SCH: Quantifying and mitigating demographic biases of machine learning in real world radiology

SCH：量化和减轻现实世界放射学中机器学习的人口统计偏差

• 批准号: 10818941
• 负责人: JEREMIAS SULAM
• 金额: $ 31.85万
• 依托单位: JOHNS HOPKINS UNIVERSITY
• 依托单位国家: 美国
• 财政年份: 2023
• 资助国家: 美国
• 起止时间: 2023-07-01 至 2026-06-30
• 项目状态: 未结题
• 来源: https://reporter.nih.gov/project-details/10818941
• 关键词: Address; Age; Algorithms; Anatomy; Area; Artificial Intelligence; Binding; Breast Cancer Detection; Certification; Chest imaging; Clinical/Radiologic; Complement; Complex; Data; Data Set; Development; Diagnosis; Diagnostic Imaging; Disease; Disparity; Engineering; Ensure; Ethics; Evaluation; Explosion; Future; Generations; Health; Health Policy; Image; Individual; Instruction; Knowledge; Location; Machine Learning; Malignant neoplasm of lung; Measures; Medical Imaging; Medical center; Medicine; Methodology; Methods; Modernization; Monitor; Outcome; Performance; Population; Procedures; Proxy; Public Health Applications Research; Race; Radiology Specialty; Recommendation; Research; Sampling; Science; Screening for cancer; Screening procedure; Series; System; Technology; Thoracic Radiography; Time; Training; Underrepresented Populations; Validation; access disparities; algorithmic bias; artificial neural network; biological sex; breast imaging; cancer diagnosis; cancer imaging; clinically relevant; deep neural network; demographics; diagnostic accuracy; diagnostic tool; disease classification; disparity reduction; health disparity; high risk; imaging modality; improved; lung cancer screening; machine learning algorithm; machine learning method; machine learning model; malignant breast neoplasm; novel; population based; prediction algorithm; predictive modeling; prevent; public health relevance; real world application; screening; screening program; tool

PROJECT SUMMARY (See instructions): The application of modern machine learning algorithms in radiology continues to grow, as these tools represent potential huge improvements in efficiency, accessibility and accuracy of diagnostic and screening tools. At the same time, these increasingly complex machine learning models can have biased predictions against individuals of under-represented demographic groups, potentially perpetuating pre-existing health disparities. Such fairness concerns are particularly important in public health applications that focus on large scale population-based screening, as in cancer screening for breast and lung cancer. In these settings, it is paramount to understand how often machine learning screening algorithms can be unfair and biased, and how to mitigate these disparities. This proposal will develop tools to quantify, correct, and analyze the biases of predictive algorithms in relation to different demographic groups in real world settings. In particular, we will develop analysis and algorithms to quantify the violation of fairness by a machine learning model in situations where information about the sensitive attribute itself (such as biological sex, race or age) are not directly observable, and we will provide algorithms that correct for their worst-case fairness violations. We will analyze our tools under distribution shifts, whereby differences in populations exist, as is common in large scale cancer screening programs. This project will also perform inference on the training samples and features most highly associated with fairness violations, thereby providing guidance on the development of solutions to prevent biased algorithms in the future. Our tools will be validated on a variety of large real-world radiology datasets spanning multiple imaging modalities, including general chest X-ray datasets that include lung cancer diagnoses (CheXpert and MIMIC-CXR), as well as the Emory Breast Cancer Imaging Dataset (EMBED) and the National Lung Cancer Screening Trial, evaluating and correcting disparities for predictive algorithms with respect to biological sex (where appropriate), race, and age. The results of this project will establish critical knowledge about the propensity of machine learning models for medical imaging diagnosis and cancer screening to be unfair and biased, as well as foundational tools to quantify and mitigate these biases in these potentially game-changing technologies.

项目摘要（请参阅说明）： 随着这些工具，现代机器学习算法在放射学中的应用不断增长 代表诊断的效率，可访问性和准确性的潜在巨大提高和 筛选工具。同时，这些日益复杂的机器学习模型可能有偏见 预测代表性不足的人群群体的个人，可能会延续 预先存在的健康差异。这样的公平关注在公共卫生中尤其重要 侧重于大规模基于人群的筛查的应用，如乳腺癌和乳腺癌筛查 肺癌。在这些设置中，了解机器学习筛查的频率至关重要 算法可能是不公平和有偏见的，以及如何减轻这些差异。该建议将发展 量化，纠正和分析预测算法的偏差的工具 现实世界中的人口群体。特别是，我们将开发分析和算法 在有关信息的情况下，在有关机器学习模型的情况下量化了有关公平性的侵犯 敏感属性本身（例如生物性别，种族或年龄）是不直接观察到的，我们将 提供对最严重的公平性违规行为正确的算法。我们将分析我们的工具 分布变化，在大规模癌症筛查中常见的人口差异存在差异 程序。该项目还将在培训样本上进行推断，并具有最高的特征 与公平违规有关，从而为制定解决方案提供指导 将来有偏见的算法。我们的工具将在各种大型现实世界放射学上得到验证 跨越多种成像方式的数据集，包括包括肺部的一般胸部X射线数据集 癌症诊断（CHEXPERT和MIMIC-CXR）以及Emory乳腺癌成像数据集 （嵌入）和国家肺癌筛查试验，评估和纠正差异 关于生物学性别（适当），种族和年龄的预测算法。结果的结果 项目将建立有关医疗机器学习模型倾向的关键知识 成像诊断和癌症筛查是不公平和有偏见的，以及量化的基础工具 并减轻这些可能改变游戏规则的技术中的这些偏见。