摘要
目的:探讨人工智能(artificial intelligence,AI)冠状动脉CT血管成像(CCTA)在冠心病诊断中的应用价值。方法:选取本院2018年3-8月,临床疑诊为冠心病的患者50例,对所有患者常规行CCTA扫描,对上传至后处理Sigovia工作站的图像分别由两位高年资医生进行手动冠状动脉分析软件的图像处理、完成报告和由冠心病智能辅助诊断工作站的自动识别图像中的血管和病变并生成报告。评价AI在冠心病中的诊断价值。结果:冠心病诊断中,与常规高年资医生相比,AI对冠心病诊断的阳性预测值为80%,阴性预测值为70%,敏感度为80%,特异度为70%。AI对冠脉病变定位识别的敏感度和特异度分别为73.91%,82.69%。AI对冠脉病变斑块性质识别的敏感度和特异度分别为54.35%,81.73%。在冠脉病变管腔狭窄程度的诊断中,AI和高年资医生对冠脉狭窄的总检出率分别为30.67%、32%,差异无统计学意义(P>0.05),AI对冠脉病变管腔狭窄程度识别的敏感度和特异度分别为60.87%,80.77%。结论:与高年资医生相比,AI辅助CCTA对冠心病的诊断具有一定的价值。
Objective:To evaluate the clinical value of artificial intelligence(AI)-based coronary computed tomography angiography(CCTA)in the diagnosis of coronary artery disease(CAD).Methods:We selected 50 patients with suspected CAD from March to August 2018 in our hospital.After routinely scanning by CCTA,the images were uploaded to the post-processing sigovia workstation and processed by two senior doctors with manual coronary artery analysis software and finished the reports.Then,the AI diagnosis workstation of CAD automatically identified the coronary artery and lesions and generated reports.At last,the diagnostic value of AI in CAD was evaluated.Results:Compared with senior doctors,the positive predictive value,negative predictive value,sensitivity and specificity of AI for diagnosis of coronary heart disease were 80%,70%,80%and 70%respectively.The sensitivity and specificity of AI for detection of coronary lesions were 73.91%and 82.69%respectively.The sensitivity and specificity of AI for the identification of coronary atherosclerotic plaques were 54.35%and 81.73%respectively.The detection rates of AI and senior doctors for coronary stenosis were 30.67%and 32%,respectively.There was no significant difference between them(P>0.05).The sensitivity and specificity of AI for diagnosis the degree of coronary artery stenosis were 60.87%and 80.77%respectively.Conclusion:AI-based CCTA had certain value in the diagnosis of CAD compared with senior doctors.
作者
黄增发
王翔
HUANG Zeng-fa;WANG Xiang(Department of Radiology,the Central Hospital of Wuhan,Tongji Medical College,Huazhong University of Science and Technology,Wuhan 430014,China)
出处
《放射学实践》
北大核心
2018年第10期1017-1021,共5页
Radiologic Practice
基金
湖北省自然科学基金面上项目(2013CFB376)
关键词
人工智能
冠状动脉CT血管成像
冠心病
诊断
Artificial intelligence
Coronary computed tomography angiography
Coronary artery disease
Diagnosis
作者简介
黄增发(1987-),男,湖北武汉人,博士,住院医师,主要从事心血管影像学研究工作;通讯作者:王翔,E-mail:wangxiang385@aliyun.com。
