Internet of Medical Things

Predicting Clinical Outcomes and Digital Phenotyping with Wearables and Machine Learning

Chenyang Lu

Cyber-Physical Systems Laboratory

Department of Computer Science and Engineering
McKelvey School of Engineering
Washington University


Abstract

Internet of Medical Things (IoMT) provides a new clinical tool for digital phenotyping and outcome prediction of patients in clinical and community settings. IoMT is driven by the rapid growth of wearable devices and machine learning algorithms. The talk will report our recent studies to monitor outpatients using wearables and develop machine learning models to predict clinical deterioration. As a case study I will present a pilot study to predict readmissions of congestive heart failure patients recently discharged from Barnes-Jewish Hospital. The results demonstrated the feasibility of continuously monitoring outpatients using wristbands. We observed high levels of patient compliance in wearing the wristbands regularly and satisfactory yield of data collection from the wristbands to a cloud-based database. Machine learning models based on multimodal data (step, sleep and heart rate) significantly outperformed the traditional clinical approach based on LACE index. We will discuss the best of practices in applying machine learning to small and unbalanced dataset common in wearable-based clinical studies. We will also provide an overview of our ongoing effort to predict surgical complications with Fitbit and our initial experience employing smart watches for mental health and mobility assessment.


Talk


Papers


Bio

Chenyang Lu is the Fullgraf Professor in the Department of Computer Science and Engineering at Washington University in St. Louis. His research interests include embedded and real-time systems, cyber-physical systems, Internet of Things, mobile health, and clinical AI. The author and co-author of over 200 research papers with over 24,000 citations and an h-index of 72, Professor Lu is Editor-in-Chief of ACM Transactions on Cyber-Physical Systems. He also served as Editor-in-Chief of ACM Transactions on Sensor Networks from 2011 to 2017 and Chair of the IEEE Technical Committee on Real-Time Systems (TCRTS) from 2018 to 2019. He received the Ph.D. degree from University of Virginia in 2001. He is a Fellow of ACM and IEEE. 


Events