Rethinking Multiple Instance Learning for Whole Slide Image Classification: A Bag-Level Classifier is a Good Instance-Level Teacher

被引：0
|
作者：
Wang, Hongyi [3 ]
Luo, Luyang [1 ]
Wang, Fang [2 ]
Tong, Ruofeng [3 ,4 ]
Chen, Yen-Wei [3 ,5 ]
Hu, Hongjie [2 ]
Lin, Lanfen [3 ]
Chen, Hao [6 ]
机构：
[1] The Hong Kong University of Science and Technology, Department of Computer Science and Engineering, Hong Kong, Hong Kong
[2] Sir Run Run Shaw Hospital, Department of Radiology, Hangzhou,310016, China
[3] Zhejiang University, College of Computer Science and Technology, Hangzhou,310063, China
[4] Zhejiang Laboratory, Research Center for Healthcare Data Science, Hangzhou,311121, China
[5] Ritsumeikan University, College of Information Science and Engineering, Osaka, Ibaraki,567-8570, Japan
[6] The Hong Kong University of Science and Technology, Division of Life Science, Department of Computer Science and Engineering, Department of Chemical and Biological Engineering, Hong Kong, Hong Kong
来源：
IEEE Transactions on Medical Imaging | 2024年 / 43卷 / 11期
关键词：
Multiple Instance Learning (MIL) has demonstrated promise in Whole Slide Image (WSI) classification. However; a major challenge persists due to the high computational cost associated with processing these gigapixel images. Existing methods generally adopt a two-stage approach; comprising a non-learnable feature embedding stage and a classifier training stage. Though it can greatly reduce memory consumption by using a fixed feature embedder pre-trained on other domains; such a scheme also results in a disparity between the two stages; leading to suboptimal classification accuracy. To address this issue; we propose that a bag-level classifier can be a good instance-level teacher. Based on this idea; we design Iteratively Coupled Multiple Instance Learning (ICMIL) to couple the embedder and the bag classifier at a low cost. ICMIL initially fixes the patch embedder to train the bag classifier; followed by fixing the bag classifier to fine-tune the patch embedder. The refined embedder can then generate better representations in return; leading to a more accurate classifier for the next iteration. To realize more flexible and more effective embedder fine-tuning; we also introduce a teacher-student framework to efficiently distill the category knowledge in the bag classifier to help the instance-level embedder fine-tuning. Intensive experiments were conducted on four distinct datasets to validate the effectiveness of ICMIL. The experimental results consistently demonstrated that our method significantly improves the performance of existing MIL backbones; achieving state-of-the-art results. The code and the organized datasets can be accessed by: https://github.com/Dootmaan/ICMIL/tree/confidence-based. © 1982-2012 IEEE;
D O I：
10.1109/TMI.2024.3404549
中图分类号：
学科分类号：
摘要：
引用
收藏
页码：3964 / 3976
相关论文
共 50 条
  • [21] Nuclei-level prior knowledge constrained multiple instance learning for breast histopathology whole slide image classification
    Wang, Xunping
    Yuan, Wei
    [J]. ISCIENCE, 2024, 27 (06)
  • [22] RoFormer for Position Aware Multiple Instance Learning in Whole Slide Image Classification
    Pochet, Etienne
    Maroun, Rami
    Trullo, Roger
    [J]. MACHINE LEARNING IN MEDICAL IMAGING, MLMI 2023, PT II, 2024, 14349 : 437 - 446
  • [23] Pseudo-Bag Mixup Augmentation for Multiple Instance Learning-Based Whole Slide Image Classification
    Liu, Pei
    Ji, Luping
    Zhang, Xinyu
    Ye, Feng
    [J]. IEEE TRANSACTIONS ON MEDICAL IMAGING, 2024, 43 (05) : 1841 - 1852
  • [24] Iterative multiple instance learning for weakly annotated whole slide image classification
    Zhou, Yuanpin
    Che, Shuanlong
    Lu, Fang
    Liu, Si
    Yan, Ziye
    Wei, Jun
    Li, Yinghua
    Ding, Xiangdong
    Lu, Yao
    [J]. PHYSICS IN MEDICINE AND BIOLOGY, 2023, 68 (15):
  • [25] DGMIL: Distribution Guided Multiple Instance Learning for Whole Slide Image Classification
    Qu, Linhao
    Luo, Xiaoyuan
    Liu, Shaolei
    Wang, Manning
    Song, Zhijian
    [J]. MEDICAL IMAGE COMPUTING AND COMPUTER ASSISTED INTERVENTION, MICCAI 2022, PT II, 2022, 13432 : 24 - 34
  • [26] Neighborhood attention transformer multiple instance learning for whole slide image classification
    Aftab, Rukhma
    Yan, Qiang
    Zhao, Juanjuan
    Yong, Gao
    Huajie, Yue
    Urrehman, Zia
    Khalid, Faizi Mohammad
    [J]. FRONTIERS IN ONCOLOGY, 2024, 14
  • [27] BEL: A BAG EMBEDDING LOSS FOR TRANSFORMER ENHANCES MULTIPLE INSTANCE WHOLE SLIDE IMAGE CLASSIFICATION
    Sens, Daniel
    Sadafi, Ario
    Casale, Francesco Paolo
    Navab, Nassir
    Marr, Carsten
    [J]. 2023 IEEE 20TH INTERNATIONAL SYMPOSIUM ON BIOMEDICAL IMAGING, ISBI, 2023,
  • [28] Instance-Level Label Propagation with Multi-Instance Learning
    Wang, Qifan
    Chechik, Gal
    Sun, Chen
    Shen, Bin
    [J]. PROCEEDINGS OF THE TWENTY-SIXTH INTERNATIONAL JOINT CONFERENCE ON ARTIFICIAL INTELLIGENCE, 2017, : 2943 - 2949
  • [29] Image- and Instance-Level Data Augmentation for Occluded Instance Segmentation
    Yu, Jun
    Du, Shenshen
    Yang, Ruiqiang
    Wang, Lei
    Chen, Minchuan
    Zhu, Qingying
    Wang, Shaojun
    Xiao, Jing
    [J]. PROCEEDINGS OF THE 6TH INTERNATIONAL WORKSHOP ON MULTIMEDIA CONTENT ANALYSIS IN SPORTS, MMSPORTS 2023, 2023, : 137 - 142
  • [30] milVAD: A bag-level MNIST modelling of voice activity detection using deep multiple instance learning
    Korkmaz, Yunus
    Boyaci, Aytug
    [J]. BIOMEDICAL SIGNAL PROCESSING AND CONTROL, 2022, 74
12345