MRI Radiomics Plus Biomarker Model Can Predict Knee OA Progression

WEDNESDAY, Aug. 27, 2025 (HealthDay News) — Longitudinal magnetic resonance imaging (MRI) radiomics features of load-bearing knee joint tissues can help predict progression of knee osteoarthritis (KOA), according to a study published online Aug. 21 in PLOS Medicine.

Ting Wang, from The Third People”s Hospital of Chengdu in China, and colleagues developed and tested the longitudinal Load-Bearing Tissue Radiomic plus Biochemical biomarker and Clinical variable Model (LBTRBC-M) to predict KOA progression. A total of 594 participants with Kellgren-Lawrence grades 1 to 3 and complete biomarker data were selected (mean age, 61.6 ± 8.9 years; 58.8 percent female; 79.3 percent White). A total of 1,753 knee MRIs were included: 594 at baseline, 575 at one-year follow-up, and 584 at two-year follow-up. The model was developed in the total development cohort (877 participants) and tested among 876 participants in the total test cohort.

The researchers found that the area under the receiver operating characteristic curves of LBTRBC-M for predicting joint space narrowing (JSN) and pain progression, JSN progression, pain progression, and nonprogression were 0.880, 0.913, 0.886, and 0.909, respectively. LBTRBC-M had overall accuracy of 70.1 percent. The prognostic accuracy of resident physicians improved from 44.7 to 49.0 percent to 64.4 to 66.5 percent with LBTRBC-M assistance.

“The integrated model using longitudinal MRI radiomics, biochemical biomarkers, and clinical KOA high-risk factors, improved the prediction performance of KOA progression. LBTRBC-M enhances the accuracy of resident physicians to predict KOA progression,” the authors write.

One author disclosed ties to the biopharmaceutical and medical technology industries.

Abstract/Full Text