| 191 | 0 | 82 |
| 下载次数 | 被引频次 | 阅读次数 |
目的 探讨人工智能（AI）细胞学阅片结合HPV分型在宫颈癌前病变（CIN2+）诊断中的效能及临床应用价值，为优化筛查策略提供依据。方法 采用回顾性配对诊断效能研究，纳入2021年3月至2025年3月在山西白求恩医院就诊的300例女性宫颈液基细胞学（LBC）标本及相应组织病理学结果。分别评估AI细胞学模型、病理医师人工阅片、HPV分型及AI+HPV联合模型在CIN2+诊断中的敏感度、特异度、阳性预测值（PPV）、阴性预测值（NPV）和曲线下面积（AUC），并采用DeLong检验和McNemar检验比较差异。同时通过决策曲线分析（DCA）评估各方法在不同风险阈值下的临床净获益。结果 AI细胞学诊断CIN2+的敏感度为87.2%，特异度80.4%,AUC为0.90(95%CI:0.86～0.94)；病理医师阅片敏感度85.0%，特异度88.9%,AUC为0.91(95%CI:0.87～0.95);HPV16/18分型敏感度最高（90.2%），但特异度最低（56.5%）,AUC为0.80(95%CI:0.75～0.84)。联合模型表现最佳，敏感度95.0%，特异度86.2%,PPV 90.0%,NPV96.2%,AUC为0.95(95%CI:0.92～0.97)，显著优于单一方法（均P<0.01）。DCA显示，在10%～40%阈值区间内，联合模型净获益始终最高。分层分析提示，联合模型在不同HPV型别、年龄及细胞学分级中均保持稳定效能，尤其在ASCUS/LSIL人群中显著提升了NPV(96.2%)，在HSIL及以上病例中敏感度达98.0%。结论 AI细胞学与HPV分型的联合诊断模式兼具高敏感度与高特异度，在减少漏诊的同时避免过度转诊，并在临床净获益方面优于单一方法，提示其在宫颈癌筛查与病理学诊断中具有应用潜力。
Abstract：Objective To evaluate the diagnostic performance and clinical application value of artificial intelligence(AI)-assisted cytologic interpretation combined with human papillomavirus(HPV) genotyping in the diagnosis of cervical intraepithelial neoplasia grade 2(CIN2+) or worse, thereby informing optimization of screening strategies. Methods A retrospective paired diagnostic efficacy study was conducted, including 300 cervical liquid-based cytology(LBC) specimens and corresponding histopathological results from women who attended Shanxi Bethune Hospital between March 2021 and March 2025. The diagnostic sensitivity, specificity, positive predictive value(PPV), negative predictive value(NPV), and area under the receiver operating characteristic curve(AUC) in the diagnosis of CIN2+ were compared among four approaches: the AI cytology model, pathologist manual review, HPV16/18 genotyping, and the AI+HPV combined model. Differences were compared using the DeLong and McNemar tests. Clinical net benefit across risk thresholds was assessed using decision curve analysis(DCA). Results For CIN2+ diagnosis, AI cytology model yielded a sensitivity of 87.2%, specificity of 80.4%, and AUC of 0.90(95% CI: 0.86–0.94). Pathologist manual review achieved a sensitivity of 85.0%, specificity of 88.9%, and AUC of 0.91(95% CI: 0.87–0.95). HPV16/18 genotyping showed the highest sensitivity(90.2%) but the lowest specificity(56.5%), with an AUC of 0.80(95% CI: 0.75–0.84). The AI+HPV combined model demonstrated the best performance, with a sensitivity of 95.0%, specificity of 86.2%, PPV of 90.0%, NPV of 96.2%, and AUC of 0.95(95% CI: 0.92–0.97), significantly outperforming any single method(all P<0.01). DCA indicated that the combined model consistently provided the greatest net benefit within the 10%–40% threshold probability range. Subgroup analysis confirmed that the combined model maintained stable performance across different HPV genotypes, age groups, and cytology categories. Notably, the combined model markedly improved NPV(96.2%) in ASC-US/LSIL cases and achieved a sensitivity of 98.0% in HSIL and above cases. Conclusion The combined diagnostic model of AI-assisted cytologic interpretation and HPV genotyping achieves both high sensitivity and specificity, reducing missed diagnoses while avoiding excessive referrals. It offers greater clinical net benefit than single-method approaches and shows promise as a multimodal strategy in cervical cancer screening and pathological diagnosis.
[1]张春丽,刘忠杰,陈欢,等.宫颈癌及宫颈上皮内瘤变患者高危型HPV感染与组织中PKP4、DSC2蛋白表达的关系[J].诊断病理学杂志, 2025, 32(7):865-869. DOI:10. 3969/j. issn. 1007-8096. 2025.07. 018.
[2]杨益,李春梅,贾蜀云,等. TruScreen联合高危HPV检测与液基细胞学联合HPV检测在宫颈癌筛查中的对比研究[J].四川大学学报（医学版）, 2025, 56(3):852-857. DOI:10. 12182/20250560609.
[3]Debeaudrap P, Kabore FN, Setha L, et al. Performance of visual inspection, partial genotyping, and their combination for the triage of women living with HIV who are screen positive for human papillomavirus:Results from the AIMA-CC ANRS 12375 multicentric screening study[J]. Int J Cancer, 2025, 156(3):598-607. DOI:10.1002/ijc. 35190.
[4]Meng Z, Zhao Z, Li B, et al. A cervical histopathology dataset for computer aided diagnosis of precancerous lesions[J]. IEEE Trans Med Imaging, 2021, 40(6):1531-1541. DOI:10. 1109/TMI. 2021.3059699.
