Decoding adolescent TMJ osteoarthritis with multimodal machine learning

Background: Early and accurate diagnosis of adolescent temporomandibular joint (TMJ) osteoarthritis (OA) is critical, as degenerative changes during growth can cause lifelong pain and deformity. This study aimed to identify key clinical and imaging predictors of adolescent TMJ-OA and to evaluate multimodal machine learning models. Methods: The diagnostic utility was evaluated in 79 adolescents (10–18 years) with TMJ pain using panoramic radiography (PR) and MRI. TMJ-OA was diagnosed based on the Diagnostic Criteria for Temporomandibular Disorders (DC/TMD). Three decision tree models were developed: Model 1 (clinical-only), Model 2 (imaging-only), and Model 3 (combined clinical and imaging). Logistic regression was used for the comparisons. Results: To ensure a robust evaluation with a small sample size (n = 79), the models were assessed using nested 5-fold cross-validation. Model 2 (imaging only) had the highest specificity (0.7714 ± 0.2321), accuracy (0.5942 ± 0.0966), and AUROC (0.719± 0.101), but a low sensitivity (0.4472 ± 0.2065). PR evidence of TMJ-OA (feature importance = 0.70; OR = 3.93) was the strongest predictor and root node in the decision tree. Model 3 (combined clinical and imaging data) showed improved sensitivity (0.6056± 0.1829), identifying PR_TMJ_OA, MRI_TMJ_ADD (anterior disc displacement), Visual Analog Scale (VAS) score, and age as key nodes (AUROC = 0.6573 ± 0.0338; OR = 2.85 for PR_TMJ_OA). Model 1 (clinical-only) had limited predictive performance (AUROC = 0.4859 ± 0.0894), with symptom duration (importance = 0.64; OR = 1.40), VAS score, and joint locking (importance = 0.20) contributing modestly. A model using PR_TMJ_OA alone achieved perfect specificity (0.9714 ± 0.0571) but low sensitivity (0.3806 ± 0.1458). Conclusions: Although PR is a meaningful screening tool for adolescent TMJ-OA, it remains insufficient as a standalone diagnostic modality. Multimodal integration of clinical and MRI findings improves diagnostic accuracy and provides interpretable, clinically aligned decision-support tools for TMJ-OA.

Temporomandibular disorders; Osteoarthritis; Adolescents; Magnetic resonance imaging; Panoramic radiography; Machine learning; Decision trees

[1] Mélou C, Sixou JL, Sinquin C, Chauvel-Lebret D. Temporomandibular disorders in children and adolescents: a review. Archives of Pediatrics. 2023; 30: 335–342.

[2] Minghelli B, Cardoso I, Porfírio M, Gonçalves R, Cascalheiro S, Barreto V, et al. Prevalence of temporomandibular disorder in children and adolescents from public schools in southern Portugal. North American Journal of Medicine and Science. 2014; 6: 126–132.

[3] Schiffman E, Ohrbach R. Executive summary of the Diagnostic Criteria for Temporomandibular Disorders for clinical and research applications. The Journal of the American Dental Association. 2016; 147: 438–445.

[4] Dworkin SF, LeResche L. Research diagnostic criteria for temporomandibular disorders: review, criteria, examinations and specifications, critique. Journal of Craniomandibular Disorders. 1992; 6: 301–355.

[5] Stoustrup P, Resnick CM, Abramowicz S, Pedersen TK, Michelotti A, Küseler A, et al. Management of orofacial manifestations of juvenile idiopathic arthritis: interdisciplinary consensus-based recommendations. Arthritis & Rheumatology. 2023; 75: 4–14.

[6] Yi J, Sun Y, Li Y, Li C, Li X, Zhao Z. Cone-beam computed tomography versus periapical radiograph for diagnosing external root resorption: a systematic review and meta-analysis. The Angle Orthodontist. 2017; 87: 328–337.

[7] Zhao R, Xiong X, Li Z, Zhang L, Yang H, Ye Z. Recent advances and educational strategies in diagnostic imaging for temporomandibular disorders: a narrative literature review. Frontiers in Neurology. 2025; 16: 1597312.

[8] Zhang T, Zhang Q, Wei J, Dai Q, Muratovic D, Zhang W, et al. Nanoparticle-enabled molecular imaging diagnosis of osteoarthritis. Materials Today Bio. 2025; 33: 101952.

