摘要
目的:探讨人工智能(AI)联合低剂量CT (LDCT)扫描在肺结节诊断中的应用价值。方法:回顾性分析2020年1月—2023年2月就诊于昌乐县人民医院的96例肺结节患者(结节124个)的临床资料,患者均行LDCT检查与AI诊断。比较AI联合LDCT诊断与单纯LDCT检测的平均阅片时间和图像质量评分;以胸腔镜病理结果为金标准,分析AI联合LDCT诊断与单纯LDCT诊断肺癌的价值及病理结果的一致性。结果AI联合LDCT检测的平均阅片时间短于单纯LDCT检测,图像质量评分高于单纯LDCT检测,差异有统计学意义(P<0.05)。124个结节经病理诊断显示恶性结节52个,良性结节72个。AI联合LDCT诊断早期肺癌的特异度、准确率和阳性预测值均高于LDCT,差异有统计学意义(P<0.05);两种检查方式的灵敏度和阴性预测值差异不显著(P> 0.05)。Kappa检验显示,LDCT诊断与病理结果的一致性尚可(Kappa值=0.443);AI联合LDCT诊断与病理结果的一致性极好(Kappa值=0.933)。结论:AI联合LDCT扫描能够提高肺癌检出率且诊断准确率与病理结果一致性较高,可作为临床筛查肺结节的常用手段。
Objective To investigate the application value of artificial intelligence(AI)combined with low-dose CT(LDCT)scanning in the diagnosis of pulmonary nodules.Methods Clinical data of 96 patients with pulmonary nodules(124 nodules)admitted to Changle County People's Hospital from January 2020 to February 2023 were retrospectively analyzed.All patients underwent LDCT examination and AI diagnosis.The average film reading time and image quality scores of AI combined with LDCT diagnosis and LDCT detection alone were compared,and the value of AI combined with LDCT diagnosis and LDCT diagnosis alone in lung cancer diagnosis and the consistency of pathological results were analyzed using thoracoscopic pathological results as the gold standard.Results The average reading time of AI combined with LDCT detection was shorter than that of LDCT detection alone,and the image quality score was higher than that of LDCT detection alone,with statistical significance(P<0.05).Among the 124 nodules,52 were malignant and 72 were benign.The specificity,accuracy and positive predictive value of AI combined with LDCT in the diagnosis of early lung cancer were higher than LDCT,and the difference was statistically significant(P<0.05).There was no significant difference in sensitivity and negative predictive value between the two methods(P>0.05).Kappa test showed that the consistency between LDCT diagnosis and pathological results was reasonable(Kappa value=0.443).The diagnosis of AI combined with LDCT was in good agreement with pathological results(Kappa=0.933).Conclusion AI combined with LDCT scan can improve the detection rate of lung cancer,and the diagnostic accuracy is consistent with the pathological results,which can be used as a common means of clinical screening lung nodules.
作者
梁付奎
傅晓琴
李彩琴
刘芸
刘山波
LIANG Fukui;FU Xiaoqin;LI Caiqin;LIU Yun;LIU Shanbo(Department of Imaging,Changle County People's Hospital,Weifang,Shandong 262400,China)
出处
《影像研究与医学应用》
2023年第15期28-30,共3页
Journal of Imaging Research and Medical Applications
关键词
肺结节
人工智能
低剂量CT
肺癌
Pulmonary nodules
Artificial intelligence AI
Low-dose CT
Lung cancer
