Pith. sign in

REVIEW 3 major objections 4 minor 60 references

FOCUS: Knowledge-enhanced Adaptive Visual Compression for Few-shot Whole Slide Image Classification

T0 review · 3 major / 4 minor · reviewed 2026-08-12 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read A three-stage adaptive compression strategy that uses pathology foundation models and LLM-generated text prompts lets few-shot whole-slide classifiers beat prior methods on lung, breast, and ovarian cancer datasets.

desk verdict A well-built three-stage visual compression pipeline for few-shot WSI classification, but the reported 4-shot gains are statistically fragile and need paired significance testing before the SOTA claim is credible. read the letter →

arxiv 2411.14743 v2 pith:DU7JDGA5 submitted 2024-11-22 cs.CV cs.AIq-bio.QM

Zhengrui Guo , Conghao Xiong , Jiabo Ma , Qichen Sun , Lishuang Feng , Jinzhuo Wang , Hao Chen This is my paper · ORCID
classification cs.CVcs.AIq-bio.QM
keywords few-shotlearningwholeslideimageclassificationmultipleinstancevisualtokencompressionpathologyfoundationmodelslanguage-guidedpromptingcomputationalcancerdiagnosis
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

FOCUS is a few-shot whole-slide-image classifier that tries to solve the data-scarcity problem by aggressively pruning away uninformative patches before classification. The paper's central claim is that a three-stage, adaptive visual compression pipeline—global redundancy removal using pathology foundation-model features, language-guided token prioritization using LLM-generated pathology descriptions, and neighbor-aware sequential filtering—lets a model trained on 4 to 16 labeled slides per cancer class outperform prior few-shot multiple-instance learning methods. On TCGA-NSCLC, CAMELYON, and UBC-OCEAN, it reports the best balanced accuracy and AUC across the three shot settings, with the largest margins at 4 shots: 81.9%, 70.1%, and 70.4% balanced accuracy on the three datasets. If the claim holds, it means that in scarce-data clinical settings, selectively discarding visual content under language guidance is more effective than aligning every patch with a text prompt. The paper also demonstrates through ablations that each compression stage and the choice of foundation model contribute to the gain.

What carries the argument

FOCUS's core mechanism is the three-stage progressive visual-token compression pipeline: (1) global redundancy removal via sliding-window cosine similarity with dynamic threshold $\tau_g=\mu(S)+\sigma(S)$; (2) language-guided prioritization using cross-modal attention between learnable-plus-LLM text prompts and patch features, selecting top-$k$ tokens by relevance rank $r_i$; and (3) sequential neighbor-aware compression that drops a token when the minimum cosine similarity to its two neighbors falls under a rising threshold $\theta_i=\theta_{base}+i\Delta\theta$. A cross-modal multi-head attention aggregator then pools the surviving tokens for prediction. This chain carries the argument because each stage's thresholds control how many patches survive, and the ablation study shows each stage adds accuracy.

What would settle it

Retain the same total number of tokens but replace the language-guided top-$k$ selection with random selection in the 4-shot UBC-OCEAN setting; if balanced accuracy does not drop, the reported gains come from compression rather than from the LLM prompts. Conversely, on mucinous-carcinoma slides with large uniform mucin pools, progressively lowering $\tau_g$ should show whether similar-but-diagnostic patches are being pruned.

Watch

Extended reading notes

Core claim

The central discovery is that in few-shot weakly-supervised whole slide classification, the bottleneck is not the aggregator but the enormous number of diagnostically irrelevant patches, and that these can be removed adaptively before aggregation. FOCUS uses a frozen pathology vision-language foundation model (CONCH) to encode patches, computes pairwise cosine similarities inside sliding windows, and discards patches whose mean similarity exceeds the dynamic threshold $\tau_g = \mu(S)+\sigma(S)$. It then ranks the surviving patches by their cross-modal attention relevance to a concatenation of learnable prompts and an LLM-generated pathology knowledge prompt, keeping the top-$k$ where $k=\min(M_{\max},\gamma N')$, and finally removes tokens whose minimum similarity to their spatial neighbors falls below an increasing threshold $\theta_i=\theta_{base}+i\Delta\theta$. The surviving, spatially coherent tokens are pooled by a text-guided multi-head attention aggregator. The paper reports consistent improvements over prior few-shot MIL methods, notably absolute balanced-accuracy gains of 1.2%, 4.3%, and 3.3% over the second-best method in the 4-shot setting on TCGA-NSCLC, CAMELYON, and UBC-OCEAN, respectively.

Load-bearing premise

The method assumes that visual redundancy in a whole slide can be read off cosine similarity between patch features, so a large, homogeneous, but diagnostically important tumor region, consisting of many mutually similar patches, could be pruned before it contributes to classification.

