北京生物医学工程

基于CT三维图像的肺结节良恶性鉴别研究

Identification of benign and malignant pulmonary nodule based on 3D CT image

作者：常莎王瑞平

单位：                      北京交通大学计算机与信息技术学院 (北京100044)        

关键词：肺部；CT图像；三维重建；良恶性结节；支持向量机

分类号：

出版年·卷·期（页码）：2013·32·1（12-16）

摘要：

目的运用计算机方法处理肺部CT图像以识别肺结节良恶性并辅助肺癌诊断，现已成为国内外研究的热点。方法提出一种基于肺部CT图像三维肺结节信息的肺结节良恶性鉴别方法。首先结合阈值分割、区域生长、形态学运算等在CT图像上分割出肺结节，进而提取每个肺结节的三维特征并优化，选择有效特征。然后，基于有效特征采用支持向量机(support vector machine, SVM)的分类算法对多维向量所描述的肺结节进行良恶性的二分类。最后从敏感性、特异性、准确率以及似然比等方面全面评估分类结果。结果实验获得敏感性为0.7776，准确性为0.7378，阳性似然比2.2410，阴性似然比0.3682，显示基于CT三维肺结节图像可以达到令人满意的肺部肿瘤良、恶性鉴别效果。结论上述结果证明了基于CT三维图像的肺结节良恶性鉴别方法的可行性。本研究对计算机辅助肺癌的诊断具有重要意义。

Objective Identification of benign and malignant pulmonary nodules by using the computer-aided detection and diagnosis system is one hot spot. Methods An identification method for benign and malignant pulmonary nodule based on the 3D information of lung CT image was proposed. First we segmented the pulmonary nodules from CT images with the methods of segmentation, regional growth, morphology, then extracted and optimized the 3D information of each pulmonary nodule. Finally we utilized the SVM classification algorithm to divide those pulmonary nodules into two categories based on the effective features. Results The classification results were assessed by sensitivity, specificity, accuracy, and likelihood ratio. From the experiment we got the results as follows, the sensitivity was 0.7776, the accuracy was 0.7378, the positive likelihood ratio was 2.2410 and the negative likelihood ratio was 0.3682. Conclusions All the results showed that the new method achieved satisfactory identification effect and was significant for the computer-aided diagnosis of lung cancer.

参考文献：

［1］崔云，马大庆.肺结节的CT计算机辅助检测和诊断的基本方法和应用［J］.中国医学影像技术，2007,23（3）：469.
Cui Yun, Ma Daqing. The basic methods and application of computer-aided detection and diagnosis of lung nodule in CT［J］.Chin J Med Imaging Technol,2007,23(3):469.
［2］李越.基于形态学及灰度熵的肺结节检测算法［D］.北京：北京交通大学,2011：24-27.
Li Yue. A detection algorithm based on morphology and gray entropy for pulmonary nodules ［D］. Beijing: Beijing Jiaotong University,2011:24-27.
［3］杨聃.三维肺部CT图像中的结节自动识别［D］.武汉：华中科技大学,2007.
Yang Dan. A thesis submitted in partial fulfillment of the requirements for the degree of master of science［D］.Wuhan: Huazhong University of Science and Technology, 2007.
［4］冯伟兴，唐墨，贺波.数字图像模式识别技术详解［M］.北京：机械工业出版社，2010:53-64.
Feng Weixing, Tang Mo, He Bo. Digital image pattern recognition programming using Visual C++［M］. Beijing: China Machine Press, 2010:53-64.
［5］Omar S. Al-Kadi, D. Watson. Texture Analysis of Aggressive and Nonaggressive Lung Tumor CE CT Images［J］. IEEE transactions on biomedical engineering, 2008,55(7):1826-1829
［6］Onur Osman, Serhat Ozekes, Osman N.Ucan.Lung nodule diagnosis using 3D template matching［J］. Computers in Biology and Medicine,2007,37:1167-1172
［7］Messay T,Hardie RC,Rogers SK. A new computationally efficient CAD system for pulmonary nodule detection in CT imagery［J］.Medical Image Analysis,2010,(14):390-406
［8］姜慧妍，何炜.基于胸部CT图像的肺癌识别方法的研究［J］.电子学报，2009，37（8）:1664-1668.
Jiang Huiyan, He Wei. Research of lung cancer recognition based on chest CT images［J］, Acta Electronica Sinica, 2009，37（8）:1664-1668.

服务与反馈：

【文章下载】【加入收藏】

提示：您还未登录，请登录！点此登录