Automated motion sensor quantification of gait and lower extremity bradykinesia

Abstract

The objective was to develop and evaluate algorithms for quantifying gait and lower extremity bradykinesia in patients with Parkinson's disease using kinematic data recorded on a heel-worn motion sensor unit. Subjects were evaluated by three movement disorder neurologists on four domains taken from the Movement Disorders Society Unified Parkinson's Disease Rating Scale while wearing the motion sensor unit. Multiple linear regression models were developed based on the recorded kinematic data and clinician scores and produced outputs highly correlated to clinician scores with an average correlation coefficient of 0.86. The newly developed models have been integrated into a home-based system for monitoring Parkinson's disease motor symptoms.