Advanced imaging reveals mired migration of neurons in Rett syndrome lab models

Using an innovative microscopy method, scientists at The Picower Institute for Learning and Memory at MIT observed how newborn neurons struggle to reach their proper places in advanced human brain tissue models of Rett syndrome, producing new insight into how developmental deficits observed in the brains of patients with the devastating disorder may emerge.

Rett syndrome, which is characterized by symptoms including severe intellectual disability and impaired social behavior, is caused by mutations in the gene MECP2. To gain new insight into how the mutation affects the early stages of human brain development, researchers in the lab of Mriganka Sur, Newton Professor of Neuroscience in MIT’s Department of Brain and Cognitive Sciences, grew 3D cell cultures called cerebral organoids, or minibrains, using cells from people with MECP2 mutations, and compared them to otherwise-identical cultures without the mutations. Then the team led by postdoc Murat Yildirim examined the development of each type of minibrain using an advanced imaging technology called third harmonic generation (THG) three-photon microscopy.

THG, which Yildirim has helped to pioneer in Sur’s lab working with MIT mechanical engineering Professor Peter So, allows for very high-resolution imaging deep into live, intact tissues without having to add any chemicals to label cells. The new study, published in eLife, is the first to use THG to image organoids, leaving them virtually undisturbed, Yildirim said. Previous organoid imaging studies have required using technologies that cannot image all the way through the 3D tissue, or methods that require killing the cultures: either slicing them into thin sections or chemically clearing and labeling them.

Three-photon microscopy employs a laser, but Yildirim and So custom engineered the lab’s microscope to apply no more power to the tissue than a cat toy laser pointer (less than 5 milliwatts).

“You should make sure you are not changing or affecting the neuronal physiology in any adverse way,” Yildirim says. “You should really keep everything intact and make sure you are not bringing something external that could be damaging. That’s why we are so careful about power (and chemical labeling).”

Even at low power, they achieved adequate signal to achieve label-free, intact imaging of fixed and live organoids. To validate that they compared their THG images to images made via more traditional chemical labeling methods.

The THG system allowed them to track the migration of newborn neurons as they made their way from the rim around open spaces in the minibrains (called ventricles) to the outer edge, which is directly analogous to the brain’s cortex. They saw that the nascent neurons in the minibrains modeling Rett syndrome moved slowly and in meandering paths compared to the faster motion in straighter lines exhibited by the same cell types in minibrains without MECP2 mutation. Sur says the consequences of such migration deficits are consistent with what scientists, including in his lab, have hypothesized is going on in fetuses with Rett syndrome.

“We know from postmortem brains and brain imaging methods that things go awry during brain development in Rett syndrome, but it has been astonishingly difficult to figure out what and why,” says Sur, who directs the Simons Center for the Social Brain at MIT. “This method has enabled us to directly visualize a key contributor.” THG images tissues without labels because it is very sensitive to changes in the refractive index of materials, Yildirim says. It therefore resolves boundaries between biological structures, such as blood vessels, cell membranes, and extracellular spaces. Because neural shapes change during their development, the team was able to also clearly see the delineation between the ventricular zone (the area around the ventricles where the newborn neurons emerge) and the cortical plate (an area that mature neurons settle into). It was also very easy to resolve various ventricles and segment them into distinct regions.

Those properties allowed the researchers to be able to see that in Rett syndrome organoids the ventricles were larger and more numerous and that the ventricular zones — the rims around the ventricles where neurons are born — were thinner. In live organoids they were able to track some of the neurons making their way toward the cortex over a few days, taking a new picture every 20 minutes, as neurons in real developing brains also attempt to do. They saw that Rett syndrome neurons achieved only about two-thirds the speed of non-mutated neurons. The paths of the Rett neurons were also significantly more wiggly. The two differences combined meant that the Rett cells barely got half as far.

“We now want to know how MECP2 influences genes and molecules that influence neuronal migration,” Sur says. “By screening Rett syndrome organoids, we have some good guesses, which we are eager to test.” Yildirim, who will launch his own lab as an assistant professor at the Cleveland Clinic’s Lerner Research Institute in September, says he has new questions based on the findings. He wants to image later in organoid development to track the consequences of the sinuous migration. He also wants to find out more about whether specific cell types struggle to migrate more or less, which could alter how cortical circuits work.

Yildirim also says he hopes to continue advancing THG three-photon microscopy, which he sees as having potential for fine-grained imaging in humans. It can be an important advantage in people, especially that the imaging method can penetrate deep into living tissue without the need for artificial labels.

In addition to Yildirim, Sur, and So, the paper’s other authors are Chloe Delepine, Danielle Feldman, Vincent Pham, Stephanie Chou, Jacque Pak Kan Ip, Alexi Nott, Li-Huei Tsai, and Guo-li Ming.

