Unsupervised Search for Ethnic Minorities' Medical Segmentation Training Set

TL;DR

This paper addresses dataset bias in medical imaging caused by demographic imbalances, proposing a greedy search strategy to select training data that improves segmentation accuracy for minority groups, promoting equitable healthcare.

Contribution

It introduces a novel training set search method that reduces racial bias in medical segmentation datasets by selecting more representative training samples.

Findings

Improved segmentation accuracy for minority groups.

Effective bias mitigation through targeted data selection.

Potential for more equitable clinical outcomes.

Abstract

This article investigates the critical issue of dataset bias in medical imaging, with a particular emphasis on racial disparities caused by uneven population distribution in dataset collection. Our analysis reveals that medical segmentation datasets are significantly biased, primarily influenced by the demographic composition of their collection sites. For instance, Scanning Laser Ophthalmoscopy (SLO) fundus datasets collected in the United States predominantly feature images of White individuals, with minority racial groups underrepresented. This imbalance can result in biased model performance and inequitable clinical outcomes, particularly for minority populations. To address this challenge, we propose a novel training set search strategy aimed at reducing these biases by focusing on underrepresented racial groups. Our approach utilizes existing datasets and employs a simple greedy algorithm to identify source images that closely match the target domain distribution. By selecting training data that aligns more closely with the characteristics of minority populations, our strategy improves the accuracy of medical segmentation models on specific minorities, i.e., Black. Our experimental results demonstrate the effectiveness of this approach in mitigating bias. We also discuss the broader societal implications, highlighting how addressing these disparities can contribute to more equitable healthcare outcomes.