基于CT图像的肺实质分割方法
2014-07-13李满
摘要:诊断肺癌的重要手段之一是高分辨率CT(High Resolution ComputedTomography,HRCT)扫描,但是医生需要丰富的阅片经验以及大量的精力时间才能阅读海量的CT图像信息。为了减少医生的精力损耗和降低漏诊率,采用计算机辅助检测成为趋势。在肺癌等肺部疾病计算机辅助诊断方法中,最核心的步骤是肺实质的分割。提出一种基于CT图像序列的新的自动肺实质分割方法,综合利用了区域生长及数学形态学开运算等算法,并通过纵向扫描方法精确定位左肺和右肺的粘连部位,从而在肺实质边界的肺结节结节容易被忽略分割及左右肺分离的难题得到了解决。对多组胸部CT序列图像的实验证明,该方法对于肺实质分割非常精确有效。
关键词:CT图像;计算机辅助诊断;区域生长;肺实质;扫描定位
中图分类号:TP391 文献标识码:A 文章编号:1009-3044(2014)05-1093-03
A Method of Lung Segmentation Based on CT Images
LI Man
(Electronic and Information Engineering, Tianjin Polytechnic University,Tianjin 300387, China)
Abstract:High-resolution CT (High Resolution ComputedTomography, HRCT) scan is an important means of diagnosis of lung cancer, however, a CT image information reading mass requires a lot of time, effort and considerable film-reading experience. To reduce energy loss of docter and cut down misdiagnosis rate,the use of computer-aided detection become a trend. Lung cancer and other lung diseases in computer-aided diagnostic methods, the most central step is the lung parenchyma segmentation. CT image sequence is proposed based on a new method for automatic segmentation of lung parenchyma, utilization of thresholding, region growing and mathematical morphology other algorithms, and through vertical scanning algorithm to pinpoint the cable around a narrow area around the lungs, an effective solution to the edge of the lung parenchyma nodule segmentation omission and so easy to separate the problem of the lung .Multiple sets of chest CT images proved the method for lung segmentation is very accurate and efficient.
Key words:CT image; computer-aided diagnosis; regional growth; lung parenchyma; scanning positioning
1 概述
隨着肺部疾病在人群中越来越流行,针对它的有关学习研究工作必须得到高度的注意。现在辅助电脑诊断肺部疾病(computer aided diagnosis CAD)系统[1]的成长是飞快的,能够很大程度上辅助医生诊断病人病情。大多数肺部疾病CA D系统都是利用计算机断层扫描图像来检测诊断肺结节,所以确定左右肺的边缘显得尤其重要,这个步骤称为“肺实质分割”[2-3]。
之前研究人员在研究肺部疾病时,已经注意到了分割肺实质的重要性。在分割肺实质之前,要对图像进行预处理,然后可以采用边界跟踪[4]、区域生长、阈值法、多尺度分析、形态学处理、基于模式分类的分割法等获得肺实质。……