Editorial extensions

If this is right

  • At 4 shots per class, FOCUS raises balanced accuracy to 81.9% on TCGA-NSCLC, 70.1% on CAMELYON, and 70.4% on UBC-OCEAN, beating the second-best method by 1.2, 4.3, and 3.3 points.
  • The advantage persists at 8 and 16 shots, with the best AUC scores of 97.2%, 94.3%, and 96.7% at 16 shots across the three datasets.
  • Each added module helps: on UBC-OCEAN at 4 shots, adding the FM-based global redundancy removal, language-guided prioritization, sequential compression, and cross-modal aggregation moves balanced accuracy from 62.3% to 70.4%.
  • CONCH, a vision-language foundation model with aligned image-text pretraining, works better as the feature extractor than UNI, GPFM, Virchow, or PLIP in this framework, indicating that cross-modal alignment in the encoder is part of the gain.
  • The framework operates on standard single-resolution WSIs, avoiding the paired multi-resolution inputs or extra reference slides required by some prior few-shot methods.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • Nothing in the paper separates the language guidance from the simple act of compression: a comparison against random token pruning with the same retained count would establish whether the LLM descriptions, rather than the compression itself, drive the 4-shot improvements.
  • The cosine-similarity redundancy proxy suggests an immediate stress test: large homogeneous tumor regions with mutually similar but diagnostically critical patches may be pruned by global or sequential stages, so the method's robustness on mucinous or diffusely growing tumors deserves direct evaluation.
  • The same recipe—off-the-shelf pretrained encoder similarity pruning, text-prompt reranking, then neighbor-aware thinning—could transfer to other gigapixel or long-document tasks where only a few class labels are available and most input tokens are irrelevant.
  • A dynamic extension in which the LLM prompt is updated per slide or per compression stage, or in which thresholds are learned rather than set by population statistics, is a natural next step the paper leaves implicit.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

3 major / 4 minor

Summary. The paper proposes FOCUS, a three-stage adaptive visual compression framework for few-shot whole-slide image classification. Using frozen pathology foundation-model features and LLM-generated text prompts, the method first discards visually redundant patches by sliding-window cosine similarity, then selects patches by cross-modal relevance to language prompts, then removes locally similar neighboring tokens with increasing thresholds, and finally aggregates the compressed tokens with a cross-modal attention head. Experiments on TCGA-NSCLC, CAMELYON, and UBC-OCEAN under 4/8/16-shot settings compare FOCUS with MIL and few-shot baselines, reporting gains in Balanced ACC, AUC, and F1, with ablations for the compression modules, foundation-model choice, and prompt source.

Significance. If the empirical claims were statistically reliable, FOCUS would make a modest but useful contribution: it is a simple, largely training-free compression strategy that leverages existing pathology foundation models and external language priors without requiring paired multi-resolution inputs or reference samples. The authors also release code, which supports reproducibility. However, the paper's central claim—that FOCUS consistently outperforms state-of-the-art few-shot MIL methods—is currently supported only by point estimates with high variance and no significance testing. The value of the method therefore depends on additional evidence that the reported margins are not noise, and on a protocol that prevents large labeled validation sets from driving few-shot model selection.

major comments (3)
  1. [Section 4.2, Table 1] The headline 4-shot Balanced ACC improvements are statistically fragile and the claim of consistent superiority is not established. In TCGA-NSCLC, FOCUS is 0.819±0.044 versus ViLa-MIL 0.807±0.043 (difference 0.012); in CAMELYON, 0.701±0.082 versus 0.658±0.036 (difference 0.043); in UBC-OCEAN, 0.704±0.088 versus 0.671±0.080 (difference 0.033). Under an unpaired comparison with ten folds, these differences correspond to roughly 0.6, 1.5, and 0.9 standard errors, respectively. The paper reports no paired significance tests, confidence intervals, or effect sizes, although the folds are presumably paired across methods. Either add paired statistical tests with multiple-comparison correction, or soften the 'consistently outperforms' claim to reflect that the observed differences are not statistically significant at the reported level of variance.
  2. [Section 4.1, Implementation Details] The evaluation protocol uses a validation set with many labels to select the best checkpoint (early stopping based on validation performance) before reporting test results. With a 6:2:2 split, the validation set contains far more labeled slides than the K-shot training set (e.g., for TCGA-NSCLC, about 200 validation slides versus 4 per class in 4-shot). In a few-shot evaluation, this large labeled validation set can act as an additional source of supervision for model selection, potentially inflating the reported few-shot performance. The authors should clarify whether all baselines are selected under the same rule, and should report results obtained without validation-based checkpoint selection, or with a validation set matched to the shot count.
  3. [Sections 3.2.1 and 3.3] The paper does not validate the key assumption that cosine similarity between patch features reliably indicates diagnostic redundancy. The compression stages discard patches whose mean or neighbor similarity exceeds thresholds (Eq. 2 and Eq. 6), but no experiment compares this selection against random token dropping, feature-only selection without language guidance, or frequency-based selection. Table 2 shows that adding each module improves Balanced ACC, but the increments are small relative to the reported standard deviations and no significance tests are provided. Consequently, the attribution of the gains to language-guided prioritization rather than to mere compression is not demonstrated. Please add ablations that isolate the effect of the cosine-similarity redundancy proxy and the language guidance component, along with sensitivity analysis for the hand-set hyperparameters w=32, gamma=0.8, and theta_base=0.7.
minor comments (4)
  1. [Introduction, Contributions] The first contribution bullet contains a grammatical error: 'We explores the untapped potential' should read 'We explore'. Also, 'utlize' in Section 4.1 should be 'utilize'.
  2. [Supplementary Material, Algorithm 1] Algorithm 1 is titled 'LA ViC-MIL', which does not match the method name FOCUS used throughout the paper; please correct the title and any internal references for consistency.
  3. [Section 4.3.3 and Appendix A.4] The explanations for why one LLM prompt outperforms another (e.g., 'comprehensive and clinically precise descriptions' versus 'overly brief descriptions') are post-hoc and unsupported by quantitative analysis; the appendix presents these as established causes rather than as hypotheses.
  4. [Related Works and Table 1] The Related Works section discusses PEMP, MSCPT, and FAST as recent few-shot MIL methods, but none of these methods is included in the experimental comparisons; the claim of outperforming 'state-of-the-art' few-shot methods is therefore limited to the subset of baselines evaluated.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: FOCUS's reported gains rest on external foundation models, external LLM-generated prompts, and a disclosed few-shot evaluation protocol rather than on any target quantity defined in terms of its own inputs.