The National Institutes of Health, The National Science Foundation, the JPB Foundation, and the Massachusetts Life Sciences Initiative provided funding for the research.

de MIT News https://ift.tt/FfJrNbW

A better way to quantify radiation damage in materials

It was just a piece of junk sitting in the back of a lab at the MIT Nuclear Reactor facility, ready to be disposed of. But it became the key to demonstrating a more comprehensive way of detecting atomic-level structural damage in materials — an approach that will aid the development of new materials, and could potentially support the ongoing operation of carbon-emission-free nuclear power plants, which would help alleviate global climate change.

A tiny titanium nut that had been removed from inside the reactor was just the kind of material needed to prove that this new technique, developed at MIT and at other institutions, provides a way to probe defects created inside materials, including those that have been exposed to radiation, with five times greater sensitivity than existing methods.

The new approach revealed that much of the damage that takes place inside reactors is at the atomic scale, and as a result is difficult to detect using existing methods. The technique provides a way to directly measure this damage through the way it changes with temperature. And it could be used to measure samples from the currently operating fleet of nuclear reactors, potentially enabling the continued safe operation of plants far beyond their presently licensed lifetimes.

The findings are reported today in the journal Science Advances in a paper by MIT research specialist and recent graduate Charles Hirst PhD ’22; MIT professors Michael Short, Scott Kemp, and Ju Li; and five others at the University of Helsinki, the Idaho National Laboratory, and the University of California at Irvine.

Rather than directly observing the physical structure of a material in question, the new approach looks at the amount of energy stored within that structure. Any disruption to the orderly structure of atoms within the material, such as that caused by radiation exposure or by mechanical stresses, actually imparts excess energy to the material. By observing and quantifying that energy difference, it’s possible to calculate the total amount of damage within the material — even if that damage is in the form of atomic-scale defects that are too small to be imaged with microscopes or other detection methods.

The principle behind this method had been worked out in detail through calculations and simulations. But it was the actual tests on that one titanium nut from the MIT nuclear reactor that provided the proof — and thus opened the door to a new way of measuring damage in materials.

The method they used is called differential scanning calorimetry. As Hirst explains, this is similar in principle to the calorimetry experiments many students carry out in high school chemistry classes, where they measure how much energy it takes to raise the temperature of a gram of water by one degree. The system the researchers used was “fundamentally the exact same thing, measuring energetic changes. … I like to call it just a fancy furnace with a thermocouple inside.”

The scanning part has to do with gradually raising the temperature a bit at a time and seeing how the sample responds, and the differential part refers to the fact that two identical chambers are measured at once, one empty, and one containing the sample being studied. The difference between the two reveals details of the energy of the sample, Hirst explains.

“We raise the temperature from room temperature up to 600 degrees Celsius, at a constant rate of 50 degrees per minute,” he says. Compared to the empty vessel, “your material will naturally lag behind because you need energy to heat your material. But if there are changes in the energy inside the material, that will change the temperature. In our case, there was an energy release when the defects recombine, and then it will get a little bit of a head start on the furnace … and that’s how we are measuring the energy in our sample.”

Hirst, who carried out the work over a five-year span as his doctoral thesis project, found that contrary to what had been believed, the irradiated material showed that there were two different mechanisms involved in the relaxation of defects in titanium at the studied temperatures, revealed by two separate peaks in calorimetry. “Instead of one process occurring, we clearly saw two, and each of them corresponds to a different reaction that’s happening in the material,” he says.

They also found that textbook explanations of how radiation damage behaves with temperature weren’t accurate, because previous tests had mostly been carried out at extremely low temperatures and then extrapolated to the higher temperatures of real-life reactor operations. “People weren’t necessarily aware that they were extrapolating, even though they were, completely,” Hirst says.

“The fact is that our common-knowledge basis for how radiation damage evolves is based on extremely low-temperature electron radiation,” adds Short. “It just became the accepted model, and that’s what’s taught in all the books. It took us a while to realize that our general understanding was based on a very specific condition, designed to elucidate science, but generally not applicable to conditions in which we actually want to use these materials.”

Now, the new method can be applied “to materials plucked from existing reactors, to learn more about how they are degrading with operation,” Hirst says.

“The single biggest thing the world can do in order to get cheap, carbon-free power is to keep current reactors on the grid. They’re already paid for, they’re working,” Short adds.  But to make that possible, “the only way we can keep them on the grid is to have more certainty that they will continue to work well.” And that’s where this new way of assessing damage comes into play.

While most nuclear power plants have been licensed for 40 to 60 years of operation, “we’re now talking about running those same assets out to 100 years, and that depends almost fully on the materials being able to withstand the most severe accidents,” Short says. Using this new method, “we can inspect them and take them out before something unexpected happens.”

