Accuracy of CNN-Based method for segmentation of the mandible

Abstract

This study aimed to evaluate the accuracy of a new automatic deep learning-based approach for fully mandible automatic segmentation from CBCTs. Forty CBCT scans from healthy patients (20 females, 20 males, mean age 23.37 ± 3.34 years) were collected and a manual mandible segmentation was carried out by using Mimics software. Twenty CBCT scans were randomly selected and used for training the artificial intelligence model file. The remaining 20 CBCT segmentation masks were used to test the accuracy of the CNN fully automatic method by comparing the segmentation volumes of the 3D models obtained with automatic and manual segmentations. The accuracy of this method was also assessed by using the DICE Score coefficient (DSC) and by the surface-to-surface matching technique. The Intraclass correlation coefficient (ICC) and Dahlberg's formula were used to test the intra-observer reliability and method error, respectively. Independent Student's t-test was used for between-groups volumetric comparison. Measurements were highly correlated with an ICC value of 0.937 while the method error was 0.24 mm 3 . An insignificant statistical difference of 0.71 (± 0,49) cm 3 was found between the methodologies (p>0.05). The matching percentage detected was 90.35% ± 1.88 (tolerance, 0.5 mm) and 96.32% ± 1.97% (tolerance 1.0 mm). All volumetric measurements were highly correlated with an ICC value of 0.937 for intra-observer evaluation and 0.919 for inter-observer evaluation. The tested technology provided an accurate fully automated segmentation of the mandible from CBCT scans performing at a much higher speed than an experienced image reader.