Application of hybrid attention guided dual-path residual learning in magnetic resonance imaging diagnosis_Journal of Biomedical Engineering

Authors：

ZHU Weipeng ¹ ,  CAI Xianfa ^1,2 , LIU Yong ³

1. College of Pharmaceutical Information Engineering, Guangdong Pharmaceutical University, Guangzhou 510006, P. R. China;
2. Imaging Department, The First Affiliated Hospital of Guangdong Pharmaceutical University, Guangzhou 510006, P. R. China;
3. School of Artificial Intelligence, Guangxi University for Nationalities, Nanning 530006, P. R. China;

Corresponding?author：

CAI Xianfa, Email: cxianfa@126.com

Keywords：

Multimodal magnetic resonance imaging; Dual-path fusion; Residual learning; Attention mechanism

DOI：

10.7507/1001-5515.202510033

Video：

Export PDF Favorites Scan Get Citation

Abstract Full text Figures/Tables Video References Cited by

Multimodal magnetic resonance imaging (MRI) often faces challenges such as insufficient model generalization ability and high misdiagnosis rates in intelligent neuroimaging diagnosis. Existing methods, including single-path classification networks such as residual networks and EfficientNet, and segmentation-oriented workflow models such as U-Net, have the following limitations when processing multi-sequence MRI data: ① single-path structures have difficulty in effectively decoupling modality-specific features; ② conventional convolutions lack adaptive enhancement in the channel and spatial dimensions. Therefore, this paper proposes a dual-path residual channel-spatial attention network (DRCSA-Net). The model learns complementary representations through a parallel dual-branch structure (channel enhancement and spatial enhancement), introduces a squeeze-and-excitation module and a channel-spatial attention module to achieve channel recalibration and spatial attention focusing, and finally integrates information and completes classification through lightweight fusion convolution and fully connected layers. To comprehensively evaluate the proposed model, five-fold cross-validation, single-/dual-path comparison, ablation experiments, and robustness analysis were conducted on four public datasets. The experimental results show that DRCSA-Net achieves high accuracy on all four public datasets and demonstrates good effectiveness and stability in the single-/dual-path comparison, ablation experiments, and robustness analysis, providing a structurally clear and stable implementation scheme for brain tumor MRI classification.

Citation： ZHU Weipeng, CAI Xianfa, LIU Yong. Application of hybrid attention guided dual-path residual learning in magnetic resonance imaging diagnosis. Journal of Biomedical Engineering, 2026, 43(4): 828-835. doi: 10.7507/1001-5515.202510033 Copy

1.
2.
3.	Mohsen S, Oraby S, Abdel-Aziz M. Deep learning and machine learning for brain tumor detection: a review, challenges, and future directions[J]. Arch Comput Methods Eng, 2025. DOI: 10.1007/s11831-025-10416-3.
4.
5.
6.	Xie S, Girshick R, Dollár P, et al. Aggregated residual transformations for deep neural networks[C]//2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), Honolulu: IEEE, 2017: 1492-1500.
7.	Tan M, Le Q V. EfficientNet: rethinking model scaling for convolutional neural networks[C]//Proceedings of the 36th International Conference on Machine Learning (ICML), Long Beach: PMLR, 2019: 6105-6114.
8.	Ronneberger O, Fischer P, Brox T. U-Net: convolutional networks for biomedical image segmentation[C]//Medical Image Computing and Computer-Assisted Intervention-MICCAI 2015, Munich: Springer, 2015: 234-241.
9.	Selvaraju R R, Cogswell M, Das A, et al. Grad-CAM: visual explanations from deep networks via gradient-based localization[C]//2017 IEEE International Conference on Computer Vision (ICCV), Venice: IEEE, 2017: 618-626.
10.
11.
12.
13.
14.	Cheng J. Brain tumor dataset[DS/OL]. V5. Figshare (2017-04-02) [2026-07-11]. https://figshare.com/articles/dataset/brain_tumor_dataset/1512427.
15.	Nickparvar M. Brain tumor MRI dataset[DS/OL]. Kaggle (2021) [2026-07-11]. https://www.kaggle.com/datasets/masoudnickparvar/brain-tumor-mri-dataset.
16.	Bhuvaji S, Kadam A, Bhumkar P, et al. Brain tumor classification (MRI)[DS/OL]. Kaggle (2025) [2026-07-11]. https://www.kaggle.com/dsv/12745533.
17.	Hashemi S M H. Crystal Clean: brain tumors MRI dataset[DS/OL]. Kaggle (2023) [2026-07-11]. https://www.kaggle.com/datasets/mohammadhossein77/brain-tumors-dataset.
18.	Ioffe S, Szegedy C. Batch normalization: accelerating deep network training by reducing internal covariate shift[C]//Proceedings of the 32nd International Conference on Machine Learning, Lille: PMLR, 2015, 37: 448-456.
19.	Ramachandran P, Zoph B, Le Q V. Searching for activation functions[C/OL]//Proceedings of the 6th International Conference on Learning Representations: Workshop Track, Vancouver: ICLR, 2018. https://openreview.net/forum?id=Hkuq2EkPf.
20.	Hu J, Shen L, Sun G. Squeeze-and-excitation networks[C]//2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), Salt Lake City: IEEE, 2018: 7132-7141.
21.	Woo S, Park J, Lee J Y, et al. CBAM: convolutional block attention module[C]//Ferrari V, Hebert M, Sminchisescu C, et al, eds. Computer Vision-ECCV 2018. Munich: Springer, 2018: 3-19.
22.	He K, Zhang X, Ren S, et al. Deep residual learning for image recognition[C]//Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR), Las Vegas: IEEE, 2016: 770-778.
23.
24.	PyTorch Contributors. ReduceLROnPlateau: PyTorch 2.5 documentation[EB/OL]. [2026-07-11]. https://docs.pytorch.org/docs/2.5/generated/torch.optim.lr_scheduler.ReduceLROnPlateau.html.
25.
26.
27.	Sobhaninia Z, Karimi N, Khadivi P, et al. Brain tumor classification by cascaded multiscale multitask learning framework based on feature aggregation[PP/OL]. v1. arXiv (2021-12-28) [2026-07-11]. https://doi.org/10.48550/arXiv.2112.14320.
28.
29.
30.
31.	Kline T L, Kitamura F, Warren D, et al. Best practices and checklist for reviewing artificial intelligence-based medical imaging papers: classification[J]. J Imaging Inform Med, 2025: 1-11.
32.
33.
34.
35.
36.
37.
38.
39.

