Related Experiment Video
Updated: Oct 21, 2025
06:49
Automated Gait Analysis in Mice with Chronic Constriction Injury
Published on: October 17, 2017
10.4K
A Validation of Supervised Deep Learning for Gait Analysis in the Cat
Charly G Lecomte1, Johannie Audet1, Jonathan Harnie1
1Department of Pharmacology-Physiology, Faculty of Medicine and Health Sciences, Centre de Recherche du CHUS, Université de Sherbrooke, Sherbrooke, QC, Canada.
Frontiers in Neuroinformatics
|September 7, 2021
Summary
DeepLabCut (DLC), a deep learning tool, accurately analyzes cat gaits without expensive markers. This cost-effective method validates kinematic variables for feline locomotion research.
Area of Science:
- Veterinary Medicine
- Biomechanics
- Animal Locomotion
Background:
- Traditional animal gait analysis relies on costly motion tracking systems and reflective markers.
- Deep learning offers a potential alternative for markerless motion tracking.
Purpose of the Study:
- To validate the accuracy of DeepLabCut (DLC) for feline gait analysis.
- To compare DLC gait analysis results with a custom-made software (Expresso).
Main Methods:
- Four adult cats underwent gait analysis on a split-belt treadmill.
- Kinematic variables were calculated using both DLC and Expresso software.
- Agreement between software was assessed using correlation coefficients.
Main Results:
- DeepLabCut (DLC) demonstrated high precision and agreement with Expresso for feline gait analysis.
- All 12 kinematic variables showed good to excellent agreement (correlation coefficients > 0.75).
- Nine variables exhibited excellent agreement (correlation coefficients > 0.9).
Conclusions:
- Deep learning, specifically DLC, is a valid and accurate tool for feline gait analysis.
- DLC provides a cost-effective, markerless alternative for measuring kinematic variables in cats.
- This method facilitates advanced research into feline locomotion.
