Structure-Enhanced Unsupervised Domain Adaptation for CT Whole-Brain Segmentation

Early and accurate identification of intracranial hemorrhage (ICH) is crucial for treatment, but the inherently low-contrast resolution of computed tomography (CT) imaging poses challenges in identification of specific cerebral regions, impacting effective and timely clinical decision-making. We propose brain structure-enhanced domain adaptation (BraSEDA), a CT-based unsupervised domain adaptation (UDA) model designed to assist in the identification of brain regions. BraSEDA framework utilizes a cross-modal instance normalization (CMIN) module for enhancing CT image structural features and creating high-quality pseudo magnetic resonance (MR) images. A multilevel CMIN architecture is also introduced for further improvement. The BraSEDA framework improved the quality of pseudo MR images in head CT to MR domain adaptation task, as reflected by the lowest-Fréchet inception distance scores $95.0\pm 12.1$ (p-value < 0.001) with and highest-BC scores $0.915\pm 0.396$ (p-value <0.01), compared to other UDA methods. It also demonstrated superior segmentation performance in other four traditional medical UDA tasks. The BraSEDA framework effectively addresses the challenge of low-contrast areas in CT brain images for ICH cases. It enables precise identification of low-contrast brain regions, making it a potential tool for assisting in the accurate and prompt initial diagnosis of ICH in emergency medical settings. The code and well-trained model has been publicly available: https://github.com/YixinChen-AI/BraSEDA.