Related Experiment Video
Updated: May 22, 2026
A Postoperative Evaluation Guideline for Computer-Assisted Reconstruction of the Mandible
Published on: January 28, 2020
Evaluation of mandibular bone in patients with type 2 diabetes mellitus using cone-beam computed tomography-based
Elifhan Alagoz1, Huseyin Alani2
1Department of Oral and Maxillofacial Radiology, Faculty of Dentistry, Bezmialem Vakif University, Fatih, Istanbul, Türkiye.
Objective:
This study aimed to evaluate the structural characteristics of the posterior mandibular alveolar bone in individuals with type 2 diabetes mellitus (T2DM) using cone-beam computed tomography (CBCT) based radiomic analysis.
Study Design:
A total of 98 CBCT scans (49 T2DM, 49 healthy controls) were retrospectively analyzed. A 7 × 7 × 7-mm region of interest (ROI) distal to the mental foramen was segmented using 3D Slicer (version 5.7.0), and radiomic features were extracted via the SlicerRadiomics extension. Interobserver reliability was confirmed by an intraclass correlation coefficient (ICC) of 0.96. Statistical analysis was performed with SPSS version, 28.0, and a logistic regression model was built to evaluate the discriminative power of features (P < .05).
Results:
Significant radiomic differences were identified between diabetic and healthy groups, including surface area, entropy, maximum, uniformity, contrast (GLCM, NGTDM), and small area emphasis (P < .05). Male subjects exhibited greater mesh volume, sphericity, energy, kurtosis, and uniformity, whereas female subjects showed elevated entropy, maximum, coarseness, contrast, and small area emphasis. Age correlated positively with sphericity (r = 0.217, P = .041) but had no independent impact on the other features. Predictive modeling achieved 61% test accuracy and an area under the curve of 0.66, with shape, first-order, and texture features being the most influential predictors.
Conclusions:
CBCT based radiomics revealed distinct structural alterations in the mandibular bone of patients with T2DM. Although predictive performance was moderate, radiomic features demonstrated potential as sensitive imaging biomarkers for early detection and risk stratification of diabetes-related bone changes.
