Digital and Quantitative Behavioral Phenotyping in Neurologic Disease

Explore digital and quantitative approaches to behavioral phenotyping in neurologic diseases through this 39-minute conference presentation by Brandon Oubre from Harvard Medical School and Massachusetts General Hospital. Learn about the limitations of current semiquantitative clinical scales used to assess motor signs in neurologic diseases, including their coarse granularity, susceptibility to floor and ceiling effects, scaling difficulties, and disconnect from natural behavior patterns. Discover two innovative research approaches addressing these challenges: the Neurobooth project, which collects multimodal time-series data across motor domains using eye tracking, audio, wearable inertial sensors, and video capture technologies to enable deep modeling of constrained behaviors and cross-domain motor coordination analysis; and efforts to capture and analyze natural, unconstrained behavior in free-living environments using wearable devices to develop ecologically valid measures of disease severity and progression. Gain insights into how these digital health technologies can support the development of more sensitive disease progression measures, early disease sign identification, and ecologically valid assessments of natural behavior, representing a significant advancement over traditional patient-reported outcome measures that are often subjective and bias-prone.