DIGITAL BIOMARKERS FOR PARKINSON'S DISEASE MONITORING: WEARABLE SENSORS, SMARTPHONE APPLICATIONS, AND MACHINE LEARNING ANALYTICS

Abstract

Parkinson's disease (PD) affects approximately 10 million people globally and is characterised by progressive motor symptoms — tremor, rigidity, bradykinesia, and postural instability — alongside a spectrum of non-motor manifestations including cognitive decline, autonomic dysfunction, and sleep disorders. Current clinical assessment relies on periodic in-clinic evaluation using rating scales such as the Movement Disorder Society Unified Parkinson's Disease Rating Scale (MDS-UPDRS), which captures a brief snapshot of motor function and is subject to inter-rater variability and recall bias. Digital biomarkers derived from wearable inertial sensors, smartwatches, smartphones, and voice analysis have emerged as objective, continuous, and ecologically valid measures of PD symptom severity that complement and potentially extend traditional clinical assessment. This review analyses 52 studies of digital biomarker development and validation for PD monitoring published between 2016 and 2024, evaluating sensor modalities,machine learning approaches, and clinical validation status. Implications for PD monitoring in Uzbekistan's neurology services are discussed.