiple neuronal partners and modulate multiple energy homeostasis functions all at the same time.”
When the researchers silenced activity in the astrocytes, they found that the mice ate less than normal.
Suzanne Dickson, a professor of neuroendocrinology at the University of Gothenburg in Sweden described the study as part of a “paradigm shift” toward the idea that glial cells have a less passive role than previously believed.
“We tend to think of glial cells as providing a support network for neuronal processes and that their activation is also important in certain forms of brain trauma or inflammation,” says Dickson, who was not involved in the research. “This study adds to the emerging evidence base that glial cells may also exert specific effects to control nerve cell function in normal physiology.”
Unknown interactions
Still unknown is how the astrocytes exert their effects on neurons. Some recent studies have suggested that glial cells can secrete chemical messengers such as glutamate and ATP; if so, these “gliotransmitters” could influence neuron activity.
Another hypothesis is that instead of secreting chemicals, astrocytes exert their effects by controlling the uptake of neurotransmitters from the space surrounding neurons, thereby affecting neuron activity indirectly.
Feng now plans to develop new research tools that could help scientists learn more about astrocyte-neuron interactions and how astrocytes contribute to modulation of appetite and feeding. He also hopes to learn more about whether there are different types of astrocytes that may contribute differently to feeding behavior, especially abnormal behavior.
“We really know very little about how astrocytes contribute to the modulation of appetite, eating, and metabolism,” he says. “In the future, dissecting out these functional difference will be critical for our understanding of these disorders.” | | 10:20a |
Prepping a robot for its journey to Mars Sarah Hensley is preparing an astronaut named Valkyrie for a mission to Mars. It is 6 feet tall, weighs 300 pounds, and is equipped with an extended chest cavity that makes it look distinctly female. Hensley spends much of her time this semester analyzing the movements of one of Valkyrie's arms.
As a fourth-year electrical engineering student at MIT, Hensley is working with a team of researchers to prepare Valkyrie, a humanoid robot also known as R5, for future space missions. As a teenager in New Jersey, Hensley loved to read in her downtime, particularly Isaac Asimov’s classic robot series. “I’m a huge science fiction nerd — and now I’m actually getting to work with a robot that’s real and not just in books. That’s like, wow.”
Hensley is studying Valkyrie for an advanced independent research program, or SuperUROP, as one of only three undergraduate students in the Robot Locomotion Group in MIT's Computer Science and Artificial Intelligence Laboratory. Most of her colleagues are graduate-level researchers and postdocs with extensive experience working on complex humanoids. The group is led by professor of electrical engineering and computer science Russ Tedrake, who successfully programmed Valkyrie’s predecessor (named Atlas) to open doors, turn valves, drill holes, climb stairs, and drive a car for the DARPA Robotics Challenge in 2015.
Valkyrie has 28 torque-controlled joints, four body cameras, and more than 200 individual sensors, Hensley says. The robot can walk, bend its joints, and turn a door handle. “This is one of the most advanced robots in the world. And it’s 20 feet from my desk,” she adds.
That’s largely because Valkyrie has a long way to go before it leaves for Mars. MIT is one of three institutions, including Northeastern University and the University of Edinburgh, that NASA selected to develop software enabling the robot to perform space-related tasks — open airlock hatches, attach and remove power cables, repair equipment, and retrieve samples. Oh yeah, and get to its feet when it falls down.
Hensley, who started in the lab over the summer, is intrigued by the challenge of harmonizing the movements of such a highly complex system. “I am trying to solve a very tricky problem,” she says. She’s working out how best to control Valkyrie’s elbow movements by comparing two potential approaches. One uses a main controller to gather information from the various motor systems within the arm, and then uses that data to make accurate movement decisions. The other approach is decentralized, and leaves it to each motor system to decide and act on its own.
Hensley gets animated discussing the alternatives. “Is it better to have multiple decision makers with access to different information? Or is better to have one decision maker choosing all of the motor inputs?” she asks. Hensley has already been accepted into a master’s degree program in electrical engineering at MIT. She hopes to be able to continue her work on Valkyrie.
Every day, Hensley leaves Tau Epsilon Phi, her co-ed fraternity house in the Back Bay; walks across the Massachusetts Ave. bridge to the Stata Center; and plants herself in front of two large monitors in the robotics lab. She analyzes a wealth of scientific literature and writes code for computer simulations of the equations that move the robot’s arm. Sometimes she gets up for peppermint tea, or to peer around the corner of her cubicle at Valkyrie.
One thing is certain, says Hensley. Pop culture fears that machines may soon prove superior to humans are laughable. When Valkyrie is turned on and moves, Hensley says, it often “kind of shivers and falls down. One thing you realize working in this lab is that we are really far away from the robot apocalypse,” she quips. “Sometimes robots work, and sometimes they don’t. That’s our challenge.” | | 11:59p |
With new model, buildings may “sense” internal damage When a truck rumbles by a building, vibrations can travel up to the structure’s roof and down again, generating transient tremors through the intervening floors and beams.