In practice, plant operators could remove a tiny sample of material from critical areas of the reactor, and analyze it to get a more complete picture of the condition of the overall reactor. Keeping existing reactors running is “the single biggest thing we can do to keep the share of carbon-free power high,” Short stresses. “This is one way we think we can do that.”

Sergei Dudarev, a fellow at the United Kingdom Atomic Energy Authority who was not associated with this work, says this “is likely going to be impactful, as it confirms, in a nice systematic manner, supported both by experiment and simulations, the unexpectedly significant part played by the small invisible defects in microstructural evolution of materials exposed to irradiation.”

The process is not just limited to the study of metals, nor is it limited to damage caused by radiation, the researchers say. In principle, the method could be used to measure other kinds of defects in materials, such as those caused by stresses or shockwaves, and it could be applied to materials such as ceramics or semiconductors as well.

In fact, Short says, metals are the most difficult materials to measure with this method, and early on other researchers kept asking why this team was focused on damage to metals. That was partly because reactor components tend to be made of metal, and also because “It’s the hardest, so, if we crack this problem, we have a tool to crack them all!”

Measuring defects in other kinds of materials can be up to 10,000 times easier than in metals, he says. “If we can do this with metals, we can make this extremely, ubiquitously applicable.” And all of it enabled by a small piece of junk that was sitting at the back of a lab.

The research team included Fredric Granberg and Kai Nordlund at the University of Helsinki in Finland; Boopathy Kombaiah and Scott Middlemas at Idaho National Laboratory; and Penghui Cao at the University of California at Irvine. The work was supported by the U.S. National Science Foundation, an Idaho National Laboratory research grant, and a Euratom Research and Training program grant.

de MIT News https://ift.tt/br0mIFH

New algorithm aces university math course questions

Multivariable calculus, differential equations, linear algebra — topics that many MIT students can ace without breaking a sweat — have consistently stumped machine learning models. The best models have only been able to answer elementary or high school-level math questions, and they don’t always find the correct solutions.

Now, a multidisciplinary team of researchers from MIT and elsewhere, led by Iddo Drori, a lecturer in the MIT Department of Electrical Engineering and Computer Science (EECS), has used a neural network model to solve university-level math problems in a few seconds at a human level.

The model also automatically explains solutions and rapidly generates new problems in university math subjects. When the researchers showed these machine-generated questions to university students, the students were unable to tell whether the questions were generated by an algorithm or a human.

This work could be used to streamline content generation for courses, which could be especially useful in large residential courses and massive open online courses (MOOCs) that have thousands of students. The system could also be used as an automated tutor that shows students the steps involved in solving undergraduate math problems.

“We think this will improve higher education,” says Drori, the work’s lead author who is also an adjunct associate professor in the Department of Computer Science at Columbia University, and who will join the faculty at Boston University this summer. “It will help students improve, and it will help teachers create new content, and it could help increase the level of difficulty in some courses. It also allows us to build a graph of questions and courses, which helps us understand the relationship between courses and their pre-requisites, not just by historically contemplating them, but based on data.”

The work is a collaboration including students, researchers, and faculty at MIT, Columbia University, Harvard University, and the University of Waterloo. The senior author is Gilbert Strang, a professor of mathematics at MIT. The research appears this week in the Proceedings of the National Academy of Sciences.

A “eureka” moment

Drori and his students and colleagues have been working on this project for nearly two years. They were finding that models pretrained using text only could not do better than 8 percent accuracy on high school math problems, and those using graph neural networks could ace machine learning course questions but would take a week to train.

Then Drori had what he describes as a “eureka” moment: He decided to try taking questions from undergraduate math courses offered by MIT and one from Columbia University that had never been seen before by a model, turning them into programming tasks, and applying techniques known as program synthesis and few-shot learning. Turning a question into a programming task could be as simple as rewriting the question “find the distance between two points” as “write a program that finds the difference between two points,” or providing a few question-program pairs as examples.

Before feeding those programming tasks to a neural network, however, the researchers added a new step that enabled it to vastly outperform their previous attempts.

In the past, they and others who’ve approached this problem have used a neural network, such as GPT-3, that was pretrained on text only, meaning it was shown millions of examples of text to learn the patterns of natural language. This time, they used a neural network pretrained on text that was also “fine-tuned” on code. This network, called Codex, was produced by OpenAI. Fine-tuning is essentially another pretraining step that can improve the performance of a machine-learning model.

The pretrained model was shown millions of examples of code from online repositories. Because this model’s training data included millions of natural language words as well as millions of lines of code, it learns the relationships between pieces of text and pieces of code.

