医学中的生成人工智能:开创性的进步还是延续历史的不准确性? 评估隐含偏见的横截面研究
Philip Sutera1, Rohini Bhatia2, Timothy Lin3
1Department of Radiation Oncology, University of Rochester Medical Center, Rochester, NY, United States.
JMIR AI
|July 3, 2025
概括
生成型人工智能图像模型显示了医学专业的偏见,与当前的劳动力数据相比,女性医生往往被低估. 数据集需要重新培训,以确保在人工智能生成的医学图像中提供多样化的表示.
科学领域:
- 人工智能的人工智能
- 医疗成像医学成像
- 健康 公平 卫生 公平
背景情况:
- 生成型人工智能 (gAI) 模型可以创建新的图像,但可能会延续训练数据中存在的历史偏见.
- 妇女在医学领域的代表性仍然不足,男性医生刻板印象仍然存在.
- 这项研究调查了各种医学领域内gAI输出中的隐性偏差.
研究的目的:
- 评估医学专业的生成人工智能 (gAI) 中隐含的性别和种族偏见.
- 将gAI生成的医生人口统计数据与当前的美国医生劳动力和居民数据进行比较.
主要方法:
- 使用DALL-E 2生成每个专业的100张图像,提示如"一位美国人[专业名称]".
- 通过医疗住院人员的共识评估了1900个gAI图像的感知性别和种族.
- 将gAI的人口分布与美国医学院协会 (AAMC) 数据进行了比较,采用千平方分析.
主要成果:
- 与医生劳动力相比,gAI在7/19种专业中代表女性过多,在6/19种专业中代表女性不足.
- 在内部医学 (18%),家庭医学 (18%) 和儿科 (27%) 观察到,女性在gAI方面的表现明显不足.
- 与现有居民相比,gAI在12/19专业中的女性代表性较低,并且在17/19专业的角色中女性代表性低于50%.
结论:
- 生成型人工智能模型产生了与居民和活跃医生劳动力相比,女性代表性不足的医生人口统计数据.
- 迫切需要对gAI数据集进行再培训,以准确地反映当前和未来医生劳动力的多样性.
- 解决gAI中的偏见对于促进医学教育和实践中的公平至关重要.
更多相关视频
03:14Augmenting Large Language Models via Vector Embeddings to Improve Domain-Specific Responsiveness
Published on: December 6, 2024
692
05:47Evidence-based Knowledge Synthesis and Hypothesis Validation: Navigating Biomedical Knowledge Bases via Explainable AI and Agentic Systems
Published on: June 13, 2025
592
相关概念视频
What is Genetic Engineering?
Overview
Stereotypes, Prejudice, and Discrimination
Humans are very diverse and although we share many similarities, we also have many differences. The social groups we belong to help form our identities (Tajfel, 1974). These differences may be difficult for some people to reconcile, which may lead to prejudice toward people who are different. Prejudice is a negative attitude and feeling toward an individual based solely on one’s membership in a particular social group (Allport, 1954; Brown, 2010). Prejudice is common against people who are...
Stereotype Content Model
The Stereotype Content Model (SCM) was first proposed by Susan Fiske and her colleagues (Fiske, Cuddy, Glick & Xu, 2002; see also Fiske, 2012 and Fiske, 2017). The SCM specifies that when someone encounters a new group, they will stereotype them based on two metrics: warmth—or that group’s perceived intent, and how likely they are to provide help or inflict harm—and competence—or their ability to carry out that objective. Depending on the warmth-competence categorization, a person will feel...
Genomic Imprinting and Inheritance
Diploid organisms inherit genetic material through chromosomes from both parents. Copies of the same gene are known as alleles. In most cases, both alleles are simultaneously expressed and allow various cellular processes to function optimally. If one of the alleles is missing or mutated, the expression of the other allele can compensate; however, this is not true for all genes.
The expression of some genes depends on which parent passed the gene to the offspring, through a phenomenon known as...
The expression of some genes depends on which parent passed the gene to the offspring, through a phenomenon known as...
Bias
Bias refers to any tendency that prevents a question from being considered unprejudiced. In research, bias occurs when one outcome or answer is selected or encouraged over others in sampling or testing. Bias can occur during any research phase, including study design, data collection, analysis, and publication.
In statistics, a sampling bias is created when a sample is collected from a population, and some members of the population are not as likely to be chosen as others (remember, each member...
In statistics, a sampling bias is created when a sample is collected from a population, and some members of the population are not as likely to be chosen as others (remember, each member...
Bias in Epidemiological Studies
Biases can arise at various stages of research, from study design and data collection to analysis and interpretation. Recognizing and addressing these biases is essential to ensure the validity and reliability of epidemiological findings.Broadly speaking, biases in epidemiology fall into three main categories: selection bias, information bias, and confounding. A more detailed description of possible biases is:
