Pith. sign in

REVIEW 4 major objections 4 minor 48 references

Aligning Knowledge Concepts to Whole Slide Images for Precise Histopathology Image Analysis

T0 review · 4 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 ConcepPath embeds GPT-4-induced expert concepts and learnable concepts into a frozen pathology vision-language space, claiming consistent gains over state-of-the-art weakly supervised methods on five TCGA whole-slide tasks.

desk verdict A promising concept-based MIL framework with informative ablations, but the headline 'significant' gains are not backed by significance testing and the concept-count tuning may be selecting on the test folds. read the letter →

arxiv 2411.18101 v1 pith:NBKKYVLJ submitted 2024-11-27 cs.CV cs.LG

Weiqin Zhao , Ziyu Guo , Yinshuang Fan , Yuming Jiang , Maximus Yeung , Lequan Yu This is my paper · ORCID
classification cs.CVcs.LG
keywords wholeslideimageanalysismultipleinstancelearningvision-languagemodelconcept-guidedaggregationGPT-4expertconceptsweaklysupervisedclassificationhistopathologyimmunotherapyresponseprediction
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

ConcepPath is a weakly supervised multiple-instance-learning framework that injects human expert knowledge into whole-slide image analysis. Instead of training only from slide labels, it asks GPT-4 to induce disease-specific histopathology concepts from medical literature and adds learnable data-driven concepts, then aligns WSI patches to both concept groups through a frozen CLIP-based pathology vision-language model. A two-stage concept-guided aggregation turns patch-level concept scores into a slide-level prediction. On five TCGA tasks (lung cancer subtyping, breast HER2 scoring, and three gastric immunotherapy-sensitive subtyping tasks) the paper reports consistent gains over seven baselines, with the largest reported jump a 6.23% AUC improvement on EBV-positive versus other gastric cancers. The claim is that expert knowledge, when grounded in literature and aligned in a pathology VLM space, is complementary to data-driven features and improves both accuracy and interpretability.

What carries the argument

The load-bearing object is the concept set itself: each concept is a text description paired with a learnable prompt prefix, embedded by the text encoder of a frozen CLIP-based pathology vision-language model. Instance-level concepts (expert plus data-driven) are matched to patch features by cosine attention weights in the first aggregation stage; bag-level expert class prompts are matched to instance-level concepts in the second stage; slide-adapters blend adapted features with original features before final class-prompt similarity. A mutual-distinctiveness loss pushes data-driven concepts away from each other so they add information rather than duplicate it.

What would settle it

Freeze the training pipeline but replace the text descriptions of the expert concepts with semantically unrelated strings (or random word salad) while keeping the same number of learnable tokens; if AUC stays at the same level as the real-concept version, then the expert-knowledge alignment is not carrying the performance. A complementary check is to have pathologists score whether the concept similarity maps localize the morphology they name: if the maps for 'keratinization' do not highlight keratinizing tumor regions, the claimed interpretable grounding is unsupported.

Watch

Extended reading notes

Core claim

The paper's central claim is that a WSI classifier can be built as a set of patch-level concept-scoring subtasks, where the concepts come from two complementary sources: expert concepts induced from medical literature by GPT-4 with manual verification, and purely learnable prompt vectors optimized on training data. In the shared embedding space of a frozen CLIP-based pathology vision-language model (QuiltNet by default, CONCH in supplementary), patch features are matched to concept embeddings, first stage aggregates patches into concept-specific bag features, second stage aggregates these into a bag representation weighted by each concept's correlation with bag-level class prompts, and slide-level prediction is the cosine similarity between adapted bag representations and class prompt embeddings. The paper reports that this two-stage, knowledge-augmented pipeline outperforms seven state-of-the-art MIL methods on all five evaluated tasks, with particularly large margins on the hardest gastric cancer immunotherapy-sensitive subtypes. It also shows that induced concepts beat directly queried concepts, and that learned concepts add 1.04–3.96 AUC points depending on task.