Many math problems can be solved using a computational graph or tree, but it is difficult to turn a problem written in text into this type of representation, Drori explains. Because this model has learned the relationships between text and code, however, it can turn a text question into code, given just a few question-code examples, and then run the code to answer the problem.

“When you just ask a question in text, it is hard for a machine-learning model to come up with an answer, even though the answer may be in the text,” he says. “This work fills in the that missing piece of using code and program synthesis.”

This work is the first to solve undergraduate math problems and moves the needle from 8 percent accuracy to over 80 percent, Drori adds.

Adding context

Turning math questions into programming tasks is not always simple, Drori says. Some problems require researchers to add context so the neural network can process the question correctly. A student would pick up this context while taking the course, but a neural network doesn’t have this background knowledge unless the researchers specify it.

For instance, they might need to clarify that the “network” in a question’s text refers to “neural networks” rather than “communications networks.” Or they might need to tell the model which programming package to use. They may also need to provide certain definitions; in a question about poker hands, they may need to tell the model that each deck contains 52 cards.

They automatically feed these programming tasks, with the included context and examples, to the pretrained and fine-tuned neural network, which outputs a program that usually produces the correct answer. It was correct for more than 80 percent of the questions.

The researchers also used their model to generate questions by giving the neural network a series of math problems on a topic and then asking it to create a new one.

“In some topics, it surprised us. For example, there were questions about quantum detection of horizontal and vertical lines, and it generated new questions about quantum detection of diagonal lines. So, it is not just generating new questions by replacing values and variables in the existing questions,” Drori says.

Human-generated vs. machine-generated questions

The researchers tested the machine-generated questions by showing them to university students. The researchers gave students 10 questions from each undergraduate math course in a random order; five were created by humans and five were machine-generated.

Students were unable to tell whether the machine-generated questions were produced by an algorithm or a human, and they gave human-generated and machine-generated questions similar marks for level of difficulty and appropriateness for the course.

Drori is quick to point out that this work is not intended to replace human professors.

“Automation is now at 80 percent, but automation will never be 100 percent accurate. Every time you solve something, someone will come up with a harder question. But this work opens the field for people to start solving harder and harder questions with machine learning. We think it will have a great impact on higher education,” he says.

The team is excited by the success of their approach, and have extended the work to handle math proofs, but there are some limitations they plan to tackle. Currently, the model isn’t able to answer questions with a visual component and cannot solve problems that are computationally intractable due to computational complexity.

In addition to overcoming these hurdles, they are working to scale the model up to hundreds of courses. With those hundreds of courses, they will generate more data that can enhance automation and provide insights into course design and curricula.

de MIT News https://ift.tt/AC6OfrI

Why it’s a problem that pulse oximeters don’t work as well on patients of color

Pulse oximetry is a noninvasive test that measures the oxygen saturation level in a patient’s blood, and it has become an important tool for monitoring many patients, including those with Covid-19. But new research links faulty readings from pulse oximeters with racial disparities in health outcomes, potentially leading to higher rates of death and complications such as organ dysfunction, in patients with darker skin.

It is well known that non-white intensive care unit (ICU) patients receive less-accurate readings of their oxygen levels using pulse oximeters — the common devices clamped on patients’ fingers. Now, a paper co-authored by MIT scientists reveals that inaccurate pulse oximeter readings can lead to critically ill patients of color receiving less supplemental oxygen during ICU stays.

The paper, “Assessment of Racial and Ethnic Differences in Oxygen Supplementation Among Patients in the Intensive Care Unit,” published in JAMA Internal Medicine, focused on the question of whether there were differences in supplemental oxygen administration among patients of different races and ethnicities that were associated with pulse oximeter performance discrepancies. 

The findings showed that inaccurate readings of Asian, Black, and Hispanic patients resulted in them receiving less supplemental oxygen than white patients. These results provide insight into how health technologies such as the pulse oximeter contribute to racial and ethnic disparities in care, according to the researchers.

The study's senior author, Leo Anthony Celi, clinical research director and principal research scientist at the MIT Laboratory for Computational Physiology, and a principal research scientist at the MIT Institute for Medical Engineering and Science (IMES), says the challenge is that health care technology is routinely designed around the majority population.

“Medical devices are typically developed in rich countries with white, fit individuals as test subjects,” he explains. “Drugs are evaluated through clinical trials that disproportionately enroll white individuals. Genomics data overwhelmingly come from individuals of European descent.”

“It is therefore not surprising that we observe disparities in outcomes across demographics, with poorer outcomes among those who were not included in the design of health care," Celi adds.

While pulse oximeters are widely used due to ease of use, the most accurate way to measure blood oxygen saturation (SaO₂) levels is by taking a sample of the patient’s arterial blood. False readings of normal pulse oximetry (SpO₂) can lead to hidden hypoxemia. Elevated bilirubin in the bloodstream and the use of certain medications in the ICU called vasopressors can also throw off pulse oximetry readings.

