Domain generalization in deep learning for contrast-enhanced imaging

TL;DR

This paper evaluates deep learning techniques for improving the generalization of contrast-enhanced image segmentation models across different clinical centers, demonstrating that single-center models can outperform multi-center models with proper methods.

Contribution

It systematically assesses various domain generalization techniques, showing that combining data augmentation and transfer learning enables single-center models to generalize effectively across centers.

Findings

Single-center models can match or surpass multi-center models with proper techniques.

Data augmentation and transfer learning improve cross-center generalization.

Methods reduce the need for extensive multi-center datasets.

Abstract

The domain generalization problem has been widely investigated in deep learning for non-contrast imaging over the last years, but it received limited attention for contrast-enhanced imaging. However, there are marked differences in contrast imaging protocols across clinical centers, in particular in the time between contrast injection and image acquisition, while access to multi-center contrast-enhanced image data is limited compared to available datasets for non-contrast imaging. This calls for new tools for generalizing single-domain, single-center deep learning models across new unseen domains and clinical centers in contrast-enhanced imaging. In this paper, we present an exhaustive evaluation of deep learning techniques to achieve generalizability to unseen clinical centers for contrast-enhanced image segmentation. To this end, several techniques are investigated, optimized and systematically evaluated, including data augmentation, domain mixing, transfer learning and domain adaptation. To demonstrate the potential of domain generalization for contrast-enhanced imaging, the methods are evaluated for ventricular segmentation in contrast-enhanced cardiac magnetic resonance imaging (MRI). The results are obtained based on a multi-center cardiac contrast-enhanced MRI dataset acquired in four hospitals located in three countries (France, Spain and China). They show that the combination of data augmentation and transfer learning can lead to single-center models that generalize well to new clinical centers not included during training. Single-domain neural networks enriched with suitable generalization procedures can reach and even surpass the performance of multi-center, multi-vendor models in contrast-enhanced imaging, hence eliminating the need for comprehensive multi-center datasets to train generalizable models.