计算机辅助工具对肺结节检测效能的研究  

Effectiveness of computer-aided detection for pulmonary nodules

作　　者：邵亚军[1] 张宝荣[1] 郭佑民[2] 任转琴 SHAO Yajun;ZHANG Baorong;GUO Youmin;REN Zhuanqin(Department of Radiology,Baoji Central Hospital,Baoji 721008,China;Department of Radiology,the First Affiliated Hospital of Xi＇an Jiaotong University,Xi＇an 710061,China)

机构地区：[1]宝鸡市中心医院医学影像科,陕西宝鸡721008 [2]西安交通大学第一附属医院影像科,陕西西安710061

出　　处：《实用放射学杂志》2018年第9期1430-1433,共4页Journal of Practical Radiology

基　　金：国家卫生计划生育委员会公益性行业科研专项资金资助项目（201402013）。

摘　　要：目的 探讨计算机辅助工具对肺结节的检测效能。方法 纳入354个肺结节利用计算机辅助工具对其检测，并对其中110个肺结节进行人工最大直径及最小直径的测量，以肺窗选取最大直径，垂直于最大直径测得最小直径。结果 本组354个肺结节检测中，计算机辅助检测工具漏判结节22个，误判结节123个，结节检出的敏感性93.8%，特异性59%；手动测量最大直径、最小直径及平均直径与计算机测量实性部分平均直径Pearson相关性系数分别是0.772、0.770、0.516。结论 计算机辅助工具在肺结节检出中敏感性高，特异性略低，检测结果需人工甄别；人工测量数据与计算机辅助系统测量子项目中实性部分平均直径显著相关。Objective To explore the efficacy of eompute^aided detection tools in detecting pulmonary nodules.Methods 354 pulmonary nodules were detected by using the computer-aided detection tool.The maximum diameter and minimum diameter of 110 pulmonary nodules were measured manually in the lung window.Results Among the 354 pulmonary nodules, the computer-aided detection tool missed 22 nodules and misdiagnosed 123 nodules with the sensitivity of 93.8% and the specificity of 59 %. The Pearson correlation coefficient of the maximum diameter, minimum diameter, and mean diameter measured in manual with by using the computer-aided detection tool were 0.772, 0.770 and 0.516, respectively.Conclusion Computer-aided detection tool have high sensitivity and low specificity in detection of pulmonary nodule,and the test results need to be screened manually.The manual measurement data is significantly correlated with the data measured by using the computer-aided detection tool for the mean diameter of the nodules.

关 键 词：肺结节 计算机体层成像 计算机辅助检测 

分 类 号：R563[医药卫生—呼吸系统] R814.42[医药卫生—内科学]