Pith. sign in

REVIEW 3 major objections 5 minor 1 cited by

Diagnostic Text-guided Representation Learning in Hierarchical Classification for Pathological Whole Slide Image

T0 review · 3 major / 5 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 PathTree turns whole-slide diagnosis into a text-guided binary tree and reports the best hierarchical F1 on three pathology datasets.

desk verdict PathTree is a well-specified text-guided hierarchical WSI classifier with consistent, if modest, gains; the SOTA claim needs stronger evaluation before it lands. read the letter →

arxiv 2411.10709 v1 pith:FIADBQKR submitted 2024-11-16 cs.CV

Jiawen Li , Qiehe Sun , Renao Yan , Yizhi Wang , Yuqiu Fu , Yani Wei , Tian Guan , Huijuan Shi
show 2 more authors
Yonghonghe He Anjia Han
This is my paper · ORCID
classification cs.CV
keywords wholeslideimagehierarchicalclassificationtext-guidedrepresentationlearningbinarydecisiontreepathologicaldiagnosisslide-levellabelsmultipleinstancevision-languagealignment
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

PathTree claims that hard-to-separate pathology classes—such as the progressive lung adenocarcinoma subtypes AAH, AIS, MIA, and IAC—are recognized more accurately when a model is forced to reason along the same coarse-to-fine decision tree a pathologist uses, rather than treating every class as an independent flat label. The paper converts each classification task into a binary tree, writes a professional pathological description for every node, and lets those texts steer how patch features extracted from a whole slide image are pooled into a global slide representation. Prediction is made by comparing the slide representation with the fine-grained text prompts, and two tree-specific losses keep the slide aligned with every node on its diagnostic path while pushing it away from sibling and unrelated classes. On three datasets—an in-house lung cryosection set, the public PANDA prostate grading set, and the public BRACS breast subtyping set—PathTree reports the best hierarchical F1 scores and competitive or best fine-grained accuracy and AUC against eight whole-slide baselines. If this is right, slide-level diagnosis can be improved without pixel-level annotation by supplying the model with the same diagnostic structure and vocabulary that pathologists use.

What carries the argument

The load-bearing mechanism is the text-guided tree-like aggregator, a recursive bottom-up fusion defined as $b_\gamma = s_\alpha b_\alpha + s_\beta b_\beta + i_\gamma$, where $i_\gamma$ is the slide embedding of parent node $\gamma$, $b_\alpha$ and $b_\beta$ are the fused embeddings of its two children, and the fusion weights are softmax scores of slide-text similarities: $s_\alpha = \frac{\exp(b_\alpha t_\alpha^\top)}{\exp(b_\alpha t_\alpha^\top)+\exp(b_\beta t_\beta^\top)}$, $s_\beta=1-s_\alpha$. This makes the text tree control which visual evidence accumulates into the root. Two losses reinforce the hierarchy: path alignment, $L_{\text{path}} = \frac{1}{|P|}\sum_{k\in P}\|g - t_k\|_2^2$, pulls the global slide feature $g$ toward every text embedding $t_k$ on its root-to-leaf path, and tree-aware matching combines triplet losses with margins ordered as $\lambda_{\text{leaf}} > \lambda_{\text{sibling}} > \lambda_{\text{parent}}$ so that sibling nodes are kept closer than unrelated leaves but farther than the parent.

What would settle it

A controlled run on any of the three datasets with the same tree structure and encoders but with the professional text prompts replaced by pathology phrases randomly assigned to nodes would settle it: if hierarchical F1 and AUC stay flat, then text semantics are not doing the work, and the reported gains come from the tree pooling itself.

Watch

Extended reading notes

Core claim

At the center of the paper is a new way to use text in weakly supervised slide classification: the diagnostic description is not just a label or a prompt that selects a feature, it is the routing rule for hierarchical pooling. PathTree begins with a hand-designed binary tree whose nodes are coarse and fine categories, each described in pathologist-verified language. Text embeddings for all nodes are passed through a bidirectional tree graph so sibling and parent semantics inform one another. The image side produces one slide embedding per tree node using two attention modules, and a text-guided aggregator fuses child embeddings into their parent with weights computed from slide-text similarity; the root embedding is the global slide feature. Classification scores are cosine similarities between that global feature and the fine-grained text prompts, and training adds a path-alignment loss and a tree-aware triplet loss. The claimed net effect is that, compared with flat multiple-instance baselines, PathTree improves hierarchical F1 on all three datasets and improves or matches fine-grained accuracy and AUC, with the detailed text prompts outperforming plain class names in ablation.

Load-bearing premise

Everything rests on the premise that the hand-built binary trees and the expert-written text descriptions actually match how pathologists distinguish the classes; if a tree groups the wrong categories or a description does not line up with the visual features, the text-guided pooling and both alignment losses will systematically misdirect the slide representation.

Editorial extensions

If this is right

  • If PathTree is right, hierarchical labels are not just extra supervision but a better description of the task, so future whole-slide models should be evaluated with hierarchical precision, recall, and F1, not only flat accuracy.
  • If the gains hold, pathologist-written text acts as a cheap source of expert knowledge that reduces the number of labeled slides needed, since the few-shot experiments show PathTree improving most at 4, 8, and 16 shots per class.
  • If the ablation is representative, prompt design directly changes clinical performance, meaning that replacing generic templates with professional pathological descriptions is a concrete way to steer model behavior.
  • If the method generalizes, any multi-class pathology task with an expert decision tree can be converted to this scheme without pixel-level annotation, including new cancer grading or subtyping tasks.

