基于CT数据的胃癌模型的三维重建和手术仿真

The incidence of gastric cancer is high in China. Surgeons have been beset always by the surgical difficulty, operation hemorrhage, long operation time, surgical complications and poor prognosis of stomach cancer. From surgery, anatomy, imaging and computer science disciplines, this project will research the key technology of three-dimensional (3D) reconstruction for abdominal organs and vascular accurately and fast based on CT data. A 3D model system will be put forward, which will includes celiac trunk and its branches, portal vein and its branches, spleen and its lobe veins, pancreas, stomach, the tumor and the lymph nodes with the pancreas in the centre. The accuracy of 3D reconstruction will be tested by statistics. Based on the 3D model system, the virtual reality technique will be applied, the force feedback system will be used, the 3D model will be printed, and then virtual operation and surgical planning will be carried out for the radical gastric surgery. The aim of this study is to reduce the operation difficulty, shorten the operation time, reduce postoperative complications, avoid vascular ligation by mistake and avoid hemorrhage caused by damage. It is to help lymph node dissect and improve the prognosis of patients with tumor in the stomach by understanding the distribution of metastatic lymph node. It is to choose the right operation way and improve the surgical safety by understanding the tumor position and the relationship between the tumor and the surrounding vessels and organs.

胃癌在我国发病率高，手术难度大、出血多、手术时间长、并发症多和病人预后差等问题一直困扰着外科医生。本课题将从外科学、解剖学、影像学和计算机学的多学科角度，研究基于CT数据的腹部器官和血管准确、快速的抽出和三维重建的关键技术，以建立胰腺为中心的腹腔干及其分支、门静脉及其分支、脾脏及其脾叶动静脉、胰腺、胃及其肿瘤和淋巴结的胃癌三维模型。通过统计学原理，检验其三维重建的准确性。基于三维模型，利用虚拟现实技术、力反馈系统，以及3D打印技术，进行胃癌的手术仿真和手术规划研究。以降低胃癌手术难度，缩短手术时间，减少手术并发症，避免血管误扎和误伤大出血；了解转移淋巴结的分布情况，提高胃癌淋巴结的清扫率，改善患者的预后；了解肿瘤在胃壁的位置及其与周围血管和器官的关系，以决定手术方式和提高手术的安全性。

胃癌在我国发病率高，手术难度大、出血多、手术时间长、并发症多和病人预后差等问题一直困扰着外科医生。本课题从外科学、解剖学、影像学和计算机学的多学科角度，研究了基于CT数据的腹部器官和血管准确、快速的抽出和三维重建的关键技术，建立了胰腺为中心的腹腔干及其分支、门静脉及其分支、脾脏及其脾叶动静脉、胰腺的胃癌三维模型。本课题在腹部器官和血管准确、快速的抽出和三维重建、外科虚拟手术的关键技术方面取得一些重要结果，我们提出了一种用于医学图像处理的高度集成的通用硬件加速器，以有效提高二维标记算法的计算性能，并降低FPGA设备的功耗；提出了边缘纹理和支持向量机分类器的全自动地标检测方法，用于CT扫描数据人体器官图像的抽出；提出了一种改进的基于Zynq的Sobel算子的医学图像边缘检测方法；提出了一种新的无监督的高维数据特征选择方法；提出了一种基于三维卷积神经网络的血管全自动分割算法；设计了一种基于Web平台的医学图像3D模型构建系统；设计了一个基于Web平台的CT扫描数据腹部器官分割平台；设计了一套基于虚拟现实头盔和力反馈装置并结合3D打印模型的恶性肿瘤外科手术系统。建立胃癌三维模型，进行胃癌根治手术的术前诊断和手术评估、手术规划和术中指导，具有重要的临床意义。