摘要
针对计算机辅助诊断的需求,提出一种基于CT图像的肺实质分割方法。首先,使用大津法实现CT图像肺实质预分割,并利用数学形态学去除噪声;然后,采用区域生长法以及小面积删除法完成肺实质分割;最后,使用改进的凸包法对分割后的肺实质CT图像进行边界修复,并将得到的掩膜图像与原图像进行数学运算,得到肺实质感兴趣区域。实验结果表明:本文提出的方法具有分割精确高、鲁棒性强、自动化程度较高的特点。
To meet the needs of computer-aided diagnosis,this paper proposes a method of lung parenchyma segmentation based on CT images.This method first uses Otsu method to realize pre-segmentation,then uses mathematical morphology to remove noise,then uses region growth method and small area deletion method to complete the segmentation of lung parenchyma,and finally,an improved convex hull method is used to perform boundary repair on the segmented picture.A mathematical operation is performed on the obtained mask image and the original image to obtain a parenchymal region of interest.The experimental results prove that the method used in this paper has the advantages of high accuracy,robustness and high degree of automation.
作者
伍冠楚
吴黎明
林耿萱
钟杨
蒋丹凤
Wu Guanchu;Wu Liming;Lin Gengxuan;Zhong Yang;Jiang Danfeng(School of Mechanical and Electrical Engineering,Guangdong University of Technology,Guangzhou 510006,China)
出处
《自动化与信息工程》
2020年第6期18-22,共5页
Automation & Information Engineering
关键词
区域生长法
凸包法
肺实质
大津法
regional growth method
convex hull method
lung parenchyma
Otsu
