人工智能辅助阅片与单纯人工...系统形态学诊断中的比对研究_杨华.pdf

169现代检验医学杂志第 38 卷第 1 期2023 年 1 月J Mod Lab Med,Vol.38,No.1,Jan.2023人工智能辅助阅片与单纯人工阅片在女性阴道微生态系统 形态学诊断中的比对研究杨华a，孙天舒b，王瑶c，徐英春c，孙宏莉c（北京协和医院 a.妇产科；b.医学科学研究中心；c.检验科，北京 100730）摘要：目的评估不同级别检验员对革兰染色涂片阴道微生态形态学评价的基线水平，分析检验员在电子阅片和镜下阅片的差异，探究使用人工智能分析系统独立进行微生态评价以及辅助检验员进行微生态评价的能力表现，评价人工智能分析系统在临床中的应用价值。方法该研究样本来源于北京协和医院中国女性人群下生殖道微生态菌群基线研究项目，收集 2021 年 5 月 2021 年 7 月女性阴道分泌物涂片共 385 例，经革兰染色和图像采集后，分别进行检验员等级考核以及人工显微镜镜下阅片、人工电子阅片、人工智能（artificial intelligence,AI）独立阅片和 AI 辅助检验员阅片。在确定镜下阅片金标准和电子阅片金标准之后，分析两种不同阅片方式在 AV 评分和 Nugent 评分的差异，比较不同级别检验员、AI，以及经 AI 辅助后，在 AV 评分和 Nugent 评分上的能力表现。结果镜下阅片和电子阅片在需氧菌性阴道炎（aerobic vaginitis，AV）和细菌性阴道病（bacterial vaginosis,BV）(含 BV 中间型)诊断的 Kappa 一致性分析分别为 0.91 和 0.93（P 0.01）。AI 独立阅片在 AV 和 BV(含 BV 中间型)诊断的准确度分别为 0.85 和 0.92，灵敏度分别为 0.86 和 0.88，Kappa 值分别为 0.62 和 0.79。初级检验员在电子阅片下的 AV 和 BV(含 BV 中间型)诊断的准确度分别为 0.850.02 和 0.890.01，灵敏度分别为 0.640.06 和 0.840.07，Kappa 值分别为 0.550.07 和 0.720.04。高级检验员在电子阅片下的 AV 和 BV(含 BV 中间型)诊断的准确度分别为 0.920.03 和 0.910.03，灵敏度分别为0.870.02 和 0.920.04，Kappa 值分别为 0.780.07 和 0.790.06。经 AI 辅助诊断后，初级检验员 AV 和 BV(含 BV中间型)诊断的 Kappa 值提升至 0.770.04 和 0.780.02，高级检验员 AV 和 BV(含 BV 中间型)诊断的 Kappa 值提升至 0.820.05 和 0.850.01。结论镜下阅片和电子阅片的一致性非常高，电子阅片或可替代镜下阅片成为一种新的阅片方式。AI 独立阅片诊断 AV 和 BV(含 BV 中间型)的能力优于普通检验员，比高级检验员略差。不同级别检验员经AI 辅助诊断后，AV 和 BV(含 BV 中间型)的诊断能力均有提升，其中初级检验员提升明显，能力接近高级检验员的水平，且各检验员之间的偏差缩小明显。整体结果表明，使用人工智能 Descartes-Image 妇科微生态辅助分析软件不仅能提升检验员诊断能力，还能减小检验员之间的偏差，使诊断结果不容易因为人为因素而出现较大波动，保证了结果的稳定性和可靠性。关键词：阴道微生态；形态学检测；人工智能；性能评估；人工智能辅助阅片中图分类号：R446.19文献标识码：A文章编号:1671-7414（2023）01-169-07doi:10.3969/j.issn.1671-7414.2023.01.032Comparative Study of Artificial Intelligence-assisted Analysis and Manual Visual Analysis in Gynecological Microbiome DiagnosisYANG Huaa,SUN Tian-shub,WANG Yaoc,XU Ying-chunc,SUN Hong-lic(a.Department of Obstetrics and Gynecology；b.Medical Science Research Center；c.Department of Laboratory Medicine,Peking Union Medical College Hospital,Beijing 100730，China)Abstract:Objective To evaluate the accuracy of morphological evaluation of vaginal microecology on Gram-stained vaginal smears by operators of differing levels of experience,determine differences between analyses using previously captured images versus live microscope images,explore the use of an artificial intelligence(AI)analysis system to independently conduct vaginal microecological evaluation,assist operators in vaginal microecological evaluation,and determine the application value of the AI analysis system in a clinical setting.MethodsA total of 385 cases of female vaginal secretion smears from May 2021 to July 2021 were collected.After gram dyeing and image acquisition,the inspectors grade assessment,manual microscope film reading,manual electronic film reading,AI independent film reading and AI auxiliary inspector film reading were conducted respectively.作者简介：杨华（1982-），女,医学博士，主治医师，研究方向：阴道微生态，E-mail:。通讯作者：孙宏莉（1971-），女，医学博士，副研究员，硕士生导师，主要研究方向：临床微生物检验，E-mail:。170现代检验医学杂志第 38 卷第 1 期2023 年 1 月J Mod Lab Med,Vol.38,No.1,Jan.2023After determining the gold standard of microscopic viewing and the gold standard of image viewing,the differences in AV score and Nugent score of two different viewing methods were analyzed,and the performance of different operators,AI,and AI-assisted performance on AV and Nugent scores were compared.ResultsThe Kappa concordance analysis of microscopic viewing and image viewing in the diagnosis of AV and BV(including BV intermediate)were 0.91 and 0.93,respectively(P 0.01).The accuracy of AI independent analysis in the diagnosis of AV and BV(including intermediate BV)were 0.85 and 0.92,sensitivity was 0.86 and 0.88 respectively,and the Kappa value was 0.62 and 0.79 respectively.The diagnostic accuracy of AV and BV(including intermediate BV)by junior operators using image viewing were 0.850.02 and 0.890.01,the sensitivity were 0.640.06 and 0.840.07,and the Kappa value were 0.550.07 and 0.720.04 respectively.The diagnostic accuracy of AV and BV(including intermediate BV)by senior operators using image viewing was 0.92 0.03 and 0.91 0.03,the sensitivity was 0.87 0.02 and 0.92 0.04 respectively,and the Kappa values was 0.78 0.07 and 0.790.06 respectively.After AI-assisted diagnosis,the Kappa values of AV and BV(including intermediate BV)diagnosed by junior operators were increased to 0.770.04 and 0.780.02,and the Kappa values of senior operators AV and BV(including intermediate BV)diagnosis were increased to 0.820.05 and 0.850.01.ConclusionThe consistency between microscopic viewing and image viewing was very high,suggesting that image viewing could replace microscopic viewing as a new viewing method.The ability of AI to independently analyze and diagnose AV and BV(including intermediate BV)was better than that of junior operators,and slightly inferior to senior operators.With AI-assisted analysis,the diagnostic capabilities for AV and BV(including intermediate BV)among both junior and senior operators improved.The performance of junior operators improved significantly and nearly approached the performance of senior operators,significantly