KneeXNeT: an ensemble-based approach for knee radiographic evaluation

Conference item

Knee osteoarthritis (OA) is the most common joint disorder and a leading cause of disability. Diagnosing OA severity typically requires expert assessment of X-ray images and is commonly based on the Kellgren-Lawrence grading system, a time-intensive process. This study aimed to develop an automated deep learning model to classify knee OA severity, reducing the need for expert evaluation. First, we evaluated ten state-of-the-art deep learning models, achieving a top accuracy of 0.69 with individual models. To address class imbalance, we employed weighted sampling, improving accuracy to 0.70. We further applied Smooth-GradCAM++ to visualize decision-influencing regions, enhancing the explainability of the best-performing model. Finally, we developed ensemble models using majority voting and a shallow neural network. Our ensemble model, KneeXNet, achieved the highest accuracy of 0.72, demonstrating its potential as an automated tool for knee OA assessment.

Authors: Srikijkasemwat, N; Kundu, SS; Wu, F; Papiez, BW

• Publication status: Published
• Peer review status: Peer reviewed
• Publisher: Springer
• Host title: Proceedings of 2024 International Conference on Medical Imaging and Computer-Aided Diagnosis (MICAD 2024)
• Pages: 407-416
• Series: Lecture Notes in Electrical Engineering
• Series number: 1372
• Publication date: 2025-04-04
• Event title: 5th International Conference on Medical Imaging and Computer-Aided Diagnosis (MICAD 2024)
• Event location: Manchester, UK
• Event website: https://www.micad.org/history_2024.html
• Event start date: 2024-11-19
• Event end date: 2024-11-21
• DOI: 10.1007/978-981-96-3863-5_37
• EISSN: 1876-1119
• ISSN: 1876-1100
• EISBN: 9789819638635
• ISBN: 9789819638628
• Language: English
• Keywords: networking and information technology r&d (NITRD); bioengineering; arthritis; chronic pain; aging; machine learning and artificial intelligence; osteoarthritis; pain research; musculoskeletal