ct图像序列中腹主动脉瘤的分割及可视化分析word格式论文.docxVIP

优秀毕业论文 精品参考文献资料 割出来。随后本文利用移动立方体（MC, Marching Cubes）算法将 分割出的 CT 三维建模，并且实现了基本的三维测量功能。通过对两个 AAA 病人的 846 幅 CT 图像进行实验，结果证明分割及重建方法快速、 准确，为实际的临床治疗研究提供参考。 关键词： 腹主动脉瘤，图像分割，主动形状模型，三维重建，区域增 长 SEGMENTATION AND VISUALIZATION OF ABDOMINAL AORTIC ANEURYSM FROM CTA IMAGES Abstract Abdominal aortic aneurysm (AAA) is a serious vascular disease caused by partly abdominal dilation which may have a fatal outcome. At the advantage of computer visualization, a Computer Aided Diagnosis (CAD) system for AAA segmentation is promising as the shape and size of AAA can forecast the possibility of AAA rupture. In addition, the reconstruction is especially valuable for practical treatment in facilitating the operation programming. AAA segmentation is composed of lumen and thrombus segmentation. The previous manual method is inefficient, therefore cannot satisfy the medical need. To reduce analysis time and increase reproducibility, fully or half automated segmentation would be of great value. The segmentation of thrombus is more difficult as the thrombus boundary can be very obscured. A novel approach based on Active Shape Model (ASM) was proposed in this paper. On the basis of statistical model in original ASM, this approach uses grey value correlation with adjacent slices and correlation combined with local intensity and gradient information for evaluating the optimal landmark positions. The experiment shows that the contour for 42 slices can be successfully found, with an effect as satisfactory as the manual segmentation from experts. On the segmentation of the lumen, a novel approach based on region growing algorithm was proposed. This approach adopts an adaptive threshold value technique to raise the robustness. In this way, all slices from one patient can be operated quickly and consecutively. The segmented CTs are then reconstructed using Marching Cube (MC) algorithm, also realizing basic measurement functions. After conducting our method in 846 CT slices from two patients, i