Self-supervised classification method for breast pathology images fusing geometric priors and multi-scale attention_Journal of Biomedical Engineering

Authors：

JIANG Liang ¹ , CAO Hui ¹ , NIE Yongqi ² ,  ZHANG Cheng ¹

1. School of Medical Information Engineering, Shandong University of Traditional Chinese Medicine, Jinan 250355, P. R. China;
2. School of Medical Imaging, Qilu Medical University, Zibo, Shandong 255000, P. R. China;

Corresponding?author：

ZHANG Cheng, Email: 13256715081@163.com

Keywords：

Breast pathology image; Self-supervised learning; Isotropic attention; Rotation equivariance; Multi-scale feature fusion

DOI：

10.7507/1001-5515.202511059

Video：

Export PDF Favorites Scan Get Citation

Abstract Full text Figures/Tables Video References Cited by

Breast whole slide image (WSI) serves as an essential basis for breast cancer diagnosis and subtype classification. However, its high resolution, multi-scale structural characteristics, and arbitrary orientation pose substantial challenges for automated analysis. Existing deep learning methods typically rely on large amounts of annotated data and struggle to handle the rotational variability and scale differences inherent in pathological images, which limits their generalization ability in cross-center settings. To address these issues, this study proposes a self-supervised learning model, IAM-BYOL, which incorporated geometric priors into the representation learning process. Building upon the BYOL framework, the model introduced an isotropic attention module (IAM). By employing discrete rotation group convolution, IAM enabled weight sharing across different rotated versions of the convolutional kernels, endowing the encoder with structural rotation equivariance. A subsequent group pooling operation converted the equivariant features into rotation-invariant isotropic representations. In addition, a multi-scale attention mechanism adjusted feature weights adaptively according to responses from different receptive fields, allowing the model to capture informative patterns ranging from nuclear-level details to tissue-level organization. Experimental results demonstrated that IAM-BYOL achieved classification accuracies of 98.74%, 99.04%, 99.01%, and 98.63% on the BreakHis dataset at 40×, 100×, 200×, and 400× magnifications, respectively, while attaining an accuracy of 93.02% on the cross-center private clinical dataset BCD. These findings indicate that introducing geometric inductive biases into pathological image representation learning can effectively enhance model robustness and generalization capability.

Citation： JIANG Liang, CAO Hui, NIE Yongqi, ZHANG Cheng. Self-supervised classification method for breast pathology images fusing geometric priors and multi-scale attention. Journal of Biomedical Engineering, 2026, 43(3): 471-478. doi: 10.7507/1001-5515.202511059 Copy

