Multiple Instance Learning
Multiple Instance Learning (MIL) is a machine learning approach addressing the challenge of classifying data sets where labels are assigned to groups (bags) of instances, rather than individual instances. Current research heavily focuses on applying MIL to whole slide image (WSI) analysis in digital pathology, employing architectures like transformers and incorporating techniques such as attention mechanisms, pseudo-labeling, and graph neural networks to improve classification accuracy and interpretability. This work is significant because it allows for efficient analysis of large, complex medical images without requiring exhaustive pixel-level annotation, potentially leading to faster and more accurate diagnoses in healthcare.
Papers
SC-MIL: Supervised Contrastive Multiple Instance Learning for Imbalanced Classification in Pathology
Dinkar Juyal, Siddhant Shingi, Syed Ashar Javed, Harshith Padigela, Chintan Shah, Anand Sampat, Archit Khosla, John Abel, Amaro Taylor-Weiner
Exploring Visual Prompts for Whole Slide Image Classification with Multiple Instance Learning
Yi Lin, Zhongchen Zhao, Zhengjie ZHU, Lisheng Wang, Kwang-Ting Cheng, Hao Chen
Pixel-Level Explanation of Multiple Instance Learning Models in Biomedical Single Cell Images
Ario Sadafi, Oleksandra Adonkina, Ashkan Khakzar, Peter Lienemann, Rudolf Matthias Hehr, Daniel Rueckert, Nassir Navab, Carsten Marr
Task-specific Fine-tuning via Variational Information Bottleneck for Weakly-supervised Pathology Whole Slide Image Classification
Honglin Li, Chenglu Zhu, Yunlong Zhang, Yuxuan Sun, Zhongyi Shui, Wenwei Kuang, Sunyi Zheng, Lin Yang
BEL: A Bag Embedding Loss for Transformer enhances Multiple Instance Whole Slide Image Classification
Daniel Sens, Ario Sadafi, Francesco Paolo Casale, Nassir Navab, Carsten Marr
Active Learning Enhances Classification of Histopathology Whole Slide Images with Attention-based Multiple Instance Learning
Ario Sadafi, Nassir Navab, Carsten Marr