Gait assessment system based on novel gait variability measures.

Area of Science:

• Biomechanics
• Medical Technology
• Machine Learning

Background:

• Gait analysis is crucial for diagnosing and monitoring neurological disorders.
• Current methods may lack objective, quantitative measures of gait variability.
• Wearable inertial measurement units (IMUs) offer potential for unobtrusive gait data collection.

Purpose of the Study:

• To develop and validate a novel gait assessment system using gait cycle trajectory variability.
• To implement a Support Vector Machine (SVM) classifier for distinguishing healthy from pathological gaits.
• To introduce a Gait Variability Index (GVI) as a quantitative measure of gait deviation from healthy patterns.

Main Methods:

• Utilized hip and knee joint angle trajectories from wearable IMUs during walking.
• Extracted gait variability features by calculating distances between individual gait cycle trajectories.
• Employed four distinct distance functions for feature calculation.
• Trained an SVM classifier to differentiate gait patterns.

Main Results:

• The developed system successfully distinguished between healthy and pathological gait patterns.
• The Gait Variability Index (GVI) was generated, quantifying the proximity of pathological gait to healthy patterns.
• Experimental validation with subjects exhibiting gait disorders showed promising results.

Conclusions:

• The proposed gait assessment system and GVI are suitable for supporting clinicians in evaluating gait performance.
• This technology can aid in monitoring gait rehabilitation progress for patients with neurological conditions.
• The system offers an objective and quantitative approach to gait variability assessment.