在癌症部位分类模型中减轻算法偏差
Abhishek Shivanna1, Adam Spannaus1, Jordan Tschida1
1Advanced Computing for Health Sciences, Oak Ridge National Laboratory, Oak Ridge, TN.
JCO clinical cancer informatics
|March 11, 2026
概括
这项研究发现,用于癌症诊断的人工智能模型在其预测中没有显著地编码种族偏见. 删除与种族相关的数据维度不会影响诊断准确度,证实了模型的公平性.
科学领域:
- 人工智能的人工智能
- 在瘤学瘤学.
- 医疗信息学 医疗信息学
背景情况:
- 人工智能 (AI) 提高了癌症诊断,但可能会延续人口偏见.
- 深度学习模型需要严格的偏见评估,以获得公平的医疗保健.
研究的目的:
- 量化AI癌症诊断模型中编码的种族信息.
- 在删除与种族相关的数据维度后评估性能变化.
主要方法:
- 在350万份癌症病理学报告上训练了一种深度学习模型.
- 用于文件嵌入的等级自我注意网络.
- 对与种族相关的维度进行训练后修剪,以评估对准确性和公平性的影响.
主要成果:
- 在癌症部位和种族预测特征之间发现了最小的重叠.
- 删除与种族相关的维度对诊断准确度的影响微不足道 (0.07%的损失).
- 没有显著的人口偏差影响临床预测.
结论:
- 从SEER数据中嵌入人工智能对于癌症部位的分类是有效的,没有显著的偏差.
- 培训后的修剪可以作为对AI模型公平性的可行审计.
相关概念视频
Mouse Models of Cancer Study
Mice have long served as models for studying human biology and pathology because of their phylogenetic and physiological similarity with humans. They are also easy to maintain and breed in the laboratory, and hence, many inbred strains are now available for research. Studies on mice have contributed immeasurably to our understanding of cancer biology.
The development of transgenic, knockout, and knock-in mice has led to an exponential increase in their use as model organisms in research,...
The development of transgenic, knockout, and knock-in mice has led to an exponential increase in their use as model organisms in research,...
Mouse Models of Cancer Study
Mice have long served as models for studying human biology and pathology because of their phylogenetic and physiological similarity with humans. They are also easy to maintain and breed in the laboratory, and hence, many inbred strains are now available for research. Studies on mice have contributed immeasurably to our understanding of cancer biology.
The development of transgenic, knockout, and knock-in mice has led to an exponential increase in their use as model organisms in research,...
The development of transgenic, knockout, and knock-in mice has led to an exponential increase in their use as model organisms in research,...
Clearance Models: Noncompartmental Models
Clearance is a pharmacokinetic parameter traditionally defined by compartment models, signifying the rate at which a drug is expelled from the body. However, a noncompartmental model offers an alternative method for assessing clearance, primarily employing empirical data obtained after administering a single drug dose.
The noncompartmental approach capitalizes on extensive sampling data, correlating the volume of distribution to systemic exposure and the administered dosage. This method enables...
The noncompartmental approach capitalizes on extensive sampling data, correlating the volume of distribution to systemic exposure and the administered dosage. This method enables...
Cancer Survival Analysis
Cancer survival analysis focuses on quantifying and interpreting the time from a key starting point, such as diagnosis or the initiation of treatment, to a specific endpoint, such as remission or death. This analysis provides critical insights into treatment effectiveness and factors that influence patient outcomes, helping to shape clinical decisions and guide prognostic evaluations. A cornerstone of oncology research, survival analysis tackles the challenges of skewed, non-normally...
