基于肺部CT图像中肺实质分割的研究进展  

Advances in pulmonary parenchyma segmentation based on computed tomography image

作　　者：张文莉 吕晓琪[1,2] 谷宇[1,2] 吴凉 张信雪 ZHANG Wenli LU Xiaoqi GU Yu WU Liang ZHANG Xinxue(Inner Mongolia Key Laboratory of Pattern Recognition and Intelligent Image Processing/School of Information Engineering, Inner Mongolia University of Science and Technology, Baotou 014010, China School of Computer Engineering and Science, Shanghai University, Shanghai 200444, China)

机构地区：[1]内蒙古科技大学信息工程学院/内蒙古自治区模式识别与智能图像处理重点实验室,内蒙古包头014010 [2]上海大学计算机工程与科学学院,上海200444

出　　处：《中国医学物理学杂志》2017年第9期902-907,共6页Chinese Journal of Medical Physics

基　　金：国家自然科学基金(61179019);内蒙古自治区自然科学基金(2015MS0604);内蒙古自治区高等学校科学研究项目(NJZY145);包头市科技计划项目(2015C2006-14)

摘　　要：肺部CT图像中肺实质的精确分割是肺部疾病诊断和治疗的一个重要步骤,也是制约计算机辅助检测技术广泛应用于肺部疾病诊断领域的主要瓶颈之一。基于近年来肺部CT图像中肺实质分割的研究进展,对其分割过程中的5个步骤进行综述,包括预处理、初分割、精分割、左右肺分离、边缘修补。最后展望了肺部CT图像中肺实质分割的发展方向。The precise segmentation of pulmonary parenchyma in chest-computed tomography image, an important step for the diagnosis and treatment of pulmonary diseases, restricts the wider application of computer-aided detection in the diagnosis of lung diseases. In recent years, pulmonary parenchyma segmentation has been greatly improved. Herein, we summarize the advances in the 5 processes of lung segmentation, including preprocessing, coarse segmentation, accurate segmentation, separation of left and right lung, border correction, and finally provide a prospect for the development of pulmonary parenchyma segmentation.

关 键 词：肺部 CT图像 肺实质 分割 综述 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]