Application of Digital Image Processing techniques for the analysis of artifacts in CBCT dental images

Cone beam computed tomography (CBCT) dental image interpretation may be hindered by the production of artifacts. An artifact is defined as a distortion in an image that does not correlate to the studied subject. We propose a model for the analysis of the artifact using digital image processing (DIP) techniques. It enables the identification of hyperdense and hypodense regions generated by the presence of materials within the dental root canal. Three standardized CBCT dental images were obtained in three stages using an Orthophos SL 3d CBCT unit. At each step the same axial slices were selected, which were registered and segmented with automatic DIP techniques. Therefore, the regions of interest corresponding to the dental tissue were obtained. To quantify the error, the hyperdense and hypodense areas (artifacts) were determined by comparing the intensity of each pixel of the teeth with different materials regarding the intensities of the image of the tooth without material.