Computers Math

From amoebas to zebras, all living things evolve. They change over time as pressures from the environment cause individuals with certain traits to become more common in a population while those with other traits become less common.
Cancer is no different. Within a growing tumor, cancer cells with the best ability to compete for resources and withstand environmental stressors will come to dominate in frequency. It’s “survival of the fittest” on a microscopic scale.
But fitness — how well suited any particular individual is to its environment — isn’t set in stone; it can change when the environment changes. The cancer cells that might do best in an environment saturated with chemotherapy drugs are likely to be different than the ones that will thrive in an environment without those drugs. So, predicting how tumors will evolve over time, especially in response to treatment, is a major challenge for scientists.
A new study by researchers at Memorial Sloan Kettering in collaboration with researchers at the University of British Columbia/BC Cancer in Canada suggests that one day it may be possible to make those predictions. The study, published June 23, 2021, in the journal Nature, was led by MSK computational biologist Sohrab Shah and BC Cancer breast cancer researcher Samuel Aparicio. The scientists showed that a machine-learning approach, built using principles of population genetics that describe how populations change over time, could accurately predict how human breast cancer tumors will evolve.
“Population genetic models of evolution match up nicely to cancer, but for a number of practical reasons it’s been a challenge to apply these to the evolution of real human cancers,” says Dr. Shah, Chief of Computational Oncology at MSK. “In this study, we show it’s possible to overcome some of those barriers.”
Ultimately, the approach could provide a means to predict whether a patient’s tumor is likely to stop responding to a particular treatment and identify the cells that are likely to be responsible for a relapse. This could mean highly tailored treatments, delivered at the optimal time, to produce better outcomes for people with cancer. More

Without GPS, autonomous systems get lost easily. Now a new algorithm developed at Caltech allows autonomous systems to recognize where they are simply by looking at the terrain around them — and for the first time, the technology works regardless of seasonal changes to that terrain.
Details about the process were published on June 23 in the journal Science Robotics, published by the American Association for the Advancement of Science (AAAS).
The general process, known as visual terrain-relative navigation (VTRN), was first developed in the 1960s. By comparing nearby terrain to high-resolution satellite images, autonomous systems can locate themselves.
The problem is that, in order for it to work, the current generation of VTRN requires that the terrain it is looking at closely matches the images in its database. Anything that alters or obscures the terrain, such as snow cover or fallen leaves, causes the images to not match up and fouls up the system. So, unless there is a database of the landscape images under every conceivable condition, VTRN systems can be easily confused.
To overcome this challenge, a team from the lab of Soon-Jo Chung, Bren Professor of Aerospace and Control and Dynamical Systems and research scientist at JPL, which Caltech manages for NASA, turned to deep learning and artificial intelligence (AI) to remove seasonal content that hinders current VTRN systems.
“The rule of thumb is that both images — the one from the satellite and the one from the autonomous vehicle — have to have identical content for current techniques to work. The differences that they can handle are about what can be accomplished with an Instagram filter that changes an image’s hues,” says Anthony Fragoso (MS ’14, PhD ’18), lecturer and staff scientist, and lead author of the Science Robotics paper. “In real systems, however, things change drastically based on season because the images no longer contain the same objects and cannot be directly compared.”
The process — developed by Chung and Fragoso in collaboration with graduate student Connor Lee (BS ’17, MS ’19) and undergraduate student Austin McCoy — uses what is known as “self-supervised learning.” While most computer-vision strategies rely on human annotators who carefully curate large data sets to teach an algorithm how to recognize what it is seeing, this one instead lets the algorithm teach itself. The AI looks for patterns in images by teasing out details and features that would likely be missed by humans.
Supplementing the current generation of VTRN with the new system yields more accurate localization: in one experiment, the researchers attempted to localize images of summer foliage against winter leaf-off imagery using a correlation-based VTRN technique. They found that performance was no better than a coin flip, with 50 percent of attempts resulting in navigation failures. In contrast, insertion of the new algorithm into the VTRN worked far better: 92 percent of attempts were correctly matched, and the remaining 8 percent could be identified as problematic in advance, and then easily managed using other established navigation techniques.
“Computers can find obscure patterns that our eyes can’t see and can pick up even the smallest trend,” says Lee. VTRN was in danger turning into an infeasible technology in common but challenging environments, he says. “We rescued decades of work in solving this problem.”
Beyond the utility for autonomous drones on Earth, the system also has applications for space missions. The entry, descent, and landing (EDL) system on JPL’s Mars 2020 Perseverance rover mission, for example, used VTRN for the first time on the Red Planet to land at the Jezero Crater, a site that was previously considered too hazardous for a safe entry. With rovers such as Perseverance, “a certain amount of autonomous driving is necessary,” Chung says, “since transmissions take seven minutes to travel between Earth and Mars, and there is no GPS on Mars.” The team considered the Martian polar regions that also have intense seasonal changes, conditions similar to Earth, and the new system could allow for improved navigation to support scientific objectives including the search for water.
Next, Fragoso, Lee, and Chung will expand the technology to account for changes in the weather as well: fog, rain, snow, and so on. If successful, their work could help improve navigation systems for driverless cars.
This project was funded by the Boeing Company, and the National Science Foundation. McCoy participated though Caltech’s Summer Undergraduate Research Fellowship program. More

