Skip to content
DraconiticResearch
Search Sources API
226,742 studies31 sourcesUpdated 34 min ago
Peer-reviewedCC BYDigital Dentistry Journal

Artificial intelligence in imaging-based diagnosis of temporomandibular disorders: A systematic review and meta-analysis of diagnostic accuracy

8 affiliations
  1. 1Division of Oral and Maxillofacial Surgery, Massachusetts General Hospital, Department of Oral and Maxillofacial Surgery, Harvard School of Dental Medicine, Boston, MA, USA; University of KU Leuven, Faculty of Medicine, Leuven, Belgium
  2. 2Division of Oral and Maxillofacial Surgery, Massachusetts General Hospital, Department of Oral and Maxillofacial Surgery, Harvard School of Dental Medicine, Boston, MA, USA; University of the Pacific, Arthur A. Dugoni School of Dentistry, San Francisco, CA, USA
  3. 3Division of Oral and Maxillofacial Surgery, Massachusetts General Hospital, Department of Oral and Maxillofacial Surgery, Harvard School of Dental Medicine, Boston, MA, USA
  4. 4Department of Orofacial Pain and Oral Medicine, Kyung Hee University Dental Hospital, Kyung Hee University School of Dentistry, Seoul, South Korea; Center for Systems Biology, Massachusetts General Hospital, Harvard Medical School, Boston, USA
  5. 5OMFS IMPATH Research Group, Department of Imaging and Pathology, Faculty of Medicine, KU Leuven, Leuven, Belgium; Department of Oral and Maxillofacial Surgery, University Hospitals Leuven, Leuven, Belgium; Department of Oral and Maxillofacial Surgery, University of Illinois Chicago, Chicago, USA
  6. 6OMFS IMPATH Research Group, Department of Imaging and Pathology, Faculty of Medicine, KU Leuven, Leuven, Belgium; Department of Oral and Maxillofacial Surgery, University Hospitals Leuven, Leuven, Belgium
  7. 7University of Ulm, Faculty of Medicine, Ulm, Germany
  8. 8OMFS IMPATH Research Group, Department of Imaging and Pathology, Faculty of Medicine, KU Leuven, Leuven, Belgium; Department of Oral and Maxillofacial Surgery, University Hospitals Leuven, Leuven, Belgium; Department of Dental Medicine, Karolinska Institute, Stockholm, Sweden

Abstract

Objective: Temporomandibular disorders (TMD) have overlapping symptoms that complicate diagnosis. Artificial intelligence (AI) techniques (e.g., deep learning, radiomics) may improve diagnostic accuracy across imaging modalities, but variability and limited generalizability require systematic evaluation. This review and meta-analysis assessed AI-based TMD diagnostic models, focusing on diagnostic accuracy across imaging modalities and methodological quality. Data sources: PubMed, Embase, Scopus, Web of Science, and the Cochrane Library were searched from January 2015 to January 2026. Study selection: In accordance with PRISMA guidelines, 791 records were screened, 37 full texts were assessed, and 30 studies were included in the systematic review; 17 were eligible for meta-analysis. Included studies were diagnostic accuracy studies and other observational designs that compared AI-based diagnosis with a reference standard and reported sensitivity and specificity. The majority followed a retrospective design. Two reviewers independently screened studies, extracted data, and assessed risk of bias (QUADAS-2). Random-effects models pooled sensitivity and specificity by imaging modality, with heterogeneity assessed using the I 2 statistic. Results: Deep learning on CBCT and MRI yielded high pooled sensitivities (both 0.94) and specificities (0.84 and 0.76, respectively). Conventional radiography (panoramic and cephalometric radiographs) demonstrated moderate performance (sensitivity 0.77; specificity 0.86). Beyond the meta-analysis, several models achieved near-expert 3D segmentation (Dice >0.95) and accurate risk stratification using multimodal data. Integration of imaging with clinical data and advanced modeling approaches improved predictive performance and interpretability. Conclusions: AI systems, particularly deep learning applied to CBCT and MRI, demonstrate high sensitivity for TMD diagnosis, although specificity estimates were more variable and the certainty of evidence was limited. This modality-stratified meta-analysis provides a benchmark for future validation studies. This is the first modality-stratified meta-analysis of pooled AI diagnostic accuracy in TMD, supporting improved diagnostic consistency and integration of imaging with clinical data.

