Application of semi-automatic cone beam computed tomography-magnetic resonance imaging image registration technique in anatomical characterisation of temporomandibular joint disc displacement: A methodological feasibility study.

OBJECTIVE This methodological study aimed to evaluate the technical feasibility and accuracy of a semi-automatic cone beam computed tomography (CBCT) and magnetic resonance imaging (MRI) registration technique for the anatomical classification of TMJ disc displacement.

METHOD A total of 61 patients (122 TMJ images) with existing clinical indications for both CBCT (suspected osseous pathology) and MRI examinations were included. Images were processed using a semi-automatic registration system (3D Slicer version 5.2.1, integrated with the Insight Segmentation and Registration Toolkit), combining rigid and elastic transformation algorithms with mutual information metrics. Manual adjustments were required in cases where automatic registration did not meet predefined accuracy thresholds. Diagnostic performance was evaluated by comparing fused CBCT-MRI images with traditional MRI-only assessments.

RESULT The registration accuracy showed high precision, with a Dice similarity coefficient of 0.89 ± 0.04 (range 0.85-0.95) and target registration error of 0.72 ± 0.14 mm (range 0.45-0.79 mm). Manual adjustment was required in 18% of cases to achieve these accuracy levels. The fused images demonstrated superior diagnostic performance, with a sensitivity of 95%, specificity of 93% and area under the receiver operating characteristic curve (AUC) of 0.94, significantly outperforming the MRI-only diagnosis (sensitivity 80%, specificity 89%, AUC 0.82; p < 0.05). Based on the disc-condyle angle and disc-condyle distance, the proposed four-grade classification system showed excellent inter-observer consistency (Cohen's Kappa = 0.85) and an intraclass correlation coefficient of 0.88.  .

CONCLUSION This methodological study demonstrates the technical feasibility of semi-automatic CBCT-MRI registration for the anatomical characterisation of TMJ disc displacement. Although the technique offers improved anatomical visualisation and measurement objectivity compared with single-modality imaging, its clinical implementation requires careful consideration of radiation exposure (particularly in young patients), cost-benefit analysis and alignment with current evidence indicating that clinical symptoms rather than disc position primarily guide TMD treatment decisions. This approach may have utility in research settings, complex diagnostic scenarios or surgical planning, where both osseous and soft tissue visualisation is already clinically indicated; however, it should not be adopted as routine diagnostic practice for patients with TMD.