Following the sequence and structure revolutions, predicting functionally relevant protein structure changes at scale remains an outstanding challenge. We introduce BioEmu, a deep learning system that emulates protein equilibrium ensembles by generating thousands of statistically independent structures per hour on a single graphics processing unit (GPU). BioEmu integrates more than 200 milliseconds of molecular dynamics (MD) simulations, static structures, and experimental protein stabilities using new training algorithms. It captures diverse functional motions-including cryptic pocket formation, local unfolding, and domain rearrangements-and predicts relative free energies with 1 kilocalorie per mole accuracy compared with millisecond-scale MD and experimental data. BioEmu provides mechanistic insights by jointly modeling structural ensembles and thermodynamic properties. This approach amortizes the cost of MD and experimental data generation, demonstrating a scalable path toward understanding and designing protein function.

In treating depression and anxiety, just over half of all clients respond. Monitoring and obtaining early client feedback can allow for rapidly adapted treatment delivery and improve outcomes. This study seeks to develop a state-of-the-art deep-learning framework for predicting clinical outcomes in internet-delivered Cognitive Behavioural Therapy (iCBT) by leveraging large-scale, high-dimensional time-series data of client-reported mental health symptoms and platform interaction data. We use de-identified data from 45,876 clients on SilverCloud Health, a digital platform for the psychological treatment of depression and anxiety. We train deep recurrent neural network (RNN) models to predict whether a client will show reliable improvement by the end of treatment using clinical measures, interaction data with the iCBT program, or both. Outcomes are based on total improvement in symptoms of depression (Patient Health Questionnaire-9, PHQ-9) and anxiety (Generalized Anxiety Disorder-7, GAD-7), as reported within the iCBT program. Using internal and external datasets, we compare the proposed models against several benchmarks and rigorously evaluate them according to their predictive accuracy, sensitivity, specificity and AUROC over treatment. Our proposed RNN models consistently predict reliable improvement in PHQ-9 and GAD-7, using past clinical measures alone, with above 87% accuracy and 0.89 AUROC after three or more review periods, outperforming all benchmark models. Additional evaluations demonstrate the robustness of the achieved models across (i) different health services; (ii) geographic locations; (iii) iCBT programs, and (iv) client severity subgroups. Results demonstrate the robust performance of dynamic prediction models that can yield clinically helpful prognostic information ready for implementation within iCBT systems to support timely decision-making and treatment adjustments by iCBT clinical supporters towards improved client outcomes.

ABSTRACT Artificial intelligence (AI) is a powerful and disruptive area of computer science, with the potential to fundamentally transform the practice of medicine and the delivery of healthcare. In this review article, we outline recent breakthroughs in the application of AI in healthcare, describe a roadmap to building effective, reliable and safe AI systems, and discuss the possible future direction of AI augmented healthcare systems.

Behavioral and Psychotic Symptoms in dementia (BPSD) are associated with negative outcomes including increased mortality and morbidity, increased cost of care and caregiver burnout. Newer technologies are leading to a greatly enhanced ability to monitor and predict these symptoms using passive sensing and sophisticated analytics that rely on signal processing and potentially machine learning. In this study we utilized a novel passive sensing approach supported by AI to demonstrate how highly specific behavioral phenomena can be detected in real time with a need for limited in‐person supervision and monitoring.

OBJECTIVES:Alzheimer's Disease (AD)-related behavioral symptoms (i.e. agitation and/or pacing) develop in nearly 90% of AD patients. In this N = 1 study, we provide proof-of-concept of detecting changes in movement patterns that may reflect underlying behavioral symptoms using a highly novel radio sensor and identifying environmental triggers.METHODS:The Emerald device is a Wi-Fi-like box without on-body sensors, which emits and processes radio-waves to infer patient movement, spatial location and activity. It was installed for 70 days in the room of patient 'E', exhibiting agitated behaviors.RESULTS:Daily motion episode aggregation revealed motor activity fluctuation throughout the data collection period which was associated with potential socio-environmental triggers. We did not detect any adverse events attributable to the use of the device.CONCLUSION:This N-of-1 study suggests the Emerald device is feasible to use and can potentially yield actionable data regarding behavioral symptom management. No active or potential device risks were encountered.

