Predictive performance of supervised machine learning models for temporomandibular joint disc displacement in adolescents diagnosed by magnetic resonance imaging

Background: Temporomandibular joint (TMJ) disc displacement is a common condition among adolescents and often requires accurate imaging-based diagnosis. This study aimed to assess the predictive performance of supervised machine learning (ML) algorithms in classifying TMJ disc displacement in adolescents based on morphological and signal intensity features derived from Magnetic Resonance Imaging (MRI). Methods: MRI data from 260 TMJs of 130 adolescent patients (aged 12–18 years) were retrospectively analyzed. Morphological variables included condylar shape, anteroposterior and mediolateral diameters, and a condylar ratio derived from these measurements. Signal intensity ratios of the superior and inferior heads of the lateral pterygoid muscle were also recorded. Disc status was initially categorized as normal, anterior displacement with reduction, or without reduction; for modeling purposes, dislocated groups were combined. ML analyses were performed using a patient-level data split to prevent information leakage, ensuring that both joints from the same patient were assigned to the same dataset. Fifteen supervised classification models were trained using group-aware cross-validation within the training set, followed by evaluation on a held-out test set. Hyperparameters were optimized exclusively within the training data. Model performance was assessed using standard classification metrics, with 95% confidence intervals estimated by patient-level bootstrap resampling. Results: Anteroposterior and mediolateral diameters differed significantly between disc status groups (p < 0.001). Among the evaluated models, MLPClassifier achieved the highest area under the receiver operating characteristic curve (ROC-AUC) (0.797), followed by Random Forest (0.777) and HistGradientBoosting (0.773). Although CatBoost ranked fourth in overall discrimination (ROC-AUC = 0.767), it demonstrated the highest recall (0.839), indicating superior sensitivity for detecting disc displacement. Ensemble and neural network-based models generally outperformed linear classifiers. Conclusions: Supervised ML algorithms showed balanced potential in diagnosing TMJ disc displacement in adolescents. Further validation in prospective clinical and multicenter studies is warranted to support clinical application.

Temporomandibular joint disorders; Magnetic resonance imaging; Machine learning; Pediatric radiology

[1] Manfredini D, Guarda-Nardini L, Winocur E, Piccotti F, Ahlberg J, Lobbezoo F. Research diagnostic criteria for temporomandibular disorders: a systematic review of axis I epidemiologic findings. Oral Surgery, Oral Medicine, Oral Pathology, Oral Radiology, and Endodontology. 2011; 112: 453–462.

[2] Cong N, Wang N, Huang S, Cheng T, Yan X. Diagnostic significance of magnetic resonance imaging in distinguishing temporomandibular disorders: a retrospective chart review. BMC Oral Health. 2021; 21: 481.

[3] Kundu H, Basavaraj P, Kote S, Singla A, Singh S. Assessment of TMJ disorders using ultrasonography as a diagnostic tool: a review. Journal of Clinical and Diagnostic Research. 2013; 7: 3116–3120.

[4] Sena MF, Mesquita KS, Santos FR, Silva FW, Serrano KV. Prevalence of temporomandibular dysfunction in children and adolescents. Revista Paulista de Pediatria. 2013; 31: 538–545.

[5] Ahmad M, Hollender L, Anderson Q, Kartha K, Ohrbach R, Truelove EL, et al. Research diagnostic criteria for temporomandibular disorders (RDC/TMD): development of image analysis criteria and examiner reliability for image analysis. Oral Surgery, Oral Medicine, Oral Pathology, Oral Radiology, and Endodontology. 2009; 107: 844–860.

[6] Manfredini D, Häggman-Henrikson B, Al Jaghsi A, Baad-Hansen L, Beecroft E, Bijelic T, et al. Temporomandibular disorders: INfORM/IADR key points for good clinical practice based on standard of care. CRANIO®. 2025; 43: 1–5.

[7] Petersson A. What you can and cannot see in TMJ imaging—an overview related to the RDC/TMD diagnostic system. Journal of Oral Rehabilitation. 2010; 37: 771–778.

[8] Klatkiewicz T, Gawriołek K, Pobudek Radzikowska M, Czajka-Jakubowska A. Ultrasonography in the diagnosis of temporomandibular disorders: a meta-analysis. Medical Science Monitor. 2018; 24: 812–817.

[9] ElShennawy EM, Hamed WM, Samir SM. Diagnostic accuracy of MRI-CBCT fused images in assessment of clinically diagnosed internal derangement of the temporomandibular joint. Oral Radiology. 2024; 40: 226–233.

[10] Tomas X, Pomes J, Berenguer J, Quinto L, Nicolau C, Mercader JM, et al. MR imaging of temporomandibular joint dysfunction: a pictorial review. Radiographics. 2006; 26: 765–781.

[11] Mercuri LG, Greene CS, Manfredini D. The temporomandibular joint disc: a complex fable about an elusive butterfly. CRANIO®. 2026; 44: 228–235.

[12] Guarda Nardini L, Meneghini M, Guido M, Baciorri F, Manfredini D. Histopathology of the temporomandibular joint disc: findings in 30 samples from joints with degenerative disease. Journal of Oral Rehabilitation. 2021; 48: 1025–1034.

