Validating computer vision algorithms against gold-standard Movement Disorder Society Unified Parkinson's Disease Rating Scales (MDS-UPDRS).
Parkinson's symptoms fluctuate dramatically between medication doses (on/off states). Semi-annual clinic visits provide only brief snapshots that fail to capture daily motor fluctuations and medication wearing-off periods.
Patients perform standardized motor tasks (finger tapping, hand opening/closing, and heel stomping) in front of their home camera. Computer vision algorithms measure movement amplitude decay and frequency slowing to calculate automated MDS-UPDRS sub-scores.
Providing neurologists with objective, longitudinal motor fluctuation charts enables precise titration of dopaminergic medications, minimizing dyskinesias and improving patient quality of life.