mlsanityData to 14 Oct 2026

Revisiting Data Challenges of Computational Pathology: A Pack-based Multiple Instance Learning Training Framework

arXiv 2509.20923 ↗First posted 25 Sep 2025cs.CVcited by 0

Not on the map. Why?

Abstract (arXiv)

Computational pathology (CPath) digitizes pathology slides into whole slide images (WSIs), enabling analysis for critical healthcare tasks such as cancer diagnosis and prognosis. However, WSIs possess extremely long sequence lengths (up to 200K), significant length variations (from 200 to 200K), and limited supervision. These extreme variations in sequence length lead to high data heterogeneity and redundancy. Conventional methods often compromise on training efficiency and optimization to preserve such heterogeneity under limited supervision. To comprehensively address these challenges, we propose a pack-based MIL framework. It packs multiple sampled, variable-length feature sequences into fixed-length ones, enabling batched training while preserving data heterogeneity. Moreover, we introduce a residual branch that composes discarded features from multiple slides into a hyperslide which is trained with tailored labels. It offers multi-slide supervision while mitigating feature loss from sampling. Meanwhile, an attention-driven downsampler is introduced to compress features in both branches to reduce redundancy. By alleviating these challenges, our approach achieves an accuracy improvement of up to 8% while using only 12% of the training time in the PANDA(UNI). Extensive experiments demonstrate that focusing data challenges in CPath holds significant potential in the era of foundation models. The code is https://github.com/FangHeng/PackMIL

Key facts. No citing papers in the map yet (data through 14 Oct 2026). Builds on 45 of its 70 references that are in the map. First posted on arXiv 25 Sep 2025.

Builds on

45 of 70 references.

References · 45

  1. 2015Batch Normalization: Accelerating Deep Network Training by Reducing Internal Covariate Shiftcited by 7,005
  2. 2024Qwen2-VL: Enhancing Vision-Language Model's Perception of the World at Any Resolutioncited by 4,234
  3. 2018Attention-based Deep Multiple Instance Learningcited by 1,095
  4. 2022Token Merging: Your ViT But Fastercited by 1,021
  5. 2020Data Efficient and Weakly Supervised Computational Pathology on Whole Slide Imagescited by 745
  6. 2021TransMIL: Transformer based Correlated Multiple Instance Learning for Whole Slide Image Classificationcited by 555
  7. 2020Dual-stream Multiple Instance Learning Network for Whole Slide Image Classification with Self-supervised Contrastive Learningcited by 478
  8. 2018HoVer-Net: Simultaneous Segmentation and Classification of Nuclei in Multi-Tissue Histology Imagescited by 388
  9. 2020Asymmetric Loss For Multi-Label Classificationcited by 381
  10. 2022Scaling Vision Transformers to Gigapixel Images via Hierarchical Self-Supervised Learningcited by 359
  11. 2022DTFD-MIL: Double-Tier Feature Distillation Multiple Instance Learning for Histopathology Whole Slide Image Classificationcited by 256
  12. 2023Patch n' Pack: NaViT, a Vision Transformer for any Aspect Ratio and Resolutioncited by 229
33 more
  1. 2020Whole Slide Images based Cancer Survival Prediction using Attention Guided Deep Multiple Instance Learning Networkscited by 178
  2. 2024Multimodal Whole Slide Foundation Model for Pathologycited by 162
  3. 2023Artificial Intelligence for Digital and Computational Pathologycited by 108
  4. 2023Interventional Bag Multi-Instance Learning On Whole-Slide Pathological Imagescited by 91
  5. 2024Morphological Prototyping for Unsupervised Slide Representation Learning in Computational Pathologycited by 72
  6. 2024Feature Re-Embedding: Towards Foundation Model-Level Performance in Computational Pathologycited by 67
  7. 2024Dynamic Graph Representation with Knowledge-aware Attention for Histopathology Whole Slide Image Analysiscited by 67
  8. 2023Attention-Challenging Multiple Instance Learning for Whole Slide Image Classificationcited by 61
  9. 2023Multiple Instance Learning Framework with Masked Hard Instance Mining for Whole Slide Image Classificationcited by 60
  10. 2024Transcriptomics-guided Slide Representation Learning in Computational Pathologycited by 51
  11. 2022Bi-directional Weakly Supervised Knowledge Distillation for Whole Slide Image Classificationcited by 43
  12. 2022DGMIL: Distribution Guided Multiple Instance Learning for Whole Slide Image Classificationcited by 40
  13. 20242DMamba: Efficient State Space Model for Image Representation with Applications on Giga-Pixel Whole Slide Image Classificationcited by 35
  14. 2025Do Multiple Instance Learning Models Transfer?cited by 29
  15. 2025ViLa-MIL: Dual-scale Vision-Language Multiple Instance Learning for Whole Slide Image Classificationcited by 28
  16. 2024Generalizable Whole Slide Image Classification with Fine-Grained Visual-Semantic Interactioncited by 27
  17. 2024Dataset Decomposition: Faster LLM Training with Variable Sequence Length Curriculumcited by 26
  18. 2023Pseudo-Bag Mixup Augmentation for Multiple Instance Learning-Based Whole Slide Image Classificationcited by 18
  19. 2020Overcoming the limitations of patch-based learning to detect cancer in whole slide imagescited by 15
  20. 2021Label Cleaning Multiple Instance Learning: Refining Coarse Annotations on Single Whole-Slide Imagescited by 13
  21. 2023CAMIL: Context-Aware Multiple Instance Learning for Cancer Detection and Subtyping in Whole Slide Imagescited by 13
  22. 2024Graph Neural Networks in Histopathology: Emerging Trends and Future Directionscited by 13
  23. 2024Rethinking Transformer for Long Contextual Histopathology Whole Slide Image Analysiscited by 12
  24. 2024AEM: Attention Entropy Maximization for Multiple Instance Learning based Whole Slide Image Classificationcited by 10
  25. 2024Benchmarking Embedding Aggregation Methods in Computational Pathology: A Clinical Data Perspectivecited by 10
  26. 2024MergeUp-augmented Semi-Weakly Supervised Learning for WSI Classificationcited by 7
  27. 2024SAM-MIL: A Spatial Contextual Aware Multiple Instance Learning Approach for Whole Slide Image Classificationcited by 6
  28. 2024Impact of Stain Variation and Color Normalization for Prognostic Predictions in Pathologycited by 6
  29. 2025Revisiting End-to-End Learning with Slide-level Supervision in Computational Pathologycited by 6
  30. 2025Fast and Accurate Gigapixel Pathological Image Classification with Hierarchical Distillation Multi-Instance Learningcited by 5
  31. 2024EMMeTT: Efficient Multimodal Machine Translation Trainingcited by 4
  32. 2023Open and reusable deep learning for pathology with WSInfer and QuPathcited by 3
  33. 2025ICFNet: Integrated Cross-modal Fusion Network for Survival Predictioncited by 3

Built on by

Not cited yet.