Load-bearing premise

The whole pipeline assumes that cosine similarity in the frozen pathology vision-language embedding space reliably indicates whether a medical concept is actually present in an image patch; if the embedding space does not align morphology to text for the target stains or tumor types, the concept guidance loses its meaning and the gains would come from the learnable adapters rather than from expert knowledge.

Editorial extensions

If this is right

  • On the five tested TCGA tasks, ConcepPath reports higher AUC and ACC than seven baselines, with a 6.23-point AUC gain for EBV vs. Others gastric cancer.
  • Induced concepts outperform directly queried concepts, so the paper concludes that grounding LLM-generated knowledge in literature and manual review matters for downstream accuracy.
  • Adding learnable data-driven concepts raises AUC by 1.04 (NSCLC), 1.16 (BRCA), and 3.96 (EBV vs. Others) points, and the optimal number grows with task difficulty, suggesting a trade-off between prior and learned knowledge.
  • Removing the bag-level guidance or the slide-adapters drops performance (e.g., 4.11 AUC points on MSI vs. Others without bag-level guidance), so both stages are claimed to contribute.
  • Concept similarity maps localize expert concepts on the slide and, per the pathologist collaborator, match established morphology, giving post-hoc interpretability beyond attention maps.

Reading between the lines

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

  • A natural extension the paper does not test is whether the same pipeline transfers to less common cancer types: the bottleneck would shift from model capacity to the quality and coverage of the medical literature used to induce concepts, so the approach is only as scalable as the concept-induction step.
  • If the 6.23% AUC gain on EBV vs. Others replicates in external cohorts, H&E-based screening could become a low-cost triage step for immune checkpoint inhibitor therapy in gastric cancer, reducing the need for IHC and PCR-based molecular tests.
  • Because the data-driven concepts are optimized prompts in the VLM space, they could be interpreted retroactively as candidate hypotheses for understudied disease mechanisms, provided the similarity maps are validated against expert annotations, which the paper leaves to future work.
  • The reported trade-off between expert and learned concepts suggests a selection rule the authors do not state: for tasks with sparse published knowledge, favor more learnable concepts; for well-characterized tasks, favor more expert concepts. This could be tested as an adaptive hyperparameter.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

4 major / 4 minor

Summary. The paper proposes ConcepPath, a weakly supervised whole-slide-image (WSI) classification framework that combines GPT-4-induced expert concepts from medical literature with learnable data-driven concepts, and aligns both to WSI patches through a frozen CLIP-based pathology vision-language model. The framework uses a two-stage concept-guided aggregation and slide-adapters before computing class-prompt similarities for prediction. The authors evaluate ConcepPath on five TCGA tasks (NSCLC subtyping, BRCA HER2 scoring, and three gastric cancer immunotherapy-related binary tasks) and report AUC/ACC improvements over seven baselines, including a 6.23% AUC gain on the EBV versus Others task. The paper also presents component ablations, comparisons of concept generation strategies, and qualitative interpretability heatmaps.

Significance. If the empirical claims hold, the paper makes a useful contribution by demonstrating that explicit human expert knowledge, induced via an LLM from medical literature, can be combined with learned concepts in a CLIP-based pathology VLM space to improve slide-level classification and provide interpretable similarity maps. The framework is novel in its decomposition of a WSI task into patch-level concept scoring subtasks, and the authors provide code, trained weights, and the induced concept lists, which are strengths for reproducibility. The component ablations (data-driven concepts, bag-level guidance, slide-adapters) are informative and support the design's internal logic. However, the central empirical claim of 'significantly outperformed' is not supported by appropriate statistical evidence, and the evaluation protocol has potential selection biases that must be addressed before the results can be accepted as stated.