More than 3,000 participants were included in the study, of whom 2,667 were white, 207 Black, 112 Hispanic, and 83 Asian — using data from the Medical Information Mart for Intensive Care version 4, or MIMIC-IV dataset. This dataset is comprised of more than 50,000 patients admitted to the ICU at Beth Israel Deaconess Medical Center, and includes both pulse oximeter readings and oxygen saturation levels detected in blood samples. MIMIC-IV also includes rates of administration of supplemental oxygen.

When the researchers compared SpO₂ levels taken by pulse oximeter to oxygen saturation from blood samples, they found that Black, Hispanic, and Asian patients had higher SpO₂ readings than white patients for a given blood oxygen saturation level measured in blood samples. The turnaround time of arterial blood gas analysis may take from several minutes up to an hour. As a result, clinicians typically make decisions based on pulse oximetry reading, unaware of its suboptimal performance in certain patient demographics.

Eric Gottlieb, the study’s lead author, a nephrologist, a lecturer at MIT, and a Harvard Medical School fellow at Brigham and Women’s Hospital, called for more research to be done, in order to better understand “how pulse oximeter performance disparities lead to worse outcomes; possible differences in ventilation management, fluid resuscitation, triaging decisions, and other aspects of care should be explored. We then need to redesign these devices and properly evaluate them to ensure that they perform equally well for all patients.”

Celi emphasizes that understanding biases that exist within real-world data is crucial in order to better develop algorithms and artificial intelligence to assist clinicians with decision-making. “Before we invest more money on developing artificial intelligence for health care using electronic health records, we have to identify all the drivers of outcome disparities, including those that arise from the use of suboptimally designed technology,” he argues. “Otherwise, we risk perpetuating and magnifying health inequities with AI.”

Celi described the project and research as a testament to the value of data sharing that is the core of the MIMIC project. “No one team has the expertise and perspective to understand all the biases that exist in real-world data to prevent AI from perpetuating health inequities,” he says. “The database we analyzed for this project has more than 30,000 credentialed users consisting of teams that include data scientists, clinicians, and social scientists.”

The many researchers working on this topic together form a community that shares and performs quality checks on codes and queries, promotes reproducibility of the results, and crowdsources the curation of the data, Celi says. “There is harm when health data is not shared,” he says. “Limiting data access means limiting the perspectives with which data is analyzed and interpreted. We've seen numerous examples of model mis-specifications and flawed assumptions leading to models that ultimately harm patients.”

de MIT News https://ift.tt/TjI0rwH

Christopher Capozzola named senior associate dean for open learning

MIT Professor Christopher Capozzola has joined MIT Open Learning as senior associate dean, effective Aug. 1. Reporting to interim Vice President for Open Learning Eric Grimson, Capozzola will oversee open education offerings including OpenCourseWare, MITx, and MicroMasters, as well as the Digital Learning Lab, Digital Learning in Residential Education, and MIT Video Productions.

Capozzola has a long history of participation in the MIT Open Learning mission. A member of the MITx Faculty Advisory Committee, Capozzola also has five courses published on OpenCourseWare (OCW), and one course, Visualizing Imperialism in the Philippines, published on both MITx and the Open Learning Library.

“Chris has proven his commitment to the mission of Open Learning through his contributions both to external learners and to MIT students, as well as through his own research and professional projects. He’s also demonstrated his ability to engage collaboratively with the MIT faculty and broader community on issues related to effective delivery of educational experiences,” says Grimson. “MIT’s open online education offerings are more relevant than they’ve ever been, reaching many millions of people around the world. Chris will provide the essential faculty attention and dedicated support needed to help Open Learning continue to reflect the full spectrum of MIT’s knowledge and teaching to the world.”

Capozzola comes to MIT Open Learning from the History Section in the School of Humanities, Arts and Social Sciences (SHASS), where he has taught since 2002. He’s the author of two books and numerous articles exploring citizenship, war, and the military in modern American history. He has served as department head since 2020 and is a MacVicar Faculty Fellow, MIT’s highest honor for undergraduate teaching. He also served as MIT’s secretary of the faculty from 2015 to 2017.

In addition to his teaching and faculty governance roles, Capozzola is an active proponent of public history. He served as a co-curator of “The Volunteers: Americans Join World War I, 1914-1919,” a multi-platform public history initiative marking the centennial of World War I. He currently serves as academic advisor for the online educational project Filipino Veterans Recognition and Education Project.

His interest in public-facing education projects has grown, he says, “because the best parts of my job involve sharing history with excited and curious audiences. It’s very clear that those audiences are enormous and global, and that learners bring their own backgrounds, questions, and interests to the kind of history we produce at MIT.”