full rationale

The paper's derivation chain is empirical rather than definitional. The three-stage compression thresholds in Eqs. (2) and (6) are computed from pathology-FM patch features and cosine similarities, not from slide labels or from the reported Balanced ACC/AUC quantities. The language prompts are generated by Claude-3.5-Sonnet and other external LLMs, and the CONCH features are frozen, so the language guidance is not fitted to the test folds. The final classifier is trained with cross-entropy on the K-shot training set, and the ablation studies in Table 2 and Figures 2-3 independently vary modules, foundation models, and prompt sources, giving the central claim independent content. Self-citations such as [14, 15, 22, 30, 50, 52, 55] appear only as related-work context and are not load-bearing; no uniqueness theorem or ansatz is imported from prior work by the same authors. The hand-chosen hyperparameters (w=32, gamma=0.8, theta_base=0.7) and validation-based early stopping are disclosed standard model-selection practices, not a fitted input renamed as a prediction. No equation or construction in the paper reduces the reported performance to the method's own inputs, so no circular step is identified.

Assumptions & free parameters 6 free parameters · 4 assumptions · 0 invented entities

The method relies on several hand-chosen hyperparameters and domain assumptions about similarity-based redundancy and cross-modal alignment. No new physical or conceptual entities are introduced. The most significant unstated dependency is the validation set used for checkpoint selection, which affects the interpretation of the few-shot results.

free parameters (6)
  • window size w = 32
    Used in global redundancy removal (Section 3.2.1); chosen by hand, sensitivity not reported.
  • compression ratio gamma = 0.8
    Controls how many tokens are kept after language-guided prioritization (Eq. 4); tuned on validation.
  • base similarity threshold theta_base = 0.7
    Starting threshold for sequential compression (Section 3.3); hand-chosen, no sensitivity analysis.
  • threshold increment Delta_theta = unspecified
    Controls how fast thresholds increase across compression stages (Eq. 6); value not given in the paper.
  • maximum sequence length Mmax = unspecified
    Caps the number of tokens in Eq. 4; value not defined.
  • number of sequential compression stages = unspecified
    The loop in Algorithm 1 iterates over stages but no count or stopping rule is specified.
assumptions (4)
  • domain assumption Cosine similarity between patch features indicates diagnostic redundancy.
    Used in Eq. 2 and Eq. 6 to remove patches; the paper does not validate this against pathologist judgment.
  • domain assumption LLM-generated text prompts contain visual features that align with CONCH patch embeddings.
    The cross-modal attention in Eq. 3 and the aggregation module rely on this alignment; no evidence is given that the alignment is reliable for these prompts.
  • domain assumption The spatial ordering of patches from CLAM preprocessing preserves local tissue context.
    The sequential compression in Section 3.3 assumes neighboring tokens are spatially adjacent and that this order is meaningful.
  • domain assumption Validation-based early stopping with a label-rich validation set is valid in the few-shot protocol.
    The experimental setup uses 20% of each dataset as validation for model selection, which may leak more label information than the K-shot training set provides.

how reviews work

0 comments
Cite this review

Pith. "Pith review of FOCUS: Knowledge-enhanced Adaptive Visual Compression for Few-shot Whole Slide Image Classification." pith.science (2026). https://pith.science/paper/DU7JDGA5

@misc{pith2026241114743,
  author       = {Pith},
  title        = {Pith review of: FOCUS: Knowledge-enhanced Adaptive Visual Compression for Few-shot Whole Slide Image Classification},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/DU7JDGA5}},
  note         = {Machine review of arXiv:2411.14743}
}
read the original abstract

