摘要
目的探讨人工智能辅助阅片的胸部CT在鉴别诊断良恶性肺结节中的临床价值。方法选取2019年1月—2021年6月青州市人民医院收治的肺结节患者50例作为研究对象,所有患者均行胸部CT检查,分别采用人工智能辅助阅片与医师阅片。以病理检查为金标准,比较两种阅片方法的肺结节检出率、诊断效能。结果50例肺结节患者经病理学诊断恶性结节42例,良性结节8例;其中人工智能辅助阅片检出阳性41例,医师阅片检出阳性40例;两组阳性检出率相比,差异无统计学意义(χ^(2)=0.065,P=0.799);人工智能辅助阅片法灵敏度、特异度、准确度及阴性预测值为95.24%、87.50%、94.00%、77.78%,高于医师阅片法的80.95%、25.00%、72.00%、20.00%,差异有统计学意义(χ^(2)=4.086、4.063、8.575、4.237,P=0.043、0.044、0.003、0.040)。结论人工智能辅助阅片在胸部CT诊断肺结节良恶性中应用价值高,可提高恶性肺结节诊断的灵敏度、特异度及准确度,利于减少漏诊、误诊发生,值得临床广泛应用。
Objective To investigate the clinical value of artificial intelligence-assisted chest CT in the differential diagnosis of benign and ma⁃lignant pulmonary nodules.Methods A total of 50 patients with pulmonary nodules admitted to Qingzhou People's Hospital from January 2019 to June 2021 were selected as the research subjects.All patients underwent chest CT examination,and artificial intelligence-assisted reading and physician reading were used respectively.Taking pathological examination as the gold standard,the detection rate and diagnostic efficiency of pulmonary nodules of the two imaging methods were compared.Results of the 50 patients with pulmonary nodules,42 were malig⁃nant nodules and 8 were benign nodules by pathological examination.Among them,41 cases were detected positive by artificial intelligence reading.and found to be positive by artificial intelligence-assisted reading,and 40 were positive by doctor's reading.The positive detection rate of the two groups was Compared,the difference was not statistically significant(χ^(2)=0.065,P=0.799);The sensitivity,specificity,accuracy and negative predictive value of the artificial intelligence-assisted reading method were 95.24%,87.50%,94.00%and 77.78%,which were higher than 80.95%,25.00%,72.00%and 20.00%of the pyhsician reading method.The difference was statistically significant(χ^(2)=4.086,4.063,8.575,4.237,P=0.043,0.044,0.003,0.040).Conclusion Artificial intelligence-assisted image reading has high application value in the diagnosis of benign and malignant pulmonary nodules by chest CT,which can improve the sensitivity,specificity and accuracy of the diagnosis of malignant pulmonary nodules,which is beneficial to reduce the occurrence of missed diagnosis and misdiagnosis,and is worthy of wide clinical application.
作者
杨扬
郭丹丹
徐鹏
王凤
YANG Yang;GUO Dandan;XU Peng;WANG Feng(Department of Imaging,Qingzhou People's Hospital,Qingzhou,Shandong Province,262500 China)
出处
《世界复合医学》
2022年第6期35-37,49,共4页
World Journal of Complex Medicine
关键词
肺结节
人工智能
胸部CT
鉴别诊断
Pulmonary nodules
Artificial intelligence
Chest CT
Differential diagnosis
