Tech

Software that purportedly reads emotions in faces is being deployed or tested for a variety of purposes, including surveillance, hiring, clinical diagnosis, and market research. But a new scientific report finds that facial movements are an inexact gauge of a person's feelings, behaviors or intentions.

"It is not possible to confidently infer happiness from a smile, anger from a scowl or sadness from a frown, as much of current technology tries to do when applying what are mistakenly believed to be the scientific facts," a group of leading experts in psychological science, neuroscience and computer science write in their comprehensive research review.

The report appears in Psychological Science in the Public Interest, a journal of the Association for Psychological Science, and is authored by Lisa Feldman Barrett of Northeastern University, Ralph Adolphs of the California Institute of Technology, Stacy Marsella of Northeastern University and the University of Glasgow, Aleix M. Martinez of The Ohio State University and Seth D. Pollak of the University of Wisconsin-Madison.

The authors note that the general public and some scientists believe that there are unique facial expressions that reliably indicate six emotion categories: anger, sadness, happiness, disgust, fear, and surprise. But in reviewing more than 1,000 published findings about facial movements and emotions, they found that typical study designs don't capture the real-life differences in the way people convey and interpret emotions on faces. A scowl or a smile can express more than one emotion depending on the situation, the individual or the culture, they say.

"People scowl when angry, on average, approximately 25 percent of the time, but they move their faces in other meaningful ways when angry," Barrett explains. "They might cry, or smile, or widen their eyes and gasp. And they also scowl when not angry, such as when they are concentrating or when they have a stomach ache. Similarly, most smiles don't imply that a person is happy, and most of the time people who are happy do something other than smile."

In a separate article in the journal, Alan Cowen and Dacher Keltner of the University of California, Berkeley; Disa Sauter, University of Amsterdam; and Jessica L. Tracy of the University of British Columbia note that most scientists agree that facial expressions are meaningful, even if they don't follow a one-to-one match with six basic emotion categories. They propose a new model for studying emotion-related responses in all their complexity and variations. This approach would measure not only facial cues, but also body movements, voice fluctuations, head movements and other indicators to capture such nuanced responses as smiles of embarrassment or sympathetic vocalizations, they say.

The report's conclusions have broad implications, according to the authorship team. The FBI and the Transportation Security Administration have trained agents in the past to assess smiling, scowling and other facial movements to identify and stop potential terrorists. Law enforcement agencies in the United States and Europe are now experimenting with technologies designed to automate emotion detection through facial scans. Some companies are experimenting with software to track the facial movements of job applicants during interviews. Such technology might be able to detect facial movements, but they do not detect the psychological meaning of those facial movements, Barrett and co-authors say.

"We thought this was an especially important issue to address because of the way so-called 'facial expressions' are being used in industry, educational and medical settings, and in national security," say Barrett and her co-authors.

Health systems in low- and middle-income countries (LMICs) are poorly prepared for the increasing number of people with high blood pressure, with more than two-thirds of people affected going without treatment, according to a new study led by researchers at Harvard T.H. Chan School of Public Health in collaboration with colleagues from more than 40 institutions around the world, including several ministries of health.

The study examined household survey data for 1 million people living in 44 LMICs and found that less than half of those with high blood pressure are properly diagnosed. Among those with the condition, only 30% are treated and only 10% have the disease under control. These proportions, however, varied widely between countries, with some performing consistently better than others even after taking into account differences in economic development.

The study will be published in The Lancet on Thursday, July 18, 2019.

"Our study shows not only that care for hypertension in these countries is severely inadequate, but also where exactly patients in each country are being lost in the care process," said Pascal Geldsetzer, postdoctoral research fellow at Harvard T.H. Chan School of Public Health and first author of the study.

For the study, researchers used a cascade of care approach, which looked at the numbers of people with hypertension who had been screened, diagnosed, treated, and controlled, to determine how well the health systems of the various countries are treating people with hypertension.

The group carried out its research using household surveys, including the World Health Organization's STEPS survey, in order to have a uniform approach when obtaining data on established risk factors. Researchers first determined how many people suffered from high blood pressure. They also determined how many of these patients had ever had their blood pressure measured prior to the survey, as well as how many were diagnosed and were taking treatment. Finally, they analyzed how many patients successfully controlled the disease with medication.

"The low proportion of patients with high blood pressure getting the treatment they need, along with the growing number of patients with high blood pressure, suggests the very urgent need for population-level prevention, especially policies that get salt and trans fat out of the food supply, promote fruits and vegetables, reduce air pollution, and address excessive consumption of tobacco and alcohol," said Lindsay Jaacks, assistant professor of global health at Harvard Chan School.