STUDY OBJECTIVE:To assess the feasibility of a noncontact radio sensor as an objective measurement tool to study postoperative recovery from endometriosis surgery. DESIGN:Prospective cohort pilot study. SETTING:Center for minimally invasive gynecologic surgery at an academically affiliated community hospital in conjunction with in-home monitoring. PATIENTS:Patients aged above 18 years who sleep independently and were scheduled to have laparoscopy for the diagnosis and treatment of suspected endometriosis. INTERVENTIONS:A wireless, noncontact sensor, Emerald, was installed in the subjects' home and used to capture physiologic signals without body contact. The device captured objective data about the patients' movement and sleep in their home for 5 weeks before surgery and approximately 5 weeks postoperatively. The subjects were concurrently asked to complete a daily pain assessment using a numeric rating scale and a free text survey about their daily symptoms. MEASUREMENTS AND MAIN RESULTS:Three women aged 23 years to 39 years and with mild to moderate endometriosis participated in the study. Emerald-derived sleep and wake times were contextualized and corroborated by select participant comments from retrospective surveys. In addition, self-reported pain levels and 1 sleep variable, sleep onset to deep sleep time, showed a significant (p <.01), positive correlation with next-day-pain scores in all 3 subjects: r = 0.45, 0.50, and 0.55. In other words, the longer it took the subject to go from sleep onset to deep sleep, the higher their pain score the following day. CONCLUSION:A patient's experience with pain is challenging to meaningfully quantify. This study highlights Emerald's unique ability to capture objective data in both preoperative functioning and postoperative recovery in an endometriosis population. The utility of this uniquely objective data for the clinician-patient relationship is just beginning to be explored.

Behavioral and Psychotic Symptoms in dementia (BPSD) are associated with negative outcomes including increased mortality and morbidity, increased cost of care and caregiver burnout. Over the past decade, technological approaches have been developed to assess and monitor BPSD in real time. However, these have relied on complex arrays of sensors and generate complex data that may not lend themselves to efficient clinical use. In this study we utilized a novel passive sensing approach supported by AI to demonstrate how highly specific behavioral phenomena can be detected in real time with limited need for in-person supervision and monitoring. Emerald is a device developed at MIT for in-home, non-intrusive patient monitoring. The device transmits low-power radio signals (100x lower power than WiFi) and monitors the reflection of these signals from people/their environment to map behavioral information. Signal data are uploaded to the cloud where customized machine learning algorithms process the data to extract gait speed, sleep patterns, spatial location, respiration Subject: We installed the device in the rooms of 4 adults with Dementia in an assisted living facility outside Boston. We measured the subjects’ behavior continuously for 3 months. Sensor data findings were compared to staff observations around specific behavior phenomena. Our goal was to demonstrate Emerald's ability to accurately capture a range of BPSD. Over the course of the study, we identified four behavioral phenomena using the sensor, that were verified by staff and validated by standardized behavioral measures. These include apathy, hypersomnia, anxiety and psychomotor agitation. This aimed at demonstrating the feasibility of an AI-backed passive sensor-based device to accurately phenotype a range of behavioral phenomena. Our initial findings demonstrate that this technology can identify movement-based behavior symptoms accurately. These findings bear validation in larger samples against standardized clinical rating scales.

