AI for Health with Wearables

ABSTRACT

Artificial intelligence (AI) has emerged as a powerful tool for solving complex health problems. AI for health is fueled by both the advancement in AI methods and the availability of diverse data. This talk will explore AI-driven precision medicine using wearables that enable unobtrusive monitoring of patients in their daily lives. To harness the full potential of wearables, it is crucial to develop machine learning (ML) models to extract reliable clinical information from noisy and incomplete sensor data. Moreover, these ML approaches need to scale effectively across a wide range of sample sizes, providing robust predictions even with limited data, while enhancing predictive power with large datasets. We will highlight three clinical studies using Fitbit wristbands as wearable instruments. First, we have established a robust feature engineering and ML pipeline specifically tailored for wearable studies with limited sample sizes. This pipeline demonstrated its effectiveness in predicting post-operative complications in a prospective clinical trial of patients undergoing pancreatic surgery. Second, we have developed WearNet, an end-to- end deep learning model designed to detect depression and anxiety disorders using wearable data. WearNet has been trained and validated on a large dataset comprising 8,996 participants, including 1,247 diagnosed with mental health disorders. Finally, we have explored multi-task ML approaches to predict individualized responses to depression therapy based on wearable data collected in a randomized controlled trial (RCT). We conclude the talk by discussing the opportunities and directions in the interdisciplinary field of AI for health, highlighting its transformative impact on healthcare outcomes.

BIOGRAPHY

Chenyang Lu is the Fullgraf Professor of Computer Science & Engineering at Washington University in St. Louis, with joint appointments in Anesthesiology, Medicine, and Neurosurgery. As the founding director of the AI for Health Institute, he leads a multidisciplinary initiative uniting AI researchers and health professionals to address critical health challenges through innovative, data-driven approaches. His current research focuses on developing machine learning models to predict health outcomes using multimodal data, advancing both precision medicine and public health. A Fellow of the ACM and IEEE, Dr. Lu received the 2022 Outstanding Technical Achievement and Leadership Award from the IEEE Technical Community on Real-Time Systems. He also serves as the editor-in-chief of ACM Transactions on Cyber-Physical Systems.