A special Lancet Series on Oral Health, published today in The Lancet, presents an "urgent need for radical reform" of oral healthcare to prioritize prevention and integrate dentistry into primary care. The series is comprised of two papers, both co-authored by Habib Benzian, DDS, MScDPH, PhD, the associate director of global health and policy for NYU College of Dentistry's World Health Organization (WHO) Collaborating Center for Quality-improvement, Evidence-based Dentistry--the only WHO Collaborating Center on oral health in the Americas.

Oral diseases, including tooth decay, gum disease, and oral cancers, are major global public health problems affecting half of the world's population, but they have gone largely unnoticed.

"This immense disease burden is an important starting point to rethink the current oral healthcare approach, which leaves large parts of the world's population without access to even basic dental care," said Benzian, who is also an adjunct professor at NYU's Colleges of Dentistry and Global Public Health. "By dedicating a special series to oral health, The Lancet is creating visibility and urgency for a long-overlooked public health crisis, and my co-authors and I offer solutions to improve global oral health and reduce inequalities."

In high-income countries, oral healthcare is dominated by treatment--including high-tech, high-cost interventions--rather than tackling the underlying causes of disease and preventing it. In low- and middle-income countries, however, dentistry is largely unavailable and unaffordable.

As part of prioritizing prevention, the Lancet Series also examines the need to reduce sugar, alcohol, and tobacco consumption to improve oral health and other chronic conditions. Sugar, in particular, is recognized as the leading cause of tooth decay. The authors call for the adoption of a range of policy changes to protect and improve health, including taxing sugary drinks and stronger regulation on advertising targeting children.

The Lancet Series also explores the problematic rift among oral healthcare, mainstream healthcare, and health policy. The series provides recommendations for system-wide reforms.

"Oral healthcare is isolated from mainstream healthcare. Thus we need a fundamentally different approach that integrates the two. A reformed oral healthcare system needs to incentivize prevention, embrace a wider team of health professionals, including primary care and mid-level providers, use evidence-based treatments, and respond to the diverse needs of local populations," said Benzian. "The growing global momentum towards universal health coverage provides a unique opportunity to reform oral healthcare. Bold efforts in Brazil and Thailand have demonstrated that major reforms are possible and can benefit oral health. Countries like France have embraced the principles of universal health coverage and are now changing their systems to include oral healthcare in innovative ways."

"Prevention is a key factor in reforming oral healthcare, and it is currently underutilized," added Richard Niederman, DMD, chair of the Department of Epidemiology & Health Promotion at the NYU College of Dentistry and director of the WHO Collaborating Center. "Our studies of cavity prevention programs for U.S. children show that these programs prevent tooth decay, improve school attendance and performance, and are cost-effective and cost-saving."

A gene linked to Alzheimer's Disease may impact cognitive health much sooner than previously realized.

The APOE gene creates a protein, apolipoprotein E, which packages cholesterol and other fats to transport them through the bloodstream. There are three versions, or alleles, of APOE. One of those is the APOE4 allele, present in about 15 percent of the population. APOE4 carriers are up to three times more likely to develop late-onset Alzheimer's disease, which occurs in people 65 and older.

It's been demonstrated the gene is linked to changes in cognitive ability that are noticeable as early as one reaches his or her 50s. But new research from UCR professor Chandra Reynolds and her colleagues holds that APOE4 starts manifesting much earlier--before adulthood.

In the journal Neurobiology of Aging, Reynolds asserts that those carrying the APOE4 gene score lower on IQ tests during childhood and adolescence. And the effect was stronger in girls than in boys.

The study involved analysis of three to four decades-old studies, the Colorado Adoption Project and the Longitudinal Twin Study, that included genotyping data from 1,321 participants when they were 6 ½ to 18 years old. Gender among participants was split almost evenly, and 92 percent of the participants were white, with 8 percent from other races. The findings are based on three IQ assessments between childhood and adolescence.

Overall, Full-scale IQ scores were lower by 1.91 points for each APOE4 allele individuals may carry. Boys scored .33 points lower on IQ tests, and women scored almost 3 points lower for each APOE4 allele. The traits most affected related to reasoning. The effect of E4 on IQ performance multiplies with each E4 allele present: a person can have up to two APOE4 alleles.

The IQ difference seems small. But long-term, Reynolds said the findings can mean fewer cognitive reserves as the APOE4 carrier ages, with the disadvantage becoming progressively magnified. Cognitive reserve is the brain's ability to navigate problems and improvise. Cognitive Reserve Theory holds that people with fewer reserves have more trouble withstanding disease as they age.

Research has also shown an association between lower childhood IQ and increased biological aging--cell and tissue damage--and cardiovascular disease before age 65.

