留言板

尊敬的读者、作者、审稿人, 关于本刊的投稿、审稿、编辑和出版的任何问题, 您可以本页添加留言。我们将尽快给您答复。谢谢您的支持!

姓名
邮箱
手机号码
标题
留言内容
验证码

可信病理人工智能:从理论到实践

周燕燕 邓杨 包骥 步宏

周燕燕, 邓杨, 包骥, 步宏. 可信病理人工智能:从理论到实践[J]. 协和医学杂志. doi: 10.12290/xhyxzz.2022-0184
引用本文: 周燕燕, 邓杨, 包骥, 步宏. 可信病理人工智能:从理论到实践[J]. 协和医学杂志. doi: 10.12290/xhyxzz.2022-0184
ZHOU Yanyan, DENG Yang, BAO Ji, BU Hong. Towards Trusted Artificial Intelligence for Pathology: From Theory to Practice[J]. Medical Journal of Peking Union Medical College Hospital. doi: 10.12290/xhyxzz.2022-0184
Citation: ZHOU Yanyan, DENG Yang, BAO Ji, BU Hong. Towards Trusted Artificial Intelligence for Pathology: From Theory to Practice[J]. Medical Journal of Peking Union Medical College Hospital. doi: 10.12290/xhyxzz.2022-0184

可信病理人工智能:从理论到实践

doi: 10.12290/xhyxzz.2022-0184
基金项目: 

成都市新型产业技术研究院技术创新项目(2017-CY02-00026-GX);四川大学华西医院临床研究孵化项目(20HXFH029);四川大学华西医院学科卓越发展1·3·5工程项目(ZYGD18012)

详细信息
    通讯作者:

    包骥,E-mail:baoji@scu.edu.cn

  • 中图分类号: R818.02;R36

Towards Trusted Artificial Intelligence for Pathology: From Theory to Practice

  • 摘要: 人工智能正在融入病理学研究的各个领域,但在临床实践中却遇到了诸多挑战,包括因医疗数据隐私保护而形成“数据孤岛”,不利于人工智能模型的训练;现有人工智能模型缺乏可解释性,导致人无法理解而难以形成人机互动;人工智能模型对多模态数据利用不足,致使其预测效能难以进一步提升等。为解决上述难题,我们建议在现有病理人工智能研究中引入最新可信人工智能技术:(1)数据安全共享,在坚持数据保护的基础上打破数据孤岛,使用联邦学习的方法、仅调用数据训练的结果而不上传数据本身,在不影响数据安全的情况下大大增加可用于训练的数据量;(2)赋予人工智能可解释性,使用图神经网络技术模拟病理医生学习病理诊断的过程,使得模型本身具有可解释性;(3)多模态信息融合,使用知识图谱技术对更为多样和全面的数据来源进行整合并深入挖掘分析,得出更准确的模型。相信通过此三方面的工作,可信病理人工智能技术可使病理人工智能达到可控可靠和明确责任,从而促进病理人工智能的发展和临床应用。
  • [1] Li BH, Hou BC, Yu WT, et al. Applications of artificial intelligence in intelligent manufacturing:a review[J]. Front Inform Technol Electron Eng,2017,18:86-96.
    [2] Rajpurkar P, Chen E, Banerjee O, et al. AI in health and medicine[J]. Nat Med, 2022,28:31-38.
    [3] European Commission. White Paper On Artificial Intelligence-A European approach to excellence and trust[EB/OL]. (2020-02-19)[2022-04-05]. https://ec.europa.eu/info/sites/info/files/commission-white-paper-artificial-intelligencefeb2020_en.pdf.
    [4] 中国信息通信研究院.可信人工智能白皮书[EB/OL]. (2021-07-09)[2022-04-05].http://www.caict.ac.cn/k-xyj/qwfb/bps/202107/P020210709319866413974.pdf.
    [5] Parwani A. Whole Slide Imaging[M]. Switzerland:Springer Nature Switzerland AG, 2022:223-236.
    [6] Willemink MJ, Koszek WA, Hardell C, et al. Preparing Medical Imaging Data for Machine Learning[J]. Radiology,2020,295:4-15.
    [7] European Commission. General Data Protection Regulation[EB/OL]. (2016-04-27)[2022-04-05]. https://gdp-r.eu/
    [8] Law Reform Commission. Hong Kong Person Date Privacy Ordinance[EB/OL]. (2012-10-01)[2022-04-05]. https://www.pcpd.org.hk/english/data_privacy_law/ordinance_at_a_Glance/ordinance.html.
    [9] Kaissis GA, Makowski MR, Rückert D, et al. Secure, privacy-preserving and federated machine learning in medical imaging[J]. Nat Machine Intel, 2020,2:305-311.
    [10] Zhou SK, Greenspan H, Davatzikos C, et al. A Review of Deep Learning in Medical Imaging:Imaging Traits, Technology Trends, Case Studies With Progress Highlights, and Future Promises[J]. Proc IEEE Inst Electr Electron Eng, 2021, 109:820-838.
    [11] 刘再毅,石镇维,梁长虹.推进联邦学习技术在医学影像人工智能中的应用[J].中华医学杂志,2022,102:318-320. Liu ZY, Shi ZW, Liang CH. Promoting the application of federated learning in medical imaging artificial intelligence[J].Zhonghua Yixue Zazhi, 2022