Reading between the lines

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

  • Beyond the paper: the same design could learn the tree from data rather than from experts, which would test whether the hierarchy itself, rather than the text encoder, drives the reported gains.
  • Beyond the paper: because prediction is computed by slide-text similarity, PathTree classes are inherently nameable; a natural extension is zero-shot evaluation on a new hospital's categories using only written descriptions and no retraining.
  • Beyond the paper: the triplet margins encode a claim about semantic distance, with parent closer than sibling; these margins could be calibrated per dataset from pathologist agreement rates, an option the paper does not explore.
  • Beyond the paper: the Nyström variant improves on SYSFL and PANDA but not BRACS, hinting that global-context aggregation helps some tasks while local high-malignancy regions matter more in breast subtyping; a testable extension is to choose the aggregation module per node type.
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

3 major / 5 minor

Summary. The paper proposes PathTree, a weakly supervised WSI classification method that organizes fine-grained classes into a binary tree and uses expert-written pathological text descriptions for each node. Patch features are extracted with PLIP, multiple slide-level representations are generated either by multi-gated attention or by multi-head Nyström attention, and a text-guided recursive aggregator fuses child embeddings along the tree path. Training combines cross-entropy on text-slide cosine similarity (Eq. 11) with a path-alignment loss (Eq. 8) and a tree-aware triplet loss (Eq. 10). The method is evaluated on an internal lung cryosection dataset, the PANDA prostate grading dataset, and the BRACS breast subtyping dataset against eight WSI baselines, with additional few-shot comparisons against Linear-Probe, CoOp, and TOP.

Significance. The core idea—encoding pathologist-derived binary diagnostic trees and using text embeddings both to guide attention and to serve as the classification weight matrix—is a reasonable and fairly original combination for hierarchical WSI classification. The paper includes a broad baseline comparison, three datasets, pathologist-verified prompt construction, and systematic ablations of text prompts, graph encoders, patch/text encoders, and loss coefficients. The hierarchical F1 improvements over the best planar baselines are directionally consistent across datasets, and the few-shot experiments are a useful addition. However, the evidence as presented is not yet sufficient to support the headline claim of consistent superiority: the hierarchical precision and recall formulas appear to be misprinted, and the evaluation protocol does not state whether folds are patient-stratified, allowing patient-level leakage to inflate the reported margins. If the metric definition is corrected and the evaluation is redone with patient-exclusive, significance-tested folds, PathTree would make a valuable contribution to weakly supervised hierarchical WSI classification.

major comments (3)
  1. [Section 4.3, Eqs. (13) and (14)] The printed formulas for H-Precision and H-Recall are identical: both have denominator Σ_j |Ct(j)|. The standard definitions use the predicted label set Σ_j |Cp(j)| in the denominator of H-Precision and the true label set in the denominator of H-Recall. As written, H-Precision and H-Recall would always be equal, which contradicts the different values reported in Table 3. Because H-F1 is the paper's headline metric, the formula must be corrected and the computations in Table 3 must be verified against the intended definitions.
  2. [Sections 4.1 and 4.4 (evaluation protocol)] The paper reports only "5-fold cross-validation" and does not state whether folds are stratified at the patient level. BRACS consists of 547 WSIs from 189 patients, and PANDA is a multi-biopsy-per-patient cohort in the public PANDA challenge. If the split is performed at the slide level, WSIs from the same patient can appear in both training and test folds. Because PathTree adds trainable components beyond the baselines (the tree-structured GAT prompt encoder and multiple attention heads), it may benefit disproportionately from such leakage. The authors should specify the split granularity and, if patient-level stratification was not used, re-run the evaluation with patient-exclusive folds.
  3. [Tables 2 and 3, Section 4.5.1] The reported margins over the best baselines are often small relative to their standard errors: for PANDA H-F1, PathTree obtains 73.29±0.50 versus CLAM's 73.22±0.37; for BRACS H-F1, 71.09±3.97 versus DSMIL's 69.72±3.30. Many fine-grained ACC differences in Table 2 also overlap within one standard error. No paired significance test is reported. Because the paper claims PathTree is "consistently competitive" and "outperforms" state-of-the-art methods, the authors should provide paired per-fold comparisons (for example, a paired t-test or corrected repeated k-fold test) and temper the conclusion accordingly if the differences are not statistically reliable.
minor comments (5)
  1. [Section 4.5.2] The text refers to "IUSP" when the intended abbreviation is "ISUP" (International Society of Urological Pathology).
  2. [Figure 9 and its caption] The figure axes are labeled 0.0 to 1.0 while the caption and surrounding text report results in percent; please clarify whether the plotted values are normalized scores or percentages.
  3. [Section 4.1 and Figure 9] The BRACS dataset is described as containing 547 WSIs in Section 4.1 and Table 1, but the Figure 9 panel header states "549 slides"; the count should be made consistent.
  4. [Abstract and Section 4.1] Minor language issues: "messages information" should be "exchanges information", and "slice-level labels" should be "slide-level labels".
  5. [Sections 3.5 and 3.6] The same text embeddings appear both as targets in the path-alignment and tree-aware losses (Eqs. 8 and 10) and as classifier weights in Eq. (11). This is a legitimate prompt-tuning formulation, but the paper should state explicitly that these losses regularize the representation toward the prediction head rather than provide fully independent external text supervision.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the slide-text alignment and slide-text similarity prediction form a standard supervised metric-learning loop, and the only self-citation is a decorative reference to an elementary binary-tree property.

full rationale

PathTree's prediction rule (Eq. 11) uses cosine similarity between a global slide embedding g and text embeddings t_j, while Eq. 8 (Lpath) and Eqs. 9-10 (Lmatch) pull g toward the true class's text embeddings during training. This is a closed training loop only in the same sense as any supervised classifier whose weights are optimized on the training set; since evaluation is 5-fold cross-validation on held-out folds, the test predictions are not forced by the training objective. The text prompts are external expert-written pathological descriptions (Section 3.1), not derived from the slide labels, and the binary-tree hierarchy is an explicit modeling choice grounded in OncoTree and pathologist practice, not a result imported from the authors' prior work. The only self-citation, 'According to the properties of the binary tree (Li et al. 2023a), we denote the number of fine-grained categories as N and the number of coarse-grained categories as N-1' (Section 3.1), asserts an elementary full-binary-tree property (internal nodes = leaves - 1); it is decorative and not load-bearing. No equation reduces to its inputs by construction, and no fitted parameter is renamed as a prediction. Concerns about patient-level splits or significance testing in Tables 2-4 are evaluation-protocol risks, not circularity.