Related

  1. Temporomandibular Joint Disc Exploration, and Its Pathologies diagnosis by using Artificial Intelligence: A Systematic Review Exploration du disque de l'articulation temporo-mandibulaire et diagnostic de ses pathologies par intelligence artificielle: Une revue systématique
    Oussama Abali, Aissa El Miad Kerkour, Abdelkrim Daoudi
    La Tunisie medicalePeer-reviewedCC BY-NC-NDFull textGeneral medicineEurope PMCPMC
  2. Diagnostic accuracy of tongue diagnosis for stroke syndrome differentiation in traditional Chinese medicine: a systematic review, meta-analysis, and narrative synthesis of artificial intelligence-assisted models.
    Yibo Huang, Jiashun Yin, Siyuan Shao +3
    Frontiers in neurologyPeer-reviewedCC BYFull textNeurologyEurope PMCCrossref
  3. Diagnostic performance of artificial intelligence in early diagnosis of mosquito-borne viral infections: Systematic review and meta-analysis
    Flavia Pennisi, Antonio Pinto, Claudia Cozzolino Cangiano +4
    PLOS Digital HealthPeer-reviewedCC BY 4.0Full textBioinformatics & computational biologyMicrobiologyPLOSPMC
  4. Generative Artificial Intelligence Models in Breast Cancer Diagnosis: A Systematic Review
    Diagnostics (Basel, Switzerland)Peer-reviewedCC BYFull textGeneral medicineEurope PMC
  5. Artificial Intelligence-Based Burn Assessment Demonstrates High Diagnostic Accuracy: Systematic Review and Meta-Analysis
    Emily Uyen Thai, Mina Bahramand, Jordyn Carroll +5
    Plastic and Reconstructive Surgery, Global OpenPeer-reviewedCC BY-NC-NDFull textEpidemiology & public healthDOAJ
BibTeX
@article{maes2026artificial,
  title = {{Artificial intelligence in imaging-based diagnosis of temporomandibular disorders: A systematic review and meta-analysis of diagnostic accuracy}},
  author = {Mathias Maes and Seema Shrestha and David A. Keith and Shruti Handa and Yeon-Hee Lee and Frederic Van Der Cruyssen and Pieter-jan Verhelst and Yannick M. Sillmann and Reinhilde Jacobs and Fernando P.S. Guastaldi},
  journal = {Digital Dentistry Journal},
  year = {2026},
  month = nov,
  doi = {10.1016/j.ddj.2026.100106},
  url = {https://doi.org/10.1016/j.ddj.2026.100106},
  note = {CC BY},
}
RIS
TY  - JOUR
TI  - Artificial intelligence in imaging-based diagnosis of temporomandibular disorders: A systematic review and meta-analysis of diagnostic accuracy
AU  - Mathias Maes
AU  - Seema Shrestha
AU  - David A. Keith
AU  - Shruti Handa
AU  - Yeon-Hee Lee
AU  - Frederic Van Der Cruyssen
AU  - Pieter-jan Verhelst
AU  - Yannick M. Sillmann
AU  - Reinhilde Jacobs
AU  - Fernando P.S. Guastaldi
PY  - 2026
DA  - 2026/11/16
JO  - Digital Dentistry Journal
DO  - 10.1016/j.ddj.2026.100106
UR  - https://doi.org/10.1016/j.ddj.2026.100106
AB  - Objective: Temporomandibular disorders (TMD) have overlapping symptoms that complicate diagnosis. Artificial intelligence (AI) techniques (e.g., deep learning, radiomics) may improve diagnostic accuracy across imaging modalities, but variability and limited generalizability require systematic evaluation. This review and meta-analysis assessed AI-based TMD diagnostic models, focusing on diagnostic accuracy across imaging modalities and methodological quality. Data sources: PubMed, Embase, Scopus, Web of Science, and the Cochrane Library were searched from January 2015 to January 2026. Study selection: In accordance with PRISMA guidelines, 791 records were screened, 37 full texts were assessed, and 30 studies were included in the systematic review; 17 were eligible for meta-analysis. Included studies were diagnostic accuracy studies and other observational designs that compared AI-based diagnosis with a reference standard and reported sensitivity and specificity. The majority followed a retrospective design. Two reviewers independently screened studies, extracted data, and assessed risk of bias (QUADAS-2). Random-effects models pooled sensitivity and specificity by imaging modality, with heterogeneity assessed using the I 2 statistic. Results: Deep learning on CBCT and MRI yielded high pooled sensitivities (both 0.94) and specificities (0.84 and 0.76, respectively). Conventional radiography (panoramic and cephalometric radiographs) demonstrated moderate performance (sensitivity 0.77; specificity 0.86). Beyond the meta-analysis, several models achieved near-expert 3D segmentation (Dice >0.95) and accurate risk stratification using multimodal data. Integration of imaging with clinical data and advanced modeling approaches improved predictive performance and interpretability. Conclusions: AI systems, particularly deep learning applied to CBCT and MRI, demonstrate high sensitivity for TMD diagnosis, although specificity estimates were more variable and the certainty of evidence was limited. This modality-stratified meta-analysis provides a benchmark for future validation studies. This is the first modality-stratified meta-analysis of pooled AI diagnostic accuracy in TMD, supporting improved diagnostic consistency and integration of imaging with clinical data.
SN  - 2950-6433
ER  -