Now researchers at MIT have developed a computational model that makes sense of such ambient vibrations, picking out key features in the noise that give indications of a building’s stability. The model may be used to monitor a building over time for signs of damage or mechanical stress. The team’s results are published online in the journal Mechanical Systems and Signal Processing.
“The broader implication is, after an event like an earthquake, we would see immediately the changes of these features, and if and where there is damage in the system,” says Oral Buyukozturk, a professor in MIT’s Department of Civil and Environmental Engineering (CEE). “This provides continuous monitoring and a database that would be like a health book for the building, as a function of time, much like a person’s changing blood pressure with age.”
Buyukozturk’s co-authors Hao Sun, a CEE postdoc who was the paper’s lead author; Aurélien Mordret, a postdoc in the Department of Earth, Atmospheric and Planetary Sciences (EAPS); Germán Prieto, the Cecil and Ida Green Career Development Assistant Professor in EAPS; and M. Nafi Toksöz, an EAPS professor.
Taking vital signs
The team tested its computational model on MIT’s Green Building — a 21-story research building made completely from reinforced concrete. The building was designed in the 1960s by architect and MIT alum I.M. Pei ’40, and stands as the tallest structure in Cambridge, Massachusetts. In 2010, Toksöz and others at MIT worked with the United States Geological Survey to outfit the Green Building with 36 accelerometers that record vibrations and movements on selected floors, from the building’s foundation to its roof.
“These sensors represent an embedded nervous system,” Buyukozturk says. “The challenge is to extract vital signs from the sensors’ data and link them to health characteristics of a building, which has been a challenge in the engineering community.”
To do this, the team first built a computer simulation of the Green Building, in the form of a finite element model — a numerical simulation that represents a large physical structure, and all its underlying physics, as a collection of smaller, simpler subdivisions. In the case of the Green Building, the researchers built a high-fidelity finite element model, then plugged various parameters into the model, including the strength and density of concrete walls, slabs, beams, and stairs in each floor.
As the model is designed, researchers should be able to introduce an excitation in the simulation — for example, a truck-like vibration — and the model would predict how the building and its various elements should respond.
“But the model uses a lot of assumptions about the building’s material, its geometry, the thickness of its elements, et cetera, which may not correspond exactly to the structure,” Buyukozturk notes. “So we are updating the model with actual measurements to be able to give better information about what may have happened to the building.”
Mining for features
To more accurately predict a building’s response to ambient vibrations, the group mined data from the Green Building’s accelerometers, looking for key features that correspond directly to a building’s stiffness or other indicators of health. To do this efficiently, the team developed a new method with the seismic interferometry concept that describes how a vibration’s pattern changes as it travels from the ground level to the roof.
“We look at the foundation level and see what motions a truck, for instance, caused there, and then how that vibration travels upward and horizontally, in speed and direction,” Buyukozturk explains.
The researchers added this equation to their model of the Green Building and ran the model multiple times, each time with a set of measurements taken by the accelerometers at a given point in time. In all, the group plugged into the model vibration measurements that were taken continuously over a two-week period in May 2015.
“We are continuously making our computational system more intelligent over time, with more data,” Buyukozturk says. “We’re confident if there is damage in the building, it will show up in our system.”
Intelligent buildings
So how has the Green Building fared since its construction more than 50 years ago?
“The building is safe, but it is subject to quite a bit of vibration, particularly in the upper floors,” Buyukozturk says. “The building, which is built on soft soil, is long in one direction and narrow in the other with stiff concrete walls on each end. Therefore, it manifests torsional movements and rocking, especially on windy days,” he says.
The team plans to verify its computational model with experiments in the lab. The researchers have constructed a 4-meter-tall replica of a building structure, which they will outfit with accelerometers. They will study the effects of ambient vibrations, as well as how the structure responds to hammer strikes and other seismic stimuli. The team is also erecting a large steel structure in Woburn, Massachusetts, about the size of a cellphone tower, and will carry out similar experiments that will ultimately help to refine the researchers’ computational model.
“I would envision that, in the future, such a monitoring system will be instrumented on all our buildings, city-wide,” says lead author Hao Sun. “Outfitted with sensors and central processing algorithms, those buildings will become intelligent, and will feel their own health in real time and possibly be resilient to extreme events.”
This research was funded, in part, by Royal Dutch Shell through the MIT Energy Initiative, and by the Kuwait-MIT Signature Project through the Kuwait Foundation for the Advancement of Sciences and the Kuwait-MIT Center for Natural Resources and the Environment. |
|