基于医学CT图像序列的主动脉夹层三维重建方法的研究

发布时间：2018-03-21 12:02

本文选题：主动脉夹层　切入点：三维重建　出处：《天津工业大学》2017年硕士论文　论文类型：学位论文

【摘要】：主动脉夹层是目前最为凶险的一类心血管疾病,其诊疗手段以腔内隔离术为主;对于主治医师需在术前掌握主动脉血管内病灶局部毗邻的立体空间关系信息,如夹层破口位置、范围、数量及分支血流受挤压情况等相关数据。而现有医学图像处理软件重建得到的主动脉夹层三维模型无法对主动脉腔体内部夹层膜的立体形态进行直观地显示,导致主治医师较难掌握主动脉夹层破口位置、大小等关键信息。因此,本文设计了一种基于医学CT图像序列的主动脉夹层三维重建方法,可以快速准确地实现对主动脉内部夹层的三维可视化显示,供医生进行临床诊断,对提高手术的成功率和挽救病人生命具有重大意义。该方法主要由图像预处理、目标分割提取、三维重建等环节组成,主要工作如下:一、根据人体胸腹腔CT图像上的噪声特点,首先利用中值滤波法进行平滑处理;然后通过拉普拉斯算子对CT图像进行锐化处理以突出主动脉及夹层的边缘特征;最后通过灰度线性映射的方法提高图像整体亮度、改善对比度,为后续主动脉及夹层的分割提取奠定基础。二、将主动形状模型算法应用于主动脉分割,实现了对人体胸腹腔主动脉的快速分割提取。随后在对夹层的提取过程中,本文提出了基于贝叶斯理论的空间连续先验模型的主动脉夹层提取算法,在Hessian矩阵提取轮廓特征像素点的基础上实现对夹层膜像素的精确识别并去除非目标像素点。三、采用基于光线投射算法的体绘制方法实现对主动脉夹层的三维重建,并通过透明化设置及颜色标记等操作,实现对主动脉内部夹层三维形态、破口位置及大小等特征的直观表示。本文提出的主动脉夹层三维重建方法在主动脉和夹层的分割过程中均充分地利用了 CT图像序列的空间连续性,通过三维空间信息辅助二维空间的分割以提高精度。本文采用Dice系数作为分割结果精度的衡量指标,经实验证实本文提出的主动脉和夹层的分割算法的精度满足实际指标要求,重建得到的主动脉夹层三维模型显示效果理想,为手术或介入治疗封闭夹层破口提供辅助,对提高手术成功率、挽救更多患者的生命具有重要的意义。
[Abstract]:Aortic dissection is the most dangerous type of cardiovascular disease at present, and its diagnosis and treatment are mainly by endovascular isolation. The range, quantity and compression of branch blood flow are related data. However, the three-dimensional model of aortic dissection reconstructed by medical image processing software can not visually display the three-dimensional shape of the internal dissecting membrane of the aortic cavity, and the three-dimensional shape of the aortic dissecting membrane can not be visualized by the existing medical image processing software. Therefore, a 3D reconstruction method of aortic dissection based on medical CT image sequence is designed, which is difficult for the attending physician to grasp the key information of aortic dissection, such as the location and size of aortic dissection. The 3D visualization of internal aortic dissection can be realized quickly and accurately, which can be used for clinical diagnosis by doctors. It is of great significance to improve the success rate of operation and save the lives of patients. The main work is as follows: firstly, according to the noise characteristics of human chest and abdominal CT images, the median filtering method is used to smooth the image. Then the CT image is sharpened by Laplacian operator to highlight the edge features of aorta and dissection. Finally, the image brightness and contrast are improved by the method of gray linear mapping. It lays the foundation for subsequent aortic and dissection segmentation. Secondly, the active shape model algorithm is applied to the aortic segmentation, and the rapid segmentation of human thoracic and peritoneal aorta is realized. Then, in the process of dissection extraction, In this paper, a spatial continuous prior model based on Bayesian theory is proposed for aortic dissection extraction. Based on the extraction of contour feature pixels by Hessian matrix, accurate recognition of dissection membrane pixels is realized and unless the target pixels are removed. The 3D reconstruction of aortic dissection is realized by volume rendering based on ray-casting algorithm, and the three-dimensional shape of aortic dissection is realized by transparent setting and color marking. The three-dimensional reconstruction method of aortic dissection in this paper makes full use of the spatial continuity of CT image sequence in the process of aortic and dissection segmentation. In this paper, the Dice coefficient is used as the index to measure the accuracy of the segmentation results. The experimental results show that the accuracy of the segmentation algorithm for aortic and dissection in this paper meets the actual requirements. The three-dimensional model of aortic dissection obtained by reconstruction is of great significance to improve the success rate of operation and save more patients' lives. It can provide assistance for the operation or interventional therapy to seal the dissecting break.
【学位授予单位】：天津工业大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：R543.1;TP391.41

【相似文献】