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

PathTree rests on manually designed hierarchical trees, expert-written text prompts, frozen PLIP embeddings, and a set of hand-set loss weights and margins. These are empirical inputs, not derived results, and they explain why the method is a contribution of engineering rather than a first-principles derivation.

free parameters (6)
  • Temperature tau (Eq. 11) = initialized 0.07, learned
    Scales cosine similarity logits; learned during training, affects prediction confidence.
  • Margin lambda_parent = 0.002
    Manually set triplet margin for parent negative text; hand-tuned.
  • Margin lambda_sibling = 0.1
    Manually set triplet margin for sibling negative text; hand-tuned.
  • Margin lambda_leaf = 0.2
    Manually set triplet margin for other fine-grained negative texts; hand-tuned.
  • Loss coefficient mu_match = 1.0 (ablated)
    Weight for tree-aware matching loss; set to 1 in main experiments, chosen by validation in ablation.
  • Loss coefficient mu_path = 1.0 (ablated)
    Weight for path alignment loss; set to 1 in main experiments, chosen by validation in ablation.
assumptions (4)
  • domain assumption The hand-designed binary tree structure per dataset accurately reflects pathologists' diagnostic steps.
    The method is built entirely around the hand-designed trees in Figure 7; if the hierarchy is wrong, the aggregation path and losses force wrong inductive biases.
  • domain assumption Expert-written diagnostic text descriptions for each tree node are semantically meaningful.
    Text embeddings from PLIP are assumed to capture pathological morphology semantics; the accuracy of the descriptions (verified by two pathologists) is a premise of text guidance.
  • domain assumption The PLIP encoder provides a suitable shared embedding space for pathology images and text.
    All patch features and text prompts are encoded by frozen PLIP; the method inherits any domain gaps, though ablation with other encoders partially tests this.
  • standard math Standard deep learning components (GAT message passing, Nystrom attention approximation, triplet and alignment losses) are valid optimization choices.
    Used without proof, standard in the literature.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Diagnostic Text-guided Representation Learning in Hierarchical Classification for Pathological Whole Slide Image." pith.science (2026). https://pith.science/paper/FIADBQKR

@misc{pith2026241110709,
  author       = {Pith},
  title        = {Pith review of: Diagnostic Text-guided Representation Learning in Hierarchical Classification for Pathological Whole Slide Image},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/FIADBQKR}},
  note         = {Machine review of arXiv:2411.10709}
}
read the original abstract

With the development of digital imaging in medical microscopy, artificial intelligent-based analysis of pathological whole slide images (WSIs) provides a powerful tool for cancer diagnosis. Limited by the expensive cost of pixel-level annotation, current research primarily focuses on representation learning with slide-level labels, showing success in various downstream tasks. However, given the diversity of lesion types and the complex relationships between each other, these techniques still deserve further exploration in addressing advanced pathology tasks. To this end, we introduce the concept of hierarchical pathological image classification and propose a representation learning called PathTree. PathTree considers the multi-classification of diseases as a binary tree structure. Each category is represented as a professional pathological text description, which messages information with a tree-like encoder. The interactive text features are then used to guide the aggregation of hierarchical multiple representations. PathTree uses slide-text similarity to obtain probability scores and introduces two extra tree specific losses to further constrain the association between texts and slides. Through extensive experiments on three challenging hierarchical classification datasets: in-house cryosectioned lung tissue lesion identification, public prostate cancer grade assessment, and public breast cancer subtyping, our proposed PathTree is consistently competitive compared to the state-of-the-art methods and provides a new perspective on the deep learning-assisted solution for more complex WSI classification.

Figures

Figures reproduced from arXiv: 2411.10709 by the authors.

Figure 1
Figure 1. The existing WSI multi-classification method is described as a pla [PITH_FULL_IMAGE:figures/full_fig_p003_1.png] view at source ↗
Figure 2
Figure 2. Overview of PathTree. The main idea is to convert challenging pathological multi-classification into hierarchical tree structures for analysis. [PITH_FULL_IMAGE:figures/full_fig_p004_2.png] view at source ↗
Figure 4
Figure 4. Schematic diagram of the two attention modules. (a) Multiple gated [PITH_FULL_IMAGE:figures/full_fig_p005_4.png] view at source ↗
Figures from the paper (7 more)
Figure 5
Figure 5. Figure 5: Graphical demonstration of path alignment learning. Its goal is to [PITH_FULL_IMAGE:figures/full_fig_p006_5.png]
Figure 6
Figure 6. Figure 6: Graphical demonstration of tree-aware matching learning, where [PITH_FULL_IMAGE:figures/full_fig_p007_6.png]
Figure 8
Figure 8. Figure 8: Average results for the weighted F1, AUC, and H-F1 over three [PITH_FULL_IMAGE:figures/full_fig_p009_8.png]
Figure 9
Figure 9. Figure 9: AUC, Weighted F1, and H-F1 results [%] on the fine-grained 7-category SYSFL, 6-category PANDA, and 7-category BRACS datasets in few-shot [PITH_FULL_IMAGE:figures/full_fig_p010_9.png]
Figure 10
Figure 10. Figure 10: AUC results [%] with different graph-based text prompt encoders on three fine-grained datasets. 4.6. Ablation Experiments 4.6.1. Effectiveness of Different Text Prompts Our proposed PathTree uses professional descriptions as prompt descriptions for learning and infere…
Figure 11
Figure 11. Figure 11: The surface plots of PathTree under optimal average Weighted F1 [PITH_FULL_IMAGE:figures/full_fig_p012_11.png]
Figure 12
Figure 12. Figure 12: Comparison of PathTree with different coefficients in the BRACS dataset. (a) Keep coefficients µp and µm equal. (b) Keep the sum of coeffi￾cients µp and µm equal to 1. aNet and PathDino, which indicates that patch encoders pre￾trained on multimodal data have greater p…

