基于Snake改进型模型的胸部CT图像分割方法  

A Segmentation Method of Thoracic CT Images Based on SnakeImproved Model

作　　者：梁计锋[1] LIANG Jifeng(College of Engineering&Technology,Xi’an Fanyi University,Xi'an 710105,China)

机构地区：[1]西安翻译学院,西安710105

出　　处：《自动化与仪器仪表》2022年第6期23-26,共4页Automation & Instrumentation

基　　金：西安翻译学院电子商务一流专业项目。

摘　　要：针对胸部CT图像分割中的一些难点,结合贪心算法的特点,引入梯度向量流(GVF)来代替传统Snake中的图像力。从添加外部约束、方程求解和边缘图生成等方面对GVF模型进行了改进。该GVF模型的扩散方程计算的边缘映射图则利用了Canny算子的边缘检测结果,从而对GVF力场的弱边界区域和小灰度变化区域进行了明显的改善。对力场扩散方程中梯度矢量流的分量分别进行归一化处理,减小了物体边界力场对曲线上各点的影响,克服了GVF模型难以解决的深凹问题,有效地控制活动等值线外部人工约束能量。改进的GVF-Snake模型能准确、反复地对胸部CT图像进行分割优化,具有良好的鲁棒性和实用性。该方法提高了胸部CT图像分割的可重复性、准确性和通用性。其应用可为医生准确观察CT图像提供良好的帮助,更对该方面病理研究提供较好依据。In view of some difficulties in chest CT image segmentation and the characteristics of greedy algorithm,gradient vector flow(GVF)is introduced to replace the traditional Snake image force.The GVF model is improved by adding external constraints,solving equations and generating edge graphs.The edge mapping diagram calculated by the diffusion equation of the GVF model uses the edge detection results of the Canny operator,so that the weak boundary region and small gray scale change region of the GVF force field are obviously improved.The components of gradient vector flow in the force field diffusion equation are normalized respectively,which reduces the influence of the object boundary force field on each point on the curve,overcomes the deep concave problem which is difficult to be solved by the GVF model,and effectively controls the artificial constrained energy outside the active isoline.The improved GVF-Snake model can accurately and repeatedly optimize the segmentation of chest CT images,and has good robustness and practicability.This method improves the repeatability,accuracy and versatility of chest CT image segmentation.Its application can provide good help for doctors to observe CT images accurately,and also provide a good basis for pathological research in this field.

关 键 词：GVF-SNAKE CT 图像分割 

分 类 号：TP273[自动化与计算机技术—检测技术与自动化装置]