Assessment of observer variation in identifying craniometric landmarks and calculating radiographic anatomic indices of the temporomandibular joint: a cross-sectional study

BACKGROUND: One of the key tasks of contemporary dentistry is the analysis of radiographic images of the temporomandibular joint. Errors arising during analysis of cone-beam computed tomography data of the temporomandibular joint may lead to incorrect radiologic interpretation and inadequate treatment planning. AIM: This work aimed to determine the extent to which observer variation influences the radiologic interpretation of temporomandibular joint status based on cone-beam computed tomography data. METHODS: Сone-beam computed tomography scans of the facial bones were obtained from 20 patients (14 women, 6 men; age range, 25–64 years) with temporomandibular joint disorders. To assess the impact of observer variation on the analysis of cone-beam computed tomography images of the temporomandibular joint, unified software and a standardized protocol for identifying craniometric landmarks were used to calculate the anatomic and topographic position of the mandibular condyle. Cone-beam computed tomography analysis was performed using our method of automated craniometry of cranial anatomical structures (Certificate of State Registration of Computer Programs No. 2017662860). The obtained data were used to analyze deviations in calculations. RESULTS: The study revealed substantial systematic and random errors in the assessment of the analyzed craniometric parameters, which are unacceptable in clinical practice. CONCLUSION: To reduce the impact of observer variation on radiologic interpretation in cone-beam computed tomography analysis of the facial bones, training programs should include not only theoretical instructions but also substantial time for workplace internship under the supervision of an experienced mentor.