Alternate Level Set Evolutions With Controlled Switch for Tooth Segmentation

Abstract

CBCT images have been widely used in digital orthodontics. In CBCT images, there are some tooth boundaries of blurriness and disappearance, and the teeth are of intensity inhomogeneity. In order to identify these boundaries and segment teeth, this paper develops an approach of alternate level set evolutions with controlled switch via slice-by-slice tooth segmentation: the user selects a voxel inside a tooth to initialize a small circle encircling the voxel, and the curve is automatically expanded until the tooth surface is formed. In each slice, the curve is expanded by two level set evolutions: the first level set evolution handles the intensity inhomogeneity and the second level set evolution solves the boundary blurriness and disappearance of the teeth. Experiments showed that the proposed method successfully segmented 12 CBCT datasets of 356 teeth and outperformed the geodesic active contour method and the popular method and the state-of-the-art method.

Automatic dental arch detection and panoramic imaging from cone beam computed tomography dataset

“…For tooth segmentation, the Dice similarity coefficient (DSC) metric is used to measure the similarity ratio between the tooth segmentation result and the ground truth. The DSC metric is defined in [31] by…”

Semi or fully automatic tooth segmentation in CBCT images: a review

“… Gao & Chae (2008) applied the SLSM for tooth root segmentation, addressing challenges posed by complex image conditions and the issue of root branching. However, when dealing with tooth crown segmentation, especially for adjacent tooth crowns ( Jiang et al, 2022 ), the performance of SLSM may be limited by certain factors.…”

“…Furthermore, Jiang et al (2022) alternately integrated the edge-based hybrid level set method with the geodesic active contour (GAC) method, controlling the evolution of these two models via a switch. Compared to several of the previously mentioned methods ( Gan et al, 2015 , 2018 ; Harrison et al, 2019 ; Jiang et al, 2019 ), this approach yields the smallest ASSD values (0.16 ± 0.08 mm).…”

Setting the regularization coefficient based on image energy in image segmentation using kernel graph cut algorithm

“…An example of this approach can be seen in active contour algorithms 12 – 14 and level sets 15 – 17 In active contour methods, the initial boundary must be determined by the user. In this method, the snake parametric model is used as a boundary energy minimizer.…”

“…To solve the above problems, level sets methods have been presented. [15][16][17] Level sets methods are divided into two categories: edge-based and region-based. 18 In edgebased methods, the energy function based on edge information is used.…”