"Our results suggest that cognitive differences associated with APOE may emerge early and become magnified later in the life course, and if so, childhood represents a key period of intervention to invest in and boost reserves," Reynolds wrote in the paper, titled "APOE effects on cognition from childhood to adolescence."

ANN ARBOR--Individuals on teams of diverse people working together can have better outcomes than those on teams with similar individuals, research as shown.

But a new study by University of Michigan and Michigan State University researchers found that the very individuals who add diversity to their science teams surprisingly do not experience positive outcomes.

Researchers examined diversity in two categories: demographic (race, sex, gender identity, sexual orientation, nationality) and scientific (career stage, academic discipline, tenure on team).

A sample of 266 participants from 105 National Science Foundation-funded environmental science teams completed questionnaires about individual and team diversity, their satisfaction with their teams and authorship practices, and perceptions of the frequency of data sharing. They also disclosed perceptions of their team climate, including team collaboration, inclusion and procedural justice, which focused on influencing team policies related to research.

Participants with more underrepresented demographic characteristics (e.g., black women, gay men not born in the U.S.) perceived their team climate--or attitudes and expectations on the team--to be more negative. This was associated with lower team satisfaction and more negative perceptions of authorship and data sharing on their teams, said study lead author Isis Settles, U-M professor of psychology and Afroamerican and African studies.

However, regardless of their own demographic characteristics, individuals on diverse teams perceived their climate more positively than individuals on more homogeneous teams.

Creating successful teams that are demographically and scientifically diverse is not a simple matter of recruiting more individuals from underrepresented groups and combining team members from a variety of backgrounds, Settles said. Diverse teams can struggle with allocation of credit, differences in perspectives and unequal power dynamics, she said.

What must happen--as the findings indicate--is improved outcomes in procedural justice, collaboration and inclusion. Team policies must be clear and openly discussed, and transparent policies and procedures must be followed to alleviate power imbalances, Settles said.

Kevin Elliott, MSU professor and co-lead author of the study, said teams must be mindful of the experiences of all members, especially those who contribute to demographic diversity.

"It is critical to provide these individuals with adequate support and recognition," he said.

Settles agreed, stating it starts with team leaders who can create norms that support the contributions of all members. This may involve creating policies and practices collaboratively and allowing for respectful conflict.

"And, of course, increasing the number of underrepresented team members can also reduce 'token effects,'" she said, noting that individuals can feel stress and social isolation because they have characteristics unique to the group.

The bottom line is that diversity benefits teams as a whole, and when combined with inclusive practices, it can help to advance the careers of underrepresented minorities in STEM, Elliott said.

Has your steering wheel been too hot to touch this summer? A new thermoelectric material reported in the journal Science could offer relief.

The widespread adoption of thermoelectric devices that can directly convert electricity into thermal energy for cooling and heating has been hindered, in part, by the lack of materials that are both inexpensive and highly efficient at room temperature.

Now researchers from the University of Houston and the Massachusetts Institute of Technology have reported the discovery of a new material that works efficiently at room temperature while requiring almost no costly tellurium, a major component of the current state-of-the-art material.

The work, described in a paper published online by Science Thursday, July 18, has potential applications for keeping electronic devices, vehicles and other components from overheating, said Zhifeng Ren, corresponding author on the work and director of the Texas Center for Superconductivity at UH, where he is also M.D. Anderson Professor of Physics.

"We have produced a new material, which is inexpensive but still performs almost as well as the traditional, more expensive material," Ren said. The researchers say future work could close the slight performance gap between their new material and the traditional material, a bismuth-tellurium based alloy.

Thermoelectric materials work by exploiting the flow of heat current from a warmer area to a cooler area, and thermoelectric cooling modules operate according to the Peltier effect, which describes the transfer of heat between two electrical junctions.

Thermoelectric materials can also be used to turn waste heat - from power plants, automobile tailpipes and other sources - into electricity, and a number of new materials have been reported for that application, which requires materials to perform at far higher temperatures.

Thermoelectric cooling modules have posed a great challenge because they have to work at cooler temperatures, where the thermoelectric figure-of-merit, or ZT, is low because it is dependent on temperature. The figure-of-merit is a metric used to determine how efficiently a thermoelectric material works.

Despite the challenge, thermoelectric cooling modules also, at least for now, offer more commercial potential, in part because they can operate for a long lifespan at cooler temperatures; thermoelectric power generation is complicated by issues related to the high temperatures at which it operates, including oxidation and thermal instability.

The market for thermoelectric cooling is growing. "The global thermoelectric module market was worth ~0.6 billion US dollars in 2018 and it is anticipated to reach ~1.7 billion US dollars by 2027," the researchers wrote.

Bismuth-tellurium alloys have been considered the best-performing material for thermal cooling for decades, but the researchers said the high cost of tellurium has limited widespread use. Jun Mao, a post-doctoral researcher at UH and first author on the paper, said the cost has recently dropped but remains about $50/kilogram. That compares to about $6/kilogram for magnesium, a primary component of the new material.

