摘要
目的评估基于深度学习的人工智能(AI)肺结节检测系统对影像实习生肺结节检出的敏感性和阅片时间的影响。方法收集胸部CT平扫500例,比较A组(5名实习生)、B组(AI)、C组(5名实习生结合AI)对肺结节的检出率、假阳性率和平均阅片时间。结果B、C两组的检出率相近,差异无统计学意义(P>0.05),A组的检出率明显低于C组,差异有统计学意义(P<0.05)。在不同位置结节检测方面,三组胸膜下结节检出率差异无统计学意义(P>0.05),A组的外周性及中心性结节检出率明显低于B、C组,差异有统计学意义(P<0.05)。对于不同大小结节的检测结果,三组大结节检出率差异无统计学意义(P>0.05),B、C组中等结节及小结节检出率明显高于A组,差异有统计学意义(P<0.05)。B、C组平均阅片时间明显低于A组,差异有统计学意义(P<0.05)。结论AI能在短时间内有效检出肺结节,使用AI能显著提高实习生对肺结节检出的敏感性并缩短阅片时间。
Objective To evaluate the effect of deep learning-based artificial intelligence(AI)pulmonary nodule detection system on the sensitivity of pulmonary nodule detection and the time of reading.Methods A total of 500 patients with chest CT scans were enrolled.The detection rate,false-positive rate and mean reading time of pulmonary nodules in group A(5 interns),group B(AI)and group C(5 interns combined with AI)were compared.Results There was no significant difference in the detection rate between group B and group A(P>0.05).The detection rate of group C was significantly higher than that of group A,with statistically significant difference(P<0.05).In the detection of nodules at different locations,there was no significant difference in the detection rate of subpleural nodules among the three groups(P>0.05).The detection rates of peripheral and central nodules in group B and C were significantly higher than those in group A,with statistically significant differences(P<0.05).In the detection of nodules of different sizes,there was no significant difference in the detection rate of greater nodules among the three groups(P>0.05).The detection rates of medium and small nodules in group B and C were significantly higher than those in group A,with statistically significant differences(P<0.05).The average reading time of group B and group C was significantly lower than that of group A,with statistically significant differences(P<0.05).Conclusions AI can effectively detect pulmonary nodules in a short time.The use of AI can significantly improve the sensitivity of interns to pulmonary nodules detection and shorten the reading time.
作者
张正华
李俊
韩丹
蔡雅倩
周小君
段慧
赵卫
ZHANG Zhenghua;LI Jun;HAN Dan;CAI Yaqian;ZHOU Xiaojun;DUAN Hui;ZHAO Wei(Department of Medical Imaging,The First Affiliated hospital of Kunming Medical University,Kunming Yunnan 650032,China)
出处
《中国继续医学教育》
2021年第21期88-92,共5页
China Continuing Medical Education
基金
云南省教育厅科学研究基金项目(2019J1229)。
关键词
教育
人工智能
肺结节
计算机体层摄影
实习生
实习带教
医学影像学
education
artificial intelligence
pulmonary nodules
computed tomography
intern
Internship teaching
medical imaging