Early Detection of Parkinson's Disease Through Predictive Analytics and Machine Learning

Millions of people worldwide are impacted by Parkinson's disease (PD), the second most prevalent neurological illness. Customized therapy, such as medication or deep brain stimulation (DBS), is essential for Parkinson's disease (PD) care. However, an accurate evaluation of the three main symptoms of the disease-bradykinesia, stiffness, and tremor-is necessary. The foundation of treatment is a clinician's diagnosis; yet, this approach only permits a cross-sectional evaluation of symptoms, which can differ hourly and are prone to inter- and intra-rater variability amongst human examiners. Although there has been a lot of interest in automating symptomatic assessment to optimize treatment plans between doctor visits, no wearable device can evaluate all three cardinal symptoms at once. Issues with stiffness measurement have hindered translation, mapping out-of-clinic muscle activity, and sensor fusion. We address all of these issues in this study using a unique wearable sensor system and machine learning methods.