DyGLNet: Hybrid global-local feature fusion with dynamic upsampling for medical image segmentation

doi:10.1016/j.patcog.2025.112792

CSpace

	DyGLNet: Hybrid global-local feature fusion with dynamic upsampling for medical image segmentation
	Zhao, Yican 1; Wang, Ce 2; Hao, You 3; Li, Lei 1; Liao, Tianli 1
	2026-05-01
发表期刊	PATTERN RECOGNITION
ISSN	0031-3203
卷号	173 页码:12
摘要	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 lightweight dynamic adaptive upsampling module (DyFusionUp) to realize highfidelity reconstruction of feature maps based on learnable offsets and 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 is available at https://github.com/YeeCan-Zhao/DyGLNet.
关键词	Medical image segmentation Feature fusion Dynamic upsampling Multi-scale
DOI	10.1016/j.patcog.2025.112792
收录类别	SCI
语种	英语
WOS研究方向	Computer Science ; Engineering
WOS类目	Computer Science, Artificial Intelligence ; Engineering, Electrical & Electronic
WOS记录号	WOS:001634957300001
出版者	ELSEVIER SCI LTD
引用统计
文献类型	期刊论文
条目标识符	http://119.78.100.204/handle/2XEOYT63/42928
专题	中国科学院计算技术研究所
通讯作者	Liao, Tianli
作者单位	1.Henan Univ Technol, Coll Informat Sci & Engn, Zhengzhou 450001, Peoples R China 2.Sun Yat Sen Univ, Sch Sci, Shenzhen 518107, Guangdong, Peoples R China 3.Chinese Acad Sci, Inst Comp Technol, Key Lab Intelligent Informat Proc, Beijing 100190, Peoples R China
推荐引用方式 GB/T 7714	Zhao, Yican,Wang, Ce,Hao, You,et al. DyGLNet: Hybrid global-local feature fusion with dynamic upsampling for medical image segmentation[J]. PATTERN RECOGNITION,2026,173:12.
APA	Zhao, Yican,Wang, Ce,Hao, You,Li, Lei,&Liao, Tianli.(2026).DyGLNet: Hybrid global-local feature fusion with dynamic upsampling for medical image segmentation.PATTERN RECOGNITION,173,12.
MLA	Zhao, Yican,et al."DyGLNet: Hybrid global-local feature fusion with dynamic upsampling for medical image segmentation".PATTERN RECOGNITION 173(2026):12.