A Symptom-Driven Imaging Protocol for Temporomandibular Disorders in Elite Athletes: Integrating AI and Conventional Modalities
Year: 2026
Type: article
Abstract: Background. Temporomandibular disorders (TMDs) affect the TMJ, masticatory muscles, and related structures. They are critical in sports medicine, as athletes face unique stressors—from facial trauma to exertional bruxism—impacting postural stability and quality of life. Despite their clinical significance, standardized imaging protocols for professional athletes are lacking. Aim. This narrative review presents an evidence-based, symptom-oriented algorithm for selecting efficient imaging modalities for TMDs in athletes. It combines conventional imaging (MRI, CBCT) with cutting-edge AI and DL screening methods to establish a contemporary diagnostic protocol. Material and methods. A literature search (2018–2026) was conducted across PubMed/MEDLINE, Scopus, WOS, and IEEE Xplore. It included studies on biomechanics, neurophysiology, sports medicine, imaging techniques (MRI, CBCT, US), and DL applications in orofacial diagnostics. Results. Diagnostic protocols should be tailored to athletes' specific clinical signs and sport-related risks. While MRI remains the gold standard for soft tissue, disc derangements, and early inflammation, CBCT offers superior resolution for osseous degeneration and fractures. Furthermore, recent DL models (e.g., TMJ-PanoNet) enable sensitive prescreening using widely available 2D panoramic radiographs. Conclusions. A symptom-oriented, AI-assisted imaging protocol improves diagnostic speed, accuracy, and treatment outcomes. Early detection of TMJ dysfunctions helps control TMD progression, reduce chronic pain, and restore biomechanical balance, potentially preventing premature career termination in professional athletes.
Source: Quality in Sport
Language: English
Subfield: Complementary and Manual Therapy
Field: Health Professions
Domain: Health Sciences
Open Access status: diamond
OpenAlex ID: W7155704246
FWCI:
Cites:
Cited by:
Related to: