A wearable automated system to quantify parkinsonian symptoms enabling closed loop deep brain stimulation

Abstract

This study presents (1) the design and validation of a wearable sensor suite for the unobtrusive capture of heterogeneous signals indicative of the primary symptoms of Parkinson’s disease; tremor, bradykinesia and muscle rigidity in upper extremity movement and (2) a model to characterise these signals as they relate to the symptom severity as addressed by the Movement Disorder Society Unified Parkinson’s Disease Rating Scale (MDS-UPDRS). The sensor suite and detection algorithms managed to distinguish between the non-mimicked and mimicked MDS-UPDRS tests on healthy subjects (p ≤ 0.15), for all the primary symptoms of Parkinson’s disease. Future trials will be conducted on Parkinsonian subjects receiving deep brain stimulation (DBS) therapy. Quantifying symptom severity and correlating severity ratings with DBS treatment will be an important step to fully automate DBS therapy.