A Privacy-Preserving Approach to Monitoring Parkinson’s Disease at the EdgeMaster Thesis
17 August 2023, by Torben Fritsch

Photo: midjourney
Abstract:
This thesis investigates using edge technology to monitor Parkinson's Disease at home while preserving patients' privacy. While healthcare costs are rising and the population is aging worldwide, new technologies are simultaneously being integrated into daily life. These devices that are located in close proximity to the data source or end-user are called edge devices. With large amounts of data generated, this offers possibilities for patient-centric healthcare services and new ways of monitoring chronic diseases.
Two central aspects when developing systems to be used in the healthcare domain are the usability of the system by the target audience and the privacy-safe storage of the generated data. Especially with a degenerative disease like Parkinson's, requirements for the usability of a system rise. These two aspects were investigated in this thesis because they were not addressed thoroughly in past studies.
A prototype was developed to show the possibilities of edge technologies in advancing the healthcare system while simultaneously considering patients' needs. It consists of a home server placed at the patient's home and a smartwatch to record data. The data is stored on the server, not accessible from the public internet, but still available over a secure channel for the doctor to analyze.
A usability and performance analysis was conducted to validate the created prototype. The results indicate that a user-friendly system that preserves patients' privacy was developed successfully. The usability study analysis shows that the system is straightforward, and users feel confident using it. The performance analysis indicates a responsive system with smaller deficits when accessing data over large distances. The prototype developed provides a basis for future research and demonstrates ways in which self-directed symptom tracking in patients' homes can be integrated into Parkinson's Disease monitoring while maintaining patients' privacy. Future work must address integrating the technology into existing healthcare infrastructure and analyzing the recorded data.
Supervised by:
Prof. Dr. Janick Edinger, Philipp Kisters