AI-Powered Gait analysis
gait analysis
Gait analysis is traditionally conducted in clinical settings by physical therapists or experts in biomechanics. Tools such as opto-electronic markers are used in opto-electronic motion capture systems to measure human motion, which is used to assess key parameters like joint angles and stride length to identify movement abnormalities. While accurate and effective, traditional analysis often relies on specialised equipment and expertise, limiting its accessibility.
A web application is thus developed to analyze walking patterns using pose estimation by Computer Vision. The app leverages on MediaPipe Pose, a Machine Learning solution provided by Google. By processing videos of individuals walking, the application detects key pose landmarks as shown in the figure on the left. To offer a comprehensive view of gait mechanics, the app calculates gait characteristics (e.g. cadence) and critical joint angles (right hip, right knee and right ankle) using the landmarks for further clinical analysis. This tool provides a quick method to estimate gait in a matter of seconds with a simple upload of an MP4 video.
gait analysis via computer vision
The Gait Cycle (Pricker, 2016)
tutorial
Here's a short tutorial video to get you started:
the application
The application is embedded below for you to try out! Do note that limitations on memory would require the videos uploaded to be short (< 15 sec).
Back-up Link: Click to enter site
Here's a short tutorial video to get you started:
App developed by Yeo Yong Yan, NUS Biomedical Engineering Sophomore (Dec 2024 - Feb 2025)
References
Pirker, W., & Katzenschlager, R. (2016). Gait disorders in adults and the elderly. Wiener Klinische Wochenschrift, 129(3–4), 81–95. https://doi.org/10.1007/s00508-016-1096-4
Robinson RO, Herzog W, Nigg BM. Use of force platform variables to quantify the effects of chiropractic manipulation on gait symmetry. J Manipulative Physiol Ther. 1987 Aug;10(4):172-6. PMID: 2958572.