Physiologic Monitoring & Digital Biomarkers in Acute and Procedural Care
MENU
Our lab is partnering with the University of North Carolina at Chapel Hill and the Digital Medicine Society (DiMe) to develop a relapse risk prediction tool for opioid use disorder using consumer sensor data. We are investigating how wearable-derived time series and digital traces of sleep, activity, and physiology may shift before relapse, and how these signals can trigger earlier support. This project offers opportunities to work on data harmonization, feature extraction, temporal modeling, and evaluation strategies that prioritize real-world deployment. We welcome students and collaborators interested in digital health analytics and intervention design.
Funded by: FDA
Lab members: Lauren Lederer, Bill Chen, Stephanie Sullivan, Jessilyn Dunn, Duncan MacKenzie
Collaborators: Digital Medicine Society (DiME), University of North Carolina Eshelman School of Pharmacy, FDA Center for Devices and Radiological Health
Perioperative (Post Surgical) Monitoring
Our lab is evaluating continuous perioperative remote monitoring using commercial wearables to capture cardiovascular, activity, sleep, and thermoregulatory signals around the time of surgery. We are investigating how post-operative physiological trajectories reflect surgical recovery and whether these signals can improve early detection of complications and prediction of readmission risk. The project it situated at the intersection of digital health analytics and clinical implementation, with opportunities to work on wearable data preprocessing, feature engineering, longitudinal modeling, and linking sensor-derived metrics with clinical outcomes. We welcome students and collaborators interested in wearable physiology and real-world deployment.
Lab members: Lauren Lederer, Jessilyn Dunn
Collaborators: Dr. Shelley Hwang, Dr. Thomas Clark Howell
InfectionWatch
Our lab is developing InfectionWatch, a device-agnostic framework that uses physiological signals from consumer wearables to detect infection earlier than symptoms alone. We integrate high resolution heart rate, activity, sleep, and related metrics from human challenge studies with real world smartwatch data linked to electronic health records (EHRs). Using machine learning and deep learning for time series modeling, we seek robust signatures of infection that generalize across devices and populations. This project involves research in digital health, data engineering, and collaborative method development.
Funded by: BARDA
Lab members: Jerry Yang, Kaileigh Moertl, Stephanie Sullivan
Collaborators: Dr. Christopher Woods
rPPG
Our lab is advancing remote photoplethysmography for non-contact vital sign monitoring by targeting a novel region of interest (ROI) beneath the eyelids, where photoplethysmographic signals are strong and consistently visible to eye-tracking cameras on AR devices. We are building a video-to-pulse pipeline that includes ROI detection, color-space transforms, and temporal signal filtering, with validation against contact PPG/ECG. This project is situated within our broader work on privacy-preserving, everyday physiological sensing and involves hands-on experience in computer vision, signal processing, and AR data collection, with opportunities to collaborate across hardware teams and clinical partners.
Lab members: Yihang Jiang, Perisa Ashar
