Quantification of condylar resorption in temporomandibular joint osteoarthritis

Abstract

Objective: This study was performed to determine the condylar morphologic variation of osteoarthritic (OA) and asymptomatic temporomandibular joints (TMJs) and to determine its correlation with pain intensity and duration.

Study design: Three-dimensional surface models of mandibular condyles were constructed from cone-beam computerized tomography images of 29 female patients with TMJ OA (Research Diagnostic Criteria for Temporomandibular Disorders group III) and 36 female asymptomatic subjects. Shape correspondence was used to localize and quantify the condylar morphology. Statistical analysis was performed with multivariate analysis of covariance analysis, using Hotelling T(2) metric based on covariance matrices, and Pearson correlation.

Results: The OA condylar morphology was statistically significantly different from the asymptomatic condyles (P < .05). Three-dimensional morphologic variation of the OA condyles was significantly correlated with pain intensity and duration.

Conclusion: Three-dimensional quantification of condylar morphology revealed profound differences between OA and asymptomatic condyles, and the extent of the resorptive changes paralleled pain severity and duration.

Figure 1

Diagram of methodology used for shape analysis that consisted of 4 steps: 1.segmentation of the CBCT volumes,2. Shape Correspondence, 3. Individual local morphologic variation, and 4. Statistical shape analysis.

Figure 2

Description of shape correspondence procedures. The segmented 3D surface models of the condyles are converted into surface meshes, and a spherical parametrization is computed for the surface meshes using an area-preserving, distortion minimizing spherical mapping. The SPHARM description is computed from the mesh and its spherical parametrization. Using the first order ellipsoid from the spherical harmonic coefficients, the spherical parametrizations are aligned to establish correspondence across all surfaces. The SPHARM description is then sampled into a triangulated surfaces (SPHARM-PDM). The condylar surfaces are represented using a subdivision level 10 (1002 points in the surface of the condyle).These SPHARM-PDM surfaces are all spatially aligned using rigid Procrustes alignment.

Figure 3

Validation of condylar 3D models constructed from CBCT: The bottom row shows the overlay of composite models constructed by 2 observers for 10 TMJ OA subjects.

Figure 4

The 3D morphological distribution of condylar shapes associated with a possible continuum of osteoarthritic changes. The vertical axis illustrates the progression (flattening, erosions and osteophytes) of degenerative change while the horizontal axis illustrates levels of severity.

Figure 5

Group comparisons using statistical shape analysis: semi-transparent overlay of composite model of asymptomatic control and the TMJ OA group, color maps, difference vectors and p-value maps of the Hotelling t-tests maps showing that statistically significant differences were noted between the groups, for the left condyle at the anterior surface of the lateral pole and posterior surface of the medial pole.

Figure 6

Local morphological correlation with pain variables. A, Local correlations with pain intensity. B, Local correlations with pain onset. C and D, statistical significance maps for A an B. The significance maps show statistically significant correlations in the superior surface of the condyle in A and the lateral and posterior surfaces of the condyle in B.