Tech

In order to exploit the properties of quantum physics technologically, quantum objects and their interaction must be precisely controlled. In many cases, this is done using light. Researchers at the University of Innsbruck and the Institute of Quantum Optics and Quantum Information (IQOQI) of the Austrian Academy of Sciences have now developed a method to individually address quantum emitters using tailored light pulses. "Not only is it important to individually control and read the state of the emitters," says Oriol Romero-Isart, "but also to do so while leaving the system as undisturbed as possible." Together with Juan Jose? Garci?a-Ripoll (IQOQI visiting fellow) from the Instituto de Fi?sica Fundamental in Madrid, Romero-Isart's research group has now investigated how specifically engineered pulses can be used to focus light on a single quantum emitter.

Self-compressing light pulse

"Our proposal is based on chirped light pulses," explains Silvia Casulleras, first author of the research paper. "The frequency of these light pulses is time-dependent." So, similar to the chirping of birds, the frequency of the signal changes over time. In structures with certain electromagnetic properties - such as waveguides - the frequencies propagate at different speeds. "If you set the initial conditions of the light pulse correctly, the pulse compresses itself at a certain distance," explains Patrick Maurer from the Innsbruck team. "Another important part of our work was to show that the pulse enables the control of individual quantum emitters." This approach can be used as a kind of remote control to address, for example, individual superconducting quantum bits in a waveguide or atoms near a photonic crystal.

Wide range of applications

In their work, now published in Physical Review Letters, the scientists show that this method works not only with light or electromagnetic pulses, but also with other waves such as lattice oscillations (phonons) or magnetic excitations (magnons). The research group led by the Innsbruck experimental physicist Gerhard Kirchmair, wants to implement the concept for superconducting qubits in the laboratory in close collaboration with the team of theorists.

CLEMSON, South Carolina -- By using laser spectroscopy in a photophysics experiment, Clemson University researchers have broken new ground that could result in faster and cheaper energy to power electronics.

This novel approach, using solution-processed perovskite, is intended to revolutionize a variety of everyday objects such as solar cells, LEDs, photodetectors for smart phones and computer chips. Solution-processed perovskite are the next generation materials for solar cell panels on rooftops, X-ray detectors for medical diagnosis, and LEDs for daily-life lighting.

The research team included a pair of graduate students and one undergraduate student who are mentored by Jianbo Gao, group leader of Ultrafast Photophysics of Quantum Devices (UPQD) group in the College of Science's Department of Physics and Astronomy.

The collaborative research was published March 12 in the high-impact journal Nature Communications. The article is titled "In-situ Observation of Trapped Carriers in Organic Metal Halide Perovskite Films with Ultra-fast Temporal and Ultra-high Energetic Resolutions."

The principal investigator was Gao, who is an assistant professor of condensed matter physics. The co-authors included graduate students Kanishka Kobbekaduwa (first author) and Pan Adhikari of the UPQD group, as well as undergraduate Lawrence Coleman, a senior in the physics department.

Other authors from Clemson were Apparao Rao, the R.A. Bowen Professor of Physics, and Exian Liu, a visiting student from China who works under Gao.

"Perovskite materials are designed for optical applications such as solar cells and LEDs," said Kobbekaduwa, a graduate student and first author of the research article. "It is important because it is much easier to synthesize compared to current silicon-based solar cells. This can be done by solution processing - whereas in silicon, you have to have different methods that are more expensive and time-consuming."

The goal of the research is to make materials that are more efficient, cheaper and easier to produce.

The unique method used by Gao's team - employing ultrafast photocurrent spectroscopy - allowed for a much higher time resolution than most methods, in order to define the physics of the trapped carriers. Here, the effort is measured in picoseconds, which are one trillionth of a second.

"We make devices using this (perovskite) material and we use a laser to shine light on it and excite the electrons within the material," Kobbekaduwa said. "And then by using an external electric field, we generate a photocurrent. By measuring that photocurrent, we can actually tell people the characteristics of this material. In our case, we defined the trapped states, which are defects in the material that will affect the current that we get."

Once the physics are defined, researchers can identify the defects - which ultimately create inefficiency in the materials. When the defects are reduced or passivated, this can result in increased efficiency, which is critical for solar cells and other devices.

As materials are created through solution processes such as spin coating or inkjet printing, the likelihood of introducing defects increases. These low temperature processes are cheaper than ultra-high temperature methods that result in a pure material. But the tradeoff is more defects in the material. Striking a balance between the two techniques can mean higher-quality and more efficient devices at lower costs.

The substrate samples were tested by shooting a laser at the material to determine how the signal propagates through it. Using a laser to illuminate the samples and collect the current made the work possible and differentiated it from other experiments that do not employ the use of an electric field.

"By analyzing that current, we are able to see how the electrons moved and how they come out of a defect," said Adhikari of the UPQD group. "It is possible only because our technique involves ultrafast time scale and in-situ devices under an electrical field. Once the electron falls into the defect, those who experiment using other techniques cannot take that out. But we can take it out because we have the electric field. Electrons have charge under the electric field, and they can move from one place to another. We are able to analyze their transport from one point to another inside the material."

That transport and the effect of material defects upon it can impact the performance of those materials and the devices in which they are used. It is all part of the important discoveries that students are making under the guidance of their mentor, creating ripples that will lead to the next great breakthrough.

"The students are not only learning; they are actually doing the work," Gao said. "I am fortunate to have talented students who - when inspired by challenges and ideas - will become influential researchers. This is all part of the important discoveries that students are making under the guidance of their mentors, creating ripples that will lead to the next great breakthrough. We are also very grateful for the strong collaborations with Shreetu Shrestha and Wanyi Nie, who are top materials scientists from Los Alamos National Laboratory."

