[1]陈 洁,李文生,张 巍.人工智能辅助系统在宫颈液基细胞学分析中的应用价值研究[J].现代检验医学杂志,2023,38(05):155-159.[doi:10.3969/j.issn.1671-7414.2023.05.029]
　CHEN Jie,LI Wensheng,ZHANG Wei.Study on the Value of Artificial Intelligence-assisted Systems in Cervical Liquid-based Cytology Analysis[J].Journal of Modern Laboratory Medicine,2023,38(05):155-159.[doi:10.3969/j.issn.1671-7414.2023.05.029]
点击复制

人工智能辅助系统在宫颈液基细胞学分析中的应用价值研究()
分享到：

《现代检验医学杂志》[ISSN:/CN:]

卷:
第38卷
期数:
2023年05期
页码:
155-159
栏目:
研究简报·实验技术
出版日期:
2023-09-15

文章信息/Info

Title:
Study on the Value of Artificial Intelligence-assisted Systems in Cervical Liquid-based Cytology Analysis
文章编号:
1671-7414（2023）05-155-05
作者:
陈 洁; 李文生; 张 巍
（陕西省人民医院病理科，西安 710068）
Author(s):
CHEN Jie; LI Wensheng; ZHANG Wei
（Department of Pathology, Shaanxi Provincial People’s Hospital, Xi’an 710068, China）
关键词:
宫颈薄层液基细胞学; 人工智能辅助系统; 沉降法液基细胞制片技术; 膜式薄层液基细胞学制片技术
分类号:
TP181；R446.8
DOI:
10.3969/j.issn.1671-7414.2023.05.029
文献标志码:
A
摘要:
目的 探讨人工智能（artificial intelligence，AI）辅助系统在宫颈薄层液基细胞学检查分析中的应用价值。方法 收集2020 年1 ～ 8 月在陕西省人民医院行宫颈细胞学检查的620 例液基病例，其中经膜式薄层液基细胞学(thinprepcytology test，TCT) 制片技术制片426 例，经沉降法液基细胞制片技术（liquid-based cytology technology，LCT）制片194 例。同时由病理科中级医师及AI 辅助系统阅片，以高级医师审核结果为最终结果进行对比分析。结果 中级医师在TCT 组及LCT 组中的分类判读结果与高级医师审核结果相比较，其差异均无统计学意义（χ2=0.594，1.014，P=0.898，0.602）。AI 辅助系统在TCT 组及LCT 组中的分类判读结果均与高级医师审核结果不一致，其差异具有统计学意义（χ2=104.267，26.349，均P<0.001）。AI 辅助系统在TCT，LCT 两种制片方式中对阳性片的检测灵敏度分别为100% 和83.33%，高于中级医师（86.36%，66.67%），其差异无统计学意义（χ2=3.220，0.444，P=0.233，1.000）；两种制片方式AI 辅助系统判读特异度为65.10% 和84.57%，均低于中级医师（99.26%，100%），差异具有统计学意义（χ2=160.931，31.424，均P<0.001）；准确度为66.90% 和84.54%，均低于中级医师（98.59%，98.97%），差异具有统计学意义（χ2=149.831，24.829，均P<0.001）。结论 AI 辅助系统在TCT及LCT两种制片方式中灵敏度与中级医师相似，有利于阳性病例的筛出，但准确度及特异度较低，需要结合高级医师审核结果以提高准确度。
Abstract:
Objective To investigate the value of artificial intelligence (AI) assisted system in the application of cervical thin layer cytology. Methods 620 liquid-based cases of cervical cytology were collected from Shaanxi Provincial People’s Hospital, of which 426 cases were produced by thinprep cytologic test (TCT) and 194 cases were produced by liquid-based cytology technology (LCT). All smears were interpreted by intermediate pathologists and AI-assisted systems, and the results were reviewed by senior doctors for comparison and analysis as the final results. Results The difference between the intermediate pathologist’s classification interpretation results in the TCT and LCT groups compared to the senior pathologist’s results was not statistically significant (χ2=0.594, 1.014, P=0.898, 0.602). The difference between the classification interpretation results of the AI-assisted system in the TCT and LCT groups compared to the senior pathologist’s results was statistically significant (χ2=104.267, 26.349, all P<0.001). The sensitivity of the AI-assisted system in detecting positive slices in both TCT and LCT production methods was 100% and 83.33% for pathologists (86.36% and 66.67%), respectively, and the difference was not statistically significant (χ2=3.220, 0.444, P=0.233, 1.000). The two the specificity of AI-assisted system interpretation was 65.10% and 84.57%, both lower than that of pathologists (99.26% ,100%), and the difference was statistically significant (χ2=160.931, 31.424, all P<0.001). The accuracy was 66.90% and 84.54%, both lower than that of pathologists (98.59%，98.97%), and the differences were statistically significant (χ2=149.831, 24.829, all P<0.001). Conclusion The AI-assisted system was similar in sensitivity to the intermediate pathologist in both TCT and LCT methods of production, facilitating the screening of positive cases, but the accuracy and specificity were low and need to be combined with senior pathologist review to improve accuracy.

