病理全切片影像（Whole Slide Image, WSI）在疾病檢測中扮演關鍵角色，然而，由
於其極高的解析度和需要專業知識進行的人工標註，導致 Pixel-level 的標註需要龐
大人力成本，同時造成深度學習模型訓練時的困難。
為了解決上述問題，我們採用多實例學習（Multiple Instance Learning，MIL）來進行
全切片影像的分析。多實例學習是一種弱監督式學習方式，僅需要 Slide-level 的標
註，即可進行全切片影像的分類與分割。
在應用於病理全切片影像的多實例學習中，一張全切片影像會被切割成許多補丁
(Patch)。為了有效利用補丁之間的相似性，我們提出了一個基於補丁特徵分群的偽
包（Cluster-based Pseudo Bags）方法，透過分群演算法聚集顏色、形狀等特徵相似的
補丁，使得多實例學習模型能夠更有效地進行補丁的特徵聚合。
我們在乳腺癌資料集 Camelyon16 與肺腺癌資料集 TCGA-NSCLS 上驗證所提出的方
法，展現出基於偽袋的多實例學習的有效性，其效能優於先前的弱監督式學習方法，
達到了最領先的成果。

Pathological Whole Slide Images (WSI) play a crucial role in disease detection. However, the exceptionally high resolution of WSI and the requirement of expert knowledge in manual annotation make a significant human labor cost for pixel-level labeling, making it challenging to train deep learning models.
To address these challenges, we adopt Multiple Instance Learning (MIL) to analyze Whole Slide Images. MIL is a form of weakly supervised learning that relies solely on slide-level annotations for classifying or segmenting WSIs.
In the application of MIL to pathological Whole Slide Images, a slide is segmented into numerous patches. To effectively leverage the similarity between patches, we propose a learnable clustering method capable of automatically capturing correlations between patches, wherein patches with similar features are clustered together. This enhances feature aggregation of patches within the MIL model.
We validate our approach on the breast cancer dataset Camelyon16 and the lung adenocarcinoma dataset TCGA-NSCLS. The results demonstrate the effectiveness of the learnable clustering method in MIL, surpassing previous weakly supervised learning methods and achieving state-of-the-art performance.

[1] Babak Ehteshami Bejnordi adn Mitko Veta, Paul Johannes van Diest, Bram van Ginneken, Nico Karssemeijer, Geert Litjens, Jeroen A. W. M. van der Laak, and the CAME LYON16 Consortium. Diagnostic assessment of deep learning algorithms for detection of lymph node metastases in women with breast cancer. The Journal of the American Medical Association, 318(22):2199–2210, 2017.
[2] Gabriele Campanella, Matthew G. Hanna, Luke Geneslaw, Allen Miraflor, Vitor Wer neck Krauss Silva, Klaus J. Busam, Edi Brogi, Victor E. Reuter, David S. Klimstra, and Thomas J. Fuchs. Clinical-grade computational pathology using weakly supervised deep learning on whole slide images. Nature Medicine, 25:1301–1309, 2019.
[3] Ting Chen, Simon Kornblith, Mohammad Norouzi, and Geoffrey Hinton. A simple framework for contrastive learning of visual representations. In Proceedings of International Conference on Machine Learning, pages 1597–1607, 2020.
[4] Veronika Cheplygina, Lauge Sorensen, David M. J. Tax, Marleen de Bruijne, and Marco Loog. Label stability in multiple instance learning. In Proceedings of International Conference on Medical Image Computing and Computer-Assisted Intervention, pages 539–546, 2015.
[5] Philip Chikontwe, Meejeong Kim, Soo Jeong Nam, Heounjeong Go, and Sang Hyun Park. Multiple instance learning with center embeddings for histopathology classification. In Proceedings of International Conference on Medical Image Computing and Computer-Assisted Intervention, pages 519–528, 2020.
[6] Thomas G. Dietterich, Richard H. Lathrop, and Tomas Lozano-Perez. Solving the multiple instance problem with axis-parallel rectangles. Artificial Intelligence, 89(1-2):31 71, 1997.
