Mechanical Circulatory Support (MCS) devices, implemented as a probabilistic deep sequence model. Existing mechanical simulators for MCS rely on oversimplifying assumptions and are insensitive to patient-specific behavior, limiting their applicability to real-world treatment scenarios. To address these shortcomings, our model Domain Adversarial Neural Process (DANP) employs a neural process architecture, allowing it to capture the probabilistic relationship between MCS pump levels and aortic pressure measurements with uncertainty. We use domain adversarial training to combine simulation data with real-world observations, resulting in a more realistic and diverse representation of potential outcomes. Empirical results with an improvement of 19% in non-stationary trend prediction establish DANP as an effective tool for clinicians to understand and make informed decisions regarding MCS patient treatment.

Mean aortic pressure (MAP) is a primary measurement for monitoring blood and O 2 delivery to major organs. Prolonged periods of hypotension, low MAP, lead to low tissue perfusion and subsequent end organ damage. Patients on mechanical circulatory support (MCS) devices, such as the Impella CP, are managed to maintain sufficient MAP for end-organ perfusion. Forecasting MAP is important for early warning of clinically concerning events, including hypotension and instability as well as device weaning. Patients presenting with cardiogenic shock as a result of acute myocardial infarction (AMI/CGS) have increased hemodynamic instability when compared to patients undergoing high-risk percutaneous coronary interventions (HRPCI). Existing deep sequence models for fore-casting often focus on the same patient cohort and cannot generalize across cohorts. In this paper, we examine how deep sequence models respond to the distribution shift of the MAP across the MCS patient cohorts during forecasting. We propose conditional RNN, a deep sequence model that learns to adapt to a different cohort by conditioning on time-invariant cohort features. Our proposed model improves the forecasting performance, achieving a 5.2 mmHg – 6.1 mmHg RMSE for cross-cohort patients.

Mean aortic pressure (MAP) is a primary measurement for monitoring blood and O 2 delivery to major organs.Prolonged periods of hypotension, low MAP, lead to low tissue perfusion and subsequent end organ damage.Patients on mechanical circulatory support (MCS) devices, such as the Impella CP, are managed to maintain sufficient MAP for end-organ perfusion.Forecasting MAP is important for early warning of clinically concerning events, including hypotension and instability as well as device weaning.Patients presenting with cardiogenic shock as a result of acute myocardial infarction (AMI/CGS) have increased hemodynamic instability when compared to patients undergoing high-risk percutaneous coronary interventions (HRPCI).Existing deep sequence models for forecasting often focus on the same patient cohort and cannot generalize across cohorts.In this paper, we examine how deep sequence models respond to the distribution shift of the MAP across the MCS patient cohorts during forecasting.We propose conditional RNN, a deep sequence model that learns to adapt to a different cohort by conditioning on time-invariant cohort features.Our proposed model improves the forecasting performance, achieving a 5.2 mmHg -6.1 mmHg RMSE for cross-cohort patients.

Background: Increasingly, drug and device clinical trials are tracking activity levels and other quality of life indices as endpoints for therapeutic efficacy. Trials have traditionally required intermittent subject visits to the clinic that are artificial, activity-intensive, and infrequent, making trend and event detection between visits difficult. Thus, there is an unmet need for wearable sensors that produce clinical quality and medical grade physiological data from subjects in the home. The current study was designed to validate the BioStamp nPoint® system (MC10 Inc., Lexington, MA, USA), a new technology designed to meet this need. Objective: To evaluate the accuracy, performance, and ease of use of an end-to-end system called the BioStamp nPoint. The system consists of an investigator portal for design of trials and data review, conformal, low-profile, wearable biosensors that adhere to the skin, a companion technology for wireless data transfer to a proprietary cloud, and algorithms for analyzing physiological, biometric, and contextual data for clinical research. Methods: A prospective, nonrandomized clinical trial was conducted on 30 healthy adult volunteers over the course of two continuous days and nights. Supervised and unsupervised study activities enabled performance validation in clinical and remote (simulated “at home”) environments. System outputs for heart rate (HR), heart rate variability (HRV) (including root mean square of successive differences [RMSSD] and low frequency/high frequency ratio), activity classification during prescribed activities (lying, sitting, standing, walking, stationary biking, and sleep), step count during walking, posture characterization, and sleep metrics including onset/wake times, sleep duration, and respiration rate (RR) during sleep were evaluated. Outputs were compared to FDA-cleared comparator devices for HR, HRV, and RR and to ground truth investigator observations for activity and posture classifications, step count, and sleep events. Results: Thirty participants (77% male, 23% female; mean age 35.9 ± 10.1 years; mean BMI 28.1 ± 3.6) were enrolled in the study. The BioStamp nPoint system accurately measured HR and HRV (correlations: HR = 0.957, HRV RMSSD = 0.965, HRV ratio = 0.861) when compared to ActiheartTM. The system accurately monitored RR (mean absolute error [MAE] = 1.3 breaths/min) during sleep when compared to a Capnostream35TM end-tidal CO2 monitor. When compared with investigator observations, the system correctly classified activities and posture (agreement = 98.7 and 92.9%, respectively), step count (MAE = 14.7, < 3% of actual steps during a 6-min walk), and sleep events (MAE: sleep onset = 6.8 min, wake = 11.5 min, sleep duration = 13.7 min) with high accuracy. Participants indicated “good” to “excellent” usability (average System Usability Scale score of 81.3) and preferred the BioStamp nPoint system over both the Actiheart (86%) and Capnostream (97%) devices. Conclusions: The present study validated the BioStamp nPoint system’s performance and ease of use compared to FDA-cleared comparator devices in both the clinic and remote (home) environments.

