基于分群和几何约束的器官外部组织快速剥离方法-《中国医学物理学杂志》

文章信息/Info

Title:: Fast extraction of tissues outside organs by clustering and geometric constraints

Author(s):: JIANG Xiangang1; ZHANG Hongbin2; 1. School of Science, East China Jiaotong University, Nanchang 330013, China 2. School of Information Engineering, East China Jiaotong University, Nanchang 330013, China

Keywords:: Keywords: tissues peeling geometric constraint clustering signed distance field

摘要:: 人体器官组织外部存在复杂的粘连和交错叠加，保留合理的外部参考组织而剥离掉部分外部组织对微创等手术提供可视化导视具有重要意义。本研究提出一种新的通过属性分群和几何约束对器官外部组织进行快速剥离方法，并采用有符号距离场运算对分割后的器官组织空洞进行有效的填充，分割和渲染的数据处理在GPU和CPU间交互操作，实现了人体器官分割和剥离的仿真显示。实验结果表明本研究方法可为各种器官的外部组织的整体和局部分割和剥离提供可视的实时显示，可为微创手术和介入治疗提供导视。

Abstract:: Abstract: Due to the complex adhesion and overlapping of tissues outside human organs, remaining reasonable reference tissues outside organs while peeling partial tissues is of great significance for providing visual guidance for minimally invasive surgery, etc. A novel method to extract the tissues outside organs by attributes clustering and geometric constraints is proposed in the study, and moreover, hole filling is effectively implemented by signed distance field operations. The data processing of segmentation and rendering is operated interactively between GPU and CPU for completing the simulation rendering of organ segmentation and peeling. The experiments show that the proposed method can realize real-time virtual displays of the segmentation and peeling of the whole and partial tissues outside different organs, which provides guidance for minimally invasive surgery and interventional therapy.