Discussion (0). Continue with ORCID to comment.

Forward citations

Cited by 1 Pith paper

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score. Full citation record

  1. Multimodal Prototype Alignment for Semi-supervised Pathology Image Segmentation

    cs.CV 2025-08 conditional novelty 4.0 of 10

    MPAMatch combines a UNI-based encoder, UniMatch-style consistency, and image/text prototype contrastive losses to improve semi-supervised pathology segmentation, reporting state-of-the-art results on four public datasets.

Reference graph

Works this paper leans on

79 extracted references · 70 canonical work pages · cited by 1 Pith paper

  1. [1]

    , author Shafique, A

    author Alfasly, S. , author Shafique, A. , author Nejat, P. , author Khan, J. , author Alsaafin, A. , author Alabtah, G. , author Tizhoosh, H. , year 2024 . title Rotation-agnostic image representation learning for digital pathology , in: booktitle Proc. IEEE/CVF Conf. Comput. Vis. Pattern Recognit. , pp. pages 11683--11693

    work page 2024

  2. [2]

    , author Anniciello, A.M

    author Brancati, N. , author Anniciello, A.M. , author Pati, P. , author Riccio, D. , author Scognamiglio, G. , author Jaume, G. , author De Pietro, G. , author Di Bonito, M. , author Foncubierta, A. , author Botti, G. , author Gabrani, M. , author Feroce, F. , author Frucci, M. , year 2022 . title Bracs: A dataset for breast carcinoma subtyping in h&e hi...

    work page 2022

  3. [3]

    o nberg, H. , author Samaratunga, H. , author Delahunt, B. , author Tsuzuki, T. , author H \

    author Bulten, W. , author Kartasalo, K. , author Chen, P.H.C. , author Str \"o m, P. , author Pinckaers, H. , author Nagpal, K. , author Cai, Y. , author Steiner, D.F. , author van Boven, H. , author Vink, R. , author Hulsbergen-van de Kaa, C. , author van der Laak, J. , author Amin, M.B. , author Evans, A.J. , author van der Kwast, T. , author Allan, R....

    work page 2022

  4. [4]

    , author Geessink, O

    author Bándi, P. , author Geessink, O. , author Manson, Q. , author Van Dijk, M. , author Balkenhol, M. , author Hermsen, M. , author Ehteshami Bejnordi, B. , author Lee, B. , author Paeng, K. , author Zhong, A. , author Li, Q. , author Zanjani, F.G. , author Zinger, S. , author Fukuta, K. , author Komura, D. , author Ovtcharov, V. , author Cheng, S. , au...

    work page 2018

  5. [5]

    , author Hanna, M.G

    author Campanella, G. , author Hanna, M.G. , author Geneslaw, L. , author Miraflor, A. , author Werneck Krauss Silva, V. , author Busam, K.J. , author Brogi, E. , author Reuter, V.E. , author Klimstra, D.S. , author Fuchs, T.J. , year 2019 . title Clinical-grade computational pathology using weakly supervised deep learning on whole slide images . journal ...

    work page 2019

  6. [6]

    , author Touvron, H

    author Caron, M. , author Touvron, H. , author Misra, I. , author J \'e gou, H. , author Mairal, J. , author Bojanowski, P. , author Joulin, A. , year 2021 . title Emerging properties in self-supervised vision transformers , in: booktitle Proc. IEEE/CVF Int. Conf. Comput. Vis. , pp. pages 9650--9660

    work page 2021

  7. [7]

    , author Cendra, F.J

    author Chan, T.H. , author Cendra, F.J. , author Ma, L. , author Yin, G. , author Yu, L. , year 2023 . title Histopathology whole slide image analysis with heterogeneous graph representation learning , in: booktitle Proc. IEEE/CVF Conf. Comput. Vis. Pattern Recognit. , pp. pages 15661--15670

    work page 2023

  8. [8]

    , author Chen, C.C

    author Chen, C.L. , author Chen, C.C. , author Yu, W.H. , author Chen, S.H. , author Chang, Y.C. , author Hsu, T.I. , author Hsiao, M. , author Yeh, C.Y. , author Chen, C.Y. , year 2021 a. title An annotation-free whole-slide training approach to pathological classification of lung cancer types using deep learning . journal Nat. Commun. volume 12 , pages 1193

    work page 2021

Show all 79 references
  1. [9]

    , author Ma, Q

    author Chen, H. , author Ma, Q. , author Lin, Z. , author Yan, J. , year 2021 b. title Hierarchy-aware label semantics matching network for hierarchical text classification , in: booktitle Proc. 59th Annu. Meeting Assoc. Comput. Linguistics & 11th Int. Joint Conf. Natural Lang...

    2021

  2. [10]

    , author Chen, C

    author Chen, R.J. , author Chen, C. , author Li, Y. , author Chen, T.Y. , author Trister, A.D. , author Krishnan, R.G. , author Mahmood, F. , year 2022 . title Scaling vision transformers to gigapixel images via hierarchical self-supervised learning , in: booktitle Proc. IEEE/...

