Parkinson's Disease (PD) owing to its highly het-erogeneous symptoms and progression is a tough problem for automated monitoring. The current method of monitoring the disease involves physical records made by the patient or the care giver about the response to orally administered dopamine precursor for disease management. Since this is indispensable data required by the medical experts for titration of medication it puts the burden of frequent record keeping on them. We aim to use a combination of Non-Contact sleep monitoring system and wearable inertial measurement units to monitor Parkinson's disease progression and response to medication. Through monitoring changes in vitals like heart rate, breath rate and sleep disturbances at night and by keeping track of body movement in the day, we aim to create a 24 hour monitoring system which can provide reliable data for medical experts for better management of the disease with minimal intrusion into the patient's everyday life.
Parkinson's Disease (PD) is a neurodegenerative disorder that affects the dopaminergic neurons in the subthalamic region of the brain. Globally, 10 million people suffer from this disease. PD prevalence is predicted to exponentially grow with the current aging population adding to the chronic disease burden already being faced by several countries . PD symptoms are broadly divided into motor and non-motor symptoms. The dopamine deficit in the substantia-nigra causes motor symptoms such as bradykinesia, rigidity, tremor and postural instability. Non motor symptoms such as dysautonomia, sleep disturbances, cognitive and behavioral issues have an additive effect on lowering the quality of life in PD patients .
The most widely used treatment is dopamine replacement therapy in the form of levodopa which patients ingest at regular intervals throughout the day. The commonly reported on-off phenomenon characterized by fluctuations of motor symptoms is a result of these medications. However, PD is a heterogeneous disorder where these symptoms and response to medication can significantly vary across the patient population. Clear understanding of a patient's "On-Off" state is crucial for a physician to accurately titrate the medication . Unfortunately, this is often misreported by patient's and their caregivers. A significant amount of research has been conducted using wearable devices in home based settings to map these motor symptom fluctuation . Artificial intelligence and Machine Learning (AI/ML) techniques have been used in the analysis of motion detection and characterization of the PD diagnosis and the patient's On and Off state. Due to the volume, complexity and variability of the data a machine learning model may prove valuable for analysis as well as to generate predictive and comparative information.
The main objective of this study is to remotely monitor symptoms of PD patients. It aims to integrate accelerometer and gyroscope sensors to capture the motor fluctuations in the form of data from the trunk, symptomatic hand and contra-lateral leg and examine its clinical utility in better. management of moderate to advanced PD. It aims to compare these reports with the observations entered into the PD Diary by the patient or caregiver. From both the acquired and observed data, the study aims to share an objective report on the motor fluctuations of these patients with physicians who can use this information to better titrate medication prescribed. Additionally, the study aims to monitor non-motor symptoms of dysautonomia and sleep disorder via the Dozee device.