肺结节图像的计算机自动检测研究进展  

Research Progress of Computer Automatic Detection of Pulmonary Nodules in Images

作　　者：陈辰[1] 苗军 齐洪钢 刘艳 郭志军[3] 刘海涛[3] 

机构地区：[1]北京信息科技大学,网络文化与数字传播北京市重点实验室,北京 [2]中国科学院大学,计算机科学与技术学院,北京 [3]华北石油管理局总医院,河北 任丘

出　　处：《计算机科学与应用》2021年第1期143-155,共13页Computer Science and Application

摘　　要：肺结节的早期筛查对肺癌的预防和治疗具有重要的意义。采用计算机断层扫描(CT)可以提高肺癌早期筛查中肺结节的检出率,是目前最为安全有效的肺癌筛查方式之一。随着CT数据日益增长,使用计算机辅助诊断(CAD)系统可以极大减少放射科医生的工作量以及降低漏检率。结合近年来肺结节检测的相关文献,对该领域的研究进展进行了综述。首先介绍了目前广泛应用的胸部CT数据集以及肺结节检测相关的评价指标。然后介绍并分析了肺结节检测框架中一些有效的方法,包括肺实质分割、候选结节检测以及假阳性剔除。最后,讨论了该领域存在的挑战以及未来发展的方向,为今后该领域的研究人员提供参考。Early screening of pulmonary nodules is of great significance for the prevention and treatment of lung cancer. Computed tomography (CT) can improve the detection rate of pulmonary nodules in early lung cancer screening, which is one of the most effective lung cancer screening methods. With the increasing of CT data, the use of computer-aided diagnosis (CAD) system can greatly reduce the workload of radiologists and reduce the rate of missed detection. This paper reviews the research progress of pulmonary nodule detection in recent years. Firstly, this paper introduces the widely used chest CT open dataset and the evaluation metrics related to the detection of pulmonary nodule. Then it introduces and analyzes some effective methods in the framework of pulmonary nodule detection, including lung parenchyma segmentation, candidate nodule detection and false positive reduction after detection. Finally, the challenges and future development of this field are discussed, which can provide reference for future researchers in this field.

关 键 词：肺结节 检测 分割 假阳性剔除 

分 类 号：R73[医药卫生—肿瘤]