Few-shot learning presents a critical solution for cancer diagnosis in computational pathology (CPath), addressing fundamental limitations in data availability, particularly the scarcity of expert annotations and patient privacy constraints. A key challenge in this paradigm stems from the inherent disparity between the limited training set of whole slide images (WSIs) and the enormous number of contained patches, where a significant portion of these patches lacks diagnostically relevant information, potentially diluting the model's ability to learn and focus on critical diagnostic features. While recent works attempt to address this by incorporating additional knowledge, several crucial gaps hinder further progress: (1) despite the emergence of powerful pathology foundation models (FMs), their potential remains largely untapped, with most approaches limiting their use to basic feature extraction; (2) current language guidance mechanisms attempt to align text prompts with vast numbers of WSI patches all at once, struggling to leverage rich pathological semantic information. To this end, we introduce the knowledge-enhanced adaptive visual compression framework, dubbed FOCUS, which uniquely combines pathology FMs with language prior knowledge to enable a focused analysis of diagnostically relevant regions by prioritizing discriminative WSI patches. Our approach implements a progressive three-stage compression strategy: we first leverage FMs for global visual redundancy elimination, and integrate compressed features with language prompts for semantic relevance assessment, then perform neighbor-aware visual token filtering while preserving spatial coherence. Extensive experiments on pathological datasets spanning breast, lung, and ovarian cancers demonstrate its superior performance in few-shot pathology diagnosis. Codes are available at https://github.com/dddavid4real/FOCUS.

Figures

Figures reproduced from arXiv: 2411.14743 by the authors.

Figure 1
Figure 1. Illustration of the proposed FOCUS framework, which consists of three key components: (a) A knowledge-enhanced adaptive visual token compression module that performs global redundancy removal via FM features and language prior-guided visual token priori￾tization, (b) A sequential visual token compression module that performs pair-wise similarity thresholding to eliminate local redundancies while preserving spatial c… view at source ↗
Figure 3
Figure 3. Performance comparison of prompts from different [PITH_FULL_IMAGE:figures/full_fig_p008_3.png] view at source ↗
Figure 2
Figure 2. Performance comparison of different foundation models [PITH_FULL_IMAGE:figures/full_fig_p008_2.png] view at source ↗

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

60 extracted references · 45 canonical work pages

  1. [1]

    Gpt-4 technical report

    Josh Achiam, Steven Adler, Sandhini Agarwal, Lama Ah- mad, Ilge Akkaya, Florencia Leoni Aleman, Diogo Almeida, Janko Altenschmidt, Sam Altman, Shyamal Anadkat, et al. Gpt-4 technical report. arXiv preprint arXiv:2303.08774 ,

    arXiv

  2. [2]

    Computing re- ceptive fields of convolutional neural networks

    Andr ´e Araujo, Wade Norris, and Jack Sim. Computing re- ceptive fields of convolutional neural networks. Distill, 4 (11):e21, 2019. 2

    work page 2019

  3. [3]

    From detection of individual metastases to classification of lymph node status at the pa- tient level: the camelyon17 challenge

    Peter Bandi, Oscar Geessink, Quirine Manson, Mar- cory Van Dijk, Maschenka Balkenhol, Meyke Hermsen, Babak Ehteshami Bejnordi, Byungjae Lee, Kyunghyun Paeng, Aoxiao Zhong, et al. From detection of individual metastases to classification of lymph node status at the pa- tient level: the camelyon17 challenge. IEEE transactions on medical imaging, 38(2):550–5...

    2018

  4. [4]

    Diagnos- tic assessment of deep learning algorithms for detection of lymph node metastases in women with breast cancer

    Babak Ehteshami Bejnordi, Mitko Veta, Paul Johannes Van Diest, Bram Van Ginneken, Nico Karssemeijer, Geert Litjens, Jeroen AWM Van Der Laak, Meyke Hermsen, Quirine F Manson, Maschenka Balkenhol, et al. Diagnos- tic assessment of deep learning algorithms for detection of lymph node metastases in women with breast cancer. Jama, 318(22):2199–2210, 2017. 6

    work page 2017

  5. [5]

    Clinical-grade computational pathology using weakly supervised deep learning on whole slide images

    Gabriele Campanella, Matthew G Hanna, Luke Geneslaw, Allen Miraflor, Vitor Werneck 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(8):1301–1309, 2019. 1, 2

    work page 2019

  6. [6]

    Pathomic fusion: an integrated framework for fusing histopathology and genomic features for cancer diag- nosis and prognosis

    Richard J Chen, Ming Y Lu, Jingwen Wang, Drew FK Williamson, Scott J Rodig, Neal I Lindeman, and Faisal Mahmood. Pathomic fusion: an integrated framework for fusing histopathology and genomic features for cancer diag- nosis and prognosis. IEEE Transactions on Medical Imag- ing, 41(4):757–770, 2020

    work page 2020

  7. [7]

    Pan-cancer in- tegrative histology-genomic analysis via multimodal deep learning

    Richard J Chen, Ming Y Lu, Drew FK Williamson, Tiffany Y Chen, Jana Lipkova, Zahra Noor, Muhammad Shaban, Maha Shady, Mane Williams, Bumjin Joo, et al. Pan-cancer in- tegrative histology-genomic analysis via multimodal deep learning. Cancer Cell, 40(8):865–878, 2022. 1

    work page 2022

  8. [8]

    Towards a general-purpose foundation model for computational pathology

    Richard J Chen, Tong Ding, Ming Y Lu, Drew FK Williamson, Guillaume Jaume, Andrew H Song, Bowen Chen, Andrew Zhang, Daniel Shao, Muhammad Shaban, et al. Towards a general-purpose foundation model for computational pathology. Nature Medicine, 30(3):850–862,

Show all 60 references
  1. [9]

    Multiple instance learn- ing with center embeddings for histopathology classifica- tion

    Philip Chikontwe, Meejeong Kim, Soo Jeong Nam, Heoun- jeong Go, and Sang Hyun Park. Multiple instance learn- ing with center embeddings for histopathology classifica- tion. In Medical Image Computing and Computer Assisted Intervention–MICCAI 2020: 23rd International Conference...

