3D tooth segmentation in cone-beam computed tomography images using distance transform. (January 2023)

• Record Type: Journal Article
• Authors: Kakehbaraei, Somayeh; Arvanaghi, Roghayyeh; Seyedarabi, Hadi; Esmaeili, Farzad; Zenouz, Ali Taghavi
• Is Part Of: Biomedical signal processing and control. Volume 79(2023)Part 1
• Publication Date: 2023-01
• Subjects: Teeth Segmentation -- Distance Transform -- Cone-Beam Computed Tomography; Signal processing -- Periodicals; Biomedical engineering -- Periodicals; Signal Processing, Computer-Assisted -- Periodicals; Image Processing, Computer-Assisted -- Periodicals; Biomedical Engineering -- Periodicals; 610.28
• DOI: 10.1016/j.bspc.2022.104122
• Languages: English
• ISSNs: 1746-8094

Abstract

Highlights: A process for extracting individual teeth from CBCT images is introduced. A parameter adaptive and morphology filtering is applied to a boundary refining and improving the algorithm. Constraints of Distance Transform (DT) are assigned based on prior knowledge of human teeth. Contacted teeth are separated by the watershed algorithm based on Morphology and DT. 3D model of teeth is obtained with triangular patches method. Abstract: Background and Objectives: Segmentation of individual teeth from Cone-Beam Computed Tomography (CBCT) images is a significant step in computer-aided systems for orthodontic planning. Teeth segmentation is still problematic because the teeth have similar intensity with the bone tissue and the variation of jawbone intensity. This study focused on improving teeth segmentation in CBCT images. Methods: In this paper, we propose a new approach for tooth segmentation in CBCT images and Three-Dimensional (3D) visualization from the segmented teeth. First, artifacts reduction is obtained by histogram adjusting and morphology filtering of teeth bone on axial slices. Second, by using Distance Transform (DT), the tooth tissue is segmented. Finally, we represent a 3D teeth view using segmented pieces based on the triangular patches method of scattered points belonging to tooth lines. Results: For all the used images, teeth boundaries are segmented with accuracy and efficiency. The results of our method are similar to the outcomes of segmentation acquired by the specialist. Conclusion: Experimental results and simulations show that the proposed method is helpful for precise and effective teeth segmentation in CBCT images.