More than one million operations are performed in Switzerland every year. A surgeon's skill has a direct impact on the outcome of the operation. Training and experience, as well as momentary fatigue and other influencing factors all play a role. At present, skill is tested by experts, either directly during an operation or by evaluating video footage. This approach is very costly and only a limited number of experts are available. Moreover, the assessment may vary and is not always fully reproducible. For some time, attempts have been made to automate and objectify the assessment of surgeons' skills.

Proof of feasibility

The key result of the study is the proof of the fundamental feasibility of an artificial intelligence (AI)-based assessment of a surgeon's skill in the context of a surgical procedure. The AI used in the study identified good or moderate surgical skill with 87 percent accuracy. This can be considered a very good finding. Lead author Joël Lavanchy explains: "What was surprising was the high degree of algorithms' accuracy with the selected method. Our method of assessing surgical skills is based on the analysis of instrument movement. Surgical instruments were identified using computer algorithms and their movement was analyzed during the time period."

Innovative, three-stage approach with AI

The research team used a newly developed, three-stage approach. The study was based on 242 videos of laparoscopic gallbladder removal procedures. The first step was to identify the instruments used. For this purpose, a convolutional neural network (CNN) was trained to recognize the instruments. In the second step, the movements were analyzed, and their patterns were extracted. In the third step, the extracted movement patterns correlated with rating results by experts using linear regression.

Broader database and in-depth training of algorithms is needed

The present study is an important first step towards assessing surgical performance. More in-depth steps are needed before the technology can be used in clinical practice. For one thing, the AI algorithms need to be trained on a broader database to further improve instrument recognition. For another thing, additional surgeries need to be investigated and, in the medium term, videos of open surgeries as well as procedures apart from the abdominal area can be addressed.

Dr. Enes Hosgor, a co-author of the study who leads the AI division at caresyntax, a medical technology company headquartered in Berlin and Boston, classifies the results as follows: "AI has mainly been used thus far to identify instruments or specific surgical phases. In our study, we now assess surgical skill based on surgical videos. In the future, the use of AI can solve problems at multiple levels: it is available on-demand peri-operatively (not dependent on a few hard-to-find experts); it is objective using algorithm-driven standards; it is comparable at a transregional level as well as surgeon level and could thus provide important support for decision-making processes at certification institutes."

AI at medical location in Bern: CAIM as an opportunity

The project provides an important indication of the future development of the use of AI in medicine. In the future, it will shift from the erstwhile evaluation of image material to the provision of expert systems. Prof. Guido Beldi, head of the study, clarifies: "The study is a first step. Now that we have demonstrated the fundamental feasibility, we can start planning assistance systems that will support surgeons during operations. For example, they will be alerted when fatigue is detected, thereby helping to prevent complications."

Colloidal suspensions of microscopic particles show complex and interesting collective behaviors. In particular, the collective dynamics of colloids is fundamental and ubiquitous for materials assembly, robotic motion, microfluidic control, and in several biological scenarios. The collective dynamics of confined colloids can be completely different from that of free colloids: for instance, confined colloids can self-organize into vortex structures, coherent motion, or different phase behaviors. On one hand, due to the complexity of colloidal suspensions, how to finely tune the collective dynamics of confined colloids remains elusive. On the other hand, since the microscale confinement is on the same length scale as the colloidal size, it is difficult to determine how the colloids interplay with each other and the geometrical constraints.

To study the colloidal collective in confinements, prior work has been focused on the microscopic visualization and simulation method, lacking direct evidence to characterize the mechanical property of colloidal interaction. Can this mechanical property be probed in a direct way or expressed as feedback of force in real-time? With the help of liquid gating technology, the answer could be yes. The leading research field "Liquid gating technology" was selected as the "2020 Top Ten Emerging Technologies In Chemistry" announced by International Union of Pure and Applied Chemistry (IUPAC). Liquid gating technology allows certain liquids to selectively open and close pores on-demand. Especially, liquid gating membranes can respond to pressure changes, which also indicate transmembrane fluid transport capability. Therefore, utilizing the pressure-driven intrusion fluids as efficient causes, the mechanics of the confined colloids can be determined in real-time. In a new research article published in the Beijing-based National Science Review, scientists at Xiamen University present a new paradigm of the liquid gating system that confines the magnetic colloidal suspension in a porous matrix. This confined magnetic colloid system (CMCS) can probe the mechanical properties of the colloidal suspension in real-time, showing the ability to allow or stop the microscale flow or dynamically manipulate the fluid transport.

Interestingly, it seems that "freedom is not free". Firstly, the colloidal suspensions are trapped by the porous matrix. However, the confined colloids are also free in their limited space because their collective dynamics is vastly controllable via the magnetic field. The collective configuration of the confined colloids is statistically and thermodynamically characterized by the colloidal entropy. Meanwhile, the interplay between the confined colloids and the interplay between the colloidal suspension and geometrical constraints are simultaneously indicated by the pressure value. Notably, the pressure change is in a linear relationship with the entropy change. Both of them are prominently affected by the geometrical constraints, packing fraction of colloids, and the strengths and directions of magnetic fields. Moreover, as a proof of concept, this system has been demonstrated for the applications of dynamic and preprogrammed fluid transport, remote drug release, microfluidic logic, and chemical reaction, enabling sustainable antifouling behavior.

Beyond the magnetic field, the reported strategy of entropy regulation of confined colloids is also applicable to other remote external stimuli, such as acoustic field, light field, electric field, and so on. This work would enlighten the exploitation for fundamental research of colloidal science, and applications ranging from fluid transport, multiphase separation, logic microfluidics, to programmable cargo transport. The findings described here would also deepen the understanding of phenomena such as swarm intelligence, cellular collective, pollutant treatment by granular particles, and stop-and-go in traffic jamming.

