Computers Math

Big data has become a big challenge for space scientists analyzing vast datasets from increasingly powerful space instrumentation. To address this, a Southwest Research Institute team has developed a machine learning tool to efficiently label large, complex datasets to allow deep learning models to sift through and identify potentially hazardous solar events. The new labeling tool can be applied or adapted to address other challenges involving vast datasets.
As space instrument packages collect increasingly complex data in ever-increasing volumes, it is becoming more challenging for scientists to process and analyze relevant trends. Machine learning (ML) is becoming a critical tool for processing large complex datasets, where algorithms learn from existing data to make decisions or predictions that can factor more information simultaneously than humans can. However, to take advantage of ML techniques, humans need to label all the data first — often a monumental endeavor.
“Labeling data with meaningful annotations is a crucial step of supervised ML. However, labeling datasets is tedious and time consuming,” said Dr. Subhamoy Chatterjee, a postdoctoral researcher at SwRI specializing in solar astronomy and instrumentation and lead author of a paper about these findings published in the journal Nature Astronomy. “New research shows how convolutional neural networks (CNNs), trained on crudely labeled astronomical videos, can be leveraged to improve the quality and breadth of data labeling and reduce the need for human intervention.”
Deep learning techniques can automate processing and interpret large amounts of complex data by extracting and learning complex patterns. The SwRI team used videos of the solar magnetic field to identify areas where strong, complex magnetic fields emerge on the solar surface, which are the main precursor of space weather events.
“We trained CNNs using crude labels, manually verifying only our disagreements with the machine,” said co-author Dr. Andrés Muñoz-Jaramillo, an SwRI solar physicist with expertise in machine learning. “We then retrained the algorithm with the corrected data and repeated this process until we were all in agreement. While flux emergence labeling is typically done manually, this iterative interaction between the human and ML algorithm reduces manual verification by 50%.”
Iterative labeling approaches such as active learning can significantly save time, reducing the cost of making big data ML ready. Furthermore, by gradually masking the videos and looking for the moment where the ML algorithm changes its classification, SwRI scientists further leveraged the trained ML algorithm to provide an even richer and more useful database.
“We created an end-to-end, deep-learning approach for classifying videos of magnetic patch evolution without explicitly supplying segmented images, tracking algorithms or other handcrafted features,” said SwRI’s Dr. Derek Lamb, a co-author specializing in evolution of magnetic fields on the surface of the Sun. “This database will be critical in the development of new methodologies for forecasting the emergence of the complex regions conducive to space weather events, potentially increasing the lead time we have to prepare for space weather.”
Story Source:
Materials provided by Southwest Research Institute. Note: Content may be edited for style and length. More

Physicists at Indiana University and the University of Tennessee have cracked the code to making microchips smaller, and the key is helium.
Microchips are everywhere, running computers and cars, and even helping people find lost pets. As microchips grow smaller, faster and capable of doing more things, the wires that conduct electricity to them must follow suit. But there’s a physical limit to how small they can become — unless they are designed differently.
“In a traditional system, as you put more transistors on, the wires get smaller,” said Paul Sokol, a professor in the IU Bloomington College of Arts and Sciences’ Department of Physics. “But under newly designed systems, it’s like confining the electrons in a one-dimensional tube, and that behavior is quite different from a regular wire.”
To study the behavior of particles under these circumstances, Sokol collaborated with a physics professor at the University of Tennessee, Adrian Del Maestro, to create a model system of electronics packed into a one-dimensional tube.
Their findings were recently published in Nature Communications.
The pair used helium to create a model system for their study because its interactions with electrons are well known, and it can be made extremely pure, Sokol said. However, there were issues with using helium in a one-dimensional space, the first being that no one had ever done it before. More

