[Development and validation of a clinical automatic diagnosis system based on diagnostic criteria for temporomandibular disorders].

OBJECTIVE

To develop a clinical automated diagnostic system for temporomandibular disorders (TMD) based on the diagnostic criteria for TMD (DC/TMD) to assist dentists in making rapid and accurate clinical diagnosis of TMD.

METHODS

Clinical and imaging data of 354 patients, who visited the Center for TMD & Orofacial Pain at Peking University Hospital of Stomatology from September 2023 to January 2024, were retrospectively collected. The study developed a clinical automated diagnostic system for TMD using the DC/TMD, built on the. NET Framework platform with branching statements as its internal structure. Further validation of the system on consistency and diagnostic efficacy compared with DC/TMD were also explored. Diagnostic efficacy of the TMD clinical automated diagnostic system for degenerative joint diseases, disc displacement with reduction, disc displacements without reduction with limited mouth opening and disc displacement without reduction without limited mouth opening was evaluated and compared with a specialist in the field of TMD. Accuracy, precision, specificity and the Kappa value were assessed between the TMD clinical automated diagnostic system and the specialist.

RESULTS

Diagnoses for various TMD subtypes, including pain-related TMD (arthralgia, myalgia, headache attributed to TMD) and intra-articular TMD (disc displacement with reduction, disc displacement with reduction with intermittent locking, disc displacement without reduction with limited opening, disc displacement without reduction without limited opening, degenerative joint disease and subluxation), using the TMD clinical automated diagnostic system were completely identical to those obtained by the TMD specialist based on DC/TMD. Both the system and the expert showed low sensitivity for diagnosing degenerative joint disease (0.24 and 0.37, respectively), but high specificity (0.96). Both methods achieved high accuracy (> 0.9) for diagnosing disc displacements with reduction and disc displacements without reduction with limited mouth opening. The sensitivity for diagnosing disc displacement without reduction without limited mouth opening was only 0.59 using the automated system, lower than the expert (0.87), while both had high specificity (0.92). The Kappa values for most TMD subtypes were close to 1, except the disc displacement without reduction without limited mouth opening, which had a Kappa value of 0.68.

CONCLUSION

This study developed and validated a reliable clinical automated diagnostic system for TMD based on DC/TMD. The system is designed to facilitate the rapid and accurate diagnosis and classification of TMD, and is expected to be an important tool in clinical scenarios.