Open AccessA Metal Artifact Reduction Method with Bone Segmentation for CBCT Images
Author(s) -
Songze Zhang,
Benxiang Jiang,
Hongcan Shi
Publication year - 2021
Publication title -
journal of physics. conference series
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.21
H-Index - 85
eISSN - 1742-6596
pISSN - 1742-6588
DOI - 10.1088/1742-6596/2024/1/012033
Subject(s) - inpainting , artificial intelligence , computer vision , artifact (error) , computer science , projection (relational algebra) , image quality , segmentation , image (mathematics) , algorithm
Nowadays, CBCT images become popular and are widely used in the dental diagnosis. The image quality is related to the accuracy of diagnosis and treatment. When metal objects such as the dental implants are present in CBCT scanning, the metal artifacts will appear and decrease the image quality. In order to reduce the metal artifacts, this paper proposes a metal artifact reduction method for the dental CBCT images. This method is based on inpainting the sinogram and mainly has three innovations. First, this method forward-projects only the region around the metal implants. Second, the bones in the forward projection region are segmented and their intensity values are set to zero before the forward projection. Finally, this method computes an image to directly correct the original CBCT image, this correction image is the back projection of the difference between the interpolated sinogram and the original sinogram. The proposed method reduces the metal artifacts and improves the image quality, it may potentially increase the diagnosis value of the CBCT images corrupted by the metal artifacts.