Poster title: Machine Learning Approaches on Multimodal and Ambulatory Data to Predict Individual Symptom Course in Adults with Obsessive-Compulsive Disorder
View Dr. Frank’s poster on the CCOS website here.
1. Can you summarize the work you presented?
Obsessive-Compulsive Disorder (OCD) is a chronic, impairing, and heterogeneous mental health condition that frequently impacts an individual's ability to function in daily tasks. While we have several treatments, such as medications, psychotherapy, and neuromodulation, we typically choose these in a trial-and-error method without ways to estimate who will respond to which treatments. This prolongs the time to effective treatment and can leave many people with residual symptoms. Also, these symptoms of OCD can fluctuate over time, even in people who respond to treatment. Our study is using a wearable sensor (Fitbit) coupled with longitudinal smartphone-based questionnaires to track behavior (such as step count, hours of sleep, daily activity levels), physiology (such as heart rate and oxygen saturation), and symptom levels in adults with OCD over 10-weeks to build machine learning models that can predict individual symptom course and who will respond to medication treatment…
Read the full interview here.