This enthusiasm extends to MIT Open Learning as well: Capozzola is eager to work with MIT faculty to leverage digital learning to be more nimble in their teaching, and supporting learners in moving smoothly through MIT's digital resources.

“What has drawn me to Open Learning from the beginning is my own curiosity about how we can teach better and differently, as well as the creativity of the people who are involved. Everything that I've worked on with Open Learning has been very collaborative,” says Capozzola. “People bring all different kinds of expertise: about technology, about the science of learning, about students at MIT and learners beyond. Only by getting everybody together and collaborating can we produce these amazing resources.”

In his new role as senior associate dean, he’s looking forward to collaborating with faculty and instructors across all of the Institute’s schools and departments, helping them to work with MIT Open Learning through every possible avenue and lowering barriers to participation.

“Open Learning is a critical component of the overall MIT mission. We need to share MIT’s knowledge with the nation and the world in the 21st century. One way to think about that is, if we’re doing something at MIT that we think advances the mission but it’s not on Open Learning, then we’re not advancing MIT’s mission,” he says. This includes offering courses through MITx and OCW, as well as working with the Residential Education team and the Digital Learning Lab to incorporate learning design and digital technologies into the classroom to improve teaching and learning at MIT.

“When it comes to technology in the classroom, I have a skeptical enthusiasm and an enthusiastic skepticism. I want to think about what it means to teach and learn at a residential university in the 2020s. We have all learned a lot of lessons about that during the pandemic, and now is a great moment to convene conversations within Open Learning, at MIT, and beyond. It’s time for thoughtful reflection about what we do and how we can engage the most people with as much of an MIT education as we can,” Capozzola says.

Another exciting opportunity Capozzola sees is guiding MIT Open Learning toward reflecting the Institute’s values and priorities as well as its knowledge. Working closely with dean for digital learning Cynthia Breazeal, who oversees MIT Open Learning’s professional and corporate education offerings and research and engagement units, Capozzola envisions developing new content and strategies that accelerate MIT’s efforts in diversity, equity, and inclusion; climate and sustainability; and more. 

“I really want Open Learning to reflect MIT. By which I mean, everyone at MIT should see themselves, their disciplines, and their high standards for teaching, learning, research represented in Open Learning,” Capozzola says. “The staff and leadership of Open Learning have worked hard over the last 10 years to do that. I’m looking forward to thinking with Open Learning and in dialog with MIT about our priorities, our values, and our next steps.”

de MIT News https://ift.tt/H5sf2py

Engineers repurpose 19th-century photography technique to make stretchy, color-changing films

Imagine stretching a piece of film to reveal a hidden message. Or checking an arm band’s color to gauge muscle mass. Or sporting a swimsuit that changes hue as you do laps. Such chameleon-like, color-shifting materials could be on the horizon, thanks to a photographic technique that’s been resurrected and repurposed by MIT engineers.

By applying a 19th-century color photography technique to modern holographic materials, an MIT team has printed large-scale images onto elastic materials that when stretched can transform their color, reflecting different wavelengths as the material is strained.

The researchers produced stretchy films printed with detailed flower bouquets that morph from warm to cooler shades when the films are stretched. They also printed films that reveal the imprint of objects such as a strawberry, a coin, and a fingerprint.

The team’s results provide the first scalable manufacturing technique for producing detailed, large-scale materials with “structural color” — color that arises as a consequence of a material’s microscopic structure, rather than from chemical additives or dyes.

“Scaling these materials is not trivial, because you need to control these structures at the nanoscale,” says Benjamin Miller, a graduate student in MIT’s Department of Mechanical Engineering. “Now that we’ve cleared this scaling hurdle, we can explore questions like: Can we use this material to make robotic skin that has a human-like sense of touch? And can we create touch-sensing devices for things like virtual augmented reality or medical training? It’s a big space we’re looking at now.”

The team’s results appear today in Nature Materials. Miller’s co-authors are MIT undergraduate Helen Liu, and Mathias Kolle, associate professor of mechanical engineering at MIT.

Hologram happenstance

Kolle’s group develops optical materials that are inspired by nature. The researchers have studied the light-reflecting properties in mollusc shells, butterfly wings, and other iridescent organisms, which appear to shimmer and shift their color due to microscopic surface structures. These structures are angled and layered to reflect light like miniature colored mirrors, or what engineers refer to as Bragg reflectors.

Groups including Kolle’s have sought to replicate this natural, structural color in materials using a variety of techniques. Some efforts have produced small samples with precise nanoscale structures, while others have generated larger samples, but with less optical precision.

