摘要
肺部CT图像病变区域检测是肺病辅助诊断技术的重要研究内容,其通过自动分析CT图像并输出病变区域的位置和尺寸等信息,帮助放射科医生做出决策,有利于肺病的早期发现与治疗.本文回顾了肺部CT图像中病变区域自动检测方法所取得的进步,并引入一个通用框架表示和描述现有方法,对2012年以来肺部病变区域辅助检测算法进行了系统性分析和性能汇总.最后讨论了目前存在的问题和有待克服的困难,探讨了未来可能的发展方向.
Automatic detection of lesion regions in lung CT images is an important research topic in computer aided diagnosis of lung diseases. The system can automatically analyze CT images, output the locations and sizes of lesion regions to help radiologists make decisions, and promote early detection and therapy of lung diseases. In this paper we review the achieved progress of automatic detection methods of lesion regions in lung CT image, and introduce a generic structure for expressing and describing existing detection methods. Furthermore, we provide a systematic analysis and comprehensive performance summary of the latest detection algorithms from 2012. Finally, we point out the challenges ahead, and discuss the future direction of computer aided detection of lung lesions.
作者
韩光辉
刘峡壁
郑光远
HAN Guang-Hui;LIU Xia-Bi;ZHENG Guang-Yuan(Beijing Laboratory of Intelligent Information, School of Computer Science, Beijing Institute of Technology, Beijing 100081)
出处
《自动化学报》
EI
CSCD
北大核心
2017年第12期2071-2090,共20页
Acta Automatica Sinica
基金
国家自然科学基金(60973059
81171407)
教育部新世纪优秀人才支持计划(NCET-10-0044)资助~~
关键词
肺部CT
肺结节
肺血管
淋巴结
计算机辅助检测
Lung CT, lung nodule, lung vessel, lymph node, computer aided detection