In addition to Ren and Mao, additional authors on the paper include Hangtian Zhu, Zihang Liu and Geethal Amila Gamage, all of the UH Department of Physics and TcSUH, and Zhiwei Ding and Gang Chen of the Department of Mechanical Engineering at the Massachusetts Institute of Technology.

They reported that the new material, comprised of magnesium and bismuth and created in a form carrying a negative charge, known as n-type, was almost as efficient as the traditional bismuth-tellurium material. That, combined with the lower cost, should expand the use of thermoelectric modules for cooling, they said.

To produce a thermoelectric module using the new material, researchers combined it with a positive-charge carrying, or p-type, version of the traditional bismuth-tellurium alloy. Mao said that allowed them to use just half as much tellurium as most current modules.

Because the cost of materials accounts for about one-third of the cost of the device, that savings adds up, he said.

The new material also more successfully maintains electrical contact than most nanostructured materials, the researchers reported.

A computational simulation suggests that insects may be capable of using the properties of light from the sky to determine their compass direction with an error of less than two degrees. Evripidis Gkanias of the University of Edinburgh, U.K., and colleagues present their findings in PLOS Computational Biology.

Several insects, including honeybees, locusts, and monarch butterflies, use the position of the sun to guide their travel. Even when the sun is not visible, these insects can sense the polarization of light in the sky and use it to estimate the sun's position. However, the precise neural processes by which insects transform properties of light from the sky into an accurate compass sense are unclear.

To explore this question, Gkanias and colleagues built a computational simulation that incorporates a hypothetical system of neurons that an insect's brain could potentially use to reconstruct the sun's position from skylight properties detected by the eye. The simulation also incorporates known physical properties of light from the sky, the layout of the insect eye, and other biological parameters determined from previous research on the insect brain.

The simulation suggests that insects should be able to use sky polarization to estimate their compass direction to an error of less than two degrees, without ambiguity and with the ability to correct for passing time. It also shows that this compass information is accurate enough to allow an insect to estimate the distance and direction back to its home, even after moving a long distance away in a random path over rough terrain.

"This highly accurate insect compass could potentially be copied for development of a component in a cheap and self-contained positioning system," Gkanias says. "Such a system could serve as an alternative to GPS for navigation of outdoor robots."

The researchers are now building a prototype physical sensor based on the simulation used in this study. They plan to test the sensor under a real sky and on a real robot.

Inventors of centuries past and scientists of today have found ingenious ways to make our lives better with magnets - from the magnetic needle on a compass to magnetic data storage devices and even MRI (magnetic resonance imaging) body scan machines.

All of these technologies rely on magnets made from solid materials. But what if you could make a magnetic device out of liquids? Using a modified 3D printer, a team of scientists at Berkeley Lab have done just that. Their findings, to be published July 19 in the journal Science, could lead to a revolutionary class of printable liquid devices for a variety of applications from artificial cells that deliver targeted cancer therapies to flexible liquid robots that can change their shape to adapt to their surroundings.

"We've made a new material that is both liquid and magnetic. No one has ever observed this before," said Tom Russell, a visiting faculty scientist at Berkeley Lab and professor of polymer science and engineering at the University of Massachusetts, Amherst who led the study. "This opens the door to a new area of science in magnetic soft matter."

Jam sessions: making magnets out of liquids

For the past seven years, Russell, who leads a program called Adaptive Interfacial Assemblies Towards Structuring Liquids in Berkeley Lab's Materials Sciences Division, has focused on developing a new class of materials - 3D-printable all-liquid structures.

One day, Russell and the current study's first author Xubo Liu came up with the idea of forming liquid structures from ferrofluids, solutions of iron-oxide particles that become strongly magnetic, but only in the presence of another magnet. "We wondered, if a ferrofluid can become temporarily magnetic, what could we do to make it permanently magnetic, and behave like a solid magnet but still look and feel like a liquid?" said Russell.

To find out, Russell and Liu - a graduate student researcher in Berkeley Lab's Materials Sciences Division and a doctoral student at the Beijing University of Chemical Technology - used a 3D-printing technique they had developed with former postdoctoral researcher Joe Forth in Berkeley Lab's Materials Sciences Division to print 1 millimeter droplets from a ferrofluid solution containing iron-oxide-nanoparticles just 20 nanometers in diameter (the average size of an antibody protein.)

Using surface chemistry and sophisticated atomic force microscopy techniques at the Molecular Foundry, co-authors Paul Ashby and Brett Helms of Berkeley Lab revealed that the nanoparticles formed a solid-like shell at the interface between the two liquids through a phenomenon called "interfacial jamming," which causes the nanoparticles to crowd at the droplet's surface, "like the walls coming together in a small room jampacked with people," said Russell.

