Ben Hassine Slim ¹ BALTI ALA ^1,2* Abid Sabeur ¹ Ben Khelifa Mohamed Moncef ² Sayadi Mounir ¹

• ¹ University of Tunis, Research Laboratory SIME, ENSIT, Tunisia, Tunis, Tunisia
• ² South University, J-AP2S Laboratory, Toulon, France, TOULON, France

Knee osteoarthritis (KOA) is a major health issue that affects millions of people worldwide. This study employs machine learning algorithms to analyze human gait based on kinematic data, aiming to enhance the diagnosis and detection of Knee osteoarthritis (KOA). By adopting this approach, we contribute to the development of more advanced and effective diagnostic methods for knee osteoarthritis (KOA), a prevalent joint condition. The proposed methodology is structured around several critical steps to optimize the model's performance. These include extracting kinematic features from the video data to capture the essential dynamics of gait, applying data filtering and reduction techniques to remove noise and enhance data quality, and selecting and calculating key gait parameters to boost the model’s predictive power. The machine learning model is then trained using these refined features, validated through cross-validation for robust performance assessment, and rigorously tested on unseen data to ensure its generalizability. Our approach yields significant improvements in classification accuracy, underscoring its potential for early and precise KOA detection. Furthermore, a comparative analysis with another model trained on the same dataset demonstrates the superiority of our method.