As the team writes, “an approach that offers both [microscale control and scalability] remains elusive, despite several potential high-impact applications.”

While puzzling over how to resolve this challenge, Miller happened to visit the MIT Museum, where a curator talked him through an exhibit on holography, a technique that produces three-dimensional images by superimposing two light beams onto a physical material.

“I realized what they do in holography is kind of the same thing that nature does with structural color,” Miller says.

That visit spurred him to read up on holography and its history, which led him back to the late 1800s, and Lippmann photography — an early color photography technique invented by Franco-Luxembourgish physicist Gabriel Lippmann, who later won the Nobel Prize in Physics for the technique.

Lippmann generated color photos by first setting a mirror behind a very thin, transparent emulsion — a material that he concocted from tiny light-sensitive grains. He exposed the setup to a beam of light, which the mirror reflected back through the emulsion. The interference of the incoming and outgoing light waves stimulated the emulsion’s grains to reconfigure their position, like many tiny mirrors, and reflect the pattern and wavelength of the exposing light.

Using this technique, Lippmann projected structurally colored images of flowers and other scenes onto his emulsions, though the process was laborious. It involved hand-crafting the emulsions and waiting for days for the material to be sufficiently exposed to light. Because of these limitations, the technique largely faded into history.

A modern twist

Miller wondered if, paired with modern, holographic materials, Lippmann photography could be sped up to produce large-scale, structurally colored materials. Like Lippmann’s emulsions, current holographic materials consist of light-sensitive molecules that, when exposed to incoming photons, can cross-link to form colored mirrors.

“The chemistries of these modern holographic materials are now so responsive that it’s possible to do this technique on a short timescale simply with a projector,” Kolle notes.

In their new study, the team adhered elastic, transparent holographic film onto a reflective, mirror-like surface (in this case, a sheet of aluminum). The researchers then placed an off-the-shelf projector several feet from the film and projected images onto each sample, including Lippman-esque bouquets.

As they suspected, the films produced large, detailed images within several minutes, rather than days, vividly reproducing the colors in the original images.

They then peeled the film away from the mirror and stuck it to a black elastic  silicone backing for support. They stretched the film and observed the colors change — a consequence of the material’s structural color: When the material stretches and thins out, its its nanoscale structures reconfigure to reflect slightly different wavelengths, for instance, changing from red to blue.  

The team found the film’s color is highly sensitive to strain. After producing an entirely red film, they adhered it to a silicone backing that varied in thickness. Where the backing was thinnest, the film remained red, whereas thicker sections strained the film, causing it to turn blue.

Similarly, they found that pressing various objects into samples of red film left detailed green imprints, caused by, say, the seeds of a strawberry and the wrinkles of a fingerprint.

Interestingly, they could also project hidden images, by tilting the film at an angle with respect to the incoming light when creating the colored mirrors. This tilt essentially caused the material’s nanostructures to reflect a red-shifted spectrum of light. For instance, green light used during material exposure and development would lead to red light being reflected, and red light exposure would give structures that reflect infrared — a wavelength that is not visible to humans. When the material is stretched, this otherwise invisible image changes color to reveal itself in red.

“You could encode messages in this way,” Kolle says.

Overall, the team’s technique is the first to enable large-scale projection of detailed, structurally colored materials.

“The beauty of this work is the fact that they have developed a simple yet extremely effective way to produce large-area photonic structures,” says Sylvia Vignolini, professor of chemistry and bio-materials at the University of Cambridge, who was not involved in the study. “This technique could be game-changing for coatings and packaging, and also for wearables.”

Indeed, Kolle notes that the new color-changing materials are easily integrated into textiles.

“Lippmann’s materials wouldn’t have allowed him to even produce a Speedo,” he says. “Now we could make a full leotard.”

Beyond fashion and textiles, the team is exploring applications such as color-changing bandages, for use in monitoring bandage pressure levels when treating conditions such as venous ulcers and certain lymphatic disorders.

This research was supported, in part, by The Gillian Reny Stepping Strong Center for Trauma Innovation at the Brigham and Women’s Hospital, the National Science Foundation, the MIT Deshpande Center for Technological Innovation, Samsung, and the MIT ME MathWorks seed fund.

de MIT News https://ift.tt/nwqM3S8

J-PAL North America launches two partnership opportunities to research social programs

J-PAL North America, a research center in the MIT Department of Economics, has opened two Evaluation Incubators: the Housing Stability Evaluation Incubator and State and Local Evaluation Incubator. J-PAL North America’s Evaluation Incubators equip partners to use randomized evaluations — the most scientifically rigorous method used to study program impact — in order to generate evidence about programs and policies that alleviate poverty. 