To make them magnetic, the scientists placed the droplets by a magnetic coil in solution. As expected, the magnetic coil pulled the iron-oxide nanoparticles toward it.

But when they removed the magnetic coil, something quite unexpected happened.

Like synchronized swimmers, the droplets gravitated toward each other in perfect unison, forming an elegant swirl. "Like little dancing droplets," said Liu.

Somehow, these droplets had become permanently magnetic. "We almost couldn't believe it," said Russell. "Before our study, people always assumed that permanent magnets could only be made from solids."

Measure by measure, it's still a magnet

All magnets, no matter how big or small, have a north pole and a south pole. Opposite poles are attracted to each other, while the same poles repel each other.

Through magnetometry measurements, the scientists found that when they placed a magnetic field by a droplet, all of the nanoparticles' north-south poles, from the 70 billion iron-oxide nanoparticles floating around in the droplet to the 1 billion nanoparticles on the droplet's surface, responded in unison, just like a solid magnet.

Key to this finding were the iron-oxide nanoparticles jamming tightly together at the droplet's surface. With just 8 nm between each of the billion nanoparticles, together they created a solid surface around each liquid droplet. Somehow, when the jammed nanoparticles on the surface are magnetized, they transfer this magnetic orientation to the particles swimming around in the core, and the entire droplet becomes permanently magnetic, just like a solid, Russell and Liu explained.

The researchers also found that the droplet's magnetic properties were preserved, even if they divided a droplet into smaller, thinner droplets about the size of a human hair, added Russell.

Among the magnetic droplets' many amazing qualities, what stands out even more, Russell noted, is that they change shape to adapt to their surroundings, morphing from a sphere to a cylinder to a pancake, or a tube as thin as a strand of hair, or even to the shape of an octopus - all without losing their magnetic properties.

The droplets' can also be tuned to switch between a magnetic mode and a nonmagnetic mode. And when their magnetic mode is switched on, their movements can be remotely controlled as directed by an external magnet, Russell added.

Liu and Russell plan to continue research at Berkeley Lab and other national labs to develop even more complex 3D-printed magnetic liquid structures, such as a liquid-printed artificial cell, or miniature robotics that move like a tiny propeller for noninvasive yet targeted delivery of drug therapies to diseased cells.

"What began as a curious observation ended up opening a new area of science," said Liu. "It's something all young researchers dream of, and I was lucky to have the chance to work with a great group of scientists supported by Berkeley Lab's world-class user facilities to make it a reality," said Liu.

The molten rock that feeds volcanoes can be stored in the Earth's crust for as long as a thousand years, a result which may help with volcanic hazard management and better forecasting of when eruptions might occur.

Researchers from the University of Cambridge used volcanic minerals known as 'crystal clocks' to calculate how long magma can be stored in the deepest parts of volcanic systems. This is the first estimate of magma storage times near the boundary of the Earth's crust and the mantle, called the Moho. The results are reported in the journal Science.

"This is like geological detective work," said Dr Euan Mutch from Cambridge's Department of Earth Sciences, and the paper's first author. "By studying what we see in the rocks to reconstruct what the eruption was like, we can also know what kind of conditions the magma is stored in, but it's difficult to understand what's happening in the deeper parts of volcanic systems."

"Determining how long magma can be stored in the Earth's crust can help improve models of the processes that trigger volcanic eruptions," said co-author Dr John Maclennan, also from the Department of Earth Sciences. "The speed of magma rise and storage is tightly linked to the transfer of heat and chemicals in the crust of volcanic regions, which is important for geothermal power and the release of volcanic gases to the atmosphere."

The researchers studied the Borgarhraun eruption of the Theistareykir volcano in northern Iceland, which occurred roughly 10,000 years ago, and was fed directly from the Moho. This boundary area plays an important role in the processing of melts as they travel from their source regions in the mantle towards the Earth's surface. To calculate how long the magma was stored at this boundary area, the researchers used a volcanic mineral known as spinel like a tiny stopwatch or crystal clock.

Using the crystal clock method, the researchers were able to model how the composition of the spinel crystals changed over time while the magma was being stored. Specifically, they looked at the rates of diffusion of aluminium and chromium within the crystals and how these elements are 'zoned'.

"Diffusion of elements works to get the crystal into chemical equilibrium with its surroundings," said Maclennan. "If we know how fast they diffuse we can figure out how long the minerals were stored in the magma."

The researchers looked at how aluminium and chromium were zoned in the crystals, and realised that this pattern was telling them something exciting and new about magma storage time. The diffusion rates were estimated using the results of previous lab experiments. The researchers then used a new method, combining finite element modelling and Bayesian nested sampling to estimate the storage timescales.

