An AI system for continuous knee osteoarthritis severity grading: An anomaly detection inspired approach with few labels

• Published in: Artificial intelligence in medicine Vol. 167; p. 103138
• Main Authors: Belton, Niamh, Lawlor, Aonghus, Curran, Kathleen M.
• Format: Journal Article
• Language: English
• Published: Netherlands Elsevier B.V 01.09.2025
• Subjects: Algorithms; Artificial Intelligence; CLIP; Contrastive learning; Deep learning; Few labels; Few shot anomaly detection; Humans; Knee Joint - diagnostic imaging; Knee osteoarthritis; Machine learning; Osteoarthritis, Knee - diagnostic imaging; Radiographic Image Interpretation, Computer-Assisted - methods; Self-supervised learning; Self-supervision; Severity of Illness Index; X-ray; Deep learning; Knee osteoarthritis; Few shot anomaly detection; Self-supervised learning; Contrastive learning; Machine learning; X-ray; Self-supervision; Artificial intelligence; Few labels; CLIP
• ISSN: 0933-3657; 1873-2860; 1873-2860
• DOI: 10.1016/j.artmed.2025.103138

The diagnostic accuracy and subjectivity of existing Knee Osteoarthritis (OA) ordinal grading systems has been a subject of on-going debate and concern. Existing automated solutions are trained to emulate these imperfect systems, whilst also being reliant on large annotated databases for fully-supervised training. This work proposes a three stage approach for automated continuous grading of knee OA that is built upon the principles of Anomaly Detection (AD); learning a robust representation of healthy knee X-rays and grading disease severity based on its distance to the centre of normality. In the first stage, SS-FewSOME is proposed, a self-supervised AD technique that learns the ‘normal’ representation, requiring only examples of healthy subjects and <3% of the labels that existing methods require. In the second stage, this model is used to pseudo label a subset of unlabelled data as ‘normal’ or ‘anomalous’, followed by denoising of pseudo labels with CLIP. The final stage involves retraining on labelled and pseudo labelled data using the proposed Dual Centre Representation Learning (DCRL) which learns the centres of two representation spaces; normal and anomalous. Disease severity is then graded based on the distance to the learned centres. The proposed methodology outperforms existing techniques by margins of up to 24% in terms of OA detection and the disease severity scores correlate with the Kellgren-Lawrence grading system at the same level as human expert performance. Code available at https://github.com/niamhbelton/SS-FewSOME_Disease_Severity_Knee_Osteoarthritis. •This work proposes a continuous Knee Osteoarthritis grading system.•We advance the original Anomaly Detection technique FewSOME.•This method uses less than 3% of the labels that existing methods require.