Related Experiment Video
Updated: Apr 23, 2026
06:06
Tissue Collection and RNA Extraction from the Human Osteoarthritic Knee Joint
Published on: July 22, 2021
6.0K
Development of Novel and Interpretable Automated Models for Predicting Total Knee Replacement in Knee Osteoarthritis:
Fanfan Zhao1, Yao Chen2, Xiaoyue Zhou3
1Center for Rehabilitation Medicine, Department of Radiology, Zhejiang Provincial People's Hospital (Affiliated People's Hospital), Hangzhou Medical College, Hangzhou, China (F.Z., L.X., C.L., D.H.).
Academic Radiology
|April 21, 2026
Summary
A new bimodal model accurately predicts knee osteoarthritis progression to total knee replacement. This model, integrating infrapatellar fat pad features and KL grading, outperforms existing methods for knee osteoarthritis (KOA) risk assessment.
Area of Science:
- Orthopedics
- Radiology
- Biomedical Engineering
Background:
- Subtle changes in the infrapatellar fat pad (IPFP) are linked to knee osteoarthritis (KOA) progression.
- Accurate prediction of KOA progression to total knee replacement (TKR) is crucial for patient management.
Purpose of the Study:
- To develop and validate a novel bimodal model for predicting TKR risk in KOA patients.
- To compare the predictive performance of the bimodal model against existing methods.
Main Methods:
- Retrospective analysis of 4039 participants from the Osteoarthritis Initiative (OAI) database.
- Development of a bimodal model integrating IPFP radiomic features and Kellgren-Lawrence (KL) grading.
- Validation using internal and external datasets, with performance evaluated by ROC-AUC and diagnostic metrics.
Main Results:
- The bimodal model achieved superior internal validation performance with an AUC of 0.903, outperforming the Fatpad-Score (AUC=0.848) and KL grading model (AUC=0.801).
- The model demonstrated high accuracy in external validation cohorts (0.918 and 1.000).
- Net reclassification improvement (NRI) tests confirmed the bimodal model's enhanced predictive capability.
Conclusions:
- The developed bimodal model offers improved prediction of TKR risk in KOA patients.
- This model provides a more accurate tool for assessing KOA progression compared to the Fatpad-score and KL grading alone.
