数据驱动的心室传导系统解剖三维重构方法研究

项目名称： 数据驱动的心室传导系统解剖三维重构方法研究

项目编号： No.61503327

项目类型： 青年科学基金项目

立项/批准年度： 2016

项目学科： 自动化技术、计算机技术

项目作者： 黎捷

作者单位： 燕山大学

项目金额： 22万元

中文摘要： 心脏是人体最重要的器官之一，作为心脏病多发部位的心室，其传导系统的三维结构，包括位置和宽度信息，决定了沿着该系统传递的电兴奋信号的传播方向，对心脏的生理病理活动具有举足轻重的影响。迄今为止，心室及传导系统的生理和病理研究已经广泛开展起来，但受成像和扫描技术的局限，大多数学者对传导系统三维几何结构的研究往往止步于近似表示，对基于解剖结构的心室传导系统的三维几何的研究却甚少涉及。本项目针对现有成果局限，以解剖数据为驱动，研究心室传导系统的三维数值重构方法，并得到具有较高精度的心室传导系统的三维几何模型。其研究内容包括：心室肌内膜层的曲面重建方法；心室解剖图的消除高光方法；心室传导系统的线结构提取方法；基于LLE的心室传导系统三维重构方法等。为心脏生理病理的高精度仿真、构建自动化虚拟心脏仿真平台、心脏病的诊断治疗及心脏药物制剂等研究提供重要的数据支撑和理论基础。

中文关键词： 生物网络仿真；生物网络可视化；生物系统动态可视化

英文摘要： Heart is one of the most significant organs in human body. As one of the tissues which heart diseases occur mostly, three dimensional construction of the ventricle conduction systems, the location and width features included, determines the spreading orientation of the electrical excitation digital along the conduction systems, and plays an important role on the physiopathological behavior of the ventricle. So far, the research on physiopathology of the ventricle and the conduction systems have been widely conducted, however, limited to the imaging and scanning technology, most scholars tend to substitute the real geometry construction for the approximated description of the conduction systems. Thus the research on the digital reconstruction based on the anatomic data is rarely involved. To overcome the existing limitation, the project aims at do the research on three dimensional digital reconstruction methods for ventricular conduction.systems resulted with the completed three dimensional geometric model with high accuracy of the ventricular conduction systems based on the anatomic data driven. The work includes: reconstruction of the endo-cardium curved surfaces from points cloud of the myocyte model; highlight-remove process on the anatomic image of the ventricle, before the line structure of the conduction systems being extracted of location and width information; then the extracted line structure is embedded onto the curved surfaces reconstructed, moreover, data-driven three dimensional digital reconstruction of the ventricle conduction systems based on deep learning is finished. It strongly provides data and the theory highly underlies fields such as physiopathologic simulation on heart with high accuracy, reconstruction of simulation platform for virtual heart, heart-disease diagonosis and pharmaceutic preparation et al.

英文关键词： Biological Network Simulation;Biological Network Visualization;Dynamic Visualizagion of Biological Systems