基于人工智能与影像医生对肺结节检出效能的应用价值

Application Value Based on Artificial Intelligence and Imaging Doctor to Pulmonary Nodule Detection Efficiency

目的探讨人工智能与影像医生对肺结节检出的应用价值。方法回顾性分析接受胸部CT检查病例100例。根据阅片对象不同分成3组,人工阅片和2款人工智能(AI)医学影像辅助诊断软件进行阅片诊断,计算比较3组肺结节的检出率、漏诊率和假阳性率等。结果100例患者筛查检出肺结节共652枚,2款AI组检出率和不同位置结节检出率均高于人工组(P均<0.05),而2款AI组漏诊率均低于人工组;2款AI组中等结节及小结节检出率高于人工组(P均<0.05),但对大结节检出率各组之间差异无统计学意义(χ^(2)=2.891,P=0.086)。结论AI对于肺结节的检出效能高于影像医生,可辅助影像医师进行肺结节的筛查及诊断工作。

Objective To explore the application value of artificial intelligence and imaging doctor in pulmonary nodule detection.Methods A retrospective analysis was performed on 100 cases of chest CT.The patients were divided into 3 groups according to the different subjects.Manual film reading and 2 artificial intelligence(AI)medical image assisted diagnosis software were used for film reading diagnosis.The detection rate,missed diagnosis rate and false positive rate of pulmonary nodules in the 3 groups were calculated and compared.Results A total of 652 pulmonary nodules were detected by screening in 100 patients.The detection rate and the detection rate of nodules at different positions in the 2 AI groups were higher than those in the artificial group(P all<0.05),while the missed diagnosis rate in the 2 AI groups was all in the artificial group.The detection rates of medium nodules and nodules in the two AI groups were higher than those in the artificial group(P all<0.05),but there was no significant difference in the detection rates of large nodules among all groups(χ^(2)=2.891,P=0.086).Conclusion AI is more effective in detecting lung nodules than imaging doctors,and can assist imaging doctors in screening and diagnosis of lung nodules.