Background: Although in-lab polysomnography (PSG) remains the gold standard for objective sleep monitoring, the use of at-home sensor systems has gained popularity in recent years. Two categories of monitoring, autonomic and limb movement physiology, arc increasingly recognized as critical for sleep disorder phenotyping, yet at-home options remain limited outside of research protocols. The purpose of this study was to validate the BiostampRC (R) sensor system for respiration, electrocardiography (ECG), and leg electromyography (EMG) against gold standard PSG recordings. Methods: We report analysis of cardiac and respiratory data from 15 patients and anterior tibialis (AT) data from 19 patients undergoing clinical PSG for any indication who simultaneously wore BiostampRC (R) sensors on the chest and the bilateral AT muscles. BiostampRC (R) is a flexible, adhesive, wireless sensor capable of capturing accelerometry, ECG, and EMG. We compared BiostampRC (R) data and feature extractions with those obtained from PSG. Results: The heart rate extracted from BiostampRC (R) ECG showed strong agreement with the PSG (cohort root mean square error of 5 beats per minute). We found the thoracic BiostampRC (R) respiratory waveform, derived from its accelerometer, accurately calculated the respiratory rate (mean average error of 0.26 and root mean square error of 1.84 breaths per minute). The AT WO signal supported periodic limb movement detection, with area under the curve of the receiver operating characteristic curve equaling 0.88. Upon completion, 88% of subjects indicated willingness to wear BiostampRC (R) at home on an exit survey. Conclusion: The results demonstrate that BiostampRC (R) is a tolerable and accurate method for capturing respiration, ECG, and AT EMG time series signals during overnight sleep when compared with simultaneous PSG recordings. The signal quality sufficiently supports analytics of clinical relevance. Larger longitudinal in-home studies are required to support the role of BiostampRC (R) in clinical management of sleep disorders involving the autonomic nervous system and limb movements.

Gait speed is a powerful clinical marker for mobility impairment in patients suffering from neurological disorders. However, assessment of gait speed in coordination with delivery of comprehensive care is usually constrained to clinical environments and is often limited due to mounting demands on the availability of trained clinical staff. These limitations in assessment design could give rise to poor ecological validity and limited ability to tailor interventions to individual patients. Recent advances in wearable sensor technologies have fostered the development of new methods for monitoring parameters that characterize mobility impairment, such as gait speed, outside the clinic, and therefore address many of the limitations associated with clinical assessments. However, these methods are often validated using normal gait patterns; and extending their utility to subjects with gait impairments continues to be a challenge. In this paper, we present a machine learning method for estimating gait speed using a configurable array of skin-mounted, conformal accelerometers. We establish the accuracy of this technique on treadmill walking data from subjects with normal gait patterns and subjects with multiple sclerosis-induced gait impairments. For subjects with normal gait, the best performing model systematically overestimates speed by only 0.01 m/s, detects changes in speed to within less than 1%, and achieves a root-mean-square-error of 0.12 m/s. Extending these models trained on normal gait to subjects with gait impairments yields only minor changes in model performance. For example, for subjects with gait impairments, the best performing model systematically overestimates speed by 0.01 m/s, quantifies changes in speed to within 1%, and achieves a root-mean-square-error of 0.14 m/s. Additional analyses demonstrate that there is no correlation between gait speed estimation error and impairment severity, and that the estimated speeds maintain the clinical significance of ground truth speed in this population. These results support the use of wearable accelerometer arrays for estimating walking speed in normal subjects and their extension to MS patient cohorts with gait impairment.

Wearable sensors have the potential to enable clinical-grade ambulatory health monitoring outside the clinic. Technological advances have enabled development of devices that can measure vital signs with great precision and significant progress has been made towards extracting clinically meaningful information from these devices in research studies. However, translating measurement accuracies achieved in the controlled settings such as the lab and clinic to unconstrained environments such as the home remains a challenge. In this paper, we present a novel wearable computing platform for unobtrusive collection of labeled datasets and a new paradigm for continuous development, deployment and evaluation of machine learning models to ensure robust model performance as we transition from the lab to home. Using this system, we train activity classification models across two studies and track changes in model performance as we go from constrained to unconstrained settings.

Sufficient range of motion of the knee joint is necessary for performing many activities of daily living. Ambulatory monitoring of knee function can provide valuable information about progression of diseases like knee osteoarthritis and recovery after surgical interventions like total knee arthroplasty. In this paper, we describe a skin-mounted, conformal, accelerometer-based system for measuring knee angle and range of motion that does not require a skilled operator to apply devices. We establish the accuracy of this technique with respect to clinical gold standard goniometric measurements on a dataset collected from normative subjects during the performance of repeated bouts of knee flexion and extension tests. Results show that knee angle and range of motion estimates are highly correlated with goniometer measurements, and track differences in knee angle and range of motion to within 1%. These results demonstrate the ability of this system to characterize knee angle and range of motion, enabling future longitudinal monitoring of knee motion in naturalistic environments.

Selon la presente invention, un dispositif electronique qu'un utilisateur porte sur lui comprend un ou plusieurs accelerometres. Le ou les accelerometres generent des informations d'acceleration sur la base d'une acceleration subie par le dispositif electronique. Le dispositif electronique comporte en outre un processeur et une ou plusieurs memoires associees, et ladite ou lesdites memoires associees incluent un code de programme d'ordinateur executable par le processeur. Le processeur, qui est configure par le code de programme d'ordinateur, amene le dispositif electronique a traiter les informations d'acceleration afin d'extraire des elements a partir des informations d'acceleration. Ce processeur, qui est configure par le code de programme d'ordinateur, amene en outre ledit dispositif electronique a traiter les elements afin de determiner l'emplacement du dispositif electronique sur l'utilisateur.