A Preliminary Study Of Automated Cervical Vertebrae Maturity Assessment System Based On CBCT Data

Objective:With the digitization and intelligence of orthodontic clinical research,the assessment of growth and development is gradually automated.The purpose of this study was to evaluate the reliability of automated cervical vertebra morphology capture based on CBCT data,and to explore the clinical application value of two methods of automated cervical vertebra maturity assessment system(including quantitative and morphological),in order to simplify the assessment of growth and development for adolescent patients.Materials and Methods:1.13 cases were selected as the research objects from the dental radiology department.Based on the CBCT data,the computer automatically intercepted the sagittal plane by three-dimensional least squares method,and then captured the second to fourth cervical vertebrae by Superpixel Segmentation.13 landmark points were selected to identify the cervical vertebrae on the automatically intercepted plane.The Auto group was automatically identified by the computer through a set of algorithms,while the Manu group was identified by an orthodontist for three times in different periods and take the average.Consistency test was performed on the two sets of data to compare the reliability of automated cervical morphology capture.2.34 individuals during growth period in the Orthodontics Department were selected as the research objects.Lateral cephalometric radiograph and CBCT taken at the same period were collected,and the CBCT sagittal sections were obtained by the method of Experiment 1.According to the parameter of quantitative and morphological identification,positioning and staging algorithms were designed to form two sets of automated cervical vertebra maturity assessment systems.2.1 Quantitative cervical vertebra maturity assessment method:The Manu group was marked and calculated by an orthodontist through lateral cephalometric radiograph for 3 times,while the Auto group was through automated quantitative cervical vertebra maturity system by computer.2.2 Morphological cervical vertebra maturity assessment method:The Manu group was assessed by an orthodontic through lateral cephalometric radiograph for 3 times,while the Auto group was through automated morphological cervical vertebra maturity system by computer.The weighted kappa test and the Gamma correlation coefficient were subsequently applied to evaluate the agreement and correlation between the Manu group and Auto group.Results:1.The automatic positioning of the cervical vertebra morphology based on the sagittal plane automatically intercepted from CBCT data has high accuracy at the 8inflection points of total 13 points,and the overall morphology of the cervical vertebra can be recognized(P>0.05).It can be used to assess the cervical vertebra maturity.2.The two set of automated cervical vertebra maturity prediction systems have strong consistency and correlation with their corresponding manual methods.The automated morphological cervical vertebra maturity system(weighted Kappa index,0.886;Gamma value,0.999)indicated a higher consistency and correlation than the automated quantitative cervical vertebra maturity system(weighted Kappa index,0.818;Gamma value,0.984).Conclusions:Two sets of automated cervical vertebrae maturity assessment system(including quantitative and morphological)based on CBCT data presented reliable outcome,high degree of automation,and clinical value.Therefore,this method can be suitable for efficient cervical vertebra maturity assessment clinically.