DyGLNet: Hybrid Global-Local Feature Fusion with Dynamic Upsampling for Medical Image Segmentation

2025-09-16

AI Summary

Medical image segmentation faces three key challenges: multi-scale lesions, ambiguous boundaries, and high computational overhead. To address these, we propose a lightweight dynamic upsampling segmentation network. Our method introduces (1) the SHDCBlock, which jointly models local details and global context; (2) DyFusionUp, a dynamic upsampling module that achieves high-fidelity feature reconstruction via learnable deformable offsets; and (3) an end-to-end lightweight architecture integrating single-head self-attention, multi-scale atrous convolutions, and adaptive upsampling. Evaluated on seven public benchmarks, our approach significantly improves boundary accuracy and small-object segmentation performance while reducing computational complexity—achieving an average 32% reduction in FLOPs. The proposed design balances accuracy and efficiency, demonstrating strong potential for clinical deployment.

Technology Category

• Computer Vision: Segmentation
• Machine Learning: Learning on the Edge & Model Compression
• Search and Optimization: Learning to Search

Application Category

• Graph Algorithms and Modeling for the Web: Efficient manipulation of static and dynamic Web-related graphs
• Search and Retrieval-Augmented AI: Efficiency and scalability of Web search engines
• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications

Abstract

Medical image segmentation grapples with challenges including multi-scale lesion variability, ill-defined tissue boundaries, and computationally intensive processing demands. This paper proposes the DyGLNet, which achieves efficient and accurate segmentation by fusing global and local features with a dynamic upsampling mechanism. The model innovatively designs a hybrid feature extraction module (SHDCBlock), combining single-head self-attention and multi-scale dilated convolutions to model local details and global context collaboratively. We further introduce a dynamic adaptive upsampling module (DyFusionUp) to realize high-fidelity reconstruction of feature maps based on learnable offsets. Then, a lightweight design is adopted to reduce computational overhead. Experiments on seven public datasets demonstrate that DyGLNet outperforms existing methods, particularly excelling in boundary accuracy and small-object segmentation. Meanwhile, it exhibits lower computation complexity, enabling an efficient and reliable solution for clinical medical image analysis. The code will be made available soon.

Problem

• Addresses multi-scale lesion variability in medical images
• Handles ill-defined tissue boundaries for accurate segmentation
• Reduces computational overhead with lightweight efficient design

Innovation

• Hybrid global-local feature fusion
• Dynamic adaptive upsampling mechanism
• Lightweight computational design