Work

A Hybrid Segmentation Method for Three-Dimensional Analysis of the Mandibular Condyle in Idiopathic Condylar Resorption

Year: 2026

Type: article

Abstract: OBJECTIVE: Accurate extraction of condylar morphology underpins the diagnosis of idiopathic condylar resorption (ICR), yet cone beam computed tomography (CBCT) segmentation is difficult in ICR because subchondral demineralization blurs the bony margin. We tested whether single-threshold segmentation accurately captures the condyle in ICR and developed a hybrid method that combines Otsu thresholding with a three-dimensional region-growing algorithm. METHODS: We retrospectively analysed CBCT scans of 23 patients with ICR (46 condyles). Manual, threshold, and hybrid segmentation were compared for accuracy (condylar volume and surface area), segmentation time, and clinical usability. Before this comparison, we validated manual segmentation, the reference standard, against optical surface scanning of 16 dry condyles. Volume, segmentation times, and surface area were compared using paired t tests, one-way ANOVA, and Friedman test; usability was assessed using the Wilcoxon signed-rank test. RESULTS: ; P < .05), whereas hybrid and manual segmentation did not differ. Regions missed by threshold segmentation showed intermediate density (mean greyscale value 418.5) and clustered in the anteromiddle condyle (41.3%), the reported initiation site of resorption. Hybrid segmentation was faster than manual and threshold segmentation (11.0 vs 20.9 and 16.5 minutes; P < .01) and scored highest for usability. CONCLUSION: Single-threshold segmentation misses early low-density subchondral changes in ICR. A hybrid Otsu-region-growing method recovers these regions with manual-level accuracy in less time, supporting CBCT as a practical tool for quantitative condylar assessment in ICR.

Source: International Dental Journal

DOI: 10.1016/j.identj.2026.111140

Authors: Xiaolei Li, Yilin Yu, Songfeng Li, Yue Xu, Han Lin

Institutions: Sun Yat-sen University, Hospital of Stomatology, Sun Yat-sen University, Temple University

Language: English

Topic: Dental Radiography and Imaging

Subfield: Oral Surgery

Field: Dentistry

Domain: Health Sciences

SDG: Good health and well-being

Open Access status: gold

Estimated APC paid: $2,550