When the coronavirus causing COVID-19 infects human cells, the cell’s protein-processing machinery makes modifications to the spike protein that render it more flexible and mobile, which could increase its ability to infect other cells and to evade antibodies, a new study from the University of Illinois Urbana-Champaign found.
The researchers created an atomic-level computational model of the spike protein and ran multiple simulations to examine the protein’s dynamics and how the cell’s modifications affected those dynamics. This is the first study to present such a detailed picture of the protein that plays a key role in COVID-19 infection and immunity, the researchers said.
Biochemistry professor Emad Tajkhorshid, postdoctoral researcher Karan Kapoor and graduate student Tianle Chen published their findings in the Proceedings of the National Academy of Sciences.
“The dynamics of a spike are very important — how much it moves and how flexible it is to search for and bind to receptors on the host cell,” said Tajkhorshid, who also is a member of the Beckman Institute for Advanced Science and Technology. “In order to have a realistic representation, you have to look at the protein at the atomic level. We hope that the results of our simulations can be used for developing new treatments. Instead of using one static structure of the protein to search for drug-binding pockets, we want to reproduce its movements and use all of the relevant shapes it adopts to provide a more complete platform for screening drug candidates instead of just one structure.”
The spike protein of SARS-CoV-2, the virus that causes COVID-19, is the protein that juts out from the surface of the virus and binds to receptors on the surface of human cells to infect them. It also is the target of antibodies in those who have been vaccinated or recovered from infection.
Many studies have looked at the spike protein and its amino acid sequence, but knowledge of its structure has largely relied on static images, Tajkhorshid said. The atomistic simulations give researchers a glimpse of dynamics that affect how the protein interacts with receptors on cells it seeks to infect and with antibodies that seek to bind to it. More

Artificial intelligence and machine learning are among the latest tools being used by cancer researchers to aid in detection and treatment of the disease.
One of the scientists working in this new frontier of cancer research is University of Colorado Cancer Center member Ryan Layer, PhD, who recently published a study detailing his research that uses big data to find cancerous mutations in cells.
“Identifying the genetic changes that cause healthy cells to become malignant can help doctors select therapies that specifically target the tumor,” says Layer, an assistant professor of computer science at CU Boulder. “For example, about 25% of breast cancers are HER2-positive, meaning the cells in this type of tumor have mutations that cause them to produce more of a protein called HER2 that helps them grow. Treatments that specifically target HER2 have dramatically increased survival rates for this type of breast cancer.”
Scientists can evaluate cell DNA to identify mutations, Layer says, but the challenge is that the human genome is massive, and mutations are a normal part of evolution.
“The human genome is long enough to fill a 1.2 million-page book, and any two people can have about 3 million genetic differences,” he says. “Finding one cancer-driving mutation in a tumor is like finding a needle in a stack of needles.”
Scanning the data
The ideal method of determining what type of cancer mutation a patient has is to compare two samples from the same patient, one from the tumor and one from healthy tissue. Such tests can be complicated and costly, however, so Layer hit upon another idea — using massive public DNA databases to look for common cell mutations that tend to be benign, so that researchers can identify rarer mutations that have the potential to be cancerous. More

Although most fundamental mathematical equations that describe electronic structures are long known, they are too complex to be solved in practice. This has hampered progress in physics, chemistry and the material sciences. Thanks to modern high-performance computing clusters and the establishment of the simulation method density functional theory (DFT), researchers were able to change this situation. However, even with these tools the modelled processes are in many cases still drastically simplified. Now, physicists at the Center for Advanced Systems Understanding (CASUS) and the Institute of Radiation Physics at the Helmholtz-Zentrum Dresden-Rossendorf (HZDR) succeeded in significantly improving the DFT method. This opens up new possibilities for experiments with ultra-high intensity lasers, as the group explains in the Journal of Chemical Theory and Computation.
In the new publication, Young Investigator Group Leader Dr. Tobias Dornheim, lead author Dr. Zhandos Moldabekov (both CASUS, HZDR) and Dr. Jan Vorberger (Institute of Radiation Physics, HZDR) take on one of the most fundamental challenges of our time: accurately describing how billions of quantum particles such as electrons interact. These so-called quantum many-body systems are at the heart of many research fields within physics, chemistry, material science, and related disciplines. Indeed, most material properties are determined by the complex quantum mechanical behavior of interacting electrons. While the fundamental mathematical equations that describe electronic structures are, in principle, long known, they are too complex to be solved in practice. Therefore, the actual understanding of e. g. elaborately designed materials has remained very limited.
This unsatisfactory situation has changed with the advent of modern high-performance computing clusters, which has given rise to the new field of computational quantum many-body theory. Here, a particularly successful tool is density functional theory (DFT), which has given unprecedented insights into the properties of materials. DFT is currently considered one of the most important simulation methods in physics, chemistry, and the material sciences. It is especially adept in describing many-electron systems. Indeed, the number of scientific publications based on DFT calculations has been exponentially increasing over the last decade and companies have used the method to successfully calculate properties of materials as accurate as never before.
Overcoming a drastic simplification
Many such properties that can be calculated using DFT are obtained in the framework of linear response theory. This concept is also used in many experiments in which the (linear) response of the system of interest to an external perturbation such as a laser is measured. In this way, the system can be diagnosed and essential parameters like density or temperature can be obtained. Linear response theory often renders experiment and theory feasible in the first place and is nearly ubiquitous throughout physics and related disciplines. However, it is still a drastic simplification of the processes and a strong limitation.
In their latest publication, the researchers are breaking new ground by extending the DFT method beyond the simplified linear regime. Thus, non-linear effects in quantities like density waves, stopping power, and structure factors can be calculated and compared to experimental results from real materials for the first time.
Prior to this publication these non-linear effects were only reproduced by a set of elaborate calculation methods, namely, quantum Monte Carlo simulations. Although delivering exact results, this method is limited to constrained system parameters, as it requires a lot of computational power. Hence, there has been a big need for faster simulation methods. “The DFT approach we present in our paper is 1,000 to 10,000 times faster than quantum Monte Carlo calculations,” says Zhandos Moldabekov. “Moreover, we were able to demonstrate across temperature regimes ranging from ambient to extreme conditions, that this comes not to the detriment of accuracy. The DFT-based methodology of the non-linear response characteristics of quantum-correlated electrons opens up the enticing possibility to study new non-linear phenomena in complex materials.”
More opportunities for modern free electron lasers
“We see that our new methodology fits very well to the capabilities of modern experimental facilities like the Helmholtz International Beamline for Extreme Fields, which is co-operated by HZDR and went into operation only recently,” explains Jan Vorberger. “With high power lasers and free electron lasers we can create exactly these non-linear excitations we can now study theoretically and examine them with unprecedented temporal and spatial resolution. Theoretical and experimental tools are ready to study new effects in matter under extreme conditions that have not been accessible before.”
“This paper is a great example to illustrate the direction my recently established group is heading to,” says Tobias Dornheim, leading the Young Investigator Group “Frontiers of Computational Quantum Many-Body Theory” installed in early 2022. “We have been mainly active in the high energy density physics community in the past years. Now, we are devoted to push the frontiers of science by providing computational solutions to quantum many-body problems in many different contexts. We believe that the present advance in electronic structure theory will be useful for researchers in a number of research fields.” More