major comments (4)
  1. [Results and Evaluation Protocol] The Abstract and Results claim that ConcepPath 'significantly outperformed' previous methods, but the manuscript reports only mean AUC/ACC over patient-level five-fold cross-validation without confidence intervals, per-fold values, or paired significance tests. On the EBV versus Others task, which has only 26 positive cases, the reported 6.23% AUC gain over the best baseline could be moved by a handful of patients. Please report per-fold results with confidence intervals and run appropriate statistical tests (e.g., DeLong's test or paired bootstrap) for the AUC comparisons, or temper the language to avoid the unsupported claim of significance.
  2. [Training Details] The Training Details state that the number of learned instance-level concepts was 'tuned from {2, 4, 6, 8, 10, 12} for each target class,' and Figure 3a reports the best per-task values (e.g., 4 for NSCLC, 8 for EBV). The manuscript does not specify whether this tuning was performed on the same test folds whose AUCs are then reported, nor does it describe a nested validation scheme. If the test folds were used to select the number, the reported results are selected maxima and the comparison against fixed-architecture baselines is biased. Please specify the tuning protocol clearly and report results for all candidate concept numbers for all tasks, ideally with nested cross-validation or a separate validation set.
  3. [Results, Baseline Models] The TOP baseline is reported as the higher performance of the author's implementation and the authors' re-implementation. Selecting the higher of two implementations for a comparison baseline introduces favorable selection and is not a pre-specified protocol. Please report both implementations separately and either justify the choice of the higher one or use a single pre-registered implementation for the comparison.
  4. [Methods, Eq. (5) and Discussion] The concept-guided aggregation (Eqs. 5–8) and final prediction (Eq. 13) rely on the assumption that cosine similarities between patch features Z and concept text embeddings C in the frozen CLIP-based pathology VLM space faithfully reflect the presence of morphological concepts. The Discussion acknowledges the domain shift between the encoder training data and downstream task data, but the paper provides no quantitative evidence about the reliability of this alignment for the target stains or cancer types. Please add an analysis that isolates this assumption, such as comparing concept similarity maps against pathologist region annotations, or reporting a variant that refines the alignment during training; otherwise it remains unclear whether the observed gains come from expert knowledge or from the learnable adapters and concepts.
minor comments (4)
  1. [Comparison of Different Expert Concept Extraction Strategies] The text contains a duplicated phrase: 'as as illustrated in Figure 2b' should read 'as illustrated in Figure 2b.'
  2. [Discussion] The model name 'PatChat' appears twice in the Discussion and should be 'PathChat' to match the cited work.
  3. [Figure 3c caption] The caption contains a duplicated phrase: 'using different CLIP-based vision-language models as ConcepPath's basic component for aligning for aligning histopathology images' should have a single 'for aligning.'
  4. [Results, Figure 2a] The caption of Figure 2a does not define 'TOP*'; the explanation is given only in the body text. Please add the definition to the caption for self-containment.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the concept pipeline is trained from external knowledge and data, not defined in terms of its outputs.

full rationale

ConcepPath's derivation chain is self-contained rather than circular. Expert instance-level concepts are induced by GPT-4 from external medical literature (Methods, Inducing Expert Concepts with LLM), and data-driven concepts are optimized from training data (Eqs. 2-4). The aggregation equations (Eqs. 5-8) compute attention and similarity weights from patch features and concept embeddings, and the final prediction (Eq. 13) is a trained similarity classifier; none of these equations injects the slide-level test label as a construction input. The number of learned concepts is a hyperparameter tuned per task, and selecting the best value may inflate reported AUC, but that is an evaluation-protocol concern, not a case of a fitted parameter being renamed a prediction. The method is benchmarked against external baseline architectures, and the claimed gains are empirical rather than derived from self-citations. No load-bearing self-citation or imported uniqueness theorem appears in the text. The Discussion's caveat about frozen encoders is a limitation, not a circular step.