    2022

  3. [11]

    , author Ding, T

    author Chen, R.J. , author Ding, T. , author Lu, M.Y. , author Williamson, D.F.K. , author Jaume, G. , author Song, A.H. , author Chen, B. , author Zhang, A. , author Shao, D. , author Shaban, M. , author Williams, M. , author Oldenburg, L. , author Weishaupt, L.L. , author Wa...

    2024

  4. [12]

    , author Lu, M.Y

    author Chen, R.J. , author Lu, M.Y. , author Shaban, M. , author Chen, C. , author Chen, T.Y. , author Williamson, D.F. , author Mahmood, F. , year 2021 c. title Whole slide images are 2d point clouds: Context-aware survival prediction using patch-based graph convolutional net...

    2021

  5. [13]

    , author Lu, M.Y

    author Chen, R.J. , author Lu, M.Y. , author Weng, W.H. , author Chen, T.Y. , author Williamson, D.F. , author Manz, T. , author Shady, M. , author Mahmood, F. , year 2021 d. title Multimodal co-attention transformer for survival prediction in gigapixel whole slide images , in...

    2021

  6. [14]

    , author Sun, Q

    author Chu, H. , author Sun, Q. , author Li, J. , author Chen, Y. , author Zhang, L. , author Guan, T. , author Han, A. , author He, Y. , year 2024 . title Retmil: Retentive multiple instance learning for histopathological whole slide image classification . journal arXiv prepr...

    2024 arXiv

  7. [15]

    , author Ocampo, P.S

    author Coudray, N. , author Ocampo, P.S. , author Sakellaropoulos, T. , author Narula, N. , author Snuderl, M. , author Feny \"o , D. , author Moreira, A.L. , author Razavian, N. , author Tsirigos, A. , year 2018 . title Classification and mutation prediction from non--small c...

    2018

  8. [16]

    , author Dong, W

    author Deng, J. , author Dong, W. , author Socher, R. , author Li, L.J. , author Li, K. , author Fei-Fei, L. , year 2009 . title Imagenet: A large-scale hierarchical image database , in: booktitle 2009 IEEE Conf. Comput. Vis. Pattern Recognit. , organization Ieee . pp. pages 248--255

    2009

  9. [17]

    , author Zou, C

    author Di, D. , author Zou, C. , author Feng, Y. , author Zhou, H. , author Ji, R. , author Dai, Q. , author Gao, Y. , year 2022 . title Generating hypergraph-based high-order representations of whole-slide histopathological images for survival prediction . journal IEEE Trans....

    2022

  10. [18]

    , author Rindtorff, N.T

    author Echle, A. , author Rindtorff, N.T. , author Brinker, T.J. , author Luedde, T. , author Pearson, A.T. , author Kather, J.N. , year 2021 . title Deep learning in cancer pathology: a new generation of clinical biomarkers . journal Br. J. Cancer volume 124 , pages 686--696

    2021

  11. [19]

    , author Veta, M

    author Ehteshami Bejnordi, B. , author Veta, M. , author Johannes van Diest, P. , author van Ginneken, B. , author Karssemeijer, N. , author Litjens, G. , author van der Laak, J.A.W.M. , author the CAMELYON16 Consortium , year 2017 . title Diagnostic assessment of deep learnin...

    2017

  12. [20]

    , author Andani, S

    author Fremond, S. , author Andani, S. , author Barkey Wolf, J. , author Dijkstra, J. , author Melsbach, S. , author Jobsen, J.J. , author Brinkhuis, M. , author Roothaan, S. , author Jurgenliemk-Schulz, I. , author Lutgens, L.C.H.W. , author Nout, R.A. , author van der Steen-...

    2023

  13. [21]

    , author Zhang, J

    author Guan, Y. , author Zhang, J. , author Tian, K. , author Yang, S. , author Dong, P. , author Xiang, J. , author Yang, W. , author Huang, J. , author Zhang, Y. , author Han, X. , year 2022 . title Node-aligned graph convolutional network for whole-slide image representatio...

    2022

  14. [22]

    , author Ying, Z

    author Hamilton, W. , author Ying, Z. , author Leskovec, J. , year 2017 . title Inductive representation learning on large graphs . journal Advances in neural information processing systems volume 30

    2017

  15. [23]

    , author Bianchi, F

    author Huang, Z. , author Bianchi, F. , author Yuksekgonul, M. , author Montine, T.J. , author Zou, J. , year 2023 . title A visual--language foundation model for pathology image analysis using medical twitter . journal Nat. Med. volume 29 , pages 2307--2316

    2023

  16. [24]

    , author Chai, H

    author Huang, Z. , author Chai, H. , author Wang, R. , author Wang, H. , author Yang, Y. , author Wu, H. , year 2021 . title Integration of patch features through self-supervised learning and transformer for survival analysis on whole slide images , in: booktitle Proc. Int. Co...

    2021

  17. [25]

    , author Tomczak, J

    author Ilse, M. , author Tomczak, J. , author Welling, M. , year 2018 . title Attention-based deep multiple instance learning , in: booktitle Proc. Int. Conf. Mach. Learn. , organization PMLR . pp. pages 2127--2136

    2018

  18. [26]

    , author Oldenburg, L

    author Jaume, G. , author Oldenburg, L. , author Vaidya, A. , author Chen, R.J. , author Williamson, D.F. , author Peeters, T. , author Song, A.H. , author Mahmood, F. , year 2024 . title Transcriptomics-guided slide representation learning in computational pathology , in: boo...