Acute cholangitis is a potentially life-threatening bacterial infection that often is associated with gallstones. Symptoms include fever, jaundice, right upper quadrant pain, and elevated liver enzymes.
While these may seem like distinctive, telltale symptoms, unfortunately, they are similar to those of a much different condition: alcohol-associated hepatitis. This challenges emergency department staff and other health care professionals who need to diagnose and treat patients with liver enzyme abnormalities and systemic inflammatory responses.
New Mayo Clinic research finds that machine-learning algorithms can help health care staff distinguish the two conditions. In an article published in Mayo Clinic Proceedings, researchers show how algorithms may be effective predictive tools using a few simple variables and routinely available structured clinical information.
“This study was motivated by seeing many medical providers in the emergency department or ICU struggle to distinguish acute cholangitis and alcohol-associated hepatitis, which are very different conditions that can present similarly,” says Joseph Ahn, M.D., a third-year gastroenterology and hepatology fellow at Mayo Clinic in Rochester. Dr. Ahn is first author of the study.
“We developed and trained machine-learning algorithms to distinguish the two conditions using some of the routinely available lab values that all of these patients should have,” Dr. Ahn says. “The machine-learning algorithms demonstrated excellent performances for discriminating the two conditions, with over 93% accuracy.”
The researchers analyzed electronic health records of 459 patients older than age 18 who were admitted to Mayo Clinic in Rochester between Jan. 1, 2010, and Dec. 31, 2019. The patients were diagnosed with acute cholangitis or alcohol-associated hepatitis.
Ten routinely available laboratory values were collected at the time of admission. After removal of patients whose data were incomplete, 260 patients with alcohol-associated hepatitis and 194 with acute cholangitis remained. These data were used to train eight machine-learning algorithms.
The researchers also externally validated the results using a cohort of ICU patients who were seen at Beth Israel Deaconess Medical Center in Boston between 2001 and 2012. The algorithms also outperformed physicians who participated in an online survey, which is described in the article.
“The study highlights the potential for machine-learning algorithms to assist in clinical decision-making in cases of uncertainty,” says Dr. Ahn. “There are many instances of gastroenterologists receiving consults for urgent endoscopic retrograde cholangiopancreatography in patients who initially deny a history of alcohol use but later turn out to have alcohol-associated hepatitis. In some situations, the inability to obtain a reliable history from patients with altered mental status or lack of access to imaging modalities in underserved areas may force providers to make the determination based on a limited amount of objective data.”
If the machine-learning algorithms can be made easily accessible with an online calculator or smartphone app, they may help health care staff who are urgently presented with an acutely ill patient with abnormal liver enzymes, according to the study.
“For patients, this would lead to improved diagnostic accuracy and reduce the number of additional tests or inappropriate ordering of invasive procedures, which may delay the correct diagnosis or subject patients to the risk of unnecessary complications,” Dr. Ahn says.
The authors are from the Division of Gastroenterology and Hepatology and the Division of Internal Medicine at Mayo Clinic in Rochester, and from the Department of Computer Science at Hanyang University in Seoul, South Korea. Co-author Yung-Kyun Noh was supported in this research by Samsung Research Funding and Incubation Center of Samsung Electronics. The authors report no competing interests.
Story Source:
Materials provided by Mayo Clinic. Original written by Jay Furst. Note: Content may be edited for style and length. More

