[Three-dimensional quantitative evaluation of condylar bone remodeling of temporomandibular joint based on cone-beam CT imaging]

Objective: To establish a three-dimensional （3D） quantitative measurement and evaluate the condylar bone remodeling of temporomandibular joint (TMJ). Methods: Pre-and post-treatment cone-beam CT (CBCT) data were obtained from 41 patients ［10 males, 31 females, mean age of (19.7±4.4) years (12-30 years old)］, who visited the Center for TMD and Orofacial Pain, Peking University School and Hospital of Stomatology from November 2014 to August 2019, and diagnosed with acute disc displacement without reduction or disc displacement with reduction, with intermittent locking and treated by manual disc reduction followed by anterior repositioning splint. First, condylar bone remodeling was evaluated according to the number of "double contour image" and qualitatively classified as no remodeling (no double contour image), partial remodeling (1-4 double contour images) and remarkable remodeling (5-6 double contour images). Then, condylar bone remodeling was quantitatively evaluated by CBCT based 3D measurement: segmenting condylar images using a semi-automatic method of the manually preliminary mark combined with watershed algorithm, reconstructing the surface models, superimposing the pre-and post-treatment condylar images and finally calculating the volumetric differences of condyle and condylar head, respectively. Results: The Kappa values of two-dimensional (2D) qualitative evaluation were 0.66-0.87, and 3D quantitative measurements of condyle and condylar head volume were also reliable, with the intraclass correlation coefficient (ICC) values of intra-observer 0.998/0.941 and inter-observer 0.999/0.942 respectively. The volumetric increment of the condyle and condylar head after treatment was (41.7±90.2) mm³ and (62.8±70.9) mm³, respectively. Eighty-two condyles were divided into three sub-groups: no remodeling (21), partial remodeling (20) and remarkable remodeling (41). Ranking of the volumetric increment of the condyle and condylar head after treatment was as follows: remarkable remodeling group>partial remodeling group>no remodeling group (P<0.05). Conclusions: The 3D quantitative measurement for evaluating condylar bone remodeling had excellent consistency and reliability, which was consistent with the qualitative classification for condylar bone remodeling. The condylar head as a region of interest was more sensitive to the volumetric changes.