    2024

  19. [27]

    , author Toutanova, L.K

    author Kenton, J.D.M.W.C. , author Toutanova, L.K. , year 2019 . title Bert: Pre-training of deep bidirectional transformers for language understanding , in: booktitle Proc. NAACL-HLT , pp. pages 4171--4186

    2019

  20. [28]

    , author Ba, J

    author Kingma, D.P. , author Ba, J. , year 2014 . title Adam: A method for stochastic optimization . journal arXiv preprint arXiv:1412.6980

    2014 arXiv

  21. [29]

    , author Welling, M

    author Kipf, T.N. , author Welling, M. , year 2017 . title Semi-supervised classification with graph convolutional networks , in: booktitle Proc. Int. Conf. Learn. Represent

    2017

  22. [30]

    , author Zhang, H

    author Kundra, R. , author Zhang, H. , author Sheridan, R. , author Sirintrapun, S.J. , author Wang, A. , author Ochoa, A. , author Wilson, M. , author Gross, B. , author Sun, Y. , author Madupuri, R. , author Satravada, B.A. , author Reales, D. , author Vakiani, E. , author A...

    2021

  23. [31]

    , author Muti, H.S

    author Laleh, N.G. , author Muti, H.S. , author Loeffler, C.M.L. , author Echle, A. , author Saldanha, O.L. , author Mahmood, F. , author Lu, M.Y. , author Trautwein, C. , author Langer, R. , author Dislich, B. , author Buelow, D.R. , author Grabsch, I.H. , author Brenner, H. ...

    2022

  24. [32]

    , author Yoon, W

    author Lee, J. , author Yoon, W. , author Kim, S. , author Kim, D. , author Kim, S. , author So, C.H. , author Kang, J. , year 2020 . title Biobert: a pre-trained biomedical language representation model for biomedical text mining . journal Bioinformatics volume 36 , pages 1234--1240

    2020

  25. [33]

    , author Li, Y

    author Li, B. , author Li, Y. , author Eliceiri, K.W. , year 2021 . title Dual-stream multiple instance learning network for whole slide image classification with self-supervised contrastive learning , in: booktitle Proc. IEEE/CVF Conf. Comput. Vis. Pattern Recognit. , pp. pag...

    2021

  26. [34]

    , author Chen, Y

    author Li, H. , author Chen, Y. , author Chen, Y. , author Yu, R. , author Yang, W. , author Wang, L. , author Ding, B. , author Han, Y. , year 2024 a. title Generalizable whole slide image classification with fine-grained visual-semantic interaction , in: booktitle Proc. IEEE...

    2024

  27. [35]

    , author Chen, Y

    author Li, J. , author Chen, Y. , author Chu, H. , author Sun, Q. , author Guan, T. , author Han, A. , author He, Y. , year 2024 b. title Dynamic graph representation with knowledge-aware attention for histopathology whole slide image analysis , in: booktitle Proc. IEEE/CVF Co...

    2024

  28. [36]

    , author Cheng, J

    author Li, J. , author Cheng, J. , author Meng, L. , author Yan, H. , author He, Y. , author Shi, H. , author Guan, T. , author Han, A. , year 2023 a. title Deeptree: Pathological image classification through imitating tree-like strategies of pathologists . journal IEEE Trans....

    2023

  29. [37]

    , author Zhao, Y

    author Li, S. , author Zhao, Y. , author Zhang, J. , author Yu, T. , author Zhang, J. , author Gao, Y. , year 2023 b. title High-order correlation-guided slide-level histology retrieval with self-supervised hashing . journal IEEE Trans. Pattern Anal. Mach. Intell

    2023

  30. [38]

    , author Li, C

    author Li, X. , author Li, C. , author Rahaman, M.M. , author Sun, H. , author Li, X. , author Wu, J. , author Yao, Y. , author Grzegorzek, M. , year 2022 . title A comprehensive review of computer-aided whole-slide image analysis: from datasets to feature extraction, segmenta...

    2022

  31. [39]

    , author Ricketts, C.J

    author Linehan, W.M. , author Ricketts, C.J. , year 2019 . title The cancer genome atlas of renal cell carcinoma: findings and clinical implications . journal Nat. Rev. Urol. volume 16 , pages 539--552

    2019

  32. [40]

    , author Bandi, P

    author Litjens, G. , author Bandi, P. , author Ehteshami Bejnordi, B. , author Geessink, O. , author Balkenhol, M. , author Bult, P. , author Halilovic, A. , author Hermsen, M. , author van de Loo, R. , author Vogels, R. , author Manson, Q.F. , author Stathonikos, N. , author ...

    2018

  33. [41]

    , author Chen, B

    author Lu, M.Y. , author Chen, B. , author Williamson, D.F.K. , author Chen, Richard J.and Liang, I. , author Ding, T. , author Jaume, G. , author Odintsov, I. , author Le, L.P. , author Gerber, G. , author Parwani, A.V. , author Zhang, A. , author Mahmood, F. , year 2024 . ti...

