Atomic partial charges are crucial parameters in molecular dynamics (MD) simulation, dictating the electrostatic contributions to intermolecular energies, and thereby the potential energy landscape. Traditionally, the assignment of partial charges has relied on surrogates of \textit{ab initio} semiempirical quantum chemical methods such as AM1-BCC, and is expensive for large systems or large numbers of molecules. We propose a hybrid physical / graph neural network-based approximation to the widely popular AM1-BCC charge model that is orders of magnitude faster while maintaining accuracy comparable to differences in AM1-BCC implementations. Our hybrid approach couples a graph neural network to a streamlined charge equilibration approach in order to predict molecule-specific atomic electronegativity and hardness parameters, followed by analytical determination of optimal charge-equilibrated parameters that preserves total molecular charge. This hybrid approach scales linearly with the number of atoms, enabling, for the first time, the use of fully consistent charge models for small molecules and biopolymers for the construction of next-generation self-consistent biomolecular force fields. Implemented in the free and open source package \texttt{espaloma\_charge}, this approach provides drop-in replacements for both AmberTools \texttt{antechamber} and the Open Force Field Toolkit charging workflows, in addition to stand-alone charge generation interfaces. Source code is available at \url{https://github.com/choderalab/espaloma_charge}.

Rapid testing is essential to fighting pandemics such as COVID-19, the disease caused by the SARS-CoV-2 virus. Exhaled human breath contains multiple volatile molecules providing powerful potential for non-invasive diagnosis of diverse medical conditions. We investigated breath detection of SARS-CoV-2 infection using cavity-enhanced direct frequency comb spectroscopy (CE-DFCS), a state-of-the-art laser spectroscopic technique capable of a real-time massive collection of broadband molecular absorption features at ro-vibrational quantum state resolution and at parts-per-trillion volume detection sensitivity. Using a total of 170 individual breath samples (83 positive and 87 negative with SARS-CoV-2 based on Reverse Transcription Polymerase Chain Reaction tests), we report excellent discrimination capability for SARS-CoV-2 infection with an area under the Receiver-Operating-Characteristics curve of 0.849(4). Our results support the development of CE-DFCS as an alternative, rapid, non-invasive test for COVID-19 and highlight its remarkable potential for optical diagnoses of diverse biological conditions and disease states.

Offering real-time identification and automated tendon measurements of the patellar tendon, plantar fascia and Achilles tendon, a new artificial intelligence (AI)-powered musculoskeletal ultrasound imaging application has received 510(k) clearance from the Food and Drug Administration (FDA). Clarius Mobile Health said the AI model identifies viewed tendons with a transparent color overlay, labels the tendon and provides subsequent measurement calipers that align with the bottom and top of the tendon at its thickest region. Users of the AI musculoskeletal ultrasound application can then adjust the measurements to facilitate clinical decision-making, according to the company. Alan Hirahara, M.D., says the new AI musculoskeletal application is "ground-breaking technology" that will assist new ultrasound users in learning musculoskeletal structures and enhance efficiency for radiologist assessment of musculoskeletal structures. "The technology will … help current users standardize how structures are measured. In research, interobserver variability exists for any measurement of structures. With the AI standardization of measurements, interobserver reliability problems will now be non-existent. I am excited to see where this technology will go …," noted Dr. Hirahara, an orthopedic surgeon in private practice in Sacramento, Calif.

Early detection of many life-threatening diseases (e.g., prostate and breast cancer) within at-risk population can improve clinical outcomes and reduce cost of care. While numerous disease-specific "screening" tests that are closer to Point-of-Care (POC) are in use for this task, their low specificity results in unnecessary biopsies, leading to avoidable patient trauma and wasteful healthcare spending. On the other hand, despite the high accuracy of Magnetic Resonance (MR) imaging in disease diagnosis, it is not used as a POC disease identification tool because of poor accessibility. The root cause of poor accessibility of MR stems from the requirement to reconstruct high-fidelity images, as it necessitates a lengthy and complex process of acquiring large quantities of high-quality k-space measurements. In this study we explore the feasibility of an ML-augmented MR pipeline that directly infers the disease sidestepping the image reconstruction process. We hypothesise that the disease classification task can be solved using a very small tailored subset of k-space data, compared to image reconstruction. Towards that end, we propose a method that performs two tasks: 1) identifies a subset of the k-space that maximizes disease identification accuracy, and 2) infers the disease directly using the identified k-space subset, bypassing the image reconstruction step. We validate our hypothesis by measuring the performance of the proposed system across multiple diseases and anatomies. We show that comparable performance to image-based classifiers, trained on images reconstructed with full k-space data, can be achieved using small quantities of data: 8% of the data for detecting multiple abnormalities in prostate and brain scans, and 5% of the data for knee abnormalities. To better understand the proposed approach and instigate future research, we provide an extensive analysis and release code.