[9] Lee YH, Jeon S, Won JH, Auh QS, Noh YK. Automatic detection and visualization of temporomandibular joint effusion with deep neural network. Scientific Reports. 2024; 14: 18865.

[10] McQueen FM, Chapman P, Pollock T, D’Souza D, Lee AC, Dalbeth N, et al. Changes in clinical disease activity are weakly linked to changes in MRI inflammation on treat-to-target escalation of therapy in rheumatoid arthritis. Arthritis Research & Therapy. 2017; 19: 241.

[11] Ozsari S, Güzel MS, Yılmaz D, Kamburoğlu K. A comprehensive review of artificial intelligence based algorithms regarding temporomandibular joint related diseases. Diagnostics. 2023; 13: 2700.

[12] Blockeel H, Devos L, Frénay B, Nanfack G, Nijssen S. Decision trees: from efficient prediction to responsible AI. Frontiers in Artificial Intelligence. 2023; 6: 1124553.

[13] Ning Y, Li S, Ong MEH, Xie F, Chakraborty B, Ting DSW, et al. A novel interpretable machine learning system to generate clinical risk scores: an application for predicting early mortality or unplanned readmission in a retrospective cohort study. PLOS Digital Health. 2022; 1: e0000062.

[14] Lee YH, Lee KM, Auh QS, Hong JP. Magnetic resonance imaging-based prediction of the relationship between whiplash injury and temporomandibular disorders. Frontiers in Neurology. 2017; 8: 725.

[15] Lee YH, Hong IK, Chun YH. Prediction of painful temporomandibular joint osteoarthritis in juvenile patients using bone scintigraphy. Clinical and Experimental Dental Research. 2019; 5: 225–235.

[16] Elreedy D, Atiya AF, Kamalov F. A theoretical distribution analysis of synthetic minority oversampling technique (SMOTE) for imbalanced learning. Machine Learning. 2024; 113: 4903–4923.

[17] Ferrer CA, Aragón E. Note on “a comprehensive analysis of synthetic minority oversampling technique (SMOTE) for handling class imbalance”. Information Sciences. 2023; 630: 322–324.

[18] Kothari SF, Baad-Hansen L, Hansen LB, Bang N, Sørensen LH, Eskildsen HW, et al. Pain profiling of patients with temporomandibular joint arthralgia and osteoarthritis diagnosed with different imaging techniques. The Journal of Headache and Pain. 2016; 17: 61.

[19] Wang XD, Zhang JN, Gan YH, Zhou YH. Current understanding of pathogenesis and treatment of TMJ osteoarthritis. Journal of Dental Research. 2015; 94: 666–673.

[20] Trevethan R. Sensitivity, specificity, and predictive values: foundations, pliabilities, and pitfalls in research and practice. Frontiers in Public Health. 2017; 5: 307.

[21] Das SK. TMJ osteoarthritis and early diagnosis. Journal of Oral Biology and Craniofacial Research. 2013; 3: 109–110.

[22] Yin Y, He S, Xu J, You W, Li Q, Long J, et al. The neuro-pathophysiology of temporomandibular disorders-related pain: a systematic review of structural and functional MRI studies. The Journal of Headache and Pain. 2020; 21: 78.

[23] Suenaga S, Nagayama K, Nagasawa T, Indo H, Majima HJ. The usefulness of diagnostic imaging for the assessment of pain symptoms in temporomandibular disorders. Japanese Dental Science Review. 2016; 52: 93–106.

[24] Chantaracherd P, John MT, Hodges JS, Schiffman EL. Temporomandibular joint disorders’ impact on pain, function, and disability. Journal of Dental Research. 2015; 94: 79S–86S.

[25] Roh HS, Kim W, Kim YK, Lee JY. Relationships between disk displacement, joint effusion, and degenerative changes of the TMJ in TMD patients based on MRI findings. Journal of Cranio-Maxillofacial Surgery. 2012; 40: 283–286.

[26] Mobilio N, Catapano S. Effect of experimental jaw muscle pain on occlusal contacts. Journal of Oral Rehabilitation. 2011; 38: 404–409.

[27] Wideman TH, Edwards RR, Walton DM, Martel MO, Hudon A, Seminowicz DA. The multimodal assessment model of pain: a novel framework for further integrating the subjective pain experience within research and practice. The Clinical Journal of Pain. 2019; 35: 212–221.