    2024

  34. [42]

    , author Chen, T.Y

    author Lu, M.Y. , author Chen, T.Y. , author Williamson, D.F. , author Zhao, M. , author Shady, M. , author Lipkova, J. , author Mahmood, F. , year 2021 a. title Ai-based pathology predicts origins for cancers of unknown primary . journal Nature volume 594 , pages 106--110

    2021

  35. [43]

    , author Williamson, D.F

    author Lu, M.Y. , author Williamson, D.F. , author Chen, T.Y. , author Chen, R.J. , author Barbieri, M. , author Mahmood, F. , year 2021 b. title Data-efficient and weakly supervised computational pathology on whole-slide images . journal Nat. Biomed. Eng. volume 5 , pages 555--570

    2021

  36. [44]

    , author Parwani, A.V

    author Niazi, M.K.K. , author Parwani, A.V. , author Gurcan, M.N. , year 2019 . title Digital pathology and artificial intelligence . journal Lancet Oncol. volume 20 , pages e253--e261

    2019

  37. [45]

    title Gpt-4 technical report

    author OpenAI , year 2023 . title Gpt-4 technical report . journal arXiv preprint arXiv:2303.08774

    2023 arXiv

  38. [46]

    , year 1975

    author Otsu, N. , year 1975 . title A threshold selection method from gray-level histograms . journal Automatica volume 11 , pages 23--27

    1975

  39. [47]

    , author Ghaffari Laleh, N

    author Perez-Lopez, R. , author Ghaffari Laleh, N. , author Mahmood, F. , author Kather, J.N. , year 2024 . title A guide to artificial intelligence for cancer researchers . journal Nat. Rev. Cancer , pages 1--15

    2024

  40. [48]

    , author Van Ginneken, B

    author Pinckaers, H. , author Van Ginneken, B. , author Litjens, G. , year 2020 . title Streaming convolutional neural networks for end-to-end learning with multi-megapixel images . journal IEEE Trans. Pattern Anal. Mach. Intell. volume 44 , pages 1581--1590

    2020

  41. [49]

    , author luo, x

    author Qu, L. , author luo, x. , author Fu, K. , author Wang, M. , author Song, Z. , year 2024 . title The rise of ai language pathologists: Exploring two-level prompt learning for few-shot weakly-supervised whole slide image classification . journal Proc. Adv. Neural Inf. Pro...

    2024

  42. [50]

    , author Kim, J.W

    author Radford, A. , author Kim, J.W. , author Hallacy, C. , author Ramesh, A. , author Goh, G. , author Agarwal, S. , author Sastry, G. , author Askell, A. , author Mishkin, P. , author Clark, J. , author Krueger, G. , author Sutskever, I. , year 2021 . title Learning transfe...

    2021

  43. [51]

    , author Babaie, M

    author Riasatian, A. , author Babaie, M. , author Maleki, D. , author Kalra, S. , author Valipour, M. , author Hemati, S. , author Zaveri, M. , author Safarpoor, A. , author Shafiei, S. , author Afshari, M. , author Rasoolijaberi, M. , author Sikaroudi, M. , author Adnan, M. ,...

    2021

  44. [52]

    , author Rodriguez, R

    author Rodriguez, J.P.M. , author Rodriguez, R. , author Silva, V.W.K. , author Kitamura, F.C. , author Corradi, G.C.A. , author de Marchi, A.C.B. , author Rieder, R. , year 2022 . title Artificial intelligence as a tool for diagnosis in digital pathology whole slide images: A...

    2022

  45. [53]

    , author Bian, H

    author Shao, Z. , author Bian, H. , author Chen, Y. , author Wang, Y. , author Zhang, J. , author Ji, X. , author zhang, y. , year 2021 . title Transmil: Transformer based correlated multiple instance learning for whole slide image classification . journal Proc. Adv. Neural In...

    2021

  46. [54]

    , author Li, C

    author Shi, J. , author Li, C. , author Gong, T. , author Zheng, Y. , author Fu, H. , year 2024 . title Vila-mil: Dual-scale vision-language multiple instance learning for whole slide image classification , in: booktitle Proc. IEEE/CVF Conf. Comput. Vis. Pattern Recognit. , pp...

    2024

  47. [55]

    , author Chen, R.J

    author Song, A.H. , author Chen, R.J. , author Ding, T. , author Williamson, D.F. , author Jaume, G. , author Mahmood, F. , year 2024 . title Morphological prototyping for unsupervised slide representation learning in computational pathology , in: booktitle Proc. IEEE/CVF Conf...

    2024

  48. [56]

    , author Jaume, G

    author Song, A.H. , author Jaume, G. , author Williamson, D.F. , author Lu, M.Y. , author Vaidya, A. , author Miller, T.R. , author Mahmood, F. , year 2023 . title Artificial intelligence for digital and computational pathology . journal Nat. Rev. Bioeng. volume 1 , pages 930--949

    2023

  49. [57]

    , author Jiang, D

    author Sun, Q. , author Jiang, D. , author Li, J. , author Yan, R. , author He, Y. , author Guan, T. , author Cheng, Z. , year 2024 . title Nciemil: Rethinking decoupled multiple instance learning framework for histopathological slide classification , in: booktitle Proc. Int. ...

    2024

  50. [58]

    , year 1972

    author Tarjan, R. , year 1972 . title Depth-first search and linear graph algorithms . journal SIAM J. Comput. volume 1 , pages 146--160

    1972

  51. [59]

    , author Shazeer, N

    author Vaswani, A. , author Shazeer, N. , author Parmar, N. , author Uszkoreit, J. , author Jones, L. , author Gomez, A.N. , author Kaiser, . , author Polosukhin, I. , year 2017 . title Attention is all you need . journal Proc. Adv. Neural Inf. Process. Syst. volume 30

    2017

  52. [60]

    , author Cucurull, G

    author Veli c kovi \'c , P. , author Cucurull, G. , author Casanova, A. , author Romero, A. , author Lio, P. , author Bengio, Y. , year 2018 . title Graph attention networks , in: booktitle Proc. Int. Conf. Learn. Represent

    2018

  53. [61]

    , author Reisenb \"u chler, D

    author Wagner, S.J. , author Reisenb \"u chler, D. , author West, N.P. , author Niehues, J.M. , author Zhu, J. , author Foersch, S. , author Veldhuizen, G.P. , author Quirke, P. , author Grabsch, H.I. , author van den Brandt, P.A. , author Hutchins, G.G. , author Richman, S.D....