Keya Medical has launched the DeepVessel FFR, a software device that utilizes deep learning to facilitate fractional flow reserve (FFR) assessment based on coronary computed tomography angiography (CCTA). Cleared by the Food and Drug Administration (FDA), the DeepVessel FFR provides a three-dimensional coronary artery tree model and estimates of FFR CT value after semi-automated review of CCTA images, according to Keya Medical. The company said the DeepVessel FFR has demonstrated higher accuracy than other non-invasive tests and suggested the software could help reduce invasive procedures for coronary angiography and stent implantation in the diagnostic workup and subsequent treatment of coronary artery disease. Joseph Schoepf, M.D., FACR, FAHA, FNASCI, the principal investigator of a recent multicenter trial to evaluate DeepVessel FFR, says the introduction of the modality in the United States dovetails nicely with recent guidelines for the diagnosis of chest pain. "I am excited to see the implementation of DeepVessel FFR. It comes together with the 2021 ACC/AHA Chest Pain Guidelines' recognition of the elevated diagnostic role of CCTA and FFR CT for the non-invasive evaluation of patients with stable or acute chest pain," noted Dr. Schoepf, a professor of Radiology, Medicine, and Pediatrics at the Medical University South Carolina.

Cenk Sidar is the cofounder and CEO of Enquire AI, combining AI, data science, and human intelligence to deliver real-time insights. In recent years, tech-celeration has changed the way humans interact in and beyond the workplace. While rapid tech adoption is considered good, it also fuels the emergence of new risks and "unknown unknowns" in an ever-changing macro landscape. As we enter 2023 on the brink of economic strife, something must balance the scales and help business leaders tackle their biggest problems. One answer lies in another tech breakthrough: Artificial intelligence is ready to perform at scale. Its full implementation cannot be predicted at this point, but it promises real-time actionable insights and offers newfound agility in an uncertain world.

Methods for sequential decision-making are often built upon a foundational assumption that the underlying decision process is stationary. This limits the application of such methods because real-world problems are often subject to changes due to external factors (passive non-stationarity), changes induced by interactions with the system itself (active non-stationarity), or both (hybrid non-stationarity). In this work, we take the first steps towards the fundamental challenge of on-policy and off-policy evaluation amidst structured changes due to active, passive, or hybrid non-stationarity. Towards this goal, we make a higher-order stationarity assumption such that non-stationarity results in changes over time, but the way changes happen is fixed. We propose, OPEN, an algorithm that uses a double application of counterfactual reasoning and a novel importance-weighted instrument-variable regression to obtain both a lower bias and a lower variance estimate of the structure in the changes of a policy's past performances. Finally, we show promising results on how OPEN can be used to predict future performances for several domains inspired by real-world applications that exhibit non-stationarity.

Stroke patients in the US could soon take advantage of cutting-edge robotics during the recovery process. The Food and Drug Administration has cleared Wandercraft's Atalante exoskeleton for use in stroke rehabilitation. The machine can help with intensive gait training, particularly for people with limited upper body mobility that might prevent using other methods. The current-generation Atalante is a self-balancing, battery-powered device with an adjustable gait that can help with early steps through to more natural walking later in therapy. While the hardware still needs to be used in a clinical setting with help from a therapist, its hands-free use lets patients reestablish their gait whether or not they can use their arms. Wandercraft plans to deliver its first exoskeletons to the US during the first quarter of the year, though it didn't name initial customers.

An AI-based home screening test to detect oral and throat cancers from saliva samples is now available in the United States with the hope of transforming oral and throat cancer detection. Based on a technology approved by the US Food and Drug Administration (FDA) as a "breakthrough device," the saliva test can detect early symptoms of oral and throat cancer with more than 90 percent accuracy. Due to a lack of effective diagnostic tools, these cancers often go undiagnosed until they have reached an advanced stage, resulting in low survival rates. In a previous study, Maria Soledad Sosa from the Icahn School of Medicine at Mount Sinai and Julio A. Aguirre-Ghiso, now at Albert Einstein College of Medicine, discovered that the ability of cancer cells to remain dormant is controlled by a protein called NR2F1. This receptor protein can enter the cell nucleus and turn numerous genes on or off to activate a program that prevents the cancer cells from proliferating.

Engineered organs that could end transplant waiting lists, abortion pills on demand and mass-marketing military drones that will revolutionize warfare are among those listed on MIT Technology Review's 10 Breakthrough Technologies of 2023. The list also includes the use of CRISPR to edit away people's problems with high cholesterol by rewriting a sliver of their DNA, artificial intelligence that makes artwork and NASA's James Webb Space Telescope, which is set to remodel our knowledge of the cosmos. The 22nd annual list features critical technological advances predicted to change how we live and work fundamentally. MIT Technology Review, owned by the Massachusetts Institute of Technology, compiled the list of companies or institutions set to develop breakthroughs and when the public can expect these innovations. Mat Honan, editor-in-chief of MIT Technology Review, said: 'Our breakthrough technologies lists are fascinating snapshots of the evolution of big tech innovation breakthroughs.