A fused attention-based hybrid model for semi-supervised medical image segmentation

Addressing the significant challenge of extensive data annotation in medical image segmentation, semi-supervised techniques have become an effective alternative for leveraging both labeled and unlabeled data. However, accurately capturing complex local structures and global contextual information remains difficult, often leading to inconsistent segmentation results. To mitigate these limitations, we propose a fused transformer-based hybrid architecture for semi-supervised medical image segmentation. The model integrates a parallel backbone comprising Deformation Convolution Blocks (DCB) and Fused Transformer Blocks (FTB), followed by a segmentation head for precise mask prediction. The DCB enhances spatial adaptability for irregular and artifact-rich regions, while the FTB—with its fused attention mechanism—captures long-range dependencies efficiently. A Ghost Layer Perceptron (GLP) embedded within the transformer further improves computational efficiency without compromising representation quality. In addition, the incorporation of a consistency loss and unsupervised contrastive learning facilitates robust feature discrimination on unlabeled data, improving generalization across modalities. Extensive experiments on four publicly available medical imaging datasets demonstrate that the proposed model achieves comparable or better accuracy than recent state-of-the-art methods, while requiring substantially fewer parameters and lower computational cost, underscoring its practicality for real-world clinical applications. • A novel fused attention (FA) based hybrid model is introduced that integrates a dual-branch backbone (Deformation CNN and Fused Transformer) along with a head for semi-supervised medical image segmentation. • The deformation convolution block (DCB) is developed to extract the small lesions and handle complex spatial features in medical images. • The Fused Transformer Block (FTB) is proposed with a FA that integrates cross-channel attention with cross-dimensional attention. It enhances the relationship between segmented features and their backgrounds by efficiently incorporating long-range dependencies across several channels and dimensions, hence capturing long-distance correlation characteristics. • The model employs a combination of the consistency loss, contrastive loss and supervised loss to ensure consistent predictions and incorporating unsupervised contrastive learning.