中国物理B ›› 2025, Vol. 34 ›› Issue (8): 80703-080703.doi: 10.1088/1674-1056/adce96

• • 上一篇    下一篇

M2ANet: Multi-branch and multi-scale attention network for medical image segmentation

Wei Xue(薛伟)1,†, Chuanghui Chen(陈创慧)1, Xuan Qi(戚轩)2,3, Jian Qin(秦健)1, Zhen Tang(唐振)1, and Yongsheng He(何永胜)2,3   

  1. 1 School of Computer Science and Technology, Anhui University of Technology, Maanshan 243032, China;
    2 Department of Radiology, Maanshan People's Hospital, Maanshan 243099, China;
    3 Maanshan Key Laboratory for Medical Image Modeling and Intelligent Analysis, Maanshan 243099, China
  • 收稿日期:2025-01-22 修回日期:2025-03-19 接受日期:2025-04-21 出版日期:2025-07-17 发布日期:2025-08-05
  • 通讯作者: Wei Xue E-mail:xuewei@ahut.edu.cn
  • 基金资助:
    Project supported by the Natural Science Foundation of the Anhui Higher Education Institutions of China (Grant Nos. 2023AH040149 and 2024AH051915), the Anhui Provincial Natural Science Foundation (Grant No. 2208085MF168), the Science and Technology Innovation Tackle Plan Project of Maanshan (Grant No. 2024RGZN001), and the Scientific Research Fund Project of Anhui Medical University (Grant No. 2023xkj122).

M2ANet: Multi-branch and multi-scale attention network for medical image segmentation

Wei Xue(薛伟)1,†, Chuanghui Chen(陈创慧)1, Xuan Qi(戚轩)2,3, Jian Qin(秦健)1, Zhen Tang(唐振)1, and Yongsheng He(何永胜)2,3   

  1. 1 School of Computer Science and Technology, Anhui University of Technology, Maanshan 243032, China;
    2 Department of Radiology, Maanshan People's Hospital, Maanshan 243099, China;
    3 Maanshan Key Laboratory for Medical Image Modeling and Intelligent Analysis, Maanshan 243099, China
  • Received:2025-01-22 Revised:2025-03-19 Accepted:2025-04-21 Online:2025-07-17 Published:2025-08-05
  • Contact: Wei Xue E-mail:xuewei@ahut.edu.cn
  • Supported by:
    Project supported by the Natural Science Foundation of the Anhui Higher Education Institutions of China (Grant Nos. 2023AH040149 and 2024AH051915), the Anhui Provincial Natural Science Foundation (Grant No. 2208085MF168), the Science and Technology Innovation Tackle Plan Project of Maanshan (Grant No. 2024RGZN001), and the Scientific Research Fund Project of Anhui Medical University (Grant No. 2023xkj122).

RichHTML

1

PDF (PC)

103

赞

摘要： Convolutional neural networks (CNNs)-based medical image segmentation technologies have been widely used in medical image segmentation because of their strong representation and generalization abilities. However, due to the inability to effectively capture global information from images, CNNs can easily lead to loss of contours and textures in segmentation results. Notice that the transformer model can effectively capture the properties of long-range dependencies in the image, and furthermore, combining the CNN and the transformer can effectively extract local details and global contextual features of the image. Motivated by this, we propose a multi-branch and multi-scale attention network (M2ANet) for medical image segmentation, whose architecture consists of three components. Specifically, in the first component, we construct an adaptive multi-branch patch module for parallel extraction of image features to reduce information loss caused by downsampling. In the second component, we apply residual block to the well-known convolutional block attention module to enhance the network's ability to recognize important features of images and alleviate the phenomenon of gradient vanishing. In the third component, we design a multi-scale feature fusion module, in which we adopt adaptive average pooling and position encoding to enhance contextual features, and then multi-head attention is introduced to further enrich feature representation. Finally, we validate the effectiveness and feasibility of the proposed M2ANet method through comparative experiments on four benchmark medical image segmentation datasets, particularly in the context of preserving contours and textures. The source code of M2ANet will be released at https://github.com/AHUT-MILAGroup/M2ANet.

关键词: medical image segmentation, convolutional neural network, multi-branch attention, multi-scale feature fusion

Abstract: Convolutional neural networks (CNNs)-based medical image segmentation technologies have been widely used in medical image segmentation because of their strong representation and generalization abilities. However, due to the inability to effectively capture global information from images, CNNs can easily lead to loss of contours and textures in segmentation results. Notice that the transformer model can effectively capture the properties of long-range dependencies in the image, and furthermore, combining the CNN and the transformer can effectively extract local details and global contextual features of the image. Motivated by this, we propose a multi-branch and multi-scale attention network (M2ANet) for medical image segmentation, whose architecture consists of three components. Specifically, in the first component, we construct an adaptive multi-branch patch module for parallel extraction of image features to reduce information loss caused by downsampling. In the second component, we apply residual block to the well-known convolutional block attention module to enhance the network's ability to recognize important features of images and alleviate the phenomenon of gradient vanishing. In the third component, we design a multi-scale feature fusion module, in which we adopt adaptive average pooling and position encoding to enhance contextual features, and then multi-head attention is introduced to further enrich feature representation. Finally, we validate the effectiveness and feasibility of the proposed M2ANet method through comparative experiments on four benchmark medical image segmentation datasets, particularly in the context of preserving contours and textures. The source code of M2ANet will be released at https://github.com/AHUT-MILAGroup/M2ANet.

Key words: medical image segmentation, convolutional neural network, multi-branch attention, multi-scale feature fusion

中图分类号:  (Image processing)

  • 07.05.Pj
42.30.Tz (Computer vision; robotic vision) 87.57.nm (Segmentation)