Modality-Specific Enhancement and Complementary Fusion for Semi-Supervised Multi-Modal Brain Tumor Segmentation
This work addresses the challenge of exploiting complementary information across MRI modalities for brain tumor segmentation, which is incremental but improves robustness in medical imaging with scarce supervision.
The paper tackled the problem of semi-supervised multi-modal brain tumor segmentation by proposing a framework with modality-specific enhancement and complementary fusion, achieving significant improvements in Dice and Sensitivity scores on the BraTS 2019 dataset under limited labeled data settings.
Semi-supervised learning (SSL) has become a promising direction for medical image segmentation, enabling models to learn from limited labeled data alongside abundant unlabeled samples. However, existing SSL approaches for multi-modal medical imaging often struggle to exploit the complementary information between modalities due to semantic discrepancies and misalignment across MRI sequences. To address this, we propose a novel semi-supervised multi-modal framework that explicitly enhances modality-specific representations and facilitates adaptive cross-modal information fusion. Specifically, we introduce a Modality-specific Enhancing Module (MEM) to strengthen semantic cues unique to each modality via channel-wise attention, and a learnable Complementary Information Fusion (CIF) module to adaptively exchange complementary knowledge between modalities. The overall framework is optimized using a hybrid objective combining supervised segmentation loss and cross-modal consistency regularization on unlabeled data. Extensive experiments on the BraTS 2019 (HGG subset) demonstrate that our method consistently outperforms strong semi-supervised and multi-modal baselines under 1\%, 5\%, and 10\% labeled data settings, achieving significant improvements in both Dice and Sensitivity scores. Ablation studies further confirm the complementary effects of our proposed MEM and CIF in bridging cross-modality discrepancies and improving segmentation robustness under scarce supervision.