Related Experiment Video
Updated: Nov 22, 2025
06:54
Clinical-oriented Three-dimensional Gait Analysis Method for Evaluating Gait Disorder
Published on: March 4, 2018
14.4K
A novel method for automatic classification of Parkinson gait severity using front-view video analysis
Taha Khan1, Ali Zeeshan2, Mark Dougherty1
1Centre for Artificial Intelligence, School of Information Technology, Halmstad University, Halmstad, Sweden.
Summary
This study presents a computer-vision method to automatically assess Parkinson's disease (PD) gait severity using video analysis. The novel framework shows promise for remote monitoring of PD symptoms.
Area of Science:
- Biomedical Engineering
- Computer Vision
- Neurology
Background:
- Gait impairment is a key symptom of Parkinson's disease (PD).
- Objective assessment of gait severity is crucial for PD management.
Purpose of the Study:
- To introduce a novel computer-vision framework for automatic classification of gait impairment severity in PD patients.
- To utilize front-view motion analysis for objective gait assessment.
Main Methods:
- Recorded 456 videos from 19 PD patients using an RGB camera.
- Rated gait performance using the Unified Parkinson's Disease Rating Scale for gait examination (UPDRS-gait).
- Developed an algorithm to detect and track subject silhouettes, extract gait features from height signals, and classify severity using a Support Vector Machine (SVM) with 10-fold cross-validation.
Main Results:
- Extracted gait features significantly differentiated between UPDRS-gait severity levels (p<0.05).
- The SVM model achieved a promising Area Under the ROC curve of 80.88% for classifying severity levels.
Conclusions:
- The developed computer-vision model is feasible for assessing Parkinson's disease gait impairment.
- This approach supports potential application for remote or home-based Parkinson's gait assessment.