"We now have really good estimates in terms of where the magma comes from in terms of depth," said Mutch. "No one's ever gotten this kind of timescale information from the deeper crust."

Calculating the magma storage time also helped the researchers determine how magma can be transferred to the surface. Instead of the classical model of a volcano with a large magma chamber beneath, the researchers say that instead, it's more like a volcanic 'plumbing system' extending through the crust with lots of small 'spouts' where magma can be quickly transferred to the surface.

A second paper by the same team, recently published in Nature Geoscience, found that that there is a link between the rate of ascent of the magma and the release of CO2, which has implications for volcano monitoring.

The researchers observed that enough CO2 was transferred from the magma into gas over the days before eruption to indicate that CO2 monitoring could be a useful way of spotting the precursors to eruptions in Iceland. Based on the same set of crystals from Borgarhraun, the researchers found that magma can rise from a chamber 20 kilometres deep to the surface in as little as four days.

Our highly mobile mammalian tongues, which allow us to swallow chewed food and suckle milk as babies, may have evolutionary origins in some of our most early mammalioform ancestors, according to a new study, which finds remarkably complex and modern mammal-like hyoid bones in a newly discovered 165-million-year-old mammaliaform species. Unlike the wide, gulping mouths and non-muscular throats of reptiles or amphibians, the complex hyoid of mammals - a small, yet critical bone in the throat - enables the dexterous tongue and complex throat movements required for swallowing chewed food and suckling liquids like milk. However, while advanced food and liquid swallowing are characteristic traits in modern mammals, the evolutionary origin of the hyolingual structures that enables them is relatively unknown. Docodonts, an early branch of the mammaliaform ancestral tree, are important in understanding the early divergence and evolution of mammaliaform traits. Chang-Fu Zhou and colleagues report on the discovery of a new Jurassic-age docodontan fossil in China, which included an exceptionally well preserved and nearly intact hyoid apparatus, which was complex and very similar to those in modern mammals. Using the mammal-like morphology of the previously undescribed Microdocodon gracilis fossil, the authors were able to identify complex hyoid structures in several other early mammaliaforms. According to the results, the development of a complex and saddle-shaped mammal-like hyoid predates the separation of the middle ear from the mandible - a key evolutionary step among early diverging mammaliaforms. And, since Microdocodon is a basal species in the docodontan clade, the authors suggest that complex hyolingual structures may have been in place even before mammals. In a related Perspective, Simone Hoffman and David Krause discuss the study in more detail.

Chemical manufacturers consume a massive amount of energy each year separating and refining feedstocks to make a wide variety of products including gasoline, plastics and food.

In a bid to reduce the amount of energy used in chemical separations, researchers at the Georgia Institute of Technology are working on membranes that could separate chemicals without using energy-intensive distillation processes.

"The vast majority of separations out in the field in a variety of industries are thermally-driven systems such as distillation, and because of that we spend an inordinate amount of energy on these separation processes - something like 10 to 15 percent of the global energy budget is spent on chemical separations," said Ryan Lively, an associate professor in Georgia Tech's School of Chemical & Biomolecular Engineering. "Separations that avoid the use of heat and a chemical phase-change are much less energy intense. In practice, using them could produce a 90 percent reduction in energy cost."

Plastic membranes are already able to separate certain molecules based on size and other differences, such as in seawater desalination. But until now, most membranes have been unable to withstand harsh solvent-rich chemical streams while also performing challenging separation tasks.

In a study published July 18 in Chemistry of Materials and sponsored by the Department of Defense and the National Science Foundation, the researchers outline a process for taking a polymer-based membrane and infusing it with a metal oxide network. The resulting membrane is far more effective at standing up to harsh chemicals without degrading.

"After placing the pre-fabricated membrane inside of our reactor, we simply expose it to metal-containing vapors that infuse themselves inside the membrane material," said Mark Losego, an assistant professor in the School of Materials Science and Engineering. "This process is called vapor phase infiltration, and it creates a uniform network of metal oxide throughout the polymer membrane. We call it a 'hybrid' membrane."

Not only was the hybrid membrane better able to withstand solvents, its chemical separation capabilities also improved.

"Some chemicals that need to be separated are very similar in terms of their size, shape and other properties, which makes them even harder to process using membranes," Lively said. "These new hybrid membrane are much more selective. They can separate chemicals that are more similar to each other."

The research team, which included graduate students Fengyi Zhang, Emily McGuinness and Yao Ma, tested the new hybrid membranes in harsh chemicals such as tetrahydrofuran, dichloromethane and chloroform, organic solvents that dissolve the pure polymer membrane in minutes. The hybrid membranes remained stable for several months during testing.