Introduction Behavioral symptoms of Alzheimer's disease (e.g. delusions, wandering, aggression, sleep disturbance) lead to increased emergency room visits, caregiver burden, and transfers to memory care facilities. Sensor technologies may hold the potential to facilitate early detection and pre-emptive intervention for these symptoms by enabling continuous passive monitoring in a way that in-person monitoring may not be able to. We present preliminary data for such an approach using a device called the Emerald, developed at MIT, which emits low-powered radio signals and can identify and track parameters related to human behavior (sleep, motion, spatial motion, and respiratory rate) based on how these waves reflect off the human body. Artificial Intelligence (AI) algorithms elicit behavioral markers from sensor data. The device does not require any contact or direct interaction by the person being monitored, thus representing true passive sensing. Methods The Emerald device was installed in the rooms of two dementia patients (N=2) with behavioral symptoms residing in an assisted living facility (ALF). Motion data was gathered continuously for a period of three months and was mapped on to spatial location and time frame. Data processing and analysis occurred simultaneously during the collection period. Additionally, study staff administered weekly standardized assessments to both the participant (MMSE) and ALF staff (NPI-NH, CMAI, PAS) to augment data collected from the Emerald. Device data was compiled and made available to the study clinician for clinical analysis and identification of emergent behavioral complications. Results In both participants, device data were used to identify specific behavioral patterns. The device detected variations in behavior by time of day, escalations in pacing, and moments of restlessness throughout the night for both participants. For one participant, clinical interpretation of device data led to the proposition that the participant was experiencing Periodic Limb Movement Disorder, which was unbeknownst to the participant or clinician prior to study participation. The device was able to identify periodic spasms, which occurred when the person was asleep, and localize these to the patient's legs. The second participant showed increase pacing, wandering, and motor agitation before being hospitalized for heightened anxiety and aggression. Device data indicates the period prior to hospitalization featured increased movement episodes relative to this participant's baseline. Conclusions We propose that behavioral phenotyping using an AI-backed passive sensing approach is feasible and safe, and that this approach can help digitally phenotype behavior symptoms in dementia. While the device merits validation against the current standard of behavior measurement in dementia, its advantages include low cost and ongoing engagement, and continuous monitoring while giving patients the option of stopping monitoring at their discretion. Further studies evaluating sensitivity and reliability are warranted to validate the clinical utility of this device. This research was funded by This project is supported by an Innovations grant from the Massachusetts Institute of Technology.

Study Objective The goal of this pilot study is to explore the relationship between sleep and pain in endometriosis patients undergoing laparoscopic surgery using novel radio-wave sensing technology. Design Sleep patterns of three endometriosis patients were monitored in their home on a nightly basis for 4 weeks prior to and 6 weeks after laparoscopic surgery. Emerald, a novel non-contact sensor developed at MIT, was installed in patients’ homes to monitor sleep. Emerald transmits low power radio signals, 1000x lower than WiFi, and uses signal reflections off the subject's body to extract their sleep stages. Participants also recorded a daily Numerical Rating Scale (NRS) of their recent Pain Level and Sleep Quality. Setting Patients were recruited from an endometriosis specialty practice at an academic community hospital. Non-contact monitoring occurred in the patients' homes using Emerald. Patients or Participants Three women ages 23-39 with pelvic pain and suspected endometriosis planning laparoscopic excisional surgery. Interventions N/A Measurements and Main Results Pearson correlation coefficients were computed between Pain Level and sleep variables computed from the Emerald device. Sleep variables include continuity (e.g. total sleep time), architecture (e.g. time in each sleep stage) and fragmentation (e.g. number of awakenings). Deep sleep onset latency, defined as the time from the sleep onset to the first epoch of deep sleep, showed a significant (p Conclusion Results indicate that increased deep sleep onset latency is a correlate of poor sleep and furthermore, poor sleep leads to higher pain sensitivity the following day. A follow-up study is needed to validate these results in a larger pelvic pain population and to evaluate whether sleep assessments should be integrated into pain treatment plans as another target for intervention.

Introduction: There is an increasing number of total hip replacements (THR) being carried out primarily on patients with a sub capital fracture reflecting the change of opinion nationally towards a more positive view on replacement as the initial choice of treatment for these fractures. A number of randomised trials have shown that THR results in better function and improvement in health-related quality of life (HRQoL) and has a lower failure rate than internal fixation.

Whiston Hospital, UK Abstracts of the spring 2005 meeting of the British Nuclear Medicine Society Manchester International Convention Centre, UK; 14–16 March, 2005