Our homes and offices are only as solid as the ground beneath them. When that solid ground turns to liquid — as sometimes happens during earthquakes — it can topple buildings and bridges. This phenomenon is known as liquefaction, and it was a major feature of the 2011 earthquake in Christchurch, New Zealand, a magnitude 6.3 quake that killed 185 people and destroyed thousands of homes.
An upside of the Christchurch quake was that it was one of the most well-documented in history. Because New Zealand is seismically active, the city was instrumented with numerous sensors for monitoring earthquakes. Post-event reconnaissance provided a wealth of additional data on how the soil responded across the city.
“It’s an enormous amount of data for our field,” said post-doctoral researcher, Maria Giovanna Durante, a Marie Sklodowska Curie Fellow previously of The University of Texas at Austin (UT Austin). “We said, ‘If we have thousands of data points, maybe we can find a trend.'”
Durante works with Prof. Ellen Rathje, Janet S. Cockrell Centennial Chair in Engineering at UT Austin and the principal investigator for the National Science Foundation-funded DesignSafe cyberinfrastructure, which supports research across the natural hazards community. Rathje’s personal research on liquefaction led her to study the Christchurch event. She had been thinking about ways to incorporate machine learning into her research and this case seemed like a great place to start.
“For some time, I had been impressed with how machine learning was being incorporated into other fields, but it seemed we never had enough data in geotechnical engineering to utilize these methods,” Rathje said. “However, when I saw the liquefaction data coming out of New Zealand, I knew we had a unique opportunity to finally apply AI techniques to our field.”
The two researchers developed a machine learning model that predicted the amount of lateral movement that occurred when the Christchurch earthquake caused soil to lose its strength and shift relative to its surroundings. More

Coming soon to a lab tabletop near you: a method of magneto-thermal imaging that offers nanoscale and picosecond resolution previously available only in synchrotron facilities.
This innovation in spatial and temporal resolution will give researchers extraordinary views into the magnetic properties of a range of materials, from metals to insulators, all from the comfort of their labs, potentially boosting the development of magnetic storage devices.
“Magnetic X-ray microscopy is a relatively rare bird,” said Greg Fuchs, associate professor of applied and engineering physics, who led the project. “The magnetic microscopies that can do this sort of spatial and temporal resolution are very few and far between. Normally, you have to pick either spatial or temporal. You can’t get them both. There’s only about four or five places in the world that have that capability. So having the ability to do it on a tabletop is really enabling spin dynamics at nanoscale for research.”
His team’s paper, “Nanoscale Magnetization and Current Imaging Using Time-Resolved Scanning-Probe Magnetothermal Microscopy,” published June 8 in the American Chemical Society’s journal Nano Letters. The lead author is postdoctoral researcher Chi Zhang.
The paper is the culmination of a nearly 10-year effort by the Fuchs group to explore magnetic imaging with magneto-thermal microscopy. Instead of blasting a material with light, electrons or X-rays, the researchers use a laser focused onto the scanning probe to apply heat to a microscopic swath of a sample and measure the resulting electrical voltage for local magnetic information.
Fuchs and his team pioneered this approach and over the years have developed an understanding of how temperature gradients evolve in time and space. More