Photovoltaics decisively contributes to sustainable energy supply. The efficiency of solar cells in directly converting light energy into electrical energy depends on the material used. Metal-halide perovskites are considered very promising materials for solar cells of the next generation. With these semiconductors named after their special crystal structure, a considerable increase in efficiency was achieved in the past years. Meanwhile, perovskite solar cells have reached an efficiency of up to 25.5 percent, which is quite close to that of silicon solar cells that are presently dominating the market. Moreover, the materials needed for perovskite solar cells are rather abundant. The solar cells can be produced easily and at low cost and they can be used for various applications. The theoretically achievable efficiency of perovskite solar cells is about 30.5 percent.

To approach this value, optoelectronic quality of perovskite semiconductors must be further increased. In principle, materials suited for photovoltaics are expected to not only absorb light, but to also emit it efficiently. This process is known as photoluminescence. The corresponding parameter, photoluminescence quantum efficiency, is perfectly suited to determine the quality of perovskite semiconductors. Together with scientists from the Center for Advanced Materials (CAM) of Heidelberg University and the Technical University of Dresden, researchers of KIT's Institute of Microstructure Technology (IMT) and Light Technology Institute (LTI) have now developed a model, by means of which photoluminescence quantum efficiency of perovskite films can be determined reliably and exactly for the first time. Their results are reported in Matter.

Materials Have More Optimization Potentials than Assumed

"With the help of our model, photoluminescence quantum efficiency under solar irradiation can be determined far more precisely," says Dr. Paul Fassl from IMT. "Photon recycling is of high importance. This is the share of photons emitted by the perovskite, which is re-absorbed and re-emitted in the thin films." The researchers applied their model to methylammonium lead triiodide (CH3NH3PbI3), one of the perovskites of highest photoluminescence quantum efficiency. So far, it has been estimated to amount to about 90 percent. Model calculations, however, revealed that it is about 78 percent. The scientists explain that previous estimations did not adequately consider the effect of light scattering and, hence, underestimated the probability of photons - the quantums of light energy - leaving the film before they are re-absorbed. "Our results show that the potential for optimization of these materials is far higher than assumed," says Dr. Ulrich W. Paetzold, Head of the Advanced Optics and Materials for Next Generation Photovoltaics Group of IMT. The team offers an open-source application based on the model, by means of which photoluminescence quantum efficiencies of various perovskite materials can be calculated.

MRI scanning can more precisely define and detect head, neck, thoracic, abdominal and spinal malformations in unborn babies, finds a large multidisciplinary study led by King's College London with Evelina London Children's Hospital, Great Ormond Street Hospital and UCL.

In the study, published today in Lancet Child and Adolescent Health, the team of researchers and clinicians demonstrate the ways that MRI scanning can show malformations in great detail, including their effect on surrounding structures. Importantly, they note that MRI is a very safe procedure for pregnant women and their babies.

They say the work is invaluable both to clinicians caring for babies before they are born and for teams planning care of the baby after delivery.

Recent research has concentrated on correcting for fetal movement in fetal brain MRI and, more recently, for imaging the fetal heart. However, there is an increasing demand to assess the entire fetus with MRI and research from King's College London School of Biomedical Engineering & Imaging Sciences at Evelina London Children's Hospital, have recently been able to develop a post-acquisition pipeline to motion correct and volume reconstruct images of the whole fetal body.

Lead researcher, Professor Mary Rutherford, from the School of Biomedical Engineering & Imaging Sciences said ultrasound remains the gold standard for fetal screening and indeed is complementary to these optimised MRI approaches for evaluating abnormalities of the fetal body.

"Until now, ultrasound has been the modality of choice to diagnose those anomalies. However, sometimes the ability of ultrasound to define the most detailed anatomy is limited. MRI scanning offers the potential to more precisely define malformations that could help support clinicians in their planning of care and counselling of parents".

MRI is commonly used in the classification of fetal brain anomalies. Although its use in fetal body anomalies is less widely adopted, advances have led to the validation of its role in the antenatal investigation of several conditions including the investigation of fetuses with spina bifida: imaging of the fetal brain along with the spinal cord is an important factor in evaluating which patients could benefit from fetal surgery.

For fetal neck masses, MRI provides a clear advantage over conventional ultrasound for assessing tumour extension and giving a 3D visualisation of the tumour's relation to the airway.

MRI may also be better than ultrasound for distinguishing between normal and abnormal lung tissue, and in making other diagnoses such as diaphragmatic hernia--particularly in late gestation, when doing so with ultrasound is challenging.

New approaches to imaging the fetal body with MRI allows both motion correction of the fetal images and volume reconstructions of body organs and defects. Researchers say this improves visualisation and therefore detection and characterisation of abnormalities.

The project brought together surgeons, fetal medicine specialists, radiologists and physicists to review the use of magnetic resonance imaging to investigate conditions in the unborn baby; this approach has already been integrated into clinical practice at Evelina London, which is part of Guy's and St Thomas' NHS Foundation Trust. Videos and images are also available for viewing.

Ongoing work is focused on a fully automated process suitable for clinical translation and wider dissemination into clinical practice.

People with high levels of emotional intelligence are less likely to be susceptible to 'fake news', according to research at the University of Strathclyde.

The study invited participants to read a series of news items on social media and to ascertain whether they were real or fictitious, briefly describing the reasons for their answers. They were also asked to complete a test to determine their levels of emotional intelligence (EQ or emotional quotient) and were asked a number of questions when considering the veracity of each news item.