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

No new physical or ontological entities are introduced. The expert concepts are textual prompts derived from medical literature, and the data-driven concepts are learned prompt vectors. The main undeclared inputs are the alignment quality of the frozen pathology VLM, the reliability of GPT-4 induction, and the per-task hyperparameter choices.

free parameters (4)
  • number of learned data-driven concepts per class = 4 for NSCLC and BRCA, up to 8 for EBV vs. Others
    Tuned from {2,4,6,8,10,12} per task; the selection procedure is not fully described and may have been chosen on the test folds.
  • alpha and beta residual blending weights in slide-adapters = not reported
    Adjustable hyperparameters in Eqs. 11 and 12; no values or tuning details are given in the main text.
  • Softmax temperature tau in Eq. 13 = not reported
    Temperature controls prediction sharpness; the value is not specified in the main text.
  • number of expert concepts per class = 26
    Chosen by the authors after manual merging and deletion of GPT-4 outputs; the choice is not derived from data.
assumptions (4)
  • domain assumption GPT-4, when prompted with selected medical literature, induces accurate and visually grounded expert concepts.
    The entire knowledge-injection step depends on this; the authors acknowledge limitations for rare or novel cancer types.
  • domain assumption The frozen CLIP-based pathology VLM embedding space aligns text and histology patches sufficiently for cosine similarity to serve as attention weights.
    Eq. 5 computes attention as Softmax(Z * C_ins^T); the encoders are frozen, and domain shift is acknowledged in the Discussion.
  • domain assumption TCGA slide labels are reliable ground truth for the five classification tasks.
    The paper relies on TCGA diagnostic and molecular labels without independent verification; this is standard practice in the field.
  • domain assumption The manually collected medical literature is representative and unbiased for each target class.
    Literature was collected via search engines and filtered by journal reputation; the authors note that divergent expert opinions remain a challenge.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Aligning Knowledge Concepts to Whole Slide Images for Precise Histopathology Image Analysis." pith.science (2026). https://pith.science/paper/NBKKYVLJ

@misc{pith2026241118101,
  author       = {Pith},
  title        = {Pith review of: Aligning Knowledge Concepts to Whole Slide Images for Precise Histopathology Image Analysis},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/NBKKYVLJ}},
  note         = {Machine review of arXiv:2411.18101}
}
read the original abstract

Due to the large size and lack of fine-grained annotation, Whole Slide Images (WSIs) analysis is commonly approached as a Multiple Instance Learning (MIL) problem. However, previous studies only learn from training data, posing a stark contrast to how human clinicians teach each other and reason about histopathologic entities and factors. Here we present a novel knowledge concept-based MIL framework, named ConcepPath to fill this gap. Specifically, ConcepPath utilizes GPT-4 to induce reliable diseasespecific human expert concepts from medical literature, and incorporate them with a group of purely learnable concepts to extract complementary knowledge from training data. In ConcepPath, WSIs are aligned to these linguistic knowledge concepts by utilizing pathology vision-language model as the basic building component. In the application of lung cancer subtyping, breast cancer HER2 scoring, and gastric cancer immunotherapy-sensitive subtyping task, ConcepPath significantly outperformed previous SOTA methods which lack the guidance of human expert knowledge.

Figures

Figures reproduced from arXiv: 2411.18101 by the authors.

Figure 1
Figure 1. Overview of ConcepPath framework. a, In real clinical processes, pathologists apply their expert knowledge to reason about histopathologic entities and factors to make a diagnosis. b, ConcepPath utilizes a large language model like GPT-4 to induce expert concepts related to diagnosis from medical literature and integrate this knowledge into an automated WSI analysis pipeline through the CLIP-based pathology vision-l… view at source ↗
Figure 2
Figure 2. Performance and expert concept generation comparison of ConcepPath. [PITH_FULL_IMAGE:figures/full_fig_p016_2.png] view at source ↗
Figure 3
Figure 3. Investigation of proposed components in ConcepPath. a, [PITH_FULL_IMAGE:figures/full_fig_p018_3.png] view at source ↗
Figures from the paper (1 more)
Figure 4
Figure 4. Figure 4: Visualizations of ConcepPath and baseline method. Instance-level expert concept sim￾ilarity maps. The slides are accurately identified as the lung squamous cell carcinoma (LUSC) and lung adenocarcinoma (LUAD) subtype, respectively. In comparison to the CLAM attention m…