    2020

  2. [10]

    Artificial intelligence and computational pathology

    Miao Cui and David Y Zhang. Artificial intelligence and computational pathology. Laboratory Investigation, 101(4): 412–422, 2021. 1

    2021

  3. [11]

    Beyond h&e: Unlocking pathological insights with polarization via self- supervised learning

    Yao Du, Jiaxin Zhuang, Xiaoyu Zheng, Jing Cong, Limei Guo, Chao He, Lin Luo, and Xiaomeng Li. Beyond h&e: Unlocking pathological insights with polarization via self- supervised learning. arXiv preprint arXiv:2503.05933, 2025. 2

    2025

  4. [12]

    The llama 3 herd of models

    Abhimanyu Dubey, Abhinav Jauhri, Abhinav Pandey, Ab- hishek Kadian, Ahmad Al-Dahle, Aiesha Letman, Akhil Mathur, Alan Schelten, Amy Yang, Angela Fan, et al. The llama 3 herd of models. arXiv preprint arXiv:2407.21783 ,

    arXiv

  5. [13]

    Fast: A dual-tier few-shot learning paradigm for whole slide image classification

    Kexue Fu, Xiaoyuan Luo, Linhao Qu, Shuo Wang, Ying Xiong, Ilias Maglogiannis, Longxiang Gao, and Manning Wang. Fast: A dual-tier few-shot learning paradigm for whole slide image classification. arXiv preprint arXiv:2409.19720, 2024. 2, 3

    2024 arXiv

  6. [14]

    Histgen: Histopathology re- port generation via local-global feature encoding and cross- modal context interaction

    Zhengrui Guo, Jiabo Ma, Yingxue Xu, Yihui Wang, Lian- sheng Wang, and Hao Chen. Histgen: Histopathology re- port generation via local-global feature encoding and cross- modal context interaction. In International Conference on Medical Image Computing and Computer-Assisted Inte...

    2024

  7. [15]

    Context matters: Query- aware dynamic long sequence modeling of gigapixel images

    Zhengrui Guo, Qichen Sun, Jiabo Ma, Lishuang Feng, Jinzhuo Wang, and Hao Chen. Context matters: Query- aware dynamic long sequence modeling of gigapixel images. arXiv preprint arXiv:2501.18984, 2025. 2

    2025 arXiv

  8. [16]

    Mscpt: Few-shot whole slide image classification with multi-scale and context- focused prompt tuning

    Minghao Han, Linhao Qu, Dingkang Yang, Xukun Zhang, Xiaoying Wang, and Lihua Zhang. Mscpt: Few-shot whole slide image classification with multi-scale and context- focused prompt tuning. arXiv preprint arXiv:2408.11505 ,

    arXiv

  9. [17]

    Computational pathology: a survey review and the way forward

    Mahdi S Hosseini, Babak Ehteshami Bejnordi, Vincent Quoc-Huy Trinh, Lyndon Chan, Danial Hasan, Xingwen Li, Stephen Yang, Taehyo Kim, Haochen Zhang, Theodore Wu, et al. Computational pathology: a survey review and the way forward. Journal of Pathology Informatics , page 100357,

  10. [18]

    Hˆ 2-mil: ex- ploring hierarchical representation with heterogeneous mul- tiple instance learning for whole slide image analysis

    Wentai Hou, Lequan Yu, Chengxuan Lin, Helong Huang, Rongshan Yu, Jing Qin, and Liansheng Wang. Hˆ 2-mil: ex- ploring hierarchical representation with heterogeneous mul- tiple instance learning for whole slide image analysis. InPro- ceedings of the AAAI conference on artificial...

    2022

  11. [19]

    A visual–language foundation model for pathology image analysis using medical twitter

    Zhi Huang, Federico Bianchi, Mert Yuksekgonul, Thomas J Montine, and James Zou. A visual–language foundation model for pathology image analysis using medical twitter. Nature medicine, 29(9):2307–2316, 2023. 2, 3, 8

    2023

  12. [20]

    Attention-based deep multiple instance learning

    Maximilian Ilse, Jakub Tomczak, and Max Welling. Attention-based deep multiple instance learning. In Inter- 9 national conference on machine learning, pages 2127–2136. PMLR, 2018. 1, 2

    2018

  13. [21]

    Model- ing dense multimodal interactions between biological path- ways and histology for survival prediction

    Guillaume Jaume, Anurag Vaidya, Richard J Chen, Drew FK Williamson, Paul Pu Liang, and Faisal Mahmood. Model- ing dense multimodal interactions between biological path- ways and histology for survival prediction. InProceedings of the IEEE/CVF Conference on Computer Vision and ...

