Development and Evaluation of AI-based Parkinson's Disease Related Motor Symptom Detection Algorithms

Abstract

Parkinson's Disease (PD) is a chronic, progressive, neurodegenerative disorder that is typically characterized by a loss of (motor) function, increased slowness and rigidity. Due to a lack of feasible biomarkers, progression cannot easily be quantified with objective measures. For the same reason, neurologists have to revert to monitoring of (motor) symptoms (i.e. by means of subjective and often inaccurate patient diaries) in order to evaluate a medication's effectiveness. Replacing or supplementing these diaries with an automatic and objective assessment of symptoms and side effects could drastically reduce manual efforts and potentially help in personalizing and improving medication regime. In turn, appearance of symptoms could be reduced and the patient's quality of life increased. The objective of this thesis is two-fold: (1) development and improvement of algorithms for detecting PD related motor symptoms and (2) to develop a software framework for time series analysis.