Detection of cemento-osseous lesions in cone beam computed tomography images using artificial intelligence algorithms

Aim: The aim of this study is to perform segmentation of cemento-osseous lesions using artificial intelligence algorithms on KIBT images and to evaluate the diagnostic performance of a diagnostic artificial intelligence model designed for the diagnosis of cemento-osseous lesions. Material and Method: For this study, CBCT images taken for various reasons in the radiology archive of Inonu University Faculty of Dentistry, Department of Oral and Maxillofacial Radiology were retrospectively reviewed. As a result of the radiographic evaluation, the images recorded in the archive with the diagnosis of cemento-osseous lesion were determined. 50 DICOM images of 50 patients were uploaded to 3D slicer software and cemento-osseous lesions were polygonally labeled and saved in NIfTI format. The nnU-Netv2-based automated algorithm for lesion segmentation was developed in the CranioCatch(CranioCatch, Eskişehir) software program using the PyTorch library in the Python framework (v3.6.1; Python Software Foundation, Wilmington, DE, USA). 80% of the data was used for training, 10% for validation and 10% for testing. The results were evaluated according to the criteria of precision, sensitivity, Dice Coefficient, Jaccard Index. Results:The precision, sensitivity, Dice Coefficient and Jaccard Index values for segmentation of cemento-osseous lesions were 0.805, 0.889, 0.839 and 0.730, respectively. Conclusion: The model we used achieved successful results in the cemento-osseous lesion segment. The results of this planned study are promising in terms of providing a guiding option for physicians in diagnosis. Keywords: Deep learning, cone beam computed tomography, cemento-osseous lesion, artificial intelligence