    2024

  14. [22]

    Label-efficient deep learning in medical image anal- ysis: Challenges and future directions

    Cheng Jin, Zhengrui Guo, Yi Lin, Luyang Luo, and Hao Chen. Label-efficient deep learning in medical image anal- ysis: Challenges and future directions. arXiv preprint arXiv:2303.12484, 2023. 1

    2023 arXiv

  15. [23]

    Dual-stream multiple instance learning network for whole slide image classifica- tion with self-supervised contrastive learning

    Bin Li, Yin Li, and Kevin W Eliceiri. Dual-stream multiple instance learning network for whole slide image classifica- tion with self-supervised contrastive learning. In Proceed- ings of the IEEE/CVF conference on computer vision and pattern recognition, pages 14318–14328, 202...

    2021

  16. [24]

    Dt-mil: deformable transformer for multi-instance learning on histopathological image

    Hang Li, Fan Yang, Yu Zhao, Xiaohan Xing, Jun Zhang, Mingxuan Gao, Junzhou Huang, Liansheng Wang, and Jian- hua Yao. Dt-mil: deformable transformer for multi-instance learning on histopathological image. InMedical Image Com- puting and Computer Assisted Intervention–MICCAI 202...

    2021

  17. [25]

    Dynamic graph repre- sentation with knowledge-aware attention for histopathology whole slide image analysis

    Jiawen Li, Yuxuan Chen, Hongbo Chu, Qiehe Sun, Tian Guan, Anjia Han, and Yonghong He. Dynamic graph repre- sentation with knowledge-aware attention for histopathology whole slide image analysis. InProceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognitio...

    2024

  18. [26]

    Interventional bag multi-instance learning on whole-slide pathological images

    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 the IEEE/CVF Conference on Computer Vision and Pattern Recognition, pages 19830–19839, 2023

    2023

  19. [27]

    Data-efficient and weakly supervised computational pathology on whole- slide images

    Ming Y Lu, Drew FK 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(6):555–570,

  20. [28]

    Visual language pre- trained multiple instance zero-shot transfer for histopathol- ogy images

    Ming Y Lu, Bowen Chen, Andrew Zhang, Drew FK Williamson, Richard J Chen, Tong Ding, Long Phi Le, Yung- Sung Chuang, and Faisal Mahmood. Visual language pre- trained multiple instance zero-shot transfer for histopathol- ogy images. In Proceedings of the IEEE/CVF conference on c...

    2023

  21. [29]

    A visual- language foundation model for computational pathology

    Ming Y Lu, Bowen Chen, Drew FK Williamson, Richard J Chen, Ivy Liang, Tong Ding, Guillaume Jaume, Igor Odintsov, Long Phi Le, Georg Gerber, et al. A visual- language foundation model for computational pathology. Nature Medicine, 30(3):863–874, 2024. 2, 3, 4, 7, 8

    2024

  22. [30]

    Towards a generalizable pathology foundation model via unified knowledge distillation

    Jiabo Ma, Zhengrui Guo, Fengtao Zhou, Yihui Wang, Yingxue Xu, Yu Cai, Zhengjie Zhu, Cheng Jin, Yi Lin Xinrui Jiang, Anjia Han, et al. Towards a generalizable pathology foundation model via unified knowledge distillation. arXiv preprint arXiv:2407.18449, 2024. 2, 3, 8

    2024 arXiv

  23. [31]

    Digital pathology and artificial intelli- gence

    Muhammad Khalid Khan Niazi, Anil V Parwani, and Metin N Gurcan. Digital pathology and artificial intelli- gence. The lancet oncology, 20(5):e253–e261, 2019. 1

    2019

  24. [32]

    The rise of ai language pathologists: Exploring two-level prompt learning for few-shot weakly-supervised whole slide image classification

    Linhao Qu, Kexue Fu, Manning Wang, Zhijian Song, et al. The rise of ai language pathologists: Exploring two-level prompt learning for few-shot weakly-supervised whole slide image classification. Advances in Neural Information Pro- cessing Systems, 36, 2024. 1, 2, 3, 6, 7

    2024

  25. [33]

    Pathology-knowledge enhanced multi-instance prompt learning for few-shot whole slide image classification

    Linhao Qu, Dingkang Yang, Dan Huang, Qinhao Guo, Rongkui Luo, Shaoting Zhang, and Xiaosong Wang. Pathology-knowledge enhanced multi-instance prompt learning for few-shot whole slide image classification. arXiv preprint arXiv:2407.10814, 2024. 2, 3

    2024 arXiv

  26. [34]

    Transmil: Transformer based correlated multiple instance learning for whole slide image classification

    Zhuchen Shao, Hao Bian, Yang Chen, Yifeng Wang, Jian Zhang, Xiangyang Ji, et al. Transmil: Transformer based correlated multiple instance learning for whole slide image classification. Advances in neural information processing systems, 34:2136–2147, 2021. 1, 2, 3, 6, 7

    2021

  27. [35]

    Vila-mil: Dual-scale vision-language multiple instance learning for whole slide image classification

    Jiangbo Shi, Chen Li, Tieliang Gong, Yefeng Zheng, and Huazhu Fu. Vila-mil: Dual-scale vision-language multiple instance learning for whole slide image classification. In Proceedings of the IEEE/CVF Conference on Computer Vi- sion and Pattern Recognition, pages 11248–11258, 20...