参考文献/References:

[1] 中国抗癌协会妇科肿瘤专业委员会. 宫颈癌诊断与治疗指南( 第四版)[J]. 中国实用妇科与产科杂志,2018, 34(6):613-622. Committee of Gynecological Oncology of Chinese Anti-Cancer Association. Guidelines for the diagnosis and treatment of cervical cancer (4th Edition)[J]. Chinese Journal of Practical Gynecology and Obstetrics, 2018, 34(6): 613-622.
[2] 黄明春. 7401 例宫颈膜式液基薄层细胞学技术的临床应用结果分析[J]. 现代检验医学杂志, 2006,21(3): 42-43. HUANG Mingchun. Analysis of the clinical results of 7401 cases of cervical membrane-based liquid-based thin-layer cytology [J]. Journal of Modern Laboratory Medicine, 2006, 21(3): 42-43.
[3] BERA K, SCHALPER K A, RIMM D L, et al. Artificial intelligence in digital pathology-new tools for diagnosis and precision oncology[J]. Nature Reviews Clinical Oncology, 2019, 16(11): 703-715.
[4] GEORGE R S, HTOO A, CHENG M, et al. Artificial intelligence in prostate cancer: definitions,current research,and future directions[J]. Urologic Oncology, 2022, 40(6): 262-270.
[5] HE Tingshan, LI Jing, WANG Peng, et al. Artificial intelligence predictive system of individual survival rate for lung adenocarcinoma[J]. Computational and Structural Biotechnology Journal, 2022, 20: 2352-2359.
[6] 何文军, 李曼, 李涛, 等．基于血细胞形态识别的自动检测系统的研发[J]. 现代检验医学杂志, 2019,34(2): 104-108. HE Wenjun, LI Man, LI Tao, et al. Study on automatic detection system base on blood cell morphology recognition[J]. Journal of Modern Laboratory Medicine, 2019, 34(2): 104-108.
[7] HOU Xin, SHEN Guangyang, ZHOU Liqiang, et al. Artificial intelligence in cervical cancer screening and diagnosis [J]. Frontiers in Oncology, 2022, 12: 851367.
[8] NAYAR R, WILBUR D C. The bethesda system for reporting cervical cytology: a historical perspective[J].Acta Cytologica, 2017, 61(4/5): 359-372.
[9] BAO Heling, SUN Xiaorong, ZHANG Yi, et al. The artificial intelligence-assisted cytology diagnostic system in large-scale cervical cancer screening: a population-based cohort study of 0.7 million women[J].Cancer Medicine, 2020, 9(18):6896-6906.
[10] 胡爱侠, 朱琳, 贺慧, 等．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[J]. 临床与实验病理学杂志, 2022, 38(1): 27-30. HU Aixia, ZHU Lin, HE Hui, et al. Value about artificial intelligence assisted analysis technology for screening cervical precancerous lesions[J]. Chinese Journal of Clinical and Experimental Pathology, 2022, 38(1): 27-30.
[11] TANG Hongping, CAI De, KONG Yanqing, et al. Cervical cytology screening facilitated by an artificial intelligence microscope: a preliminary study[J]. Cancer Cytopathology, 2021, 129(9): 693-700.
