迈向机器学习在医疗保健领域的实际应用：外部验证、持续监测和随机临床试验。
Toward real-world deployment of machine learning for health care: External validation, continual monitoring, and randomized clinical trials.
作者信息
Yuan Han
机构信息
Centre for Quantitative Medicine Duke-NUS Medical School Singapore Singapore.
出版信息
Health Care Sci. 2024 Oct 14;3(5):360-364. doi: 10.1002/hcs2.114. eCollection 2024 Oct.
Abstract
In this commentary, we elucidate three indispensable evaluation steps toward the real-world deployment of machine learning within the healthcare sector and demonstrate referable examples for diagnostic, therapeutic, and prognostic tasks. We encourage researchers to move beyond retrospective and within-sample validation, and step into the practical implementation at the bedside rather than leaving developed machine learning models in the dust of archived literature.
摘要
在本评论中，我们阐明了在医疗保健领域实际部署机器学习所需的三个必不可少的评估步骤，并展示了诊断、治疗和预后任务的可参考示例。我们鼓励研究人员超越回顾性和样本内验证，步入床边的实际应用，而不是将已开发的机器学习模型尘封在存档文献中。
相似文献
1
Toward real-world deployment of machine learning for health care: External validation, continual monitoring, and randomized clinical trials.迈向机器学习在医疗保健领域的实际应用：外部验证、持续监测和随机临床试验。
Health Care Sci. 2024 Oct 14;3(5):360-364. doi: 10.1002/hcs2.114. eCollection 2024 Oct.
2
DEPLOYR: a technical framework for deploying custom real-time machine learning models into the electronic medical record.DEPLOYR：一个将定制的实时机器学习模型部署到电子病历中的技术框架。
J Am Med Inform Assoc. 2023 Aug 18;30(9):1532-1542. doi: 10.1093/jamia/ocad114.
3
The future of Cochrane Neonatal.考克兰新生儿协作网的未来。
Early Hum Dev. 2020 Nov;150:105191. doi: 10.1016/j.earlhumdev.2020.105191. Epub 2020 Sep 12.
4
Development and validation of a machine learning-based predictive model for assessing the 90-day prognostic outcome of patients with spontaneous intracerebral hemorrhage.基于机器学习的预测模型评估自发性脑出血患者 90 天预后结局的开发与验证。
J Transl Med. 2024 Mar 4;22(1):236. doi: 10.1186/s12967-024-04896-3.
5
Clinical deployment environments: Five pillars of translational machine learning for health.临床部署环境：健康领域转化型机器学习的五大支柱
Front Digit Health. 2022 Aug 19;4:939292. doi: 10.3389/fdgth.2022.939292. eCollection 2022.
6
Error Consistency for Machine Learning Evaluation and Validation with Application to Biomedical Diagnostics.用于机器学习评估与验证的误差一致性及其在生物医学诊断中的应用
Diagnostics (Basel). 2023 Apr 1;13(7):1315. doi: 10.3390/diagnostics13071315.
7
Machine learning in medicine: a practical introduction.医学中的机器学习：实用入门
BMC Med Res Methodol. 2019 Mar 19;19(1):64. doi: 10.1186/s12874-019-0681-4.
8
Machine Learning Prognostic Models for Gastrointestinal Bleeding Using Electronic Health Record Data.基于电子健康记录数据的机器学习预测胃肠道出血模型
Am J Gastroenterol. 2020 Aug;115(8):1199-1200. doi: 10.14309/ajg.0000000000000720.
9
Continual Learning for Activity Recognition.持续学习的活动识别。
Annu Int Conf IEEE Eng Med Biol Soc. 2022 Jul;2022:2416-2420. doi: 10.1109/EMBC48229.2022.9871690.
10
Predictive models for clinical decision making: Deep dives in practical machine learning.临床决策预测模型：实用机器学习深度剖析。
J Intern Med. 2022 Aug;292(2):278-295. doi: 10.1111/joim.13483. Epub 2022 Apr 25.
