Unsupervised Cross-Modality Adaptation for Medical Image Segmentation via Image and Feature Alignment

This study proposes a novel method for unsupervised cross-modality adaptation in medical image segmentation, utilizing image and feature alignment. Specifically, the approach employs a deeply synergistic strategy between image-level and feature-level alignments to optimize the similarity between two modalities. The effectiveness of the proposed method is validated through experiments on different datasets. Overall, the results demonstrate that our method outperforms existing approaches in terms of segmentation accuracy and robustness across different modalities.