# PARKA AI: A Sensor-Integrated Mobile Application for Parkinson’s Disease Monitoring and Self-Management

## TL;DR

PARKA AI is a mobile app that uses Apple Watch data and self-reports to help Parkinson’s patients manage their condition and communicate better with healthcare providers.

## Contribution

PARKA AI introduces a novel iOS app integrating sensor data and LLMs to improve Parkinson’s disease self-management and patient–HCP communication.

## Key findings

- PARKA AI combines Apple Watch HealthKit data with self-reported logs to generate patient-friendly summaries.
- The app uses a large language model to process data and create concise reports for healthcare providers.
- The design emphasizes accessibility and engagement through data visualizations and personalized tools.

## Abstract

Parkinson’s disease (PD), a progressive neurodegenerative disorder affecting over 10 million people worldwide, necessitates continuous symptom monitoring to optimize treatment and enhance quality of life. Effective communication between patients and healthcare providers (HCPs) is vital but often hindered by fragmented data and cognitive impairments. PARKA AI, a novel iOS application, leverages Apple Watch HealthKit data (e.g., tremor detection, mobility metrics, heart rate, and sleep patterns) and integrates it with self-reported logs (e.g., mood, medication adherence) to empower PD self-management and improve patient–HCP interactions. Employing a human-centered design approach, we developed a high-fidelity prototype using a large language model (LLM)— Google Gemini 1.5 Flash—to process and analyze self-reports and objective sensor-derived data from Apple Healthkit to generate patient-friendly summaries and concise HCP reports. PARKA AI provides accessible data visualizations, personalized self-management tools, and streamlined HCP reports to foster engagement and communication. This paper outlines the derived design requirements, prototype features, and illustrative use cases to show how LLMs can be used in digital health tools. Future work will focus on real-world usability testing to validate the application’s efficacy and accessibility.