Researchers have developed a simple lab-based technique that allows them to look inside lithium-ion batteries and follow lithium ions moving in real time as the batteries charge and discharge, something which has not been possible until now.
Using the low-cost technique, the researchers identified the speed-limiting processes which, if addressed, could enable the batteries in most smartphones and laptops to charge in as little as five minutes.
The researchers, from the University of Cambridge, say their technique will not only help improve existing battery materials, but could accelerate the development of next-generation batteries, one of the biggest technological hurdles to be overcome in the transition to a fossil fuel-free world. The results are reported in the journal Nature.
While lithium-ion batteries have undeniable advantages, such as relatively high energy densities and long lifetimes in comparison with other batteries and means of energy storage, they can also overheat or even explode, and are relatively expensive to produce. Additionally, their energy density is nowhere near that of petrol. So far, this makes them unsuitable for widespread use in two major clean technologies: electric cars and grid-scale storage for solar power.
“A better battery is one that can store a lot more energy or one that can charge much faster — ideally both,” said co-author Dr Christoph Schnedermann, from Cambridge’s Cavendish Laboratory. “But to make better batteries out of new materials, and to improve the batteries we’re already using, we need to understand what’s going on inside them.”
To improve lithium-ion batteries and help them charge faster, researchers need to follow and understand the processes occurring in functioning materials under realistic conditions in real time. Currently, this requires sophisticated synchrotron X-ray or electron microscopy techniques, which are time-consuming and expensive. More

Stem cell therapy is at the cutting edge of regenerative medicine, but until now researchers and clinicians have had to painstakingly evaluate stem cell quality by looking at each cell individually under a microscope. Now, researchers from Japan have found a way to speed up this process, using the power of artificial intelligence (AI).
In a study published in February in Stem Cells, researchers from Tokyo Medical and Dental University (TMDU) reported that their AI system, called DeepACT, can identify healthy, productive skin stem cells with the same accuracy that a human can.
Stem cells are able to develop into several different kinds of mature cells, which means they can be used to grow new tissues in cases of injury or disease. Keratinocyte (skin) stem cells are used to treat inherited skin diseases and to grow sheets of skin that is used to repair large burns.
“Keratinocyte stem cells are one of the few types of adult stem cells that grow well in the lab. The healthiest keratinocytes move more quickly than less healthy cells, so they can be identified by the eye using a microscope,” explains Takuya Hirose, one of the lead authors of the study. “However, this method is time-consuming, labor-intensive, and error-prone.”
To address this, the researchers aimed to develop a system that would identify and track the movement of these stem cells automatically.
“We trained this system through a process called ‘deep learning’ using a library of sample images,” says the co-lead author, Jun’ichi Kotoku. “Then we tested it on a new group of images and found that the results were very accurate compared with manual analysis.”
In addition to detecting individual stem cells, the DeepACT system also calculates the ‘motion index’ of each colony, which indicates how fast thecells at the central region of the colony move compared with those at the marginal region. The colonies with the highest motion index were much more likely than the colonies with lower motion index to grow well, making them good candidates for generating sheets of new skin for transplantation to burn patients.
“DeepACT is a powerful new way to perform accurate quality control of human keratinocyte stem cells and will make this process both more reliable and more efficient,” states Daisuke Nanba, senior author.
Given that skin transplants can fail if they contain too many unhealthy or unproductive stem cells, being able to quickly and easily identify the most suitable cells would be a considerable clinical advantage. Automated quality control could also be valuable for industrial stem cell manufacturing, to help ensure a stable cell supply and lower production costs.
Story Source:
Materials provided by Tokyo Medical and Dental University. Note: Content may be edited for style and length. More

