Mumbai Hospital Pioneers AI Tool to Forecast Parkinson's Gait Freezing
In a groundbreaking initiative ahead of World Parkinson's Day on April 11, doctors in Mumbai are developing an artificial intelligence (AI)-based tool designed to predict whether and when newly diagnosed Parkinson's patients might develop freezing of gait. This common and highly disabling symptom causes patients to feel as if their feet are stuck to the ground, severely impacting mobility and independence.
Collaborative Effort to Combat a Debilitating Symptom
Jaslok Hospital on Pedder Road has embarked on this innovative project in collaboration with the Paris Brain Institute in France. The tool, expected to take approximately two years to complete, targets freezing of gait, which affects up to 75%-80% of patients with advanced Parkinson's disease. This condition can lead to sudden difficulties in starting to walk, turning, or navigating narrow spaces, often resulting in falls, injuries, and a significant loss of independence.
Addressing a Growing Health Concern in India
Senior neurosurgeon Dr. Paresh Doshi, who is leading the study, highlighted the increasing prevalence of Parkinson's disease in India, including among younger patients. "There is no cure for Parkinson's disease, but our tool will help patients estimate how rapidly their disease will progress, allowing for better treatment planning," he explained. The team chose to focus on freezing of gait due to its profound impact on patients' lives, including reduced mobility, social isolation, and heightened anxiety and depression from fear of falling.
How the AI System Works
The AI system will analyze routine videos of patients walking, which can be recorded even on a mobile phone, utilizing computer vision and machine learning technologies. It aims to detect subtle changes in gait and body balance to predict both the risk of freezing of gait and its potential onset. Unlike traditional methods, this tool will not require expensive gait analyzers costing between Rs 80 lakh and Rs 1 crore, wearable devices, or high-cost tests, making it more accessible.
Phased Development and Data Analysis
The project involves a multi-phase approach:
- Initial Phase: Researchers will analyze over 400 videos of patients taken over the past 30 years using machine learning, examining gait through more than 35 points. Jaslok Hospital's neurosurgery department holds clinical and video data from more than 750 patients.
- Model Building: In the first phase, a model will be built using retrospective data from more than 150 patients who developed freezing of gait.
- Prospective Testing: In the second phase, the tool will be tested prospectively on 337 patients over up to three years to validate its accuracy and effectiveness.
Goals and Future Implications
The ultimate goal is to create an open-access, low-cost tool that enables clinicians to identify high-risk patients earlier, plan treatments more effectively, and reduce falls and disability associated with Parkinson's disease. This innovation represents a significant step forward in managing a condition that affects millions worldwide, offering hope for improved quality of life through advanced technology and international collaboration.
