A novel classification system, termed the Sivan classification, was developed to enhance the diagnosis of jaw lesions by utilizing visual volumetric analysis of three-dimensional Cone Beam Computed Tomography (CBCT) images. This classification groups lesions into ten categories, primarily divided into hypovolumetric, hypervolumetric, and normovolumetric groups. To validate this system, 10 raters-comprising 5 general dentists and 5 oral radiology specialists-assessed the CBCT images and diagnosed the lesions using the Sivan classification. Eight raters repeated the process after one month to assess consistency. The overall agreement between raters, quantified using kappa statistics, was 0.82, indicating excellent consistency. Hypervolumetric and normovolumetric lesions demonstrated the highest agreement (kappa 0.84 and 0.82, respectively), while hypovolumetric lesions showed substantial agreement (kappa 0.77). Pairwise interrater agreement ranged from 76 to 93%, with kappa values between 0.75 and 0.87. Intrarater reliability was equally strong, with kappa values between 0.79 and 0.89.These results suggest that the Sivan classification provides a robust and reliable framework for diagnosing jaw lesions using CBCT volumetric analysis, surpassing traditional diagnostic methods in accuracy and consistency.
Keywords: Classification; Cone-beam computed tomography; Jaw neoplasms.
© 2024. The Author(s).