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

48 extracted references · 44 canonical work pages

  1. [1]

    Weakly supervised deep learning for whole slide lung cancer image analysis.IEEE transactions on cybernetics50, 3950–3962 (2019)

    Wang, X.et al. Weakly supervised deep learning for whole slide lung cancer image analysis.IEEE transactions on cybernetics50, 3950–3962 (2019). 12

    work page 2019

  2. [2]

    M., Rong, R., Zhan, X

    Wang, S., Yang, D. M., Rong, R., Zhan, X. & Xiao, G. Pathology image analysis using segmentation deep learning algorithms.The American journal of pathology189, 1686–1698 (2019)

    work page 2019

  3. [3]

    et al.Classificationandmutationpredictionfromnon–smallcelllungcancerhistopathology images using deep learning.Nature medicine 24, 1559–1567 (2018)

    Coudray,N. et al.Classificationandmutationpredictionfromnon–smallcelllungcancerhistopathology images using deep learning.Nature medicine 24, 1559–1567 (2018)

    work page 2018

  4. [4]

    & Sun, J.Deep residual learning for image recognition, 770–778 (2016)

    He, K., Zhang, X., Ren, S. & Sun, J.Deep residual learning for image recognition, 770–778 (2016)

    work page 2016

  5. [5]

    Riasatian, A.et al.Fine-tuning and training of densenet for histopathology image representation using tcga diagnostic slides.Medical Image Analysis70, 102032 (2021)

    work page 2021

  6. [6]

    Chen,R.J. et al. Scaling vision transformers to gigapixel images via hierarchical self-supervised learning, 16144–16155 (2022)

    work page 2022

  7. [7]

    Wang, X. et al. Transformer-based unsupervised contrastive learning for histopathological image classification. Medical image analysis81, 102559 (2022)

    work page 2022

  8. [8]

    & Welling, M.Attention-based deep multiple instance learning, 2127–2136 (PMLR, 2018)

    Ilse, M., Tomczak, J. & Welling, M.Attention-based deep multiple instance learning, 2127–2136 (PMLR, 2018)

    work page 2018

Show all 48 references
  1. [9]

    Hou, W. et al. H2-mil: Exploring hierarchical representation with heterogeneous multiple instance learning for whole slide image analysis, 933–941 (2022)

    2022

  2. [10]

    Guan, Y. et al. Node-aligned graph convolutional network for whole-slide image representation and classification, 18813–18823 (2022)

    2022

  3. [11]

    H., Cendra, F

    Chan, T. H., Cendra, F. J., Ma, L., Yin, G. & Yu, L.Histopathology whole slide image analysis with heterogeneous graph representation learning, 15661–15670 (2023)

    2023

  4. [12]

    & Song, Z

    Qu, L., Luo, X., Fu, K., Wang, M. & Song, Z. The rise of ai language pathologists: Exploring two- level prompt learning for few-shot weakly-supervised whole slide image classification.arXiv preprint arXiv:2305.17891 (2023)

    2023 arXiv

  5. [13]

    & Rajpoot, N.Multiple instance captioning: Learning representations from histopathology textbooks and articles, 16549–16559 (2021)

    Gamper, J. & Rajpoot, N.Multiple instance captioning: Learning representations from histopathology textbooks and articles, 16549–16559 (2021)

    2021

  6. [14]

    Lu, M. Y. et al. Visual language pretrained multiple instance zero-shot transfer for histopathology images, 19764–19775 (2023)

    2023

  7. [15]

    Radford, A. et al. Learning transferable visual models from natural language supervision, 8748–8763 (PMLR, 2021)