Researchers found that those who identified the types of news correctly were most likely to score highly in the EQ tests. There was a similar correlation between correct identification and educational attainment.

The study, by researchers in Strathclyde's School of Psychological Sciences & Health and School of Government & Public Policy, has been published in the journal PLOS ONE.

Dr Tony Anderson, Senior Teaching Fellow in Psychology at Strathclyde and partner in the research, said: "Fake news on social media is now a matter of considerable public and governmental concern. Research on dealing with this issue is still in its infancy but recent studies have started to focus on the psychological factors which might make some individuals less susceptible to fake news.

"We assessed whether people were better able to disregard the emotionally charged content of such items and better equipped to assess the veracity of the information. We found that, while distinguishing real news content from fake was challenging, on average participants were more likely to make the correct decision than not.

"Previous research has shown that people can be trained to enhance their own EQ levels. This should help them to discern with a greater degree of accuracy which news is reliable and which is misleading."

Participants were presented with real and fabricated news stories on issues including health, crime, wealth inequality and the environment. Fictitious items featured aspects including emotive language, brief information and a lack of attributed sources.

Comments from people who incorrectly believed fabricated stories were real included: "I have personal experience of this"; "My kids are in this position so I completely get this"; "The graph shows it all" and "The commenter on the post has the same thoughts as me." Those who correctly identified fictitious stories made comments including: "There is emotive/condescending language in the blurb"; "Fearmongering article with no data"; "The source is not an official scientific or governmental source" and "Comes across as more of a rant."

(BOSTON) -- There is a great need to generate various types of cells for use in new therapies to replace tissues that are lost due to disease or injuries, or for studies outside the human body to improve our understanding of how organs and tissues function in health and disease. Many of these efforts start with human induced pluripotent stem cells (iPSCs) that, in theory, have the capacity to differentiate into virtually any cell type in the right culture conditions. The 2012 Nobel Prize awarded to Shinya Yamanaka recognized his discovery of a strategy that can reprogram adult cells to become iPSCs by providing them with a defined set of gene-regulatory transcription factors (TFs). However, progressing from there to efficiently generating a wide range of cell types with tissue-specific differentiated functions for biomedical applications has remained a challenge.

While the expression of cell type-specific TFs in iPSCs is the most often used cellular conversion technology, the efficiencies of guiding iPSC through different "lineage stages" to the fully functional differentiated state of, for example, a specific heart, brain, or immune cell currently are low, mainly because the most effective TF combinations cannot be easily pinpointed. TFs that instruct cells to pass through a specific cell differentiation process bind to regulatory regions of genes to control their expression in the genome. However, multiple TFs must function in the context of larger gene regulatory networks (GRNs) to drive the progression of cells through their lineages until the final differentiated state is reached.

Now, a collaborative effort led by George Church, Ph.D. at Harvard's Wyss Institute for Biologically Inspired Engineering and Harvard Medical School (HMS), and Antonio del Sol, Ph.D., who leads Computational Biology groups at CIC bioGUNE, a member of the Basque Research and Technology Alliance, in Spain, and at the Luxembourg Centre for Systems Biomedicine (LCSB, University of Luxembourg), has developed a computer-guided design tool called IRENE, which significantly helps increase the efficiency of cell conversions by predicting highly effective combinations of cell type-specific TFs. By combining IRENE with a genomic integration system that allows robust expression of selected TFs in iPSCs, the team demonstrated their approach to generate higher numbers of natural killer cells used in immune therapies, and melanocytes used in skin grafts, than other methods. In a scientific first, generated breast mammary epithelial cells, whose availability would be highly desirable for the repopulation of surgically removed mammary tissue. The study is published in Nature Communications.

"In our group, the study naturally built on the 'TFome' project, which assembled a comprehensive library containing 1,564 human TFs as a powerful resource for the identification of TF combinations with enhanced abilities to reprogram human iPSCs to different target cell types," said Wyss Core Faculty member Church. "The efficacy of this computational algorithm will boost a number of our tissue engineering efforts at the Wyss Institute and HMS, and as an open resource can do the same for many researchers in this burgeoning field." Church is the lead of the Wyss Institute's Synthetic Biology platform, and Professor of Genetics at HMS and of Health Sciences and Technology at Harvard and MIT.

Tooling up

Several computational tools have been developed to predict combinations of TFs for specific cell conversions, but almost exclusively these are based on the analysis of gene expression patterns in many cell types. Missing in these approaches was a view of the epigenetic landscape, the organization of the genome itself around genes and on the scale of entire chromosome sections which goes far beyond the sequence of the naked genomic DNA.

"The changing epigenetic landscape in differentiating cells predicts areas in the genome undergoing physical changes that are critical for key TFs to gain access to their target genes. Analyzing these changes can inform more accurately about GRNs and their participating TFs that drive specific cell conversions," said co-first author Evan Appleton, Ph.D. Appleton is a Postdoctoral Fellow in Church's group who joined forces with Sascha Jung, Ph.D., from del Sol's group in the new study. "Our collaborators in Spain had developed a computational approach that integrated those epigenetic changes with changes in gene expression to produce critical TF combinations as an output, which we were in an ideal position to test."

The team used their computational "Integrative gene Regulatory Network model" (IRENE) approach to reconstruct the GRN controlling iPSCs, and then focused on three target cell types with clinical relevance to experimentally validate TF combinations prioritized by IRENE. To deliver TF combinations into iPSCs, they deployed a transposon-based genomic integration system that can integrate multiple copies of a gene encoding a TF into the genome, which allows all factors of a combination to be stably expressed. Transposons are DNA elements that can jump from one position of the genome to another, or in this case from an exogenously provided piece of DNA into the genome.