引用本文的文献
1
Rethinking Domain-Specific Pretraining by Supervised or Self-Supervised Learning for Chest Radiograph Classification: A Comparative Study Against ImageNet Counterparts in Cold-Start Active Learning.通过监督学习或自监督学习对胸部X光片分类进行特定领域预训练的再思考：与冷启动主动学习中的ImageNet对应模型的比较研究
Health Care Sci. 2025 Apr 6;4(2):110-143. doi: 10.1002/hcs2.70009. eCollection 2025 Apr.
2
Reflecting on Achievements and Renewing Our Vision.回顾成就，重塑愿景。
Health Care Sci. 2025 Jan 28;4(1):1-3. doi: 10.1002/hcs2.127. eCollection 2025 Feb.
3
Leveraging anatomical constraints with uncertainty for pneumothorax segmentation.利用带有不确定性的解剖学约束进行气胸分割。
Health Care Sci. 2024 Dec 15;3(6):456-474. doi: 10.1002/hcs2.119. eCollection 2024 Dec.
本文引用的文献
1
Clinical domain knowledge-derived template improves post hoc AI explanations in pneumothorax classification.临床领域知识衍生模板可提高气胸分类事后人工智能解释的质量。
J Biomed Inform. 2024 Aug;156:104673. doi: 10.1016/j.jbi.2024.104673. Epub 2024 Jun 9.
2
Zero-shot interpretable phenotyping of postpartum hemorrhage using large language models.使用大语言模型对产后出血进行零样本可解释表型分析。
NPJ Digit Med. 2023 Nov 30;6(1):212. doi: 10.1038/s41746-023-00957-x.
3
TransformEHR: transformer-based encoder-decoder generative model to enhance prediction of disease outcomes using electronic health records.TransformEHR：基于转换器的编解码器生成模型，用于使用电子健康记录增强疾病结局预测。
Nat Commun. 2023 Nov 29;14(1):7857. doi: 10.1038/s41467-023-43715-z.
4
Machine learning in clinical practice: Evaluation of an artificial intelligence tool after implementation.临床实践中的机器学习：实施后人工智能工具的评估。
Emerg Med Australas. 2024 Feb;36(1):118-124. doi: 10.1111/1742-6723.14325. Epub 2023 Sep 28.
5
Algorithmic fairness in artificial intelligence for medicine and healthcare.人工智能在医学和医疗保健中的算法公平性。
Nat Biomed Eng. 2023 Jun;7(6):719-742. doi: 10.1038/s41551-023-01056-8. Epub 2023 Jun 28.
6
AutoPrognosis 2.0: Democratizing diagnostic and prognostic modeling in healthcare with automated machine learning.自动预后2.0：通过自动化机器学习实现医疗保健中诊断和预后建模的普及。
PLOS Digit Health. 2023 Jun 22;2(6):e0000276. doi: 10.1371/journal.pdig.0000276. eCollection 2023 Jun.
7
Development and internal-external validation of statistical and machine learning models for breast cancer prognostication: cohort study.统计和机器学习模型在乳腺癌预后预测中的开发和内外验证：队列研究。
BMJ. 2023 May 10;381:e073800. doi: 10.1136/bmj-2022-073800.
8
Blinded, randomized trial of sonographer versus AI cardiac function assessment.超声医师与人工智能心脏功能评估的盲法、随机试验。
Nature. 2023 Apr;616(7957):520-524. doi: 10.1038/s41586-023-05947-3. Epub 2023 Apr 5.
9
Implemented machine learning tools to inform decision-making for patient care in hospital settings: a scoping review.实施机器学习工具以在医院环境中为患者护理决策提供信息：范围综述。
BMJ Open. 2023 Feb 7;13(2):e065845. doi: 10.1136/bmjopen-2022-065845.
10
Implementation of Machine Learning Pipelines for Clinical Practice: Development and Validation Study.用于临床实践的机器学习管道的实施：开发与验证研究。
JMIR Med Inform. 2022 Dec 16;10(12):e37833. doi: 10.2196/37833.
Zotero 插件