Background: Precise volume assessment of bone defects is crucial for planning reconstructive surgery involving bone grafting. Additionally, detecting the volume of a bone defect is a critical step in assessing its dimensional changes during the follow-up phase. CBCT (Cone Beam Computed Tomography) segmentation is about separating precise structural elements from adjacent anatomy to visualize certain anatomical structure or pathology.
Aim: To evaluate the accuracy of manual CBCT segmentation in the volumetric analysis of simulated bone defects.
Materials and methods: Twenty-one bone defects were created in bovine rib blocks. Osteolytic defects were made by perforating the buccal plate of the bone. The Blocks were scanned with an i-CAT CBCT machine. DICOM (Digital Imaging and Communication in Medicine) data were transferred to Mimics software for segmentation and volumetric analysis. To evaluate the accuracy of CBCT segmentation in volume analysis, the results were compared with the physical measurements (gold standard) of the defects using the Wilcoxon-Signed rank test. The intra-class correlation coefficient and Cronbach's alpha reliability coefficient were used to assess the examiners' agreement.
Results: Wilcoxon-signed rank test comparisons showed no significant difference between the volumetric analysis of bone defects using CBCT segmentation and the gold standard measurements. The CBCT segmentation of the defects using MIMICS software showed good inter-examiner agreement, with an ICC value of (0.87).
Conclusion: CBCT segmentation using MIMICS software offers an accurate, user-friendly, and non-invasive option for assessing the volume of bone defects.