Suppr超能文献

提高对医学机器学习中潜在偏差的认识:数据马拉松的经验

Raising awareness of potential biases in medical machine learning: Experience from a Datathon.

作者信息

Hochheiser Harry, Klug Jesse, Mathie Thomas, Pollard Tom J, Raffa Jesse D, Ballard Stephanie L, Conrad Evamarie A, Edakalavan Smitha, Joseph Allan, Alnomasy Nader, Nutman Sarah, Hill Veronika, Kapoor Sumit, Claudio Eddie Pérez, Kravchenko Olga V, Li Ruoting, Nourelahi Mehdi, Diaz Jenny, Taylor W Michael, Rooney Sydney R, Woeltje Maeve, Celi Leo Anthony, Horvat Christopher M

机构信息

Department of Biomedical Informatics, University of Pittsburgh, Pittsburgh, Pennsylvania, United States of America.

UPMC Intensive Care Unit Service Center, UPMC, Pittsburgh, Pennsylvania, United States of America.

出版信息

PLOS Digit Health. 2025 Jul 11;4(7):e0000932. doi: 10.1371/journal.pdig.0000932. eCollection 2025 Jul.

Abstract

OBJECTIVE

To challenge clinicians and informaticians to learn about potential sources of bias in medical machine learning models through investigation of data and predictions from an open-source severity of illness score.

METHODS

Over a two-day period (total elapsed time approximately 28 hours), we conducted a datathon that challenged interdisciplinary teams to investigate potential sources of bias in the Global Open Source Severity of Illness Score. Teams were invited to develop hypotheses, to use tools of their choosing to identify potential sources of bias, and to provide a final report.

RESULTS

Five teams participated, three of which included both informaticians and clinicians. Most (4/5) used Python for analyses, the remaining team used R. Common analysis themes included relationship of the GOSSIS-1 prediction score with demographics and care related variables; relationships between demographics and outcomes; calibration and factors related to the context of care; and the impact of missingness. Representativeness of the population, differences in calibration and model performance among groups, and differences in performance across hospital settings were identified as possible sources of bias.

DISCUSSION

Datathons are a promising approach for challenging developers and users to explore questions relating to unrecognized biases in medical machine learning algorithms.

摘要

目的

通过对开源疾病严重程度评分的数据和预测进行调查,促使临床医生和信息专家了解医学机器学习模型中潜在的偏差来源。

方法

在为期两天的时间里(总耗时约28小时),我们举办了一场数据马拉松,要求跨学科团队调查全球开源疾病严重程度评分中潜在的偏差来源。邀请各团队提出假设,使用他们选择的工具来识别潜在的偏差来源,并提供一份最终报告。

结果

五个团队参与其中,其中三个团队既有信息专家又有临床医生。大多数团队(4/5)使用Python进行分析,其余团队使用R。常见的分析主题包括GOSSIS-1预测评分与人口统计学和护理相关变量的关系;人口统计学与结果之间的关系;校准以及与护理背景相关的因素;以及数据缺失的影响。人群的代表性、不同组之间校准和模型性能的差异以及不同医院环境下性能的差异被确定为可能的偏差来源。

讨论

数据马拉松是一种很有前景的方法,可促使开发者和用户探索与医学机器学习算法中未被识别的偏差相关的问题。

相似文献

1
Raising awareness of potential biases in medical machine learning: Experience from a Datathon.
PLOS Digit Health. 2025 Jul 11;4(7):e0000932. doi: 10.1371/journal.pdig.0000932. eCollection 2025 Jul.
2
Raising awareness of potential biases in medical machine learning: Experience from a Datathon.
medRxiv. 2024 Nov 2:2024.10.21.24315543. doi: 10.1101/2024.10.21.24315543.
5
Are Current Survival Prediction Tools Useful When Treating Subsequent Skeletal-related Events From Bone Metastases?
Clin Orthop Relat Res. 2024 Sep 1;482(9):1710-1721. doi: 10.1097/CORR.0000000000003030. Epub 2024 Mar 22.
6
Health professionals' experience of teamwork education in acute hospital settings: a systematic review of qualitative literature.
JBI Database System Rev Implement Rep. 2016 Apr;14(4):96-137. doi: 10.11124/JBISRIR-2016-1843.
7
Is It Possible to Develop a Patient-reported Experience Measure With Lower Ceiling Effect?
Clin Orthop Relat Res. 2025 Apr 1;483(4):693-703. doi: 10.1097/CORR.0000000000003262. Epub 2024 Oct 25.
10
Audit and feedback: effects on professional practice.
Cochrane Database Syst Rev. 2025 Mar 25;3(3):CD000259. doi: 10.1002/14651858.CD000259.pub4.

引用本文的文献

1
Leveraging Datathons to Teach AI in Undergraduate Medical Education: Case Study.
JMIR Med Educ. 2025 Apr 16;11:e63602. doi: 10.2196/63602.

本文引用的文献

1
Algorithmic individual fairness and healthcare: a scoping review.
JAMIA Open. 2024 Dec 30;8(1):ooae149. doi: 10.1093/jamiaopen/ooae149. eCollection 2025 Feb.
2
Tackling algorithmic bias and promoting transparency in health datasets: the STANDING Together consensus recommendations.
Lancet Digit Health. 2025 Jan;7(1):e64-e88. doi: 10.1016/S2589-7500(24)00224-3. Epub 2024 Dec 18.
3
Measuring and Reducing Racial Bias in a Pediatric Urinary Tract Infection Model.
AMIA Jt Summits Transl Sci Proc. 2024 May 31;2024:488-497. eCollection 2024.
6
Guiding Principles to Address the Impact of Algorithm Bias on Racial and Ethnic Disparities in Health and Health Care.
JAMA Netw Open. 2023 Dec 1;6(12):e2345050. doi: 10.1001/jamanetworkopen.2023.45050.
7
Considerations for addressing bias in artificial intelligence for health equity.
NPJ Digit Med. 2023 Sep 12;6(1):170. doi: 10.1038/s41746-023-00913-9.
8
Using measures of race to make clinical predictions: Decision making, patient health, and fairness.
Proc Natl Acad Sci U S A. 2023 Aug 29;120(35):e2303370120. doi: 10.1073/pnas.2303370120. Epub 2023 Aug 22.
9
Bias in artificial intelligence algorithms and recommendations for mitigation.
PLOS Digit Health. 2023 Jun 22;2(6):e0000278. doi: 10.1371/journal.pdig.0000278. eCollection 2023 Jun.
10
AI Reporting Guidelines: How to Select the Best One for Your Research.
Radiol Artif Intell. 2023 Apr 5;5(3):e230055. doi: 10.1148/ryai.230055. eCollection 2023 May.

文献AI研究员

20分钟写一篇综述,助力文献阅读效率提升50倍。

立即体验

用中文搜PubMed

大模型驱动的PubMed中文搜索引擎

马上搜索

文档翻译

学术文献翻译模型,支持多种主流文档格式。

立即体验