Volume 13 ,Issue 2 ,August 2026 ,Pages 1-12
Ayat A. Arkan 1 ; Dena N. Mohammad 2
1 Department of diagnosis, oral and maxillofacial radiology College of Dentistry, University of Sulaimani, Sulaymaniyah , Iraq.
2 Oral Diagnosis Department, College of Dentistry, University of Sulaimani, Sulaymaniyah 46001, Iraq.
Objective: The study aimed to compare the diagnostic accuracy of different imaging modalities for detecting intraosseous jaw lesions, using histopathological diagnosis as the reference standard, and to assess interobserver agreement in response to initial radiographic reports.
Methods: A total of 108 patients diagnosed with intraosseous jaw lesions were included. Radiographic images (Cone Beam Computed Tomography, Computed Tomography, and Orthopantomography) and reports with the related histopathological diagnosis were collected. Two radiologists, blinded to the histopathological results, reviewed the images using a (semiquantitative) scoring system. Kappa coefficients were calculated to evaluate interobserver agreement.
Results: The majority of lesions were radiolucent (79.6%), with cystic lesions (50.0%) being the most common, particularly radicular cysts (24.1%). Well-defined borders were predominant (62.0%). Most lesions were related to the inferior alveolar canal (IAC) (20.4%), followed by the maxillary sinus (11.1%). Both observers showed substantial agreement in response to the extension of the lesions and relation to the tooth (0.701 and 0.660, respectively). CBCT showed the highest diagnostic accuracy between the two observers in detecting score 3 of the differential diagnosis (81.8%), which represents a correct first-ranked differential diagnosis corresponding to the histopathological findings.
Conclusions: CBCT has the best accuracy in the diagnosis of intraosseous jaw lesions, particularly for lesions' density, anatomical extension, and relation to the tooth. Still, certain radiographic interpretation parameters showed non-agreement between observers. Evidence-based diagnosis and treatment planning require standardized reporting and proper interobserver training.