    2024

  28. [36]

    Mor- phological prototyping for unsupervised slide representation learning in computational pathology

    Andrew H Song, Richard J Chen, Tong Ding, Drew FK Williamson, Guillaume Jaume, and Faisal Mahmood. Mor- phological prototyping for unsupervised slide representation learning in computational pathology. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Re...

    2024

  29. [37]

    Song, Richard J

    Andrew H. Song, Richard J. Chen, Guillaume Jaume, Anurag Jayant Vaidya, Alexander Baras, and Faisal Mah- mood. Multimodal prototyping for cancer survival predic- tion. In Proceedings of the 41st International Conference on Machine Learning, pages 46050–46073. PMLR, 2024. 1

    2024

  30. [38]

    A comprehensive survey of few-shot learning: Evolution, applications, challenges, and opportu- nities

    Yisheng Song, Ting Wang, Puyu Cai, Subrota K Mondal, and Jyoti Prakash Sahoo. A comprehensive survey of few-shot learning: Evolution, applications, challenges, and opportu- nities. ACM Computing Surveys, 55(13s):1–40, 2023. 1

    2023

  31. [39]

    Review the cancer genome atlas (tcga): an im- measurable source of knowledge

    Katarzyna Tomczak, Patrycja Czerwi ´nska, and Maciej Wiz- nerowicz. Review the cancer genome atlas (tcga): an im- measurable source of knowledge. Contemporary Oncol- ogy/Wsp´ołczesna Onkologia, 2015(1):68–77, 2015. 6

    2015

  32. [40]

    Deep learning in histopathology: the path to the clinic

    Jeroen Van der Laak, Geert Litjens, and Francesco Ciompi. Deep learning in histopathology: the path to the clinic. Na- ture medicine, 27(5):775–784, 2021. 1

    2021

  33. [41]

    A foundation model for clinical-grade computational pathology and rare cancers detection

    Eugene V orontsov, Alican Bozkurt, Adam Casson, George Shaikovski, Michal Zelechowski, Kristen Severson, Eric Zimmermann, James Hall, Neil Tenenholtz, Nicolo Fusi, et al. A foundation model for clinical-grade computational pathology and rare cancers detection. Nature medicine ...

    2024

  34. [42]

    Scl-wc: Cross-slide contrastive learning for weakly-supervised whole-slide image classifi- cation

    Xiyue Wang, Jinxi Xiang, Jun Zhang, Sen Yang, Zhongyi Yang, Ming-Hui Wang, Jing Zhang, Wei Yang, Junzhou Huang, and Xiao Han. Scl-wc: Cross-slide contrastive learning for weakly-supervised whole-slide image classifi- cation. Advances in neural information processing systems, 3...

    2022

  35. [43]

    A pathology foundation model for can- cer diagnosis and prognosis prediction

    Xiyue Wang, Junhan Zhao, Eliana Marostica, Wei Yuan, Ji- etian Jin, Jiayu Zhang, Ruijiang Li, Hongping Tang, Kanran Wang, Yu Li, et al. A pathology foundation model for can- cer diagnosis and prognosis prediction. Nature, pages 1–9,

  36. [44]

    Dsnet: A dual-stream framework for weakly- supervised gigapixel pathology image analysis

    Tiange Xiang, Yang Song, Chaoyi Zhang, Dongnan Liu, Mei Chen, Fan Zhang, Heng Huang, Lauren O’Donnell, and Wei- dong Cai. Dsnet: A dual-stream framework for weakly- supervised gigapixel pathology image analysis. IEEE Trans- actions on Medical Imaging, 41(8):2180–2190, 2022. 1, 2

    2022

  37. [45]

    Conghao Xiong, Hao Chen, Joseph J. Y . Sung, and Irwin King. Diagnose like a pathologist: transformer-enabled hierarchical attention-guided multiple instance learning for whole slide image classification. In Proceedings of the Thirty-Second International Joint Conference on Ar...

    2023

  38. [46]

    Conghao Xiong, Hao Chen, Hao Zheng, Dong Wei, Yefeng Zheng, Joseph J. Y . Sung, and Irwin King. Mome: Mix- ture of multimodal experts for cancer survival prediction. In International Conference on Medical Image Computing and Computer-Assisted Intervention , pages 318–328. Spri...

