Equitable modelling of brain imaging by counterfactual augmentation with morphologically constrained 3D deep generative models

TL;DR

This paper introduces Countersynth, a 3D deep generative model that creates biologically plausible, label-driven brain image modifications to improve fairness and accuracy in medical imaging tasks, especially under data imbalance.

Contribution

The paper presents Countersynth, a novel morphologically constrained 3D generative model for synthesizing counterfactual brain images to enhance equitable discriminative modeling.

Findings

Achieves state-of-the-art fidelity in synthetic brain image generation.

Improves model performance and fairness across demographic groups.

Outperforms existing augmentation methods in benchmarks.

Abstract

We describe Countersynth, a conditional generative model of diffeomorphic deformations that induce label-driven, biologically plausible changes in volumetric brain images. The model is intended to synthesise counterfactual training data augmentations for downstream discriminative modelling tasks where fidelity is limited by data imbalance, distributional instability, confounding, or underspecification, and exhibits inequitable performance across distinct subpopulations. Focusing on demographic attributes, we evaluate the quality of synthesized counterfactuals with voxel-based morphometry, classification and regression of the conditioning attributes, and the Fr\'{e}chet inception distance. Examining downstream discriminative performance in the context of engineered demographic imbalance and confounding, we use UK Biobank magnetic resonance imaging data to benchmark CounterSynth augmentation against current solutions to these problems. We achieve state-of-the-art improvements, both in overall fidelity and equity. The source code for CounterSynth is available online.