1.	Murtaza G, Shuib L, Abdul Wahab A W, et al. Deep learning-based breast cancer classification through medical imaging modalities: state of the art and research challenges. Artif Intell Rev, 2020, 53: 1655-1720.
2.	Peng J, Wang Y. Medical image segmentation with limited supervision: a review of deep network models. IEEE Access, 2021, 9: 36827-36851.
3.	Liu C, Dong Y, Xiang W, et al. A comprehensive study on robustness of image classification models: benchmarking and rethinking. Int J Comput Vis, 2025, 133(2): 567-589.
4.	Rani V, Kumar M, Gupta A, et al. Self-supervised learning for medical image analysis: a comprehensive review. Evol Syst, 2024, 15: 1607-1633.
5.	Qu L, Liu S, Liu X, et al. Towards label-efficient automatic diagnosis and analysis: a comprehensive survey of advanced deep learning-based weakly-supervised, semi-supervised and self-supervised techniques in histopathological image analysis. Phys Med Biol, 2022, 67(20): 20TR01.
6.	丁維龍, 劉津龍, 朱峰龍. 基于Transformer的乳腺組織病理圖像自監督分類方法. 計算機輔助設計與圖形學學報, 2025, 37(8): 1346-1358.
7.	姜良, 張程, 曹慧, 等. 基于深度學習的乳腺病理圖像診斷研究進展. 生物醫學工程學雜志, 2024, 41(5): 1072-1077,1084.
8.	Sabit H, Yadav A K, Salimy S, et al. Integrating multi-omics and artificial intelligence for personalized breast cancer management: a guide to clinicians. Cancer Lett, 649, 2026: 218468.
9.	Rahman T, Baras A S, Chellappa R. Evaluation of a task-specific self-supervised learning framework in digital pathology relative to transfer learning approaches and existing foundation models. Mod Pathol, 2025, 38(1): 100636.
10.	Campanella G, Chen S, Singh M, et al. A clinical benchmark of public self-supervised pathology foundation models. Nat Commun, 2025, 16(1): 3640.
11.	劉世峰, 王欣. 基于對比學習的醫學圖像分類改進方法. 吉林大學學報(信息科學版), 2024, 42(5): 881-888.
12.	Li L, Pan H, Liang Y, et al. PMFN-SSL: self-supervised learning-based progressive multimodal fusion network for cancer diagnosis and prognosis. Knowl Based Syst, 2024, 289: 111502.
13.	Liu H, Zhang Y, Luo J. Contrastive learning-based histopathological features infer molecular subtypes and clinical outcomes of breast cancer from unannotated whole slide images. Comput Biol Med, 2024, 170: 107997.
14.	Sani M M, Royat A, Baghshah M S. Classification of breast cancer histopathology images using a modified supervised contrastive learning method. Med Biol Eng Comput, 2025, 63(3): 721-731.
15.	Jiang L, Zhang C, Zhang H, et al. A lightweight spatially-aware classification model for breast cancer pathology images. Biocybern Biomed Eng, 2024, 44(3): 586-608.
16.	Kohler M, Kohler B. Analysis of convolutional neural network image classifiers in a rotationally symmetric model. IEEE Trans Inf Theory, 2023, 69(8): 5203-5218.
17.	Pacini M, Santin G, Lepri B, et al. On universality classes of equivariant networks. Adv Neural Inf Process Syst, 2026, 38: 61689-61719.
18.	何孫偉, 李秀娟, 謝元忠, 等. 人工智能技術在乳腺癌新輔助化療病理完全緩解預測中的研究現狀與挑戰. 生物醫學工程學雜志, 2025, 42(5): 1076-1084.
19.	Grill J B, Strub F, Altché F, et al. Bootstrap your own latent-a new approach to self-supervised learning. Adv Neural Inf Process Syst, 2020, 33: 21271-21284.
20.	Si J, Jia W, Jiang H. Breast cancer histopathology images classification through multi-view augmented contrastive learning and pre-learning knowledge distillation. IEEE Access, 2024, 12: 25359-25371.
21.	Wu J, Yuan T, Zeng J, et al. A medically assisted model for precise segmentation of osteosarcoma nuclei on pathological images. IEEE J Biomed Health Inform, 2023, 27(8): 3982-3993.
22.	Spanhol F A, Oliveira L S, Petitjean C, et al. A dataset for breast cancer histopathological image classification. IEEE Trans Biomed Eng, 2015, 63(7): 1455-1462.
23.	Zhou P, Xie X, Lin Z, et al. Towards understanding convergence and generalization of AdamW. IEEE Trans Pattern Anal Mach Intell, 2024, 46(9): 6486-6493.
24.	Obi J C. A comparative study of several classification metrics and their performances on data. World J Adv Eng Technol Sci, 2023, 8(1): 308-314.
25.	Liu X, Zhang F, Hou Z, et al. Self-supervised learning: generative or contrastive. IEEE Trans Knowl Data Eng, 2021, 35(1): 857-876.
26.	Gui J, Chen T, Zhang J, et al. A survey on self-supervised learning: algorithms, applications, and future trends. IEEE Trans Pattern Anal Mach Intell, 2024, 46(12): 9052-9071.
27.	Fashi P A, Hemati S, Babaie M, et al. A self-supervised contrastive learning approach for whole slide image representation in digital pathology. J Pathol Inform, 2022, 13: 100133.
28.	Voigt B, Fischer O, Schilling B, et al. Investigation of semi-and self-supervised learning methods in the histopathological domain. J Pathol Inform, 2023, 14: 100305.
29.	Yao D, Shao Y. A data efficient transformer based on Swin Transformer. Vis Comput, 2024, 40(4): 2589-2598.
30.	Tang J, Zhang T, Gong Z, et al. High precision cervical precancerous lesion classification method based on ConvNeXt. Bioengineering, 2023, 10(12): 1424.
31.	Torpey D, Klein R. DeepSet SimCLR: self-supervised deep sets for improved pathology representation learning. Pattern Recognit Lett, 2024, 186: 64-70.
32.	Pham T X, Niu A, Zhang K, et al. Self-supervised visual representation learning via residual momentum. IEEE Access, 2023, 11: 116706-116720.
33.	Maurya R, Pandey N N, Dutta M K, et al. FCCS-Net: breast cancer classification using multi-level fully convolutional-channel and spatial attention-based transfer learning approach. Biomed Signal Process Control, 2024, 94: 106258.
34.	Sreelekshmi V, Pavithran K, Nair J J. SwinCNN: an integrated Swin transformer and CNN for improved breast cancer grade classification. IEEE Access, 2024, 12: 68697-68710.
35.	Yan T, Chen G, Zhang H, et al. Convolutional neural network with parallel convolution scale attention module and ResCBAM for breast histology image classification. Heliyon, 2024, 10(10): e30889.
36.	Nair L S, Amarnath K R, Nair J J. Advancing breast cancer detection: SE-Conformer framework for malignancy detection in histopathology images. IEEE Access, 2025, 13: 100105-100118.

