Related Experiment Video
Updated: Jun 5, 2026
08:05
Design and Analysis for Fall Detection System Simplification
Published on: April 6, 2020
Diagnosing fatigue in gait patterns by support vector machines and self-organizing maps
Daniel Janssen1, Wolfgang I Schöllhorn, Karl M Newell
1Training and Movement Science, University of Mainz, Albert Schweitzer Strasse 22, 55099 Mainz, Germany. djanssen@uni-mainz.de
Human Movement Science
|January 4, 2011
Summary
Machine learning models accurately identified individuals and detected leg fatigue from gait patterns. Support vector machines and self-organizing maps show promise for analyzing biomechanical changes during exercise.
Area of Science:
- Biomechanics
- Machine Learning
- Exercise Physiology
Background:
- Gait analysis is crucial for understanding neuromuscular control.
- Leg fatigue significantly alters gait patterns.
- Automated classification of gait changes is needed.
Purpose of the Study:
- To train and test Support Vector Machines (SVM) and Self-Organizing Maps (SOM) for gait pattern classification.
- To differentiate gait before, during, and after leg exhaustion.
- To assess inter-individual gait recognition and fatigue detection.
Main Methods:
- Collected ground reaction forces for 18 gait cycles from 9 participants.
- Induced leg exhaustion using isokinetic exercises with added weights (44.4±8.8kg).
- Analyzed gait data using time courses and deviations from average patterns.
Main Results:
- Achieved 100% accuracy in recognizing individual gait patterns (inter-individual recognition).
- Reached 98.1% accuracy in detecting leg fatigue.
- SOMs provided a visual representation of fatigue development.
Conclusions:
- SVM and SOM are effective tools for classifying gait patterns related to fatigue.
- High accuracy in person and fatigue recognition demonstrates the potential of these methods.
- This approach offers a novel way to visualize and quantify exercise-induced fatigue.
