AI-Based Video App for At-Home Monitoring of Motor Functions in PD Patients

Investigators:
Hamed Tabkhi, ForesightCares Inc.
Mona Azarbayjani, ForesightCares Inc.
Sanjay Iyer, Memory & Movement Charlotte

MassAITC Cohort: Year 3 (Aging)

Project Accomplishments: The pilot project focused on developing and validating an AI-based visual assessment (AVA) app for at-home monitoring of motor function in older adults with Parkinson’s Disease (PD). The project aimed to customize the app for user-friendly, independent use by PD patients and their caregivers, and to test its feasibility in real-world settings. Through collaboration with neurologists, physical therapists, and PD patients, the team identified key clinical needs and usability challenges. Two studies were conducted: the first gathered insights from 14 PD patients and their caregivers, while the second involved usability testing of the app prototype. Feedback led to a comprehensive redesign of the app, including simplified navigation, larger fonts, and voice guidance. A four-week pilot program demonstrated the app’s reliability, with 80% of participants completing assessments independently. 

The app achieved a high System Usability Scale (SUS) score of 84.6 and strong user satisfaction, with participants praising its intuitive design and expressing interest in continued use. The AI technology was enhanced to improve the accuracy of motor assessments such as the Chair Rise, Timed Up and Go, and Standing Balance tests, all processed in real time on iPad devices. The AVA app is positioned as a scalable, cost-effective tool for empowering PD patients and caregivers to manage symptoms and monitor disease progression from home. 

Initial Proposal Abstract: Limited access to movement disorder specialists, particularly in rural and underserved areas, is a pressing challenge for Parkinson’s disease (PD) patients. Doctors need to understand the varying symptoms of PD to adjust treatments accurately. Home-based monitoring emerges as a potent solution, offering real-time and comprehensive insights into the disease’s progression. Such monitoring is especially crucial in rural settings where infrequent clinical visits might miss critical insights into the disease’s progression, making it essential to bridge this care gap.

This pilot is centered around the usability of AI and empowering caregivers to perform routine AI-based assessments independently. Grounded on our existing AVA Minimum Viable Product (MVP), this pilot primarily seeks to evaluate AVA’s potential for home-based monitoring by caregivers without any intervention or changes in the treatment plan. Partnering with “Memory & Movement,” a clinic specializing in PD, we aim to assess the feasibility of AVA in real-world settings. We will involve 35 older adult PD patients and their caregivers in this evaluation, ensuring inclusivity by including 15 participants from underserved and rural communities.

The proposed effort includes enhancing the app with a user-centric design based on feedback from older PD patients and caregivers, improving AI for precise hand movement detection, and redesigning the user interface to achieve a User-Centric Customization of the AVA App for PD Patients and Their Caregivers: This aim refines the current AVA into a highly functional technology with a visually appealing and user-centric interface that facilitates seamless motor function assessment for individuals with PD.

Outcomes: