整合临床及多参数MRI信息的前列腺癌CAD系统:诊断效能研究  

Computer-aided diagnosis system for the detection of prostate cancer based on clinical information and multiparametric MRI:analysis of the diagnostic efficacy

作　　者：高歌[1] 胡娟[2] 王成彦[3] 王鹤[1] 杨学东[4] 张珏[3,5] 王霄英[1] 

机构地区：[1]北京大学第一医院医学影像科,北京100034 [2]昆明医科大学第一附属医院医学影像科,昆明650032 [3]北京大学前沿交叉学科研究院,北京100871 [4]中国中医科学院广安门医院放射科,北京100053 [5]北京大学工学院,北京100871

出　　处：《放射学实践》2016年第12期1143-1145,共3页Radiologic Practice

摘　　要：目的:将临床信息和多参数MRI(mpMRI)数据输入前列腺计算机辅助诊断(CAD)系统,研究其诊断效能。方法:选取本院前列腺MRI数据库中连续266例患者的病例资料,所有患者均以定性诊断为目的行前列腺MRI检查,临床资料和病理结果完整,且完成随访。由1位高年资泌尿系统影像诊断医师根据前列腺影像报告和数据系统(PI-RADS)进行阅片,以PI-RADS评分为mpMRI信息,输入CAD系统。选取患者资料中的年龄、T-PSA、F/T-PSA、直肠指诊、超声表现等作为临床信息,输入CAD系统。采用受试者工作特征曲线(ROC)分析,以最终诊断结果为金标准,比较mpMRI数据、临床信息、mpMRI+临床信息作为输入项时,CAD输出结果的诊断效能。结果:以临床信息为输入项,CAD诊断前列腺癌的ROC下面积(AUC)为0.888;以mpMRI信息作为输入项,CAD诊断前列腺的AUC为0.980;将mpMRI和临床信息整合作为输入项时,诊断效能最高,AUC为0.986,相应的诊断敏感度、特异度和符合率分别为93.65、96.15和94.12。结论:整合临床信息和mpMRI信息输入前列腺CAD系统,可得到最高的诊断效能。Objective：To evaluate the diagnostic efficacy of computer-aided diagnosis （CAD） system in detection of prostate cancer （PCa）. Methods： 266 consecutive patients were recruited from institutional prostate MRI database. All the patients were suspected of PCa,who underwent mpMRI （T2 WI,DWI,MRS and DCE） examination with subsequent ultrasound guided biopsy within 3 months. PI-RADS scores given by uroradiologists were input as mpMRI information for CAD. Clinical data including age,T-PSA,F/T-PSA,digital rectal examination （DRE） and ultrasound findings were input as clinical information for CAD,the output of CAD was the probability of PCa. Taking the clinical outcome as reference standard, the clinical information,mpMRI information or integration of clinical and mpMRI information were input into the CAD respectively,and the corresponding diagnostic efficacy of CAD was studied using ROC analysis. Results： Based on clinical information,the area under the ROC （AUC） of the CAD system for diagnosis of PCa was 0. 888;based on mpMRI,the AUC was 0. 980;and based on clinical information combined with mpMRI, the diagnostic efficacy was the highest with AUC of 0. 986 ,and the corresponding diagnostic sensitivity, specificity and accuracy were 93.65,96.15 and 94.12. Conclusion： The input of CAD system in detection of PCa should consist of both the clinical information and the information of mpMRI.

关 键 词：诊断 计算机辅助 磁共振成像 前列腺肿瘤 前列腺特异性抗原 

分 类 号：R445.2[医药卫生—影像医学与核医学] R737.25[医药卫生—诊断学]