全自动脑血肿CT影像分割算法研究  

Research on automatic segmentation algorithm of CT image of cerebral hematoma

作　　者：杨峰 Yang Feng(Affiliated Hospital of North China University of Technology Tangshan,Hebei 063000)

机构地区：[1]华北理工大学附属医院,河北省唐山市063000

出　　处：《现代科学仪器》2020年第6期106-110,共5页Modern Scientific Instruments

摘　　要：文章提出全自动脑血肿CT影像分割法的主要目的是为了缩短明确脑血肿患者病灶区域的时间,以加快脑部手术计划制定的进程,避免传统半自动或手动分割法处理脑血肿CT医学影像对临床时间的影响。基于FM分割法(Fast Marching,快速行进)的完善设计了一种三维智能脑血肿CT影像分割算法,以阙值处理的方式实现了种子点集的自动选取,以兴趣区域的选择建立了三维分割算法,在ROI金字塔迭代分割的基础上深入精简FM分割法的计算步骤。总之,通过一系列的改进措施,脑血肿CT影像中病灶区域的具体位置能够在短时间内得以确定,脑血肿手术计划也因此可以快速制定。In order to reduce the influence of semi-automatic or manual segmentation on clinical time in the process of brain surgery planning,this paper proposes a 3D intelligent CT image segmentation algorithm based on the improvement of FM(fast marching).The method of threshold processing is used to create seed point set intelligently instead of manual selection of seed points;algorithm steps are simplified by selecting region of interest;FM algorithm is further simplified on the basis of pyramid iterative segmentation of ROI to obtain segmentation results in the shortest time.The verification results show that the proposed algorithm can intelligently,accurately and quickly complete the segmentation of CT images of cerebral hematoma.

关 键 词：分割算法 CT影像 FM法 迭代分割 

分 类 号：TP3[自动化与计算机技术—计算机科学与技术]