Applying Machine Learning to Gait Analysis Data for Hip Osteoarthritis Diagnosis

Ghidotti, Anna; Regazzoni, Daniele; Rizzi, Caterina; Fiorentino, Gennaro

doi:10.3233/SHTI250178

Abstract

Background:

Hip osteoarthritis (OA) is a degenerative joint disease that affects approximately 25% of individuals over their lifetime, with prevalence expected to rise due to population aging. Gait analysis is recognized as a valuable tool for understanding the biomechanical impact of OA, providing insights into how patients adapt to pain and dysfunction.

Objectives:

This study aims to investigate gait differences between hip OA patients and healthy controls using markerless systems and machine learning techniques to classify patients based on gait data.

Methods:

Thirty-six hip OA patients and thirteen healthy controls were analyzed, and machine learning models were trained using gait parameters.

Results:

The Support Vector Classifier model achieved the highest classification F1-score (96%), demonstrating the potential of combining gait analysis with machine learning to support hip OA diagnosis.

Conclusion:

The findings of the present study suggest that machine learning applied to gait analysis data could enhance diagnosis and management of OA in clinical practice.