"Our research team composed of scientists from the LCSB and CIC bioGUNE has a long-standing expertise in developing computational methods to facilitate cell conversion. IRENE is an additional resource in our toolbox and one for which experimental validation has demonstrated it substantially increased efficiency in most tested cases," corresponding author Del Sol, who is Professor at LCSB and CIC bioGUNE. "Our fundamental research should ultimately benefit patients, and we are thrilled that IRENE could enhance the production of cell sources readily usable in therapeutic applications, such as cell transplantation and gene therapies."

Validating the computer-guided design tool in cells

The researchers chose human mammary epithelial cells (HMECs) as a first cell type. Thus far HMECs are obtained from one tissue environment, dissociated, and transplanted to one where breast tissue has been resected. HMECs generated from patients' cells, via an intermediate iPSC stage, could provide a means for less invasive and more effective breast tissue regeneration. One of the combinations that was generated by IRENE enabled the team to convert 14% of iPSCs into differentiated HMECs in iPSC-specific culture media, showing that the provided TFs were sufficient to drive the conversion without help from additional factors.

The team then turned their attention to melanocytes, which can provide a source of cells in cellular grafts to replace damaged skin. This time they performed the cell conversion in melanocyte destination medium to show that the selected TFs work under culture conditions optimized for the desired cell type. Two out of four combinations were able to increase the efficiency of melanocyte conversion by 900% compared to iPSCs grown in destination medium without the TFs. Finally, the researchers compared combinations of TFs prioritized by IRENE to generate natural killer (NK) cells with a state-of-the-art differentiation method based on cell culture conditions alone. Immune NK cells have been found to improve the treatment of leukemia. The researchers' approach outperformed the standard with five out of eight combinations increasing the differentiation of NK cells with critical markers by up to 250%.

"This novel computational approach could greatly facilitate a range of cell and tissue engineering efforts at the Wyss Institute and many other sites around the world. This advance should greatly expand our toolbox as we strive to develop new approaches in regenerative medicine to improve patients' lives," said Wyss Founding Director Donald Ingber, M.D., Ph.D., who is also the Judah Folkman Professor of Vascular Biology at HMS and Boston Children's Hospital, and Professor of Bioengineering at the Harvard John A. Paulson School of Engineering and Applied Sciences.

BUFFALO, N.Y. - University at Buffalo computer scientists have developed a tool that automatically identifies deepfake photos by analyzing light reflections in the eyes.

The tool proved 94% effective in experiments described in a paper accepted at the IEEE International Conference on Acoustics, Speech and Signal Processing to be held in June in Toronto, Canada.

"The cornea is almost like a perfect semisphere and is very reflective," says the paper's lead author, Siwei Lyu, PhD, SUNY Empire Innovation Professor in the Department of Computer Science and Engineering. "So, anything that is coming to the eye with a light emitting from those sources will have an image on the cornea.

"The two eyes should have very similar reflective patterns because they're seeing the same thing. It's something that we typically don't typically notice when we look at a face," says Lyu, a multimedia and digital forensics expert who has testified before Congress.

The paper, "Exposing GAN-Generated Faces Using Inconsistent Corneal Specular Highlights," is available on the open access repository arXiv.

Co-authors are Shu Hu, a third-year computer science PhD student and research assistant in the Media Forensic Lab at UB, and Yuezun Li, PhD, a former senior research scientist at UB who is now a lecturer at the Ocean University of China's Center on Artificial Intelligence.

Tool maps face, examines tiny differences in eyes

When we look at something, the image of what we see is reflected in our eyes. In a real photo or video, the reflections on the eyes would generally appear to be the same shape and color.

However, most images generated by artificial intelligence - including generative adversary network (GAN) images - fail to accurately or consistently do this, possibly due to many photos combined to generate the fake image.

Lyu's tool exploits this shortcoming by spotting tiny deviations in reflected light in the eyes of deepfake images.

To conduct the experiments, the research team obtained real images from Flickr Faces-HQ, as well as fake images from http://www.thispersondoesnotexist.com, a repository of AI-generated faces that look lifelike but are indeed fake. All images were portrait-like (real people and fake people looking directly into the camera with good lighting) and 1,024 by 1,024 pixels.

The tool works by mapping out each face. It then examines the eyes, followed by the eyeballs and lastly the light reflected in each eyeball. It compares in incredible detail potential differences in shape, light intensity and other features of the reflected light.

'Deepfake-o-meter,' and commitment to fight deepfakes

While promising, Lyu's technique has limitations.

For one, you need a reflected source of light. Also, mismatched light reflections of the eyes can be fixed during editing of the image. Additionally, the technique looks only at the individual pixels reflected in the eyes - not the shape of the eye, the shapes within the eyes, or the nature of what's reflected in the eyes.

Finally, the technique compares the reflections within both eyes. If the subject is missing an eye, or the eye is not visible, the technique fails.

Lyu, who has researched machine learning and computer vision projects for over 20 years, previously proved that deepfake videos tend to have inconsistent or nonexistent blink rates for the video subjects.

In addition to testifying before Congress, he assisted Facebook in 2020 with its deepfake detection global challenge, and he helped create the "Deepfake-o-meter," an online resource to help the average person test to see if the video they've watched is, in fact, a deepfake.

He says identifying deepfakes is increasingly important, especially given the hyper-partisan world full of race-and gender-related tensions and the dangers of disinformation - particularly violence.

"Unfortunately, a big chunk of these kinds of fake videos were created for pornographic purposes, and that (caused) a lot of ... psychological damage to the victims," Lyu says. "There's also the potential political impact, the fake video showing politicians saying something or doing something that they're not supposed to do. That's bad."

UNIVERSITY PARK, Pa. -- Did you ever wonder how social networking applications like Facebook and LinkedIn make recommendations on the people you should friend or pages you should follow?

