Characterizing neuroimaging 'brain-behavior' model performance bias in rural populations

项目摘要 近五分之一的美国人口居住在农村地区，其中大约五分之一的人居住在农村地区。 居民患有精神疾病。虽然这些精神疾病的发病率与城市地区相似， 生活在农村地区的人面临着不成比例的负面精神病后果负担。现代进步， 精神病学研究集中在使用机器学习和人类神经成像来预测诊断 和治疗结果。然而，最近的证据表明，机器学习模型本身可能 通过绩效偏见推动健康差距。具体而言，在2008年创建的临床决策模型 大多数人口可能在创建期间代表性不足的人口中表现不佳 该模型（例如，如果患者在农村，选择正确治疗的可能性较小）。鉴于几乎所有 精神病学研究中的神经成像“大脑行为”预测模型是从收集的数据中产生的， 在人口密集的大都市地区，这项研究将评估“大脑行为”模型在农村的表现偏差 人口。它还将调查消除这种偏见的方法，这种偏见在农村地区造成了进一步的健康差距。 人口。在目标1中，我将使用来自青少年大脑和认知研究中的9，811名个体的神经成像数据。 发展研究，以创建一个“大脑行为”的认知预测模型。在目标2中，我将评估这个模型 为城乡绩效偏差，并追求战略，以减少模型偏差。这项研究将具有重要意义 理解医疗保健中的算法如何驱动健康差异以及我们如何减少 通过设计在代表性不足的人群中公平表现的模型来消除这些差异。

PROJECT SUMMARY Nearly one-fifth of the Unites States population resides in a rural region, and approximately one-fifth of those residents suffers from a mental illness. While these rates of mental illness are similar to urban areas, individuals living in rural regions face a disproportionate burden of negative psychiatric outcomes. Modern advances in psychiatric research have focused on using machine learning and human neuroimaging to predict diagnoses and treatment outcomes. However, recent evidence suggests that machine learning models themselves may drive health disparities through performance bias. Specifically, clinical decision-making models created in majority populations may not perform as well in populations that were underrepresented during the creation of the model (e.g., poorer likelihood of choosing the correct treatment if patients are rural). Given that virtually all neuroimaging ‘brain-behavior’ predictive models in psychiatry research are generated from data collected in highly populated metropolitan areas, this study will evaluate ‘brain-behavior’ models for performance bias in rural populations. It will also investigate means of eliminating this bias that creates further health disparities in rural populations. In Aim 1, I will use neuroimaging data from 9,811 individuals in the Adolescent Brain and Cognitive Development Study to create a ‘brain-behavior’ predictive model of cognition. In Aim 2, I will evaluate this model for urban-rural performance bias and pursue strategies to reduce model bias. This study will have important implications for understanding how algorithms in healthcare drive health disparities and how we can reduce these disparities by designing models that perform equitably within underrepresented populations.