    2021

  8. [16]

    & Zou, J

    Huang, Z., Bianchi, F., Yuksekgonul, M., Montine, T. & Zou, J. Leveraging medical twitter to build a visual–language foundation model for pathology ai.bioRxiv 2023–03 (2023)

    2023

  9. [17]

    Ikezogwo, W. O. et al. Quilt-1m: One million image-text pairs for histopathology. arXiv preprint arXiv:2306.11207 (2023)

    2023 arXiv

  10. [18]

    Lu, M. Y. et al. Towards a visual-language foundation model for computational pathology.arXiv preprint arXiv:2307.12914 (2023)

    2023 arXiv

  11. [19]

    Truhn, D., Reis-Filho, J. S. & Kather, J. N. Large language models should be used as scientific reasoning engines, not knowledge databases.Nature Medicine1–2 (2023)

    2023

  12. [20]

    Gpt-4 is here: what scientists think.Nature 615, 773 (2023)

    Sanderson, K. Gpt-4 is here: what scientists think.Nature 615, 773 (2023)

    2023

  13. [21]

    Echle, A. et al. Deep learning in cancer pathology: a new generation of clinical biomarkers.British journal of cancer124, 686–696 (2021)

    2021

  14. [22]

    Mandair, D., Reis-Filho, J. S. & Ashworth, A. Biological insights and novel biomarker discovery through deep learning approaches in breast cancer histopathology.NPJ Breast Cancer9, 21 (2023)

    2023

  15. [23]

    Wolff, A.et al.American society of clinical oncology; college of american pathologists. recommendations for human epidermal growth factor receptor 2 testing in breast cancer: American society of clinical oncology/collegeofamericanpathologistsclinicalpracticeguidelineupdate. J ...

    2013

  16. [24]

    Lu, W. et al. Slidegraph+: Whole slide image level graphs to predict her2 status in breast cancer. Medical Image Analysis80, 102486 (2022)

    2022

  17. [25]

    Kelly, R. J. Immunotherapy for esophageal and gastric cancer.American society of clinical oncology educational book37, 292–300 (2017)

    2017

  18. [26]

    & Ushiku, T

    Hinata, M. & Ushiku, T. Detecting immunotherapy-sensitive subtype in gastric cancer using histologic image-based deep learning.Scientific reports 11, 22636 (2021)

    2021

  19. [27]

    Thirty years of epstein-barr virus-associated gastric carcinoma.Virchows Archiv 476, 353–365 (2020)

    Fukayama, M.et al. Thirty years of epstein-barr virus-associated gastric carcinoma.Virchows Archiv 476, 353–365 (2020)

    2020

  20. [28]

    L., Lohse, C

    Grogg, K. L., Lohse, C. M., Pankratz, V. S., Halling, K. C. & Smyrk, T. C. Lymphocyte-rich gastric cancer: associations with epstein-barr virus, microsatellite instability, histology, and survival.Modern 13 pathology 16, 641–651 (2003)

    2003

  21. [29]

    Arai, T. et al. Frequent microsatellite instability in papillary and solid-type, poorly differentiated adenocarcinomas of the stomach.Gastric Cancer 16, 505–512 (2013)

    2013

  22. [30]

    Yao, J., Zhu, X., Jonnagaddala, J., Hawkins, N. J. & Huang, J. Whole slide images based cancer survival prediction using attention guided deep multiple instance learning networks.Medical Image Anal. 65, 101789 (2020). URL https://doi.org/10.1016/j.media.2020.101789

    2020

  23. [31]

    Lu, M. Y.et al. Data-efficient and weakly supervised computational pathology on whole-slide images. Nature biomedical engineering5, 555–570 (2021)

    2021

  24. [32]

    Zheng, Y., Gindra, R., Betke, M., Beane, J. E. & Kolachalama, V. B. A deep learning based graph- transformer for whole slide image classification.medRxiv 2021–10 (2021)