Behind the scenes are machine learning models that classify nodes based on the data they contain about users -- for example, their level of education, location or political affiliation. The models then use these classifications to recommend people and pages to each user. But there is significant bias in the recommendations made by these models -- known as graph neural networks (GNNs) -- as they rely on user features that are highly related to sensitive attributes such as gender or skin color.

Recognizing that the majority of users are reluctant to publicize their sensitive attributes, researchers at the Penn State College of Information Sciences and Technology have developed a novel framework which estimates sensitive attributes to help GNNs make fair recommendations.

The team found that their model, called FairGNN, maintains high performance on node classification using limited, user-supplied sensitive information, while at the same time reducing bias.

"It has been widely reported that people tend to build relationships with those sharing the same sensitive attributes such as ages and regions," said Enyan Dai, doctoral candidate in informatics and lead author on the research paper. "There are some existing machine learning models that aim to eliminate bias, but they require people's sensitive attributes to make them fair and accurate. We are proposing to apply another model based on the very few sensitive attributes that we have (and instead look at other provided information) which could provide us very good insight to give fair predictions toward sensitive attributes such as your gender and skin color."

The researchers trained their model with two real-world datasets: user profiles on Pokec, a popular social network in Slovakia, similar to Facebook and Twitter; and a dataset of approximately 400 NBA basketball players. In the Pokec dataset, they treated the region in which each user was from as the sensitive attribute, and set the classification task to predict the working field of the users. In the NBA data, they identified players as those in the U.S. and those overseas, using location as the sensitive attribute with the classification task to predict whether the salary of each player is over the median.

They then used the same datasets to test their model with other state-of-the-art methods for fair classification. First, they evaluated FairGNN in terms of fairness and classification performance. Then, they performed "ablation studies" -- which remove certain components of the model to test the significance of each component to the overall system -- to further strengthen the model. They then tested whether FairGNN is effective when different amounts of sensitive attributes are provided in the training set.

"Our experiment shows that the classification performance doesn't decrease," said Suhang Wang, assistant professor of information sciences and technology and principal investigator on the project. "But in terms of fairness, we can make the model much more fair."

According to the researchers, their framework could make an impact for other real-world use cases.

"Our findings could be useful in applications, such as job applicant rankings, crime detection or in financial loan applications," said Wang. "But those are domains where we don't want to introduce bias. So we want to give accurate predictions while maintaining fairness."

Added Dai, "[If] this fair machine learning model could be introduced in these applications, we will have more fair data and this problem would be gradually dissolved."

Quantum dots are manmade nanoparticles of semiconducting material comprising only a few thousand atoms. Because of the small number of atoms, a quantum dot’s properties lie between those of single atoms or molecules and bulk material with a huge number of atoms. By changing the nanoparticles’ size and shape, it is possible to fine-tune their electronic and optical properties – how electrons bond and move through the material, and how light is absorbed and emitted by it.

Thanks to increasingly refined control of the nanoparticles’ size and shape, the number of commercial applications has grown. Those already available include lasers, LEDs, and TVs with quantum dot technology.

However, there is a problem that can impair the efficiency of devices or appliances using this nanomaterial as an active medium. When light is absorbed by a material, the electrons are promoted to higher energy levels, and when they return to their fundamental state, each one can emit a photon back to the environment. In conventional quantum dots the electron’s return trip to its fundamental state can be disturbed by various quantum phenomena, delaying the emission of light to the exterior.

The imprisonment of electrons in this way, known as the “dark state”, retards the emission of light, in contrast with the path that lets them return quickly to the fundamental state and hence to emit light more efficiently and directly (“bright state”).

This delay can be shorter in a new class of nanomaterial made from perovskite, which is arousing considerable interest among researchers in materials science as a result (read more at: agencia.fapesp.br/32682/). 

A study conducted by researchers in the Chemistry and Physics Institutes of the University of Campinas (UNICAMP) in the state of São Paulo, Brazil, in collaboration with scientists at the University of Michigan in the United States, made strides in this direction by providing novel insights into the fundamental physics of perovskite quantum dots. An article on the study is published in Science Advances.

“We used coherent spectroscopy, which enabled us to analyze separately the behavior of the electrons in each nanomaterial in an ensemble of tens of billions of nanomaterials. The study is groundbreaking insofar as it combines a relatively new class of nanomaterials – perovskite – with an entirely novel detection technique,” Lázaro Padilha Junior, principal investigator for the project on the Brazilian side, told Agência FAPESP.

FAPESP supported the study via a Young Investigator Grant and a Regular Research Grant awarded to Padilha.

“We were able to verify the energy alignment between the bright state [associated with triplets] and the dark state [associated with singlets], indicating how this alignment depends on the size of the nanomaterial. We also made discoveries regarding the interactions between these states, opening up opportunities for the use of these systems in other fields of technology, such as quantum information,” Padilha said.

“Owing to the crystal structure of perovskite, the level of bright energy divides into three, forming a triplet. This provides various paths for excitation and for the electrons to return to the fundamental state. The most striking result of the study was that by analyzing the lifetimes of each of the three bright states and the characteristics of the signal emitted by the sample we obtained evidence that the dark state is present but located at a higher energy level than two of the three bright states. This means that when light is shone on the sample the excited electrons are trapped only if they occupy the highest bright level and are then shifted to the dark state. If they occupy the lower bright levels, they return to the fundamental state more efficiently.”

To study how electrons interact with light in these materials, the group used multidimensional coherent spectroscopy (MDCS), in which a burst of ultrashort laser pulses (each lasting about 80 femtoseconds, or 80 quadrillionths of a second) is beamed at a sample of perovskite chilled to minus 269 degrees Celsius.

