Causal methods improve fairness and accuracy in medical image augmentation
To do($x$) or not to do($x$): Medical Image Counterfactuals for Dataset Augmentation
Machine LearningArtificial Intelligence
Summary
Medical image analysis can suffer because the training data isn't balanced, causing biased results. The authors studied different ways to create synthetic medical images to add to training data and reduce bias. They compared three methods that change images in different ways, including a causal method that models cause-and-effect relationships. Their experiments showed that using the causal approach leads to better and fairer medical image models. This helps guide how to generate synthetic medical images effectively.
What this means in practice
- •For hospital data teams: Generate synthetic medical images with causal editing to improve diagnostic model fairness across patient subgroups.
- •For medical imaging tool developers: Incorporate causally aware synthetic data generation into training pipelines to enhance model robustness and reduce bias.
Authors
Yasin Ibrahim, Robin J. Evans, Konstantinos Kamnitsas
Abstract
Medical image analysis is often hindered by biased datasets, which can lead to biased models and limited clinical applicability. A promising strategy for mitigating such biases is to augment training data with synthetic images. Counterfactual (CF) generation is one such strategy, though the term is used in two different senses: in some works, CFs are produced through causality-based interventions derived from structural causal models, whereas in others, they are produced by non-causal image edits or conventional conditional generative models, such as altering anatomy or adding pathologies. In this work, we study this distinction and evaluate its practical consequences for medical image augmentation. We compare three conditioning strategies: $\textit{Deterministic}$, which changes selected variables while holding the remaining variables fixed; $\textit{Undirected}$, which updates variables according to learned statistical associations without assigning causal directions; and $\textit{Causal}$, which propagates interventions along a directed causal graph. We analyse how these choices affect the resulting images, and explore when causally grounded methods improve dataset augmentation or bring limited benefit. In particular, we assess downstream performance and fairness, where fairness refers to reduced sensitivity to dataset biases across sensitive subgroups. Our experiments demonstrate that using a causal approach to synthetic training data generation can lead to tangible benefits, with these insights offering valuable guidance to machine learning practitioners for the effective design of data generation protocols.