The researchers also tested separating two chemicals very close in size. The hybrid membranes were able to differentiate aromatic molecules that differed in size by as little as 0.2 nanometers.

"One of the most exciting things about this work was how straightforward this process is from a manufacturing perspective," Losego said. "We're essentially taking pre-made membranes and applying a treatment to them. That's something that would be very simple to translate to an industrial scale."

Future research on the membranes will involve looking at how to fine tune the oxide infusions and make new types of hybrid membranes capable of separating a variety of other chemicals.

AMHERST, Mass. - Conventional magnets are hard and rigid but have made great contributions to society and to modern industry, says materials scientist Thomas Russell of the University of Massachusetts Amherst. But this award-winning innovator dreamed of more - what if magnets could be soft, flowable as liquid and malleable to conform to a limited space?

In an article in this week's Science, he and first author Xubo Liu from Beijing University of Chemical Technology, with others at Lawrence Berkeley National Laboratory and the University of California, Berkeley, report on a simple way they developed to transform paramagnetic ferrofluids - plain metal particles in suspension - into a magnetic state. The new ferromagnetic liquid droplets "represent a milestone for the further development of magnetic materials," Russell says.

This means that by applying an external magnetic field, scientists can control liquid devices made this way, like waving Harry Potter's wand, he suggests, "which opens promising research and application areas such as liquid actuators, liquid robotics and active-matter delivery."

As the polymer scientist explains, he, Liu and the team used iron oxide nanoparticles in a special oil-polymer mixture to transform paramagnetic ferrofluid into the ferromagnetic state at room temperature. Because of nanoparticle-polymer mix interactions, the resulting ultra-soft droplet has magnetic properties similar to solid magnets but with liquid characteristics.

At nano scale, traditional ferromagnetic materials become magnetic only in the presence of a magnetic field. Based on these special physical properties, ferrofluids are already used in electrical devices, medical applications, mechanical engineering and materials science research, he notes.

Russell, who is also a visiting professor at the Berkeley National Lab, adds that the technique extends scientific knowledge of magnetic materials, and should encourage research into the deep-seated mechanism of how liquid magnets form. "This will facilitate the development of relative advanced instruments and new material theories," he predicts. "These amazing liquid magnetic materials will attract attention in biology, physics and chemistry."

A thousand years ago, he reflects, European voyagers used compasses they made from magnetite dug from the Earth to explore and discover new continents. For centuries, people learned to build smart magnetic devices to improve quality of life. "Such jumps in science and technology are always followed by a sudden emergence of a new material or theory," Russell notes.

He and colleagues hope that the new, reconfigurable ferromagnetic liquid droplets they describe will provide more such possibilities, such as magnetically actuated liquid robotics, liquid vessels for delivering active matter and information technology with programmable liquid droplet patterns.

Seeing was believing until technology reared its mighty head and gave us powerful and inexpensive photo-editing tools. Now, realistic videos that map the facial expressions of one person onto those of another, known as deepfakes, present a formidable political weapon.

But whether it's the benign smoothing of a wrinkle in a portrait, or a video manipulated to make it look like a politician saying something offensive, all photo editing leaves traces for the right tools to discover.

Research led by Amit Roy-Chowdhury's Video Computing Group at the University of California, Riverside has developed a deep neural network architecture that can identify manipulated images at the pixel level with high precision. Roy-Chowdhury is a professor of electrical and computer engineering and the Bourns Family Faculty Fellow in the Marlan and Rosemary Bourns College of Engineering.

A deep neural network is what artificial intelligence researchers call computer systems that have been trained to do specific tasks, in this case, recognize altered images. These networks are organized in connected layers; "architecture" refers to the number of layers and structure of the connections between them.

Objects in images have boundaries and whenever an object is inserted or removed from an image, its boundary will have different qualities than the boundaries of objects in the image naturally. For example, someone with good Photoshop skills will do their best to make the inserted object looks as natural as possible by smoothing these boundaries.

While this might fool the naked eye, when examined pixel by pixel, the boundaries of the inserted object are different. For example, they are often smoother than the natural objects. By detecting boundaries of inserted and removed objects, a computer should be able to identify altered images.

The researchers labeled nonmanipulated images and the relevant pixels in boundary regions of manipulated images in a large dataset of photos. The aim was to teach the neural network general knowledge about the manipulated and natural regions of photos. They tested the neural network with a set of images it had never seen before, and it detected the altered ones most of the time. It even spotted the manipulated region.

"We trained the system to distinguish between manipulated and nonmanipulated images, and now if you give it a new image it is able to provide a probability that that image is manipulated or not, and to localize the region of the image where the manipulation occurred," Roy-Chowdhury said.

