摘要
提出一种基于区域生长算法的肝脏CT影像分割方法,以减少在肝脏医学影像分割中由于对感兴趣区域分割不完整或过度分割导致的临床诊断失误问题。采用区域生长算法的邻接连续阈值法,实现对肝脏CT图像的分割,选取合适的种子点并在各个方向按照提前规定好的生长准则对各个像素点进行筛选分类,分割出感兴趣的区域,使用近百张不同的肝脏CT图像对算法进行训练和测试,完善在此技术之上开发出的系统功能模块。经过测试后,其分割成功率较高,几乎可以准确快速地分割出肝脏CT图像中感兴趣的区域,为临床医学中肝脏处病灶诊断提供可靠依据,避免诊断失误,具有实际应用意义。
A method of liver CT image segmentation based on regional growth algorithm is proposed to reduce the problem of clinical diagnostic errors due to incomplete or excessive segmentation of the region of interest in liver medical image segmentation.The adjacent contiguous continuous threshold method of the regional growth algorithm is used to realize the segmentation of liver CT images,the appropriate seed point is selected,and each pixel is screened and classified in each direction according to the growth criteria specified in advance.The region of interest is finally segmented.Nearly 100 different liver CT images are used to train and test the algorithm,and the system functional modules developed on top of this technology are improved.After testing,the segmentation success rate is relatively high,which can almost achieve the accurate and rapid segmentation of the region of interest in the liver CT image.Accurate segmentation of the region of interest in liver CT images provides a reliable basis for the diagnosis of liver lesions in clinical medicine,and avoids diagnosis errors,and has practical application significance.
作者
胡紫睿
刘倩
Hu Zirui;Liu Qian(School of Information Engineering,Ningxia University,Yinchuan 750021,China)
出处
《黑龙江科学》
2024年第6期88-92,共5页
Heilongjiang Science
基金
宁夏回族自治区大学生创新创业训练计划基金资助项目(202310749343)
宁夏回族自治区重点研发计划项目“数字生态体系下公共法律服务系统的研究与应用”(2022BEG03073)。
关键词
CT图像
区域生长算法
图像分割
边缘检测
CT image
Area growth algorithm
Image segmentation
Edge detection