“The pulses irradiate the sample at tightly controlled intervals. By modifying the intervals and detecting the light emitted by the sample as a function of the interval, we can analyze the electron-light interaction and its dynamics with high temporal precision, mapping the typical interaction times, the energy levels with which they couple, and the interactions with other particles,” Padilha said.

The MDCS technique can be used to analyze billions of nanoparticles at the same time and to distinguish between different families of nanoparticles present in the sample.

The experimental system was developed by a team led by Steven Cundiff, principal investigator for the study at the University of Michigan. Some of the measurements were made by Diogo Almeida, a former member of Cundiff’s team and now at UNICAMP’s ultrafast spectroscopy laboratory with a postdoctoral fellowship from FAPESP under Padilha’s supervision.

Quantum dots were synthesized by Luiz Gustavo Bonato, a PhD candidate at UNICAMP’s Chemistry Institute. “The care Bonato took in preparing the quantum dots and his protocol were fundamentally important, as evidenced by their quality and size, and by the properties of the nanometric material,” said Ana Flávia Nogueira, co-principal investigator for the study in Brazil. Nogueira is a professor at the Chemistry Institute (IQ-UNICAMP) and principal investigator for Research Division 1 at the Center for Innovation in New Energies (CINE), an Engineering Research Center (ERC) established by FAPESP and Shell.

“The results obtained are very important since knowledge of the optical properties of the material and how its electrons behave opens up opportunities for the development of new technologies in semiconductor optics and electronics. The incorporation of perovskite is highly likely to be the most distinctive feature of the next generation of television sets,” Nogueira said.

BOSTON -- SARS-CoV-2, the virus that causes COVID-19, has mutated throughout the pandemic. New variants of the virus have arisen throughout the world, including variants that might possess increased ability to spread or evade the immune system. Such variants have been identified in California, Denmark, the U.K., South Africa and Brazil/Japan. Understanding how well the COVID-19 vaccines work against these variants is vital in the efforts to stop the global pandemic, and is the subject of new research from the Ragon Institute of MGH, MIT and Harvard and Massachusetts General Hospital.

In a study recently published in Cell, Ragon Core Member Alejandro Balazs, PhD, found that the neutralizing antibodies induced by the Pfizer and Moderna COVID-19 vaccines were significantly less effective against the variants first described in Brazil/Japan and South Africa. Balazs's team used their experience measuring HIV neutralizing antibodies to create similar assays for COVID-19, comparing how well the antibodies worked against the original strain versus the new variants.

"We were able to leverage the unique high-throughput capacity that was already in place and apply it to SARS-CoV-2," says Balazs, who is also an assistant professor of Medicine at Harvard Medical School and assistant investigator in the Department of Medicine at MGH. "When we tested these new strains against vaccine-induced neutralizing antibodies, we found that the three new strains first described in South Africa were 20-40 times more resistant to neutralization, and the two strains first described in Brazil and Japan were five to seven times more resistant, compared to the original SARS-CoV-2 virus."

Neutralizing antibodies, explains Balazs, work by binding tightly to the virus and blocking it from entering cells, thus preventing infection. Like a key in a lock, this binding only happens when the antibody's shape and the virus's shape are perfectly matched to each other. If the shape of the virus changes where the antibody attaches to it - in this case, in SARS-CoV-2's spike protein - then the antibody may no longer be able to recognize and neutralize the virus as well. The virus would then be described as resistant to neutralization.

"In particular," says Wilfredo Garcia-Beltran, MD, PhD, a resident physician in the Department of Pathology at MGH and first author of the study, "we found that mutations in a specific part of the spike protein called the receptor binding domain were more likely to help the virus resist the neutralizing antibodies." The three South African variants, which were the most resistant, all shared three mutations in the receptor binding domain. This may contribute to their high resistance to neutralizing antibodies.

Currently, all approved COVID-19 vaccines work by teaching the body to produce an immune response, including antibodies, against the SARS-CoV-2 spike protein. While the ability of these variants to resist neutralizing antibodies is concerning, it doesn't mean the vaccines won't be effective.

"The body has other methods of immune protection besides antibodies," says Balazs. "Our findings don't necessarily mean that vaccines won't prevent COVID, only that the antibody portion of the immune response may have trouble recognizing some of these new variants."

Like all viruses, SARS-CoV-2 is expected to continue to mutate as it spreads. Understanding which mutations are most likely to allow the virus to evade vaccine-derived immunity can help researchers develop next-generation vaccines that can provide protection against new variants. It can also help researchers develop more effective preventative methods, such as broadly protective vaccines that work against a wide variety of variants, regardless of which mutations develop.

MIAMI--A new study lead by scientists at the University of Miami (UM) Rosenstiel School of Marine and Atmospheric Science demonstrates that under realistic environmental conditions oil drifting in the ocean after the DWH oil spill photooxidized into persistent compounds within hours to days, instead over long periods of time as was thought during the 2010 Deepwater Horizon oil spill. This is the first model results to support the new paradigm of photooxidation that emerged from laboratory research.

After an oil spill, oil droplets on the ocean surface can be transformed by a weathering process known as photooxidation, which results in the degradation of crude oil from exposure to light and oxygen into new by-products over time. Tar, a by-product of this weathering process, can remain in coastal areas for decades after a spill. Despite the significant consequences of this weathering pathway, photooxidation was not taken into account in oil spill models or the oil budget calculations during the Deepwater Horizon spill.