In a recent article, Kessler Foundation scientists advocated for the incorporation of virtual reality (VR) technology in cognitive rehabilitation research in multiple sclerosis (MS). They presented a conceptual framework supporting VR as an adjuvant to traditional cognitive rehabilitation and exercise training for MS, theorizing that VR could strengthen the effects of traditional rehabilitative therapies by increasing sensory input and promoting multisensory integration and processing.
MS and exercise researchers Carly L.A. Wender, PhD, John DeLuca, PhD, and Brian M. Sandroff, PhD, authored the review, “Developing the rationale for including virtual reality in cognitive rehabilitation and exercise training approaches for managing cognitive dysfunction in MS,” which was published open access on April 3, 2022 by NeuroSci as part of the Special Issue Cognitive Impairment and Neuropsychiatric Dysfunctions in Multiple Sclerosis.
Current pharmacological therapies for MS are not effective for cognitive dysfunction, a common consequence of MS that affects the daily lives of many individuals. This lack of efficacy underscores the need to consider other approaches to managing these disabling cognitive deficits.
The inclusion of VR technology in rehabilitation research and care for MS has the potential not only to improve cognition but to facilitate the transfer of those cognitive gains to improvements in everyday function, according to Brian Sandroff, PhD, senior research scientist in the Center for Neuropsychology and Neuroscience Research at Kessler Foundation. “With VR, we can substantially increase engagement and the volume of sensory input,” he foresees. “And by promoting multisensory integration and processing, VR can augment the effects of the two most promising nonpharmacological treatments — cognitive rehabilitation and exercise.”
Virtual environments are flexible and varied, enabling investigators to control the range and progression of cognitive challenges, with the potential for greater adaptations and stronger intervention effects. VR also allows for the incorporation of cognitive rehabilitation strategies into exercise training sessions, which may support a more direct approach to improving specific cognitive domains through exercise prescriptions. The application of VR to stroke research has shown more improvement in motor outcomes compared with traditional therapy, as well as greater neural activation in the affected area of the brain, suggesting that greater gains may persist over time.
Dr. Sandroff emphasized the largely conceptual advantages for the use of VR to treat cognitive dysfunction in individuals with MS. “More clinical research is needed to explore the efficacy of combining VR with cognitive rehabilitation and/or exercise training, and the impact on everyday functioning on individual with MS,” Dr. Sandroff concluded. “The conceptual framework we outline includes examples of ways immersive and interactive VR can be incorporated into MS clinical trials that will form the basis for larger randomized clinical trials.”
Story Source:
Materials provided by Kessler Foundation. Note: Content may be edited for style and length. More

Explanation methods that help users understand and trust machine-learning models often describe how much certain features used in the model contribute to its prediction. For example, if a model predicts a patient’s risk of developing cardiac disease, a physician might want to know how strongly the patient’s heart rate data influences that prediction.
But if those features are so complex or convoluted that the user can’t understand them, does the explanation method do any good?
MIT researchers are striving to improve the interpretability of features so decision makers will be more comfortable using the outputs of machine-learning models. Drawing on years of field work, they developed a taxonomy to help developers craft features that will be easier for their target audience to understand.
“We found that out in the real world, even though we were using state-of-the-art ways of explaining machine-learning models, there is still a lot of confusion stemming from the features, not from the model itself,” says Alexandra Zytek, an electrical engineering and computer science PhD student and lead author of a paper introducing the taxonomy.
To build the taxonomy, the researchers defined properties that make features interpretable for five types of users, from artificial intelligence experts to the people affected by a machine-learning model’s prediction. They also offer instructions for how model creators can transform features into formats that will be easier for a layperson to comprehend.
They hope their work will inspire model builders to consider using interpretable features from the beginning of the development process, rather than trying to work backward and focus on explainability after the fact. More