人工智能辅助分析技术在子宫颈细胞癌前病变筛查中的应用价值  

Value about artificial intelligence assisted analysis technology for screening cervical precancerous lesions

作　　者：胡爱侠[1] 朱琳 贺慧[1] 胡金星 王榭延 李道宏 靳钰炜 洪帆 孔令非[1] 杨志明 温洲 HU Ai-xia;ZHU Lin;HE Hui;HU Jin-xing;WANG Xie-yan;LI Dao-hong;JIN Yu-wei;HONG Fan;KONG Ling-fei;YANG Zhi-ming;WEN Zhou(Department of Pathology,Henan Provincial People’s Hospital,Zhengzhou 450003,China;iDeepWise Artificial Intelligence Robot Technology Co.,Ltd,Beijing 100085,China)

机构地区：[1]河南省人民医院病理科,郑州450003 [2]深思考人工智能机器人科技(北京)有限公司,北京100085

出　　处：《临床与实验病理学杂志》2022年第1期27-30,共4页Chinese Journal of Clinical and Experimental Pathology

基　　金：河南省重大科技专项项目(161100311400)。

摘　　要：目的探讨人工智能辅助分析技术在子宫颈细胞癌前病变筛查中的应用价值。方法使用人工智能细胞学辅助阅片系统对86000例涂片进行分析,并与原诊断结果进行对比分析,统计其特异性、敏感性等指标,同时分析人工智能细胞学辅助阅片系统的分级诊断结果。结果人工智能细胞学辅助阅片系统对子宫颈液基涂片的特异性可达74.16%,敏感性高达98.77%,同时对意义不明的非典型鳞状细胞(atypical squamous cells of unknown significance,ASCUS)/非典型鳞状细胞,不除外高度鳞状上皮内病变(atypical squamous cells,excluding highly squamous intraepithelial lesions,ASC-H)/低度鳞状上皮内病变(low-grade squamous intraepithelial lesions,LSIL)细胞具有更高的敏感性。结论人工智能辅助分析技术实现了子宫颈细胞学智能化阅片,细胞学辅助阅片系统的应用有效地提高了病理医师的工作效率和诊断准确率,降低了漏诊风险。Purpose To explore the value of artificial intelligence assisted analysis for cervical precancerous lesions screening.Methods 86000 cervical smear samples were screened through the artificial intelligence auxiliary analysis system,and the results were analyzed with the previous archive results.The accuracy,specificity,sensitivity and negative rate were calculated.At the same time,the grading diagnosis results of AI assisted technology were analyzed.Results The specificity of artificial intelligence assisted screening system on thinprep cytologic test smear could reached 74.16%,and the sensitivity was as high as 98.77%.At the same time,the artificial intelligence auxiliary analysis system had higher sensitivity and specificity in atypical squamous cells of unknown significance/atypical squamous cells,excluding highly squamous intraepithelial lesions/low-grade squamous intraepithelial lesions.Conclusion Artificial intelligence-assisted analysis technology is used to realize the intelligent screening of cervical cytology.The application of the auxiliary screening system effectively improves the doctor’s work efficiency and diagnosis accuracy,and reduces the risk of missed diagnosis.

关 键 词：子宫颈液基细胞学 人工智能辅助分析技术 子宫颈细胞癌前病变筛查 

分 类 号：R737.33[医药卫生—肿瘤]