    ,102:318-320.
    [12] Yang D, Xu Z, Li WQ, et al. Federated Semi-Supervised Learning for COVID Region Segmentation in Chest CT using Multi-National Data from China, Italy, Japan[J]. Med Image Anal, 2021,70:e101992.
    [13] Lu MY, Chen RJ, Kong DH, et al. Federated learning for computational pathology on gigapixel whole slide images[J]. Med Image Anal, 2022,76:e102298.
    [14] Du MN, Liu NH, Hu X. Techniques for Interpretable Machine Learning[J]. Commun ACM,2020,63:68-77.
    [15] Selvaraju RR, Cogswell M, Das A, et al. Grad-CAM:Visual Explanations from Deep Networks via Gradient-Based Localization[J]. Int J Comput Vis, 2020,128:336-359.
    [16] Yu K, Wang F, Berry GJ, et al. Classifying non-small cell lung cancer types and transcriptomic subtypes using convolutional neural networks[J]. J Am Med Inform Assoc, 2020,27:757-769.
    [17] Sousa I, Vellasco M, Silva E. Local Interpretable Model-Agnostic Explanations for Classification of Lymph Node Metastases[J]. Sensors, 2019,19:1-18.
    [18] Saporta A, Gui XT, Agrawal A, et al. Deep learning saliency maps do not accurately highlight diagnostically relevant regions for medical image interpretation[J].MedRxiv,2021. https://doi.org/10.1101/2021.02.28.21252634.
    [19] Ehsan U, Passi S, Liao QV, et al. The Who in Explainable AI:How AI Background Shapes Perceptions of AI Explanations[J]. Arxiv, 2021. https://arxiv.org/abs/2107.13509.
    [20] Li X, Dvornek NC, Zhou Y, et al. Efficient Interpretation of Deep Learning Models Using Graph Structure and Cooperative Game Theory:Application to ASD Biomarker Discovery[J]. Inf Process Med Imaging, 2019, 11492:718-730.
    [21] Li X, Zhou Y, Dvornek NC, et al. Efficient Shapley Explanation for Features Importance Estimation Under Uncertainty[J]. Med Image Comput Assist Interv, 2020, 12261:792-801.
    [22] Sarder SP. From What to Why, the Growing Need for a Focus Shift Toward Explainability of AI in Digital Pathology[J]. Front Physiol, 2022,12:e821217.
    [23] Hegde N, Hipp JD, Liu Y, et al. Similar Image Search for Histopathology:SMILY[J]. NPJ Digit Med,2019,2:56-65.
    [24] Li X, Duncan J. BrainGNN:Interpretable Brain Graph Neural Network for fMRI Analysis[J]. Med Image Anal,2021,74,e102233.
    [25] Li X, Zhou Y, Dvornek NC, et al. Pooling Regularized Graph Neural Network for fMRI Biomarker Analysis[J]. Med Image Comput Assist Interv, 2020, 12267:625-635.
    [26] Amit S. Introducing the knowledge graph[R]. America:Official Blog of Google, 2012.
    [27] 崔洁.面向乳腺肿瘤诊断的知识图谱及辅助决策研究[D].上海:东华大学,2018.
  • 加载中
计量
  • 文章访问数:  5
  • HTML全文浏览量:  0
  • PDF下载量:  8
  • 被引次数: 0
出版历程
  • 收稿日期:  2022-04-06
  • 网络出版日期:  2022-06-11

目录

    /

    返回文章
    返回