1. Murtaza G, Shuib L, Abdul Wahab A W, et al. Deep learning-based breast cancer classification through medical imaging modalities: state of the art and research challenges. Artif Intell Rev, 2020, 53: 1655-1720.
2. Peng J, Wang Y. Medical image segmentation with limited supervision: a review of deep network models. IEEE Access, 2021, 9: 36827-36851.
3. Liu C, Dong Y, Xiang W, et al. A comprehensive study on robustness of image classification models: benchmarking and rethinking. Int J Comput Vis, 2025, 133(2): 567-589.
4. Rani V, Kumar M, Gupta A, et al. Self-supervised learning for medical image analysis: a comprehensive review. Evol Syst, 2024, 15: 1607-1633.
5. Qu L, Liu S, Liu X, et al. Towards label-efficient automatic diagnosis and analysis: a comprehensive survey of advanced deep learning-based weakly-supervised, semi-supervised and self-supervised techniques in histopathological image analysis. Phys Med Biol, 2022, 67(20): 20TR01.
6. 丁維龍, 劉津龍, 朱峰龍. 基于Transformer的乳腺組織病理圖像自監督分類方法. 計算機輔助設計與圖形學學報, 2025, 37(8): 1346-1358.
7. 姜良, 張程, 曹慧, 等. 基于深度學習的乳腺病理圖像診斷研究進展. 生物醫學工程學雜志, 2024, 41(5): 1072-1077,1084.
8. Sabit H, Yadav A K, Salimy S, et al. Integrating multi-omics and artificial intelligence for personalized breast cancer management: a guide to clinicians. Cancer Lett, 649, 2026: 218468.
9. Rahman T, Baras A S, Chellappa R. Evaluation of a task-specific self-supervised learning framework in digital pathology relative to transfer learning approaches and existing foundation models. Mod Pathol, 2025, 38(1): 100636.
10. Campanella G, Chen S, Singh M, et al. A clinical benchmark of public self-supervised pathology foundation models. Nat Commun, 2025, 16(1): 3640.
11. 劉世峰, 王欣. 基于對比學習的醫學圖像分類改進方法. 吉林大學學報(信息科學版), 2024, 42(5): 881-888.
12. Li L, Pan H, Liang Y, et al. PMFN-SSL: self-supervised learning-based progressive multimodal fusion network for cancer diagnosis and prognosis. Knowl Based Syst, 2024, 289: 111502.
13. Liu H, Zhang Y, Luo J. Contrastive learning-based histopathological features infer molecular subtypes and clinical outcomes of breast cancer from unannotated whole slide images. Comput Biol Med, 2024, 170: 107997.
14. Sani M M, Royat A, Baghshah M S. Classification of breast cancer histopathology images using a modified supervised contrastive learning method. Med Biol Eng Comput, 2025, 63(3): 721-731.
15. Jiang L, Zhang C, Zhang H, et al. A lightweight spatially-aware classification model for breast cancer pathology images. Biocybern Biomed Eng, 2024, 44(3): 586-608.
16. Kohler M, Kohler B. Analysis of convolutional neural network image classifiers in a rotationally symmetric model. IEEE Trans Inf Theory, 2023, 69(8): 5203-5218.
17. Pacini M, Santin G, Lepri B, et al. On universality classes of equivariant networks. Adv Neural Inf Process Syst, 2026, 38: 61689-61719.
18. 何孫偉, 李秀娟, 謝元忠, 等. 人工智能技術在乳腺癌新輔助化療病理完全緩解預測中的研究現狀與挑戰. 生物醫學工程學雜志, 2025, 42(5): 1076-1084.
19. Grill J B, Strub F, Altché F, et al. Bootstrap your own latent-a new approach to self-supervised learning. Adv Neural Inf Process Syst, 2020, 33: 21271-21284.