The researchers are working on still images for now, but they point out that this can also help them detect deepfake videos.

"If you can understand the characteristics in a still image, in a video it's basically just putting still images together one after another," Roy-Chowdhury said. "The more fundamental challenge is probably figuring out whether a frame in a video is manipulated or not."

Even a single manipulated frame would raise a red flag. But Roy-Chowdhury thinks we still have a long way to go before automated tools can detect deepfake videos in the wild.

"It's a challenging problem," Roy-Chowdhury said. "This is kind of a cat and mouse game. This whole area of cybersecurity is in some ways trying to find better defense mechanisms, but then the attacker also finds better mechanisms."

He said completely automated deepfake detection might not be achievable in the near future.

"If you want to look at everything that's on the internet, a human can't do it on the one hand, and an automated system probably can't do it reliably. So it has to be a mix of the two," Roy-Chowdhury said.

Deep neural network architectures can produce lists of suspicious videos and images for people to review. Automated tools can reduce the amount of data that people -- like Facebook content moderators -- have to sift through to determine if an image has been manipulated.

For this use, the tools are right around the corner.

"That probably is something that these technologies will contribute to in a very short time frame, probably in a few years," Roy-Chowdhury said.

Leukemia stem cells protect themselves against the immune defense by suppressing a target molecule for killer cells. This protective mechanism can be tricked with drugs. In the journal "Nature", scientists from Basel, Tübingen and Heidelberg describe the new therapeutic approaches that can possibly be derived from these results.

Patients with acute myeloid leukemia (AML) often relapse after apparently successful treatment. Leukemia stem cells that survive the therapy are responsible for the return of the disease. Scientists can partially explain this phenomenon: The stem cells have protective mechanisms that make them resistant to chemotherapy. But how do they manage to escape the immune defense?

A team of scientists from the Universities of Basel and Tübingen, the German Cancer Research Center (DKFZ), the Heidelberg Institute for Stem Cell Technology and Experimental Medicine (HI-STEM) and the German Cancer Consortium (DKTK) have now investigated this phenomenon and discovered a surprising mechanism.

The researchers analyzed the leukemia cells of 175 AML patients and found that the cancer stem cells suppress the NKG2D-L proteins on their surface. These proteins enable the natural killer cells (NK cells) to recognize damaged and infected cells as well as cancer cells and kill them if necessary. In this way, the leukemia stem cells escape destruction by the immune system. Leukemia cells without stem cell properties, on the other hand, present these target molecules on their surface and are therefore kept in check by the NK cells.

In mice to which AML cells from patients had been transferred, the researchers showed that while the normal AML cells (without stem cell properties) were controlled by the NK cells, the NKG2D-L-negative leukemia stem cells escaped the killer squad. "Such a connection between stem cell properties and the ability to escape the immune system was unknown until now," said Claudia Lengerke from the University Hospital of Basel and the University of Basel. "An essential mechanism of this immune resistance in leukemia stem cells is apparently the suppression of danger signals such as NKG2D-L on the cell surface," added Helmut Salih from the University Hospital of Tübingen and the German Cancer Consortium DKTK.

What is behind this extraordinary protective mechanism? The scientists noticed that leukemia stem cells produce a particularly high amount of PARP1, an enzyme that apparently blocks NKG2D-L production. Preclinical experiments with mice to which human leukemia cells had been transferred showed that PARP1 actually plays an important role in immune escape: If these animals were treated with drugs that inhibit PARP1, the leukemia stem cells again expressed NKG2D-L on their surface - and were then recognized and eliminated by NK cells.

Cancer therapies involving the immune system have been successfully applied for many years in the form of allogeneic stem cell transplantation in AML patients in certain disease situations. In recent years, further novel immunotherapeutic approaches have been developed which are currently being clinically tested. "Our results show how cancer stem cells cleverly trick the immune system. The elucidation of the underlying mechanism now makes it possible to counterattack," says Andreas Trumpp, German Cancer Research Center and HI-STEM.

The results provide the basis for the possibility of combating malignant leukemia stem cells by combining PARP inhibitors with active NK cells. The scientists involved are now planning to evaluate this approach in a clinical study.

What The Study Did: This multicenter study used in vivo β-amyloid cerebrospinal fluid, a biomarker of Alzheimer disease, and positron emission tomography findings to track progression of Alzheimer disease over six years among study participants.

Authors: Niklas Mattsson, M.D., Ph.D., and Oskar Hansson, M.D., Ph.D., of Lund University in Malmo, Sweden, are the corresponding authors.

(doi:10.1001/jamaneurol.2019.2214)

Editor's Note: The article includes conflict of interest and funding/support disclosures. Please see the article for additional information, including other authors, author contributions and affiliations, financial disclosures, funding and support, etc.