Colormap augmentation: a novel method for cross-modality domain generalization

TL;DR

This paper introduces a new method called CmapAug to improve the generalization of deep learning models for medical image segmentation across different imaging modalities.

Contribution

The novel CmapAug method combines standard and color-based augmentations to address domain shift in cross-modality settings.

Findings

CmapAug achieved a maximum Dice score of 83.2% in liver segmentation.

The method outperformed baseline models in cross-modality domain generalization.

Augmentation strategies effectively mitigate domain shift without requiring target domain data.

Abstract

Domain generalization plays a crucial role in analyzing medical images from diverse clinics, scanner vendors, and imaging modalities. Existing methods often require substantial computational resources to train a highly generalized segmentation network, presenting challenges in terms of both availability and cost. The goal of this work is to evaluate a novel, yet simple and effective method for enhancing the generalization of deep learning models in segmentation across varying modalities.

Eight augmentation methods will be applied individually to a source domain dataset in order to generalize deep learning models. These models will then be tested on completely unseen target domain datasets from a different imaging modality and compared against a lower baseline model. By leveraging standard augmentation techniques, extensive intensity augmentations, and carefully chosen color transformations, we aim to address the domain shift problem, particularly in the cross-modality setting.

Our novel CmapAug method, when combined with standard augmentation techniques, resulted in a substantial improvement in the Dice Score, outperforming the baseline. While the baseline struggled to segment the liver structure in some test cases, our selective combination of augmentation methods achieved Dice scores as high as 83.2%.

Our results highlight the general effectiveness of the tested augmentation methods in addressing domain generalization and mitigating the domain shift problem caused by differences in imaging modalities between the source and target domains. The proposed augmentation strategy offers a simple yet powerful solution to this challenge, with significant potential in clinical scenarios where annotated data from the target domain are limited or unavailable.

Figures