20. Si J, Jia W, Jiang H. Breast cancer histopathology images classification through multi-view augmented contrastive learning and pre-learning knowledge distillation. IEEE Access, 2024, 12: 25359-25371.
21. Wu J, Yuan T, Zeng J, et al. A medically assisted model for precise segmentation of osteosarcoma nuclei on pathological images. IEEE J Biomed Health Inform, 2023, 27(8): 3982-3993.
22. Spanhol F A, Oliveira L S, Petitjean C, et al. A dataset for breast cancer histopathological image classification. IEEE Trans Biomed Eng, 2015, 63(7): 1455-1462.
23. Zhou P, Xie X, Lin Z, et al. Towards understanding convergence and generalization of AdamW. IEEE Trans Pattern Anal Mach Intell, 2024, 46(9): 6486-6493.
24. Obi J C. A comparative study of several classification metrics and their performances on data. World J Adv Eng Technol Sci, 2023, 8(1): 308-314.
25. Liu X, Zhang F, Hou Z, et al. Self-supervised learning: generative or contrastive. IEEE Trans Knowl Data Eng, 2021, 35(1): 857-876.
26. Gui J, Chen T, Zhang J, et al. A survey on self-supervised learning: algorithms, applications, and future trends. IEEE Trans Pattern Anal Mach Intell, 2024, 46(12): 9052-9071.
27. Fashi P A, Hemati S, Babaie M, et al. A self-supervised contrastive learning approach for whole slide image representation in digital pathology. J Pathol Inform, 2022, 13: 100133.
28. Voigt B, Fischer O, Schilling B, et al. Investigation of semi-and self-supervised learning methods in the histopathological domain. J Pathol Inform, 2023, 14: 100305.
29. Yao D, Shao Y. A data efficient transformer based on Swin Transformer. Vis Comput, 2024, 40(4): 2589-2598.
30. Tang J, Zhang T, Gong Z, et al. High precision cervical precancerous lesion classification method based on ConvNeXt. Bioengineering, 2023, 10(12): 1424.
31. Torpey D, Klein R. DeepSet SimCLR: self-supervised deep sets for improved pathology representation learning. Pattern Recognit Lett, 2024, 186: 64-70.
32. Pham T X, Niu A, Zhang K, et al. Self-supervised visual representation learning via residual momentum. IEEE Access, 2023, 11: 116706-116720.
33. Maurya R, Pandey N N, Dutta M K, et al. FCCS-Net: breast cancer classification using multi-level fully convolutional-channel and spatial attention-based transfer learning approach. Biomed Signal Process Control, 2024, 94: 106258.
34. Sreelekshmi V, Pavithran K, Nair J J. SwinCNN: an integrated Swin transformer and CNN for improved breast cancer grade classification. IEEE Access, 2024, 12: 68697-68710.
35. Yan T, Chen G, Zhang H, et al. Convolutional neural network with parallel convolution scale attention module and ResCBAM for breast histology image classification. Heliyon, 2024, 10(10): e30889.
36. Nair L S, Amarnath K R, Nair J J. Advancing breast cancer detection: SE-Conformer framework for malignancy detection in histopathology images. IEEE Access, 2025, 13: 100105-100118.

Journal of Biomedical Engineering

Self-supervised classification method for breast pathology images fusing geometric priors and multi-scale attention

Abstract Full text Figures/Tables Video References Cited by

Previous Article

Next Article

Format

Content