CT图像肺结节计算机辅助检测与分类系统设计(英文)  

Design of Computer-aided Detection and Classification of Lung Nodules Using CT Images

作　　者：聂生东[1] 郑斌[2] 李雯[1] 

机构地区：[1]上海理工大学医疗器械学院,上海200093 [2]匹兹堡大学放射系

出　　处：《系统仿真学报》2007年第5期935-944,共10页Journal of System Simulation

基　　金：上海市教委重点项目(06ZZ33);上海市重点学科建设项目(P0502)

摘　　要：近年来,肺癌已经成为癌症中的主要杀手,早期发现肺癌是降低死亡率的最有效的途径。在传统的二维胸片和三维CT片上,早期肺癌一般表现为肺结节。因此,肺癌的早期诊断归结为对肺结节的检测上面。而通过人工阅片的方式检测肺结节是非常费时费力的。计算机辅助检测技术(CAD)具有检测肺结节的潜力,它把计算机图像处理的结果作为放射科医生的辅助参考,它能够检测出放射科医生易于漏检的微、小结节。因此,利用CAD增加了早期诊断肺癌的机会。基于近年来国内外在肺结节计算机辅助诊断和检测方面的研究进展,提出了一个研发CAD的基本思想和框架。Lung cancer is promising techniques to pulmonary nodules on the leading cause of cancer deaths. Early detection is the most improve the change of survival. Lung cancer shows up as conventional two-dimensional chest radiographs and three-dimensional CT scans, so, in order to find lung cancer at an early stage, radiologist should find lung nodule as earlier as possible. Computer-aided detection （CAD） schemes have the potential to increase the chance of early diagnosis, which regards the output result of CAD as a ＇second opinion＇ to assist radiologist in interpreting images. Based on the review of computer-aided detection and diagnosis of lung nodules using CT at home and abroad in recent years, a basic design idea of developing CAD was proposed.

关 键 词：肺结节 计算机辅助检测 分类 早期诊断 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]