Researchers have developed an artificial intelligence (AI)-based brain age prediction model to quantify deviations from a healthy brain-aging trajectory in patients with mild cognitive impairment, according to a study published in Radiology: Artificial Intelligence. The model has the potential to aid in early detection of cognitive impairment at an individual level.
Amnestic mild cognitive impairment (aMCI) is a transition phase from normal aging to Alzheimer’s disease (AD). People with aMCI have memory deficits that are more serious than normal for their age and education, but not severe enough to affect daily function.
For the study, Ni Shu, Ph.D., from State Key Laboratory of Cognitive Neuroscience and Learning, Beijing Normal University, in Beijing, China, and colleagues used a machine learning approach to train a brain age prediction model based on the T1-weighted MR images of 974 healthy adults aged from 49.3 to 95.4 years. The trained model was applied to estimate the predicted age difference (predicted age vs. actual age) of aMCI patients in the Beijing Aging Brain Rejuvenation Initiative (616 healthy controls and 80 aMCI patients) and the Alzheimer’s Disease Neuroimaging Initiative (589 healthy controls and 144 aMCI patients) datasets.
The researchers also examined the associations between the predicted age difference and cognitive impairment, genetic risk factors, pathological biomarkers of AD, and clinical progression in aMCI patients.
The results showed that aMCI patients had brain-aging trajectories distinct from the typical normal aging trajectory, and the proposed brain age prediction model could quantify individual deviations from the typical normal aging trajectory in these patients. The predicted age difference was significantly associated with individual cognitive impairment of aMCI patients in several domains, specifically including memory, attention and executive function.
“The predictive model we generated was highly accurate at estimating chronological age in healthy participants based on only the appearance of MRI scans,” the researchers wrote. “In contrast, for aMCI, the model estimated brain age to be greater than 2.7 years older on average than the patient’s chronological age.”
The model further showed that progressive aMCI patients exhibit more deviations from typical normal aging than stable aMCI patients, and the use of the predicted age difference score along with other AD-specific biomarkers could better predict the progression of aMCI. Apolipoprotein E (APOE) ε4 carriers showed larger predicted age differences than non-carriers, and amyloid-positive patients showed larger predicted age differences than amyloid-negative patients.
Combining the predicted age difference with other biomarkers of AD showed the best performance in differentiating progressive aMCI from stable aMCI.
“This work indicates that predicted age difference has the potential to be a robust, reliable and computerized biomarker for early diagnosis of cognitive impairment and monitoring response to treatment,” the authors concluded.
Story Source:
Materials provided by Radiological Society of North America. Note: Content may be edited for style and length. More

Machine learning algorithms do a lot for us every day — send unwanted email to our spam folder, warn us if our car is about to back into something, and give us recommendations on what TV show to watch next. Now, we are increasingly using these same algorithms to make environmental predictions for us.
A team of researchers from the University of Minnesota, University of Pittsburgh, and U.S. Geological Survey recently published a new study on predicting flow and temperature in river networks in the 2021 Society for Industrial and Applied Mathematics (SIAM) International Conference on Data Mining (SDM21) proceedings. The study was funded by the National Science Foundation (NSF).
The research demonstrates a new machine learning method where the algorithm is “taught” the rules of the physical world in order to make better predictions and steer the algorithm toward physically meaningful relationships between inputs and outputs.
The study presents a model that can make more accurate river and stream temperature predictions, even when little data is available, which is the case in most rivers and streams. The model can also better generalize to different time periods.
“Water temperature in streams is a ‘master variable’ for many important aquatic systems, including the suitability of aquatic habitats, evaporation rates, greenhouse gas exchange, and efficiency of thermoelectric energy production,” said Xiaowei Jia, a lead author of the study and assistant professor in the University of Pittsburgh’s Department of Computer Science at University in the School of Computing and Information. “Accurate prediction of water temperature and streamflow also aids in decision making for resource managers, for example helping them to determine when and how much water to release from reservoirs to downstream rivers.
A common criticism of machine learning is that the predictions aren’t rooted in physical meaning. That is, the algorithms are just finding correlations between inputs and outputs, and sometimes those correlations can be “spurious” or give false results. The model often won’t be able to handle a situation where the relationship between inputs and outputs changes. More