    2021

  25. [33]

    Advances in neural information processing systems34, 2136–2147 (2021)

    Shao, Z.et al.Transmil: Transformer based correlated multiple instance learning for whole slide image classification. Advances in neural information processing systems34, 2136–2147 (2021)

    2021

  26. [34]

    Language in a bottle: Language model guided concept bottlenecks for interpretable image classification, 19187–19197 (2023)

    Yang, Y.et al. Language in a bottle: Language model guided concept bottlenecks for interpretable image classification, 19187–19197 (2023)

    2023

  27. [35]

    Learning concise and descriptive attributes for visual recognition, 3090–3100 (2023)

    Yan, A.et al. Learning concise and descriptive attributes for visual recognition, 3090–3100 (2023)

    2023

  28. [36]

    Microsatelliteinstabilityingastriccancer:Molecularbases,clinical perspectives,andnewtreatmentapproaches(2018)

    Ratti,M.,Lampis,A.&Hahne,J.C. Microsatelliteinstabilityingastriccancer:Molecularbases,clinical perspectives,andnewtreatmentapproaches(2018). URLhttps://pubmed.ncbi.nlm.nih.gov/30173350/

    2018

  29. [37]

    Benchmarking pathclip for pathology image analysis.Journal of Imaging Informatics in Medicine1–17 (2024)

    Zheng, S.et al. Benchmarking pathclip for pathology image analysis.Journal of Imaging Informatics in Medicine1–17 (2024)

    2024

  30. [38]

    Lu, M. Y.et al. A multimodal generative ai copilot for human pathology.Nature 1–3 (2024)

    2024

  31. [39]

    Zhou, K., Yang, J., Loy, C. C. & Liu, Z. Learning to prompt for vision-language models.International Journal of Computer Vision130, 2337–2348 (2022)

    2022

  32. [40]

    Clip-adapter: Better vision-language models with feature adapters (2021)

    Gao, P.et al. Clip-adapter: Better vision-language models with feature adapters (2021). 2110.04544

    2021 arXiv

  33. [41]

    & Lozano-Pérez, T.Attention is all you need, 570–576 (1998)

    Maron, O. & Lozano-Pérez, T.Attention is all you need, 570–576 (1998)

    1998

  34. [42]

    Hou, L. et al. Patch-based convolutional neural network for whole slide tissue image classification, 2424–2433 (2016)

    2016

  35. [43]

    & Zhou, Z.-H.Deep miml network, Vol

    Feng, J. & Zhou, Z.-H.Deep miml network, Vol. 31 (2017)

    2017

  36. [44]

    Clinical-grade computational pathology using weakly supervised deep learning on whole slide images.Nature medicine 25, 1301–1309 (2019)

    Campanella, G.et al. Clinical-grade computational pathology using weakly supervised deep learning on whole slide images.Nature medicine 25, 1301–1309 (2019)

    2019

  37. [45]

    Xu, G. et al. Camel: A weakly supervised learning framework for histopathology image segmentation, 10682–10691 (2019)

    2019

  38. [46]

    Weakly-supervised learning for lung carcinoma classification using deep learning

    Kanavati, F.et al. Weakly-supervised learning for lung carcinoma classification using deep learning. Scientific reports 10, 9297 (2020)

    2020

  39. [47]

    Zhu, W., Lou, Q., Vang, Y. S. & Xie, X.Deep multi-instance networks with sparse label assignment for whole mammogram classification, 603–611 (Springer, 2017)

    2017

  40. [48]

    An image of {cancer subtype B}

    Li, B., Li, Y. & Eliceiri, K. W.Dual-stream multiple instance learning network for whole slide image classification with self-supervised contrastive learning, 14318–14328 (2021). 14 GPT-4 Medical Literature Pathologist Microscope Human Diagnosis Related Concepts … �𝒚𝒚 Ours: LL...

    2021

Pith tools

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