Prediction of knee pain improvement over two years for knee osteoarthritis using a dynamic nomogram based on MRI-derived radiomics: a proof-of-concept study

• DOI: 10.1016/j.joca.2022.10.014
• Publication Date: 2022-11-02T17:52:45Z
• Authors: T. Lin; S. Peng; S. Lu; S. Fu; D. Zeng; J. Li; T. Chen; T. Fan; C. Lang; S. Feng; J. Ma; C. Zhao; B. Antony; F. Cicuttini; X. Quan; Z. Zhu; C. Ding
• Keywords: Nomogram; Proof of concept; Knee pain

Objectives To develop and validate a nomogram to detect improved knee pain in osteoarthritis (OA) by integrating magnetic resonance imaging (MRI) radiomics signature of subchondral bone clinical characteristics. Methods Participants were selected from the Vitamin D Effects on Osteoarthritis (VIDEO) study. The primary outcome was 20% improvement score over 2 years participants administrated either vitamin or placebo. Radiomics features characteristics 216 extracted analyzed. randomly split into training validation cohorts at ratio 8:2. Least absolute shrinkage selection operator (LASSO) regression used select generate signatures. optimal indicators fitted using multivariable logistic model. Results showed favorable discrimination performance [AUCtraining, 0.79 (95% CI: 0.72–0.79), AUCvalidation, 0.83 0.70–0.96)] as well good calibration. Additional contributing value fusion statistically significant (NRI, 0.23; IDI, 0.14, P < 0.001 cohort NRI, 0.29; 0.18, 0.05 validating cohort). Decision curve analysis confirmed usefulness nomogram. Conclusion radiomics-based comprising MR variables achieves predictive efficacy accuracy differentiating among OA patients. This proof-of-concept study provides promising way predict clinically meaningful outcomes.