[28] dos Anjos Pontual ML, Freire JS, Barbosa JM, Frazão MA, dos Anjos Pontual A. Evaluation of bone changes in the temporomandibular joint using cone beam CT. Dentomaxillofacial Radiology. 2012; 41: 24–29.

[29] Kaimal S, Ahmad M, Kang W, Nixdorf D, Schiffman EL. Diagnostic accuracy of panoramic radiography and MRI for detecting signs of TMJ degenerative joint disease. General Dentistry. 2018; 66: 34–40.

[30] Costa Dias S, Habre C, Di Paolo PL, d’Angelo P, Augdal TA, Angenete OW, et al. ESR essentials: juvenile idiopathic arthritis; what every radiologist needs to know—practice recommendations by the European Society of Paediatric Radiology. European Radiology. 2026; 36: 1261–1271.

[31] Maita I, Dhillon A, Friesen R, Almeida FT. Diagnostic value of panoramic radiographs in the assessment of degenerative joint disease: a retrospective study. International Dental Journal. 2025; 75: 100910.

[32] Wu S, Zhang D, Xia S, Shen P, Yang C. Effect of temporomandibular joint anterior disc displacement on condylar height in different age groups. BMC Oral Health. 2025; 25: 1100.

[33] Kumar R, Pallagatti S, Sheikh S, Mittal A, Gupta D, Gupta S. Correlation between clinical findings of temporomandibular disorders and MRI characteristics of disc displacement. The Open Dentistry Journal. 2015; 9: 273–281.

[34] González-González AM, Herrero AJ. A systematic review of temporomandibular disorder diagnostic methods. CRANIO®. 2024; 42: 348–360.

[35] Lee Y-H, Won JH, Kim S, Auh QS, Noh Y-K. Advantages of deep learning with convolutional neural network in detecting disc displacement of the temporomandibular joint in magnetic resonance imaging. Scientific Reports. 2022; 12: 11352.

[36] Rodríguez DM, Cuéllar MP, Morales DP. Concept logic trees: enabling user interaction for transparent image classification and human-in-the-loop learning. Applied Intelligence. 2024; 54: 3667–3679.

[37] Hassan R, Nguyen N, Finserås SR, Adde L, Strümke I, Støen R. Unlocking the black box: enhancing human-AI collaboration in high-stakes healthcare scenarios through explainable AI. Technological Forecasting and Social Change. 2025; 219: 124265.

[38] Frantsve-Hawley J, Abt E, Carrasco-Labra A, Dawson T, Michaels M, Pahlke S, et al. Strategies for developing evidence-based clinical practice guidelines to foster implementation into dental practice. The Journal of the American Dental Association. 2022; 153: 1041–1052.

[39] Kweon HH, Lee JH, Youk TM, Lee BA, Kim YT. Panoramic radiography can be an effective diagnostic tool adjunctive to oral examinations in the national health checkup program. Journal of Periodontal & Implant Science. 2018; 48: 317–325.

[40] Im GI. The concept of early osteoarthritis and its significance in regenerative medicine. Tissue Engineering and Regenerative Medicine. 2022; 19: 431–436.

[41] Wei Y, Qian H, Zhang X, Wang J, Yan H, Xiao N, et al. Progress in multi-omics studies of osteoarthritis. Biomarker Research. 2025; 13: 26.

[42] Clarke E, Varela L, Jenkins RE, Lozano-Andrés E, Cywińska A, Przewozny M, et al. Proteome and phospholipidome interrelationship of synovial fluid-derived extracellular vesicles in equine osteoarthritis: an exploratory ‘multi-omics’ study to identify composite biomarkers. Biochemistry and Biophysics Reports. 2024; 37: 101635.

[43] Liu Y, Zhang S, Liu K, Hu X, Gu X. Advances in drug discovery based on network pharmacology and omics technology. Current Pharmaceutical Analysis. 2024; 21: 33–43.

[44] Wang Z, Zhao Y, Zhang L. Emerging trends and hot topics in the application of multi-omics in drug discovery: a bibliometric and visualized study. Current Pharmaceutical Analysis. 2024; 21: 20–32.

[45] Chustecki M. Benefits and risks of AI in health care: narrative review. Interactive Journal of Medical Research. 2024; 13: e53616.

[46] Alowais SA, Alghamdi SS, Alsuhebany N, Alqahtani T, Alshaya AI, Almohareb SN, et al. Revolutionizing healthcare: the role of artificial intelligence in clinical practice. BMC Medical Education. 2023; 23: 689.