The UM Rosenstiel School research team developed the first oil-spill model algorithm that tracks the dose of solar radiation oil droplets receive as they rise from the deep sea and are transported at the ocean surface. The authors found that the weathering of oil droplets by solar light occurred within hours to days, and that roughly 75 percent of the photooxidation during the Deepwater Horizon oil spill occurred on the same areas where chemical dispersants were sprayed from aircraft. Photooxidized oil is known to reduce the effectiveness of aerial dispersants.

"Understanding the timing and location of this weathering process is highly consequential. said Claire Paris, a UM Rosenstiel School faculty and senior author of the study. "It helps directing efforts and resources on fresh oil while avoiding stressing the environment with chemical dispersants on oil that cannot be dispersed."

"Photooxidized compounds like tar persist longer in the environment, so modeling the likelihood of photooxidation is critically important not only for guiding first response decisions during an oil spill and restoration efforts afterwards, but it also needs to be taken into account on risk assessments before exploration activities" added Ana Carolina Vaz, assistant scientist at UM's Cooperative Institute for Marine and Atmospheric Studies and lead author of the study.

Researchers at Tufts University School of Engineering have created light-activated composite devices able to execute precise, visible movements and form complex three-dimensional shapes without the need for wires or other actuating materials or energy sources. The design combines programmable photonic crystals with an elastomeric composite that can be engineered at the macro and nano scale to respond to illumination.

The research provides new avenues for the development of smart light-driven systems such as high-efficiency, self-aligning solar cells that automatically follow the sun's direction and angle of light, light-actuated microfluidic valves or soft robots that move with light on demand. A "photonic sunflower," whose petals curl towards and away from illumination and which tracks the path and angle of the light, demonstrates the technology in a paper that appears March 12th, 2021 in Nature Communications.

Color results from the absorption and reflection of light. Behind every flash of an iridescent butterfly wing or opal gemstone lie complex interactions in which natural photonic crystals embedded in the wing or stone absorb light of specific frequencies and reflect others. The angle at which the light meets the crystalline surface can affect which wavelengths are absorbed and the heat that is generated from that absorbed energy.

The photonic material designed by the Tufts team joins two layers: an opal-like film made of silk fibroin doped with gold nanoparticles (AuNPs), forming photonic crystals, and an underlying substrate of polydimethylsiloxane (PDMS), a silicon-based polymer. In addition to remarkable flexibility, durability, and optical properties, silk fibroin is unusual in having a negative coefficient of thermal expansion (CTE), meaning that it contracts when heated and expands when cooled. PDMS, in contrast, has a high CTE and expands rapidly when heated. As a result, when the novel material is exposed to light, one layer heats up much more rapidly than the other, so the material bends as one side expands and the other contracts or expands more slowly.

"With our approach, we can pattern these opal-like films at multiple scales to design the way they absorb and reflect light. When the light moves and the quantity of energy that's absorbed changes, the material folds and moves differently as a function of its relative position to that light," said Fiorenzo Omenetto, corresponding author of the study and the Frank C. Doble Professor of Engineering at Tufts.

Whereas most optomechanical devices that convert light to movement involve complex and energy-intensive fabrication or setups, "We are able to achieve exquisite control of light-energy conversion and generate 'macro motion' of these materials without the need for any electricity or wires," Omenetto said.

The researchers programmed the photonic crystal films by applying stencils and then exposing them to water vapor to generate specific patterns. The pattern of surface water altered the wavelength of absorbed and reflected light from the film, thus causing the material to bend, fold and twist in different ways, depending on the geometry of the pattern, when exposed to laser light.

The authors demonstrated in their study a "photonic sunflower," with integrated solar cells in the bilayer film so that the cells tracked the light source. The photonic sunflower kept the angle between the solar cells and the laser beam nearly constant, maximizing the cells' efficiency as the light moved. The system would work as well with white light as it does with laser light. Such wireless, light-responsive, heliotropic (sun-following) systems could potentially enhance light-to-energy conversion efficiency for the solar power industry. The team's demonstrations of the material also included a butterfly whose wings opened and closed in response to light and a self-folding box.

Osaka, Japan - Scientists from the Institute of Scientific and Industrial Research at Osaka University fabricated nanopores in silicon dioxide, that were only 300 nm, in diameter surrounded by electrodes. These nanopores could prevent particles from entering just by applying a voltage, which may permit the development of sensors that can detect very small concentrations of target molecules, as well as next-generation DNA sequencing technology.

Nanopores are tiny holes that are wide enough for just a single molecule or particle to pass through. The motion of nanoparticles through these holes can usually be detected as an electrical signal, which makes them a promising platform for novel single-particle sensors. However, control of the motion of the particles has been a challenge so far.

Scientists at Osaka University used integrated nanoelectromechanical systems technology to produce solid-state nanopores, only 300 nm wide, with circular platinum gate electrodes surrounding the openings that can prevent nanoparticles from passing through. This is accomplished by selecting the correct voltage that pulls ions in the solution to create a countervailing flow that blocks the entry of the nanoparticle.

"Single-nanoparticle motions could be controlled via the voltage applied to the surrounding gate electrode, when we fine-tuned the electroosmotic flow via the surface electric potential," first author Makusu Tsutsui says. After the particle has been trapped at the nanopore opening, a subtle force imbalance between the electrophoretic attraction and the hydrodynamic drag can then be created. At that time, the particles can be pulled in extremely slowly, which may allow long polymers, like DNA, to be threaded through at the correct speed for sequencing.

"The present method can not only enable better sensing accuracy of sub-micrometer objects, such as viruses, but also provides a method for protein structural analysis," senior author Tomoji Kawai says. While nanopores have already been used to determine the identity of various target molecules based on the current generated, the technology demonstrated in this project may allow for wider range of analytes to be tested this way. For example, small molecules, such as proteins and micro-RNA segments that need to be pulled in at a very controlled speed, may also be detected.