[13] Legrell PE, Reibel J, Nylander K, Hörstedt P, Isberg A. Temporomandibular joint condyle changes after surgically induced non-reducing disk displacement in rabbits: a macroscopic and microscopic study. Acta Odontologica Scandinavica. 1999; 57: 290–300.

[14] Jha N, Lee KS, Kim YJ. Diagnosis of temporomandibular disorders using artificial intelligence technologies: a systematic review and meta-analysis. PLOS ONE. 2022; 17: e0272715.

[15] Duyan Yüksel H, Orhan K, Evlice B, Kaya Ö. Evaluation of temporomandibular joint disc displacement with MRI-based radiomics analysis. Dentomaxillofacial Radiology. 2025; 54: 19–27.

[16] Kao ZK, Chiu NT, Wu HH, Chang WC, Wang DH, Kung YY, et al. Classifying temporomandibular disorder with artificial intelligent architecture using magnetic resonance imaging. Annals of Biomedical Engineering. 2023; 51: 517–526.

[17] Lasek J, Nurzynska K, Piórkowski A, Strzelecki M, Obuchowicz R. Deep learning for ultrasonographic assessment of temporomandibular joint morphology. Tomography. 2025; 11: 27.

[18] Nagamatsu-Sakaguchi C, Maekawa K, Ono T, Yanagi Y, Minakuchi H, Miyawaki S, et al. Test-retest reliability of MRI-based disk position diagnosis of the temporomandibular joint. Clinical Oral Investigations. 2012; 16: 101–108.

[19] Soydan Çabuk D, Etöz M, Akgün İE, Doğan S, Öztürk E, Coşgunarslan A. The evaluation of lateral pterygoid signal intensity changes related to temporomandibular joint anterior disc displacement. Oral Radiology. 2021; 37: 74–79.

[20] Manfredini D. Etiopathogenesis of disk displacement of the temporomandibular joint: a review of the mechanisms. Indian Journal of Dental Research. 2009; 20: 212–221.

[21] Bueno CH, Pereira DD, Pattussi MP, Grossi PK, Grossi ML. Gender differences in temporomandibular disorders in adult populational studies: a systematic review and meta-analysis. Journal of Oral Rehabilitation. 2018; 45: 720–729.

[22] Minervini G, Franco R, Marrapodi MM, Fiorillo L, Cervino G, Cicciù M. Prevalence of temporomandibular disorders in children and adolescents evaluated with Diagnostic Criteria for Temporomandibular Disorders: a systematic review with meta-analysis. Journal of Oral Rehabilitation. 2023; 50: 522–530.

[23] Olate S, Ravelo V, Alister JP, Netto HD, Haidar ZS, Sacco R. Early treatment of unilateral condylar hyperplasia in adolescents: preliminary results. Journal of Clinical Medicine. 2023; 12: 3408.

[24] Tallents RH, Katzberg RW, Murphy W, Proskin H. Magnetic resonance imaging findings in asymptomatic volunteers and symptomatic patients with temporomandibular disorders. Journal of Prosthetic Dentistry. 1996; 75: 529–533.

[25] Kurita H, Ohtsuka A, Kobayashi H, Kurashina K. Alteration of the horizontal mandibular condyle size associated with temporomandibular joint internal derangement in adult females. Dentomaxillofacial Radiology. 2002; 31: 373–378.

[26] Katsavrias EG, Halazonetis DJ. Condyle and fossa shape in Class II and Class III skeletal patterns: a morphometric tomographic study. American Journal of Orthodontics and Dentofacial Orthopedics. 2005; 128: 337–346.

[27] Alexiou K, Stamatakis H, Tsiklakis K. Evaluation of the severity of temporomandibular joint osteoarthritic changes related to age using cone beam computed tomography. Dentomaxillofacial Radiology. 2009; 38: 141–147.

[28] Orhan K, Driesen L, Shujaat S, Jacobs R, Chai X. Development and validation of a magnetic resonance imaging-based machine learning model for TMJ pathologies. BioMed Research International. 2021; 2021: 6656773.

[29] Taborri J, Molinaro L, Russo L, Palmerini V, Larion A, Rossi S. Comparison of machine learning algorithms fed with mobility-related and baropodometric measurements to identify temporomandibular disorders. Sensors. 2024; 24: 3646.

[30] Yıldız NT, Kocaman H, Yıldırım H, Canlı M. An investigation of machine learning algorithms for prediction of temporomandibular disorders by using clinical parameters. Medicine. 2024; 103: e39912.

[31] Zhang Y, Zhu T, Zheng Y, Xiong Y, Liu W, Zeng W, et al. Machine learning-based medical imaging diagnosis in patients with temporomandibular disorders: a diagnostic test accuracy systematic review and meta-analysis. Clinical Oral Investigations. 2024; 28: 186.

[32] Farook TH, Dudley J. Automation and deep (machine) learning in temporomandibular joint disorder radiomics: a systematic review. Journal of Oral Rehabilitation. 2023; 50: 501–521.