Integrating deep learning-based radiographic features with demographic variables for hip osteoarthritis prediction: development and validation using individual participant data from the World COACH Consortium
Year: 2026
Type: article
Abstract: OBJECTIVE: To evaluate whether hybrid models integrating deep learning (DL) features from baseline radiographs and demographic information can predict incident radiographic hip osteoarthritis (RHOA) and to compare their performance to models using radiographic or demographic information in isolation. DESIGN: Data were pooled from seven prospective cohorts within the Worldwide Collaboration on OsteoArthritis prediction for the Hip (World COACH) consortium. Incident RHOA was defined as hips without RHOA at baseline that developed RHOA within four to eight years. From baseline radiographs, radiographic features were extracted with a DL-based feature extractor that was either unoptimized, optimized with baseline RHOA grades, or optimized with follow-up RHOA grades. The baseline RHOA grade acted as conventional reference. Baseline age, body mass index, and birth-assigned sex were considered as demographic predictors. Using internal and leave-one-cohort-out cross-validation, we assessed the predictive performance of models using only demographic predictors, only radiographic predictors, or both modalities combined (hybrid). RESULTS: In the pooled dataset of 25,090 hips, 5.1% developed incident RHOA in four to eight years. Both the radiographic-only and hybrid models including follow-up-optimized radiographic features showed the best discriminative performance (AUC: 0.81 ± 0.01), whereas the demographic-only model showed an AUC of 0.71 (0.01) during internal cross-validation. The performance of the hybrid model including follow-up-optimized features dropped to an average AUC of 0.64 ± 0.09 during leave-one-cohort-out cross-validation. CONCLUSIONS: Hybrid models combining DL-based features from baseline radiographs with demographic predictors show promise for RHOA risk identification, though demographics provided limited complementary information and performance declined in leave-one-cohort-out cross-validation.
Source: Osteoarthritis Imaging
Authors: Myrthe A. van den Berg, Gijs van Tulder, Harbeer Ahedi, Nigel K. Arden, Sita M.A. Bierma-Zeinstra, Fleur Boel, Cindy G. Boer, Michiel M.A. van Buuren, Flavia M. Cicuttini, Timothy F. Cootes, David T. Felson, Willem Paul Gielis, Joshua Heerey, Joanne Kemp, Graeme Jones, Stefan Kluzek, Nancy E. Lane, Claudia Lindner, Joyce B.J. van Meurs, Andrea B. Mosler, Amanda E. Nelson, Michael C. Nevitt, Edwin H. Oei, Jos Runhaar, Harrie Weinans, Jesse H. Krijthe, Rintje Agricola
Institutions: Erasmus MC, Delft University of Technology, University of Tasmania, Sydney Orthopaedic Research Institute, Menzies Institute for Medical Research, University of Oxford, Erasmus University Rotterdam, Monash University, University of Manchester, Boston University, University Medical Center Utrecht, La Trobe University, University of Nottingham, National Institute for Health and Care Research, NIHR Oxford Biomedical Research Centre, University of California, Davis, University of North Carolina at Chapel Hill, University of California System
Language: English
Topic: Hip disorders and treatments
Subfield: Surgery
Field: Medicine
Domain: Health Sciences
Open Access status: hybrid
Estimated APC paid: $1,400
OpenAlex ID: W7213406037