Auto-Threshold Bone Segmentation Based on CT Image and Its Application on CTA Bone-Subtraction

摘要

CTA technology is characterized by the higher clinically practical value in the inspection of the vascular diseases compared with other similar technologies. The bone-subtraction is the key method to improve the quality of CTA subtraction image and promotion of CTA technology. In this paper, a bone-subtraction method of the 3D CTA was proposed. The method includes a bone segmentation algorithm with automatic threshold and automatic seed point. Meanwhile, by combining with other algorithms, for example, maximization mutual information registration algorithm, the bone and other influenced factors were removed from the subtraction image, and the vascular part was completely retained as well. In the experimental assessment, the accuracy of the automatic bone segmentation algorithm was evaluated by the 3D CT images in different positions; meanwhile, the effect of the bone-subtraction method was evaluated by the 25 sets of 3D CTA images, and was also compared with the ordinary registration subtraction method. The experiments showed that the auto-threshold bone segmentation algorithm in this method could correctly segment the bone. And this bone-subtraction method was better than ordinary registration subtraction method on the effect of total CTA images. Therefore, it is significant for the promotion of the bone-subtraction CTA technology.