    2024

  39. [47]

    Conghao Xiong, Yi Lin, Hao Chen, Hao Zheng, Dong Wei, Yefeng Zheng, Joseph J. Y . Sung, and Irwin King. Takt: Target-aware knowledge transfer for whole slide image clas- sification. In International Conference on Medical Image Computing and Computer-Assisted Intervention, pages 503–

  40. [48]

    A whole-slide foundation model for digital pathology from real-world data

    Hanwen Xu, Naoto Usuyama, Jaspreet Bagga, Sheng Zhang, Rajesh Rao, Tristan Naumann, Cliff Wong, Zelalem Gero, Javier Gonz ´alez, Yu Gu, et al. A whole-slide foundation model for digital pathology from real-world data. Nature, pages 1–8, 2024. 2, 3

    2024

  41. [49]

    Multimodal optimal transport- based co-attention transformer with global structure con- sistency for survival prediction

    Yingxue Xu and Hao Chen. Multimodal optimal transport- based co-attention transformer with global structure con- sistency for survival prediction. In Proceedings of the IEEE/CVF International Conference on Computer Vision , pages 21241–21251, 2023. 1

    2023

  42. [50]

    A multimodal knowledge-enhanced whole-slide pathology foundation model

    Yingxue Xu, Yihui Wang, Fengtao Zhou, Jiabo Ma, Shu Yang, Huangjing Lin, Xin Wang, Jiguang Wang, Li Liang, Anjia Han, et al. A multimodal knowledge-enhanced whole-slide pathology foundation model. arXiv preprint arXiv:2407.15362, 2024. 2, 3

    2024 arXiv

  43. [51]

    Towards better understanding and better generalization of few-shot classification in histology images with contrastive learning

    Jiawei Yang, Hanbo Chen, Jiangpeng Yan, Xiaoyu Chen, and Jianhua Yao. Towards better understanding and better generalization of few-shot classification in histology images with contrastive learning. arXiv preprint arXiv:2202.09059,

    arXiv

  44. [52]

    Mambamil: En- hancing long sequence modeling with sequence reordering in computational pathology

    Shu Yang, Yihui Wang, and Hao Chen. Mambamil: En- hancing long sequence modeling with sequence reordering in computational pathology. In International Conference on Medical Image Computing and Computer-Assisted Interven- tion, pages 296–306. Springer, 2024. 1, 2

    2024

  45. [53]

    Prototypical multiple instance learning for predicting lymph node metastasis of breast cancer from whole-slide pathological images

    Jin-Gang Yu, Zihao Wu, Yu Ming, Shule Deng, Yuanqing Li, Caifeng Ou, Chunjiang He, Baiye Wang, Pusheng Zhang, and Yu Wang. Prototypical multiple instance learning for predicting lymph node metastasis of breast cancer from whole-slide pathological images. Medical Image Analysis...

    2023

  46. [54]

    Dtfd- mil: Double-tier feature distillation multiple instance learn- ing for histopathology whole slide image classification

    Hongrun Zhang, Yanda Meng, Yitian Zhao, Yihong Qiao, Xiaoyun Yang, Sarah E Coupland, and Yalin Zheng. Dtfd- mil: Double-tier feature distillation multiple instance learn- ing for histopathology whole slide image classification. In Proceedings of the IEEE/CVF conference on comp...

    2022

  47. [55]

    Prototypical information bottlenecking and dis- entangling for multimodal cancer survival prediction

    Yilan Zhang, Yingxue Xu, Jianqi Chen, Fengying Xie, and Hao Chen. Prototypical information bottlenecking and dis- entangling for multimodal cancer survival prediction. arXiv preprint arXiv:2401.01646, 2024. 1

    2024 arXiv

  48. [56]

    Diffusion-based virtual staining from polarimetric mueller matrix imaging

    Xiaoyu Zheng, Jing Wen, Jiaxin Zhuang, Yao Du, Jing Cong, Limei Guo, Chao He, Lin Luo, and Hao Chen. Diffusion-based virtual staining from polarimetric mueller matrix imaging. arXiv preprint arXiv:2503.01352, 2025. 2

    2025 arXiv

  49. [57]

    A graph-transformer for whole slide image classifica- tion

    Yi Zheng, Rushin H Gindra, Emily J Green, Eric J Burks, Margrit Betke, Jennifer E Beane, and Vijaya B Kolacha- lama. A graph-transformer for whole slide image classifica- tion. IEEE transactions on medical imaging , 41(11):3003– 3015, 2022. 1, 2

    2022

  50. [58]

    Cross-modal translation and alignment for survival analysis

    Fengtao Zhou and Hao Chen. Cross-modal translation and alignment for survival analysis. In Proceedings of the IEEE/CVF International Conference on Computer Vision , pages 21485–21494, 2023. 1

    2023

  51. [59]

    Learning to prompt for vision-language models

    Kaiyang Zhou, Jingkang Yang, Chen Change Loy, and Ziwei Liu. Learning to prompt for vision-language models. In- ternational Journal of Computer Vision, 130(9):2337–2348,

  52. [2022]

    large polygonal cells with clear cytoplasm

    4 11 F OCU S: Knowledge-enhanced Adaptive Visual Compression for Few-shot Whole Slide Image Classification Supplementary Material A. Pathology Prior Knowledge Prompt This section lists the language prompt generated by LLMs for each dataset, including TCGA-NSCLC, CAME- LYON, an...

Pith tools

Reviewed August 12, 2026 · model on record in the stance chip above.