[7] Alexey Dosovitskiy, Lucas Beyer, Alexander Kolesnikov, Dirk Weissenborn, Xiaohua Zhai, Thomas Unterthiner, Mostafa Dehghani, Matthias Minderer, Georg Heigold, Syl vain Gelly, Jakob Uszkoreit, and Neil Houlsby. An image is worth 16x16 words: Transformers for image recognition at scale. In Proceedings of International Conference on Learning Representations, 2021.
[8] Kaiming He, Xiangyu Zhang, Shaoqing Ren, and Jian Sun. Deep residual learning for image recognition. In Proceedings of IEEE Conference on Computer Vision and Pattern Recognition, pages 770–778, 2016.
[9] Le Hou, Dimitris Samaras, Tahsin M. Kurc, Yi Gao, James E. Davis, and Joel H. Saltz. Patch-based convolutional neural network for whole slide tissue image classification. In Proceedings of IEEE Conference on Computer Vision and Pattern Recognition, pages 2424–2433, 2016.
[10] Wentai Hou, Lequan Yu, Chengxuan Lin, Helong Huang, Rongshan Yu, Jing Qin, and Liansheng Wang. H2-mil: Exploring hierarchical representation with heterogeneous multiple instance learning for whole slide image analysis. In Proceedings of AAAI Conference on Artificial Intelligence, pages 933–941, 2022.
[11] Maximilian Ilse, Jakub M. Tomczak, and Max Welling. Attention-based deep multiple instance learning. In Proceedings of International Conference on Machine Learning, 2018.
[12] Diederik P. Kingma and Jimmy Ba. Adam: A method for stochastic optimization. In Proceedings of International Conference on Learning Representations, 2015.
[13] Bin Li, Yin Li, and Kevin W. Eliceiri. Dual-stream multiple instance learning network for whole slide image classification with self-supervised contrastive learning. In Proceedings of IEEE Conference on Computer Vision and Pattern Recognition, pages 14318–14328, 2021.
[14] Tiancheng Lin, Zhimiao Yu, Hongyu Hu, Yi Xu, and Chang Wen Chen. Interventional bag multi-instance learning on whole-slide pathological images. In Proceedings of IEEE Conference on Computer Vision and Pattern Recognition, pages 19830–19839, 2023.
[15] Ming Y. Lu, Drew F. K. Williamson, Tiffany Y. Chen, Richard J. Chen, Matteo Barbieri, and Faisal Mahmood. Data efficient and weakly supervised computational pathology on whole slide images. Nature Biomedical Engineering, 5:555–570, 2021.
[16] The Cancer Genome Atlas Research Network. Comprehensive molecular profiling of lung adenocarcinoma. Nature, 511:543–550, 2014.
[17] Comprehensive Genomic Characterization of Squamous Cell Lung Cancers. Comprehensive molecular profiling of lung adenocarcinoma. Nature, 489:519–525, 2012.
[18] Gwenole Quellec, Guy Cazuguel, Beatrice Cochener, and Mathieu Lamard. Multiple instance learning for medical image and video analysis. IEEE Reviews in Biomedical Engineering, 10:213–234, 2017.
[19] Olaf Ronneberger, Philipp Fischer, and Thomas Brox. U-net: Convolutional networks for biomedical image segmentation. In Proceedings of International Conference on Medical Image Computing and Computer-Assisted Intervention, pages 234–241, 2015.
[20] Zhuchen Shao, Hao Bian, Yang Chen, Yifeng Wang, Jian Zhang, Xiangyang Ji, and Yongbing Zhang. Transmil: Transformer based correlated multiple instance learning for whole slide image classification. In Proceedings of International Conference on Neural Information Processing Systems, pages 6742–6751, 2021.
[21] Yash Sharma, Aman Shrivastava, Lubaina Ehsan, Christopher A. Moskaluk, Sana Syed, and Donald E. Brown. Cluster-to-conquer: A framework for end-to-end multi-instance learning for whole slide image classification. In Proceedings of Medical Imaging with Deep Learning, 2021.
[22] Paul Tourniaire, Marius Ilie, Paul Hofman, Nicholas Ayache, and Herve Delingette. Attention-based multiple instance learning with mixed supervision on the camelyon16 dataset. In Proceedings of the MICCAI Workshop on Computational Pathology, pages 216–226, 2021.
