摘要
目的探讨基于深度学习(deep learning,DL)的冠状动脉CT血管成像(coronary computed to mography angiography,CCTA)的冠状动脉疾病报告与数据系统(coronary artery disease reporting and data sys tem,CAD RADSTM)冠心病诊断中的临床应用价值。方法选取我院2018年9-12月,临床疑诊为冠心病的患者226例,常规行CCTA扫描,按照CAD RADSTM报告系统,分别由两位高年资医师和基于深度学习的自动软件对患者进行分级报告。结果基于深度学习对CCTA的CAD RADS分级和医师对CCTA的CAD RADS分级评价差异无统计学意义(P=0.84),一致性检验结果显示两者的Kappa值为0.687(P<0.05),一致性较好。结论基于深度学习的CCTA图像的CAD RADS标准化报告能较准确评价疑似冠心病患者,与诊断医师具有较好的一致性。
Objective To evaluatethe clinical value of deep learning based Coronary Artery Disease Reporting and Data System,(CAD RADSTM)in coronary computed tomography angiography(CCTA).Methods We selected 226 patients with suspected coronary artery disease(CAD)from September to December 2018 in our hospital.CCTA scans were performed routinely.According to CAD RADSTM,two senior doctors and DL based auto mated softwarewere used to classify the reports.Results There was no significant difference between the DL based CAD RADS grading and doctors basedCAD RADS grading in CCTA(P=0.84).The consistency test showed that the Kappa value of the two groups was 0.687(P<0.05),which indicated that DL based CAD RADS grading is in substantial agreement with doctors based CAD RADS grading in CCTA.Conclusion The standardized report of DL based CAD RADS in CCTA images can be an reference foraccurately evaluating suspected patients with CAD,and it alsohas ahigh level ofconsistency with the diagnostic physicians.
作者
肖建伟
张树桐
谢元亮
胡芸
陈燕浩
陈发祥
王翔
XIAO Jianwei;ZHANG Shutong;XIE Yuanliang;HU Yun;CHEN Yanhao;CHEN Faxiang;WANG Xiang(Department of Radiology,The Central Hospital of Wuhan,Tongji Medical College,Huazhong Univer sity of Science and Technology,Wuhan 430030,China)
出处
《实用医学杂志》
CAS
北大核心
2019年第20期3227-3230,共4页
The Journal of Practical Medicine
基金
湖北省自然科学基金面上项目(编号：2013CFB376)
关键词
深度学习
冠状动脉CT血管成像
冠心病
冠状动脉疾病报告与数据系统
deep learning
coronary computed tomography angiography
coronary artery disease
coronary artery disease reporting and data system
