Multi-scale conv-attention U-Net for medical image segmentation DOI Creative Commons
Linqiang Pan, Chengxue Zhang, Jingbo Sun

et al.

Scientific Reports, Journal Year: 2025, Volume and Issue: 15(1)

Published: April 8, 2025

U-Net-based network structures are widely used in medical image segmentation. However, effectively capturing multi-scale features and spatial context information of complex organizational remains a challenge. To address this, we propose novel structure based on the U-Net backbone. This model integrates Adaptive Convolution (AC) module, Multi-Scale Learning (MSL) Conv-Attention module to enhance feature expression ability segmentation performance. The AC dynamically adjusts convolutional kernel through an adaptive layer. enables extract different shapes scales adaptively, further improving its performance scenarios. MSL is designed for fusion. It aggregates fine-grained high-level semantic from resolutions, creating rich connections between encoding decoding processes. On other hand, incorporates efficient attention mechanism into skip connections. captures global using low-dimensional proxy high-dimensional data. approach reduces computational complexity while maintaining effective channel extraction. Experimental validation CVC-ClinicDB, MICCAI 2023 Tooth, ISIC2017 datasets demonstrates that our proposed MSCA-UNet significantly improves accuracy robustness. At same time, it lightweight outperforms existing methods.

Language: Английский

S-Net: A Novel Shallow Network for Enhanced Detail Retention in Medical Image Segmentation DOI
Qinghua Shang,

Guangshuo Wang,

Xiao Hua Wang

et al.

Computer Methods and Programs in Biomedicine, Journal Year: 2025, Volume and Issue: unknown, P. 108730 - 108730

Published: March 1, 2025

Language: Английский

Citations

0
MSHV-Net: A Multi-Scale Hybrid Vision Network for Skin Image Segmentation DOI
Haicheng Qu, Yi Gao,

Qingling Jiang

et al.

Digital Signal Processing, Journal Year: 2025, Volume and Issue: unknown, P. 105166 - 105166

Published: March 1, 2025

Language: Английский

Citations

0
Dual-Filter Cross Attention and Onion Pooling Network for Enhanced Few-Shot Medical Image Segmentation DOI Creative Commons
Lina Ni, Yang Liu, Zekun Zhang

et al.

Sensors, Journal Year: 2025, Volume and Issue: 25(7), P. 2176 - 2176

Published: March 29, 2025

Few-shot learning has demonstrated remarkable performance in medical image segmentation. However, existing few-shot segmentation (FSMIS) models often struggle to fully utilize query information, leading prototype bias and limited generalization ability. To address these issues, we propose the dual-filter cross attention onion pooling network (DCOP-Net) for FSMIS. DCOP-Net consists of a stage stage. During stage, introduce (DFCA) module avoid entanglement between background features support foreground features, effectively integrating into prototypes. Additionally, design an (OP) that combines eroding mask operations with masked average generate multiple prototypes, preserving contextual information mitigating bias. In present parallel threshold perception (PTP) robust thresholds differentiation self-reference regularization (QSR) strategy enhance model accuracy consistency. Extensive experiments on three publicly available datasets demonstrate outperforms state-of-the-art methods, exhibiting superior capabilities.

Language: Английский

Citations

0
Grouped multi-scale vision transformer for medical image segmentation DOI Creative Commons
Zexuan Ji,

Zheng Chen,

Xiao Ma

et al.

Scientific Reports, Journal Year: 2025, Volume and Issue: 15(1)

Published: April 1, 2025

Abstract Medical image segmentation plays a pivotal role in clinical diagnosis and pathological research by delineating regions of interest within medical images. While early approaches based on Convolutional Neural Networks (CNNs) have achieved significant success, their limited receptive field constrains ability to capture long-range dependencies. Recent advances Vision Transformers (ViTs) demonstrated remarkable improvements leveraging self-attention mechanisms. However, existing ViT-based models often struggle effectively multi-scale variations single attention layer, limiting capacity model complex anatomical structures. To address this limitation, we propose Grouped Multi-Scale Attention (GMSA), which enhances feature representation grouping channels performing at different scales layer. Additionally, introduce Inter-Scale (ISA) facilitate cross-scale fusion, further improving performance. Extensive experiments the Synapse, ACDC, ISIC2018 datasets demonstrate effectiveness our model, achieving state-of-the-art results segmentation. Our code is available at: https://github.com/Chen2zheng/ScaleFormer .

Language: Английский

Citations

0
Multi-scale conv-attention U-Net for medical image segmentation DOI Creative Commons
Linqiang Pan, Chengxue Zhang, Jingbo Sun

et al.

Scientific Reports, Journal Year: 2025, Volume and Issue: 15(1)

Published: April 8, 2025

U-Net-based network structures are widely used in medical image segmentation. However, effectively capturing multi-scale features and spatial context information of complex organizational remains a challenge. To address this, we propose novel structure based on the U-Net backbone. This model integrates Adaptive Convolution (AC) module, Multi-Scale Learning (MSL) Conv-Attention module to enhance feature expression ability segmentation performance. The AC dynamically adjusts convolutional kernel through an adaptive layer. enables extract different shapes scales adaptively, further improving its performance scenarios. MSL is designed for fusion. It aggregates fine-grained high-level semantic from resolutions, creating rich connections between encoding decoding processes. On other hand, incorporates efficient attention mechanism into skip connections. captures global using low-dimensional proxy high-dimensional data. approach reduces computational complexity while maintaining effective channel extraction. Experimental validation CVC-ClinicDB, MICCAI 2023 Tooth, ISIC2017 datasets demonstrates that our proposed MSCA-UNet significantly improves accuracy robustness. At same time, it lightweight outperforms existing methods.

Language: Английский

Citations

0