[23] Ming Tu, Jing Huang, Xiaodong He, and Bowen Zhou. Multiple instance learning with graph neural networks. In arXiv:1906.04881, 2019.
[24] Ashish Vaswani, Noam Shazeer, Niki Parmar, Jakob Uszkoreit, Llion Jones, Aidan N. Gomez, Łukasz Kaiser, and Illia Polosukhin. Attention is all you need. In Proceedings of Conference on Neural Information Processing Systems, 2017.
[25] Xiaolong Wang, Ross Girshick, Abhinav Gupta, and Kaiming He. Non-local neural networks. In Proceedings of IEEE Conference on Computer Vision and Pattern Recognition, 2018.
[26] Xinggang Wang, Yongluan Yan, Peng Tang, Xiang Bai, and Wenyu Liu. Revisiting multiple instance neural networks. Pattern Recognition, 74:15–24, 2018.
[27] Xiyue Wang, Sen Yang, Jun Zhang, Minghui Wang, Jing Zhang, Wei Yang, Jun zhou Huang, and Xiao Han. Transformer-based unsupervised contrastive learning for histopathological image classification. Medical Image Analysis, 81, 2022.
[28] Jinxi Xiang, Xiyue Wang, Jun Zhang, Sen Yang, Xiao Han, and Wei Yang. Exploring low-rank property in multiple instance learning for whole slide image classification. In Proceedings of International Conference on Learning Representations, 2023.
[29] Yunyang Xiong, Zhanpeng Zeng, Rudrasis Chakraborty, Mingxing Tan, Glenn Fung, Yin Li, and Vikas Singh. Nystromformer: A nystrom-based algorithm for approximating self-attention. In Proceedings of AAAI Conference on Artificial Intelligence, 2021.
[30] Yongluan Yan, Xinggang Wang, Xiaojie Guo, Jiemin Fang, Wenyu Liu, and Junzhou Huang. Deep multi-instance learning with dynamic pooling. In Proceedings of Asian Conference on Machine Learning, 2018.
[31] Hongrun Zhang, Yanda Meng, Yitian Zhao, Yihong Qiao, Xiaoyun Yang, Sarah E. Coupland, and Yalin Zheng. Dtfd-mil: Double-tier feature distillation multiple instance learning for histopathology whole slide image classification. In Proceedings of IEEE Conference on Computer Vision and Pattern Recognition, pages 18802–18812, 2022.
[32] Yu Zhao, Zhenyu Lin, Kai Sun, Yidan Zhang, Junzhou Huang, Liansheng Wang, and Jianhua Yao. Setmil: Spatial encoding transformer-based multiple instance learning for pathological image analysis. In Proceedings of International Conference on Medical Image Computing and Computer-Assisted Intervention, pages 66–76, 2022.
[33] Yu Zhao, Fan Yang, Yuqi Fang, Hailing Liu, Niyun Zhou, Jun Zhang, Jiarui Sun, Sen Yang, Bjoern Menze, Xinjuan Fan, and Jianhua Yao. Predicting lymph node metastasis using histopathological images based on multiple instance learning with deep graph convolution. In Proceedings of IEEE Conference on Computer Vision and Pattern Recognition, pages 4837–4846, 2020.
[34] Zhen Zhu, Mengde Xu, Song Bai, Tengteng Huang, and Xiang Bai. Asymmetric non local neural networks for semantic segmentation. In Proceedings of IEEE International Conference on Computer Vision, 2019.
[35] Zhonghang Zhu, Lequan Yu, Wei Wu, Rongshan Yu, Defu Zhang, and Liansheng Wang. Murcl: Multi-instance reinforcement contrastive learning for whole slide image classification. IEEE Transactions on Medical Imaging, 42(5):1337–1348, 2023.
[36] Kelly H. Zou, Simon K. Warfield, Aditya Bharatha, Clare M.C. Tempany, Michael R. Kaus, Steven J. Haker, William M. Wells III, Ferenc A. Jolesz, and Ron Kikinis. Statistical validation of image segmentation quality based on a spatial overlap index: Scientific reports. Academic Radiology, 11(2):178–189, 2004.