[12] 李雪, 石中月, 杨志明, 等．人工智能辅助分析在宫颈液基薄层细胞学检查中的应用价值[J]．首都医科大学学报, 2020, 41(3): 360-363. LI Xue, SHI Zhongyue, YANG Zhiming, et al. Value about artificial intelligence-assisted liquid-based thin-layer cytology for cytology cervical cancer screening[J]. Journal of Capital Medical University, 2020, 41(3): 360-363.
[13] GUPTA R, SODHANI P, MEHROTRA R, et al. Cervical high-grade squamous intraepithelial lesion on conventional cytology: cytological patterns, pitfalls, and diagnostic clues[J]. Diagnostic Cytopathology, 2019, 47(12): 1267-1276.
[14] WON K H, LEE J Y, CHO H Y, et al. Impact of age on the false negative rate of human papillomavirus DNA test in patients with atypical squamous cells of undetermined significance[J]. Obstetrics Gynecology Science, 2015, 58(2): 117-123.
[15] WATSON M, BENARD V, LIN L, et al. Provider management of equivocal cervical cancer screening results among underserved women, 2009-2011: follow-up of atypical squamous cells of undetermined significance[J]. Cancer Causes Control, 2015, 26(5):759-764.
[16] LEE Y, LEE C, PARK I A, et al. Cytomorphological features of hyperchromatic crowded groups in Liquid-Based cervicovaginal cytology: a single institutional experience[J]. Journal of Pathology and Translational Medicine, 2019, 53(6): 393-398.
[17] GUPTA S, SODHANI P, CHACHRA K L, et al. Outcome of “atypical squamous cells” in a cervical cytology screening program:implications for follow up in resource limited settings[J]. Diagnostic Cytopathology, 2007, 35(11): 677-680.
[18] GOMES DE OLIVEIRA G, ELEUT?RIO R M N, SILVEIRA GON?ALVES A K, et al. Atypical squamous cells in liquid-based cervical cytology: microbiology, inflammatory infiltrate, and human papillomavirus-DNA testing[J]. Acta Cytol, 2018, 62(1)：28-33.
[19] OBERMEYER Z, EMANUEL E J. Predicting the future-big data,machine learning,and clinical medicine[J].The New England Journal of Medicine, 2016, 375(13): 1216-1219.
[20] 《宫颈液基细胞学人工智能辅助诊断数据集标注规范与质量控制专家共识（2022 版）》编写组. 宫颈液基细胞学人工智能辅助诊断数据集标注规范与质量控制专家共识(2022 版)[J]. 中华病理学杂志,2022, 51(12): 1205-1209. The Drafting Group of Chinese Expert Consensus on Label Criterion and Quality Control of Artificial Intelligence-assisted Liquid-based Cervical Cytology Diagnosis Dataset (2022 version). Expert consensus on label criterion and quality control of artificial intelligence-assisted liquid-based cervical cytology diagnosis dataset (2022 version) [J]. Chinese Journal of Pathology, 2022, 51(12): 1205-1209.

备注/Memo

备注/Memo:
基金项目： 陕西省自然科学基础研究计划（2021JQ-914）：BRAF 基因突变及VE-1 蛋白表达在甲状腺针吸检诊断的研究。陕西省人民医院科技人才支持计划“菁英人才”项目（2021JY-50）：BRAF 基因在甲状腺乳头状癌亚型中的表达差异研究。
作者简介：陈洁（1990-）女，硕士研究生，主治医师，研究方向：临床病理诊断学，E-mail：381189166@qq.com。
通讯作者：张巍（1985-）女，硕士，主治医师，研究方向：肿瘤病理诊断，E-mail：zhangwei1113@163.com。
更新日期/Last Update: 2023-09-15