Evaluation of supervised machine learning models in predicting temporomandibular joint disc displacement on 3T magnetic resonance imaging

Key Points

• Machine learning models effectively identify MRI-based morphometric patterns related to TMJ disc displacement, enhancing diagnostics.
• The study highlights significant patterns detected by ML, indicating improved prediction reliability for TMJ issues on MRI scans.
• Assessment involved evaluating diverse machine learning models against traditional radiologic standards for TMJ disc assessment.
• Although promising, the findings suggest that clinical diagnosis still relies on established standards, indicating ML's supportive role rather than a replacement.

Abstract