Evaluation Incubators offer organizations and government agencies the opportunity to expand the base of evidence on solutions to pressing policy questions. Both incubators provide selected partners with technical assistance, training, flexible funding, and connections to academic researchers to build their capacity to generate evidence and utilize data to drive decision-making. Partners may also have the opportunity to leverage these resources to carry out a randomized evaluation in collaboration with J-PAL-affiliated researchers.

Growing housing instability necessitates innovation and evidence 

In the United States, nearly 600,000 people experience homelessness on a given night, and over 1.4 million people access shelter in a given year. Homelessness disproportionately affects Black people, LGBTQ individuals, people with severe mental illness, veterans, survivors of domestic violence, and members of several other marginalized communities. Rigorous evidence on strategies to reduce homelessness and foster housing stability is critical to ensure that people have a safe and stable place to live.

Randomized evaluations have already demonstrated their capacity to transform the field. For example, rigorous evidence on Housing First programs — which provide housing with no preconditions — shifted narratives and norms of service provision when results showed marked reductions in chronic homelessness. 

As the scope and complexity of housing instability grow in the United States, so too does the need for rigorous research to identify the most effective strategies to end homelessness. Current partners are focusing their research specifically on cash transfers. 

Nicole Moler, impact analyst at Compass Family Services, notes the organization’s excitement about the potential to implement and rigorously evaluate a cash transfer program with support from J-PAL North America: “We’ve had the opportunity to give small amounts of funds to some clients before and have seen the difference it can make, but haven’t had the data to back up our observations. We are excited to explore randomized evaluation because we want to really know if it will work at a larger scale. At the same time, Compass has a desire to make sure the evaluation is done carefully, ethically, and rigorously. This is where J-PAL’s expertise comes in.” 

While cash transfers are a promising solution, unanswered policy questions range widely, from how to address the shortage of housing supply to what bundle of services are most effective and for whom. As such, J-PAL North America’s Housing Stability Evaluation Incubator invites letters of interest from service providers seeking to evaluate any programs aiming to reduce homelessness or foster long-term housing stability. 

Continuums of care, community-based organizations, and other service providers that wish to learn more about this opportunity can contact the Homelessness and Housing Stability team directly or see below for more information. Interested government agencies seeking to address homelessness can direct their inquiries and letters of interest to the State and Local Evaluation Incubator.

State and local governments are uniquely positioned to create and use evidence that addresses social challenges 

State and local governments play a central role in building rigorous evidence to inform poverty alleviation and promote well-being. They make decisions about funding public schools and community colleges, consider how prescription monitoring practices can be adjusted to prevent opioid overprescription, and take on new practices to address disparities in the criminal justice system, to name just a few examples. Researchers and state and local leaders have shared key areas where new or additional research at the state and local level is best positioned to address barriers to mobility from poverty. This kind of evidence is critical to informing how state and local governments fund and develop these programs. 

State agencies, county authorities, and city offices can also act as catalysts of innovation by testing new policy approaches to foster upward mobility and community well-being. In turn, decision-makers can use this evidence to improve and scale effective policies and programs to reach more people. For example, California’s Shasta County Superior Court partnered with J-PAL North America to better understand how to reduce failure to appear (FTA) — when a defendant does not attend a scheduled court hearing — in their district. FTA can be costly for those summoned to court. Even for minor offenses, an FTA can lead to additional fines, and in some cases, an arrest warrant, which can have serious long-term effects on an individual's record. 

Shawn Watts of the Shasta County Superior Court speaks to how critical their partnership with J-PAL North America was in building evidence to address this challenge: “Prior to this project, the court had no experience with randomized evaluations. The education we received on this research tool was instrumental to our understanding of how we should approach solutions to our issues. This study showed us that texting could reduce FTAs in our general population, especially if we had more reliable cell phone numbers for our defendants.”  

This year, to maximize the impacts of the American Rescue Plan, investing in effective evidence-based programming is needed more than ever as communities rebuild from the pandemic. Any state or local government agency interested in testing a promising social program, including those focused on housing stability and homelessness prevention, is encouraged to apply to the State and Local Evaluation Incubator. 

Information for prospective applicants

Interested organizations are encouraged to submit a letter of interest by Oct. 17 to either the Housing Stability Evaluation Incubator or State and Local Evaluation Incubator. If you believe you may be eligible for both opportunities, please apply for the State and Local Evaluation Incubator. Detailed instructions on how to apply to the incubators can be found on their respective webpages. 

J-PAL North America will host a webinar on Aug. 25 at 1:30 p.m. ET to provide more information on both Evaluation Incubators, review the application process, and answer questions. Please contact Laina Sonterblum with any questions about the Housing Stability Evaluation Incubator and Mera Cronbaugh with any questions regarding the State and Local Evaluation Incubator.

de MIT News https://ift.tt/Iy953Zz