[5]Allahqoli L, LaganàAS, Mazidimoradi A, et al. Diagnosis of cervical cancer and pre-cancerous lesions by artificial intelligence:A systematic review[J]. Diagnostics(Basel), 2022, 12(11):2771. DOI:10.3390/diagnostics12112771.
[6]温永琴,张若愚,李先蕾,等.人工智能构建宫颈液基薄层细胞涂片质控模型对比研究[J].西安交通大学学报（医学版）, 2025,46(3):544-550. DOI:10. 7652/jdyxb202503024.
[7]Aslanimehr M, Nemati S, Sadeghi H, et al. High-risk HPV genotypes in women with abnormal cytology:A 12-year retrospective study[J].Infect Agent Cancer, 2025, 20(1):34. DOI:10. 1186/s13027-025-00664-0.
[8]Wang T, Zhang H, Liu Y, et al. Updates in cervical cancer screening guidelines, the bethesda system for reporting cervical cytology,and clinical management recommendations[J]. J Clin Transl Pathol,2023,3(2):75-83. DOI:10. 14218/jctp. 2023. 00004.
[9]王红,刘茵,徐慧芳,等.基于前瞻性队列的高危型人乳头瘤病毒分型检测在宫颈癌筛查中的效果评价[J].中华肿瘤杂志, 2025,47(5):435-442. DOI:10. 3760/cma. j. cn112152-20240219-00079.
[10]Goldstein DY, Yang T, Lucic D, et al. Multi-center evaluation of the Alinity m HR HPV assay with liquid-based cytology cervical specimens in the United States[J]. Microbiol Spectr, 2025, 13(3):e0191824. DOI:10. 1128/spectrum. 01918-24.
[11]张兰兰,郭英,张义磊,等.人工阅片联合人工智能TBS辅助阅片在宫颈液基细胞学中的应用体会[J].诊断病理学杂志, 2024,31(11):1082-1085. DOI:10. 3969/j. issn. 1007-8096. 2024. 11. 012.
[12]Verhoef L, Bleeker MCG, Polman N, et al. Colposcopy referrals and CIN3 detection after triage by host cell DNA methylation and/or HPV genotyping in HPV positive women with low-grade cytology from a population-based Dutch primary HPV screening trial[J]. Int J Cancer,2025, 156(5):1065-1073. DOI:10. 1002/ijc. 35289.
[13]蒋圣露,高霏.人工智能辅助新技术在宫颈癌筛查中的新进展[J].中国病原生物学杂志, 2025, 20(3):401-404. DOI:10. 13350/j. cjpb.250327.
[14]Hou X, Shen G, Zhou L, et al. Artificial intelligence in cervical cancer screening and diagnosis[J]. Front Oncol, 2022, 12:851367. DOI:10. 3389/fonc. 2022. 851367.
[15]陈福连.“三阶梯”筛查对宫颈腺癌的诊断价值[D].福建:福建医科大学, 2022.
[16]吴晓娟,张艺宝,任鸿儒,等.基于卷积神经网络的宫颈癌调强放疗剂量预测方法[J].中国医学物理学杂志, 2025, 42(4):421-428.DOI:10. 3969/j. issn. 1005-202X. 2025. 04. 001.
[17]Akash RS, Islam R, Badhon SSI, et al. CerviXpert:A multi-structural convolutional neural network for predicting cervix type and cervical cell abnormalities[J]. Digit Health, 2024, 10:20552076241295440.DOI:10. 1177/20552076241295440.
[18]Beckhaus T, Kachuri L, Nakase T, et al. Genome-wide association analyses of HPV16 and HPV18 seropositivity identify susceptibility loci for cervical cancer[J]. J Med Virol, 2025, 97(2):e70195. DOI:10. 1002/jmv. 70195.
[19]Murillo R, Gamboa O, Hernández G, et al. Accuracy of combined molecular and morphology-based triage for HPV-positive women in routine cervical cancer screening services from Colombia[J]. Prev Med, 2021, 153:106801. DOI:10. 1016/j. ypmed. 2021. 106801.
[20]苑蕾蕾,兰什,王逸群.不同年龄女性HR-HPV感染现状及液基细胞学TBS+DNA双智能诊断系统在宫颈癌筛查中的应用探讨[J].循证医学, 2024, 24(5):295-303. DOI:10. 12019/j. issn. 1671-5144.202404033.
[21]Liu L, Liu J, Su Q, et al. Performance of artificial intelligence for diagnosing cervical intraepithelial neoplasia and cervical cancer:A systematic review and meta-analysis[J]. EClinicalMedicine, 2024,80:102992. DOI:10. 1016/j. eclinm. 2024. 102992.
[22]Wu H, Huang L, Chen X, et al. Establishment of a novel risk stratification system integrating clinical and pathological parameters for prognostication and clinical decision-making in early-stage cervical cancer[J]. Cancer Med, 2024, 13(22):e70394. DOI:10. 1002/cam4.70394.
[23]薛鹏.人工智能技术在宫颈癌筛查、分流和诊断中的应用研究[D].中国医学科学院, 2023.
[24]Xue P, Dang L, Kong LH, et al. Deep learning enabled liquid-based cytology model for cervical precancer and cancer detection[J]. Nat Commun, 2025, 16(1):3506. DOI:10. 1038/s41467-025-58883-3.
基本信息:
中图分类号:R737.33
引用信息:
[1]尚海霞,史晓峰,于宏鑫,等.人工智能细胞学阅片结合HPV分型在宫颈病变病理学诊断中的应用[J].诊断病理学杂志,2025,32(12):1560-1565.
基金信息:
2024年度山西省高质量发展研究项目（SXGZL202449）
2025-10-29
2025-10-29
2025-10-29