1.
2.
3. Mohsen S, Oraby S, Abdel-Aziz M. Deep learning and machine learning for brain tumor detection: a review, challenges, and future directions[J]. Arch Comput Methods Eng, 2025. DOI: 10.1007/s11831-025-10416-3.
4.
5.
6. Xie S, Girshick R, Dollár P, et al. Aggregated residual transformations for deep neural networks[C]//2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), Honolulu: IEEE, 2017: 1492-1500.
7. Tan M, Le Q V. EfficientNet: rethinking model scaling for convolutional neural networks[C]//Proceedings of the 36th International Conference on Machine Learning (ICML), Long Beach: PMLR, 2019: 6105-6114.
8. Ronneberger O, Fischer P, Brox T. U-Net: convolutional networks for biomedical image segmentation[C]//Medical Image Computing and Computer-Assisted Intervention-MICCAI 2015, Munich: Springer, 2015: 234-241.
9. Selvaraju R R, Cogswell M, Das A, et al. Grad-CAM: visual explanations from deep networks via gradient-based localization[C]//2017 IEEE International Conference on Computer Vision (ICCV), Venice: IEEE, 2017: 618-626.
10.
11.
12.
13.
14. Cheng J. Brain tumor dataset[DS/OL]. V5. Figshare (2017-04-02) [2026-07-11]. https://figshare.com/articles/dataset/brain_tumor_dataset/1512427.
15. Nickparvar M. Brain tumor MRI dataset[DS/OL]. Kaggle (2021) [2026-07-11]. https://www.kaggle.com/datasets/masoudnickparvar/brain-tumor-mri-dataset.
16. Bhuvaji S, Kadam A, Bhumkar P, et al. Brain tumor classification (MRI)[DS/OL]. Kaggle (2025) [2026-07-11]. https://www.kaggle.com/dsv/12745533.
17. Hashemi S M H. Crystal Clean: brain tumors MRI dataset[DS/OL]. Kaggle (2023) [2026-07-11]. https://www.kaggle.com/datasets/mohammadhossein77/brain-tumors-dataset.
18. Ioffe S, Szegedy C. Batch normalization: accelerating deep network training by reducing internal covariate shift[C]//Proceedings of the 32nd International Conference on Machine Learning, Lille: PMLR, 2015, 37: 448-456.
19. Ramachandran P, Zoph B, Le Q V. Searching for activation functions[C/OL]//Proceedings of the 6th International Conference on Learning Representations: Workshop Track, Vancouver: ICLR, 2018. https://openreview.net/forum?id=Hkuq2EkPf.
20. Hu J, Shen L, Sun G. Squeeze-and-excitation networks[C]//2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), Salt Lake City: IEEE, 2018: 7132-7141.
21. Woo S, Park J, Lee J Y, et al. CBAM: convolutional block attention module[C]//Ferrari V, Hebert M, Sminchisescu C, et al, eds. Computer Vision-ECCV 2018. Munich: Springer, 2018: 3-19.
22. He K, Zhang X, Ren S, et al. Deep residual learning for image recognition[C]//Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR), Las Vegas: IEEE, 2016: 770-778.
23.
24. PyTorch Contributors. ReduceLROnPlateau: PyTorch 2.5 documentation[EB/OL]. [2026-07-11]. https://docs.pytorch.org/docs/2.5/generated/torch.optim.lr_scheduler.ReduceLROnPlateau.html.
25.
26.
27. Sobhaninia Z, Karimi N, Khadivi P, et al. Brain tumor classification by cascaded multiscale multitask learning framework based on feature aggregation[PP/OL]. v1. arXiv (2021-12-28) [2026-07-11]. https://doi.org/10.48550/arXiv.2112.14320.
28.
29.
30.
31. Kline T L, Kitamura F, Warren D, et al. Best practices and checklist for reviewing artificial intelligence-based medical imaging papers: classification[J]. J Imaging Inform Med, 2025: 1-11.
32.
33.
34.
35.
36.
37.
38.
39.

Journal of Biomedical Engineering

Application of hybrid attention guided dual-path residual learning in magnetic resonance imaging diagnosis

Abstract Full text Figures/Tables Video References Cited by

Previous Article

Next Article

Format

Content