    2023

  54. [62]

    , author Khosla, A

    author Wang, D. , author Khosla, A. , author Gargeya, R. , author Irshad, H. , author Beck, A.H. , year 2016 . title Deep learning for identifying metastatic breast cancer . journal arXiv preprint arXiv:1606.05718

    2016 arXiv

  55. [63]

    , author Ma, S

    author Wang, W. , author Ma, S. , author Xu, H. , author Usuyama, N. , author Ding, J. , author Poon, H. , author Wei, F. , year 2023 a. title When an image is worth 1,024 x 1,024 words: A case study in computational pathology . journal arXiv preprint arXiv:2312.03558

    2023 arXiv

  56. [64]

    , author Yu, L

    author Wang, Z. , author Yu, L. , author Ding, X. , author Liao, X. , author Wang, L. , year 2023 b. title Shared-specific feature learning with bottleneck fusion transformer for multi-modal whole slide image analysis . journal IEEE Trans. Med. Imaging volume 42 , pages 3374--3383

    2023

  57. [65]

    , author Zhang, J

    author Xiang, J. , author Zhang, J. , year 2022 . title Exploring low-rank property in multiple instance learning for whole slide image classification , in: booktitle Proc. Int. Conf. Learn. Represent

    2022

  58. [66]

    , author Song, Y

    author Xiang, T. , author Song, Y. , author Zhang, C. , author Liu, D. , author Chen, M. , author Zhang, F. , author Huang, H. , author O’Donnell, L. , author Cai, W. , year 2022 . title Dsnet: A dual-stream framework for weakly-supervised gigapixel pathology image analysis . ...

    2022

  59. [67]

    , author Chen, H

    author Xiong, C. , author Chen, H. , author Sung, J.J. , author King, I. , year 2023 . title Diagnose like a pathologist: transformer-enabled hierarchical attention-guided multiple instance learning for whole slide image classification , in: booktitle Proc. Int. Joint Conf. Ar...

    2023

  60. [68]

    o mformer: A nystr \

    author Xiong, Y. , author Zeng, Z. , author Chakraborty, R. , author Tan, M. , author Fung, G. , author Li, Y. , author Singh, V. , year 2021 . title Nystr \"o mformer: A nystr \"o m-based algorithm for approximating self-attention , in: booktitle Proc. AAAI Conf. Artif. Intel...

    2021

  61. [69]

    , author Hu, W

    author Xu, K. , author Hu, W. , author Leskovec, J. , author Jegelka, S. , year 2018 . title How powerful are graph neural networks? journal arXiv preprint arXiv:1810.00826

    2018 arXiv

  62. [70]

    , author Chen, L

    author Yang, H. , author Chen, L. , author Cheng, Z. , author Yang, M. , author Wang, J. , author Lin, C. , author Wang, Y. , author Huang, L. , author Chen, Y. , author Peng, S. , author Ke, Z. , author Li, W. , year 2021 . title Deep learning-based six-type classifier for lu...

    2021

  63. [71]

    , author Wang, Y

    author Yang, S. , author Wang, Y. , author Chen, H. , year 2024 . title Mambamil: Enhancing long sequence modeling with sequence reordering in computational pathology . journal arXiv preprint arXiv:2403.06800

    2024 arXiv

  64. [72]

    , author Meng, Y

    author Zhang, H. , author Meng, Y. , author Zhao, Y. , author Qiao, Y. , author Yang, X. , author Coupland, S.E. , author Zheng, Y. , year 2022 . title Dtfd-mil: Double-tier feature distillation multiple instance learning for histopathology whole slide image classification , i...

    2022

  65. [73]

    , author Gao, J

    author Zhang, Y. , author Gao, J. , author Zhou, M. , author Wang, X. , author Qiao, Y. , author Zhang, S. , author Wang, D. , year 2023 . title Text-guided foundation model adaptation for pathological image classification , in: booktitle Proc. Int. Conf. Med. Image Comput. Co...

    2023

  66. [74]

    , author Chen, P

    author Zhang, Z. , author Chen, P. , author Shi, X. , author Yang, L. , year 2019 . title Text-guided neural network training for image recognition in natural scenes and medicine . journal IEEE Trans. Pattern Anal. Mach. Intell. volume 43 , pages 1733--1745

    2019

  67. [75]

    , author Gindra, R.H

    author Zheng, Y. , author Gindra, R.H. , author Green, E.J. , author Burks, E.J. , author Betke, M. , author Beane, J.E. , author Kolachalama, V.B. , year 2022 . title A graph-transformer for whole slide image classification . journal IEEE Trans. Med. Imaging volume 41 , pages...

    2022

  68. [76]

    , author Ma, C

    author Zhou, J. , author Ma, C. , author Long, D. , author Xu, G. , author Ding, N. , author Zhang, H. , author Xie, P. , author Liu, G. , year 2020 . title Hierarchy-aware global model for hierarchical text classification , in: booktitle Proc. 58th Annu. Meeting Assoc. Comput...

    2020

  69. [77]

    , author Yang, J

    author Zhou, K. , author Yang, J. , author Loy, C.C. , author Liu, Z. , year 2022 a. title Conditional prompt learning for vision-language models , in: booktitle Proc. IEEE/CVF Conf. Comput. Vis. Pattern Recognit. , pp. pages 16816--16825

    2022

  70. [78]

    , author Yang, J

    author Zhou, K. , author Yang, J. , author Loy, C.C. , author Liu, Z. , year 2022 b. title Learning to prompt for vision-language models . journal Int. J. Comput. Vis. volume 130 , pages 2337--2348

    2022

  71. [79]

    write newline

    " write newline "" before.all 'output.state := FUNCTION n.dashify 't := "" t empty not t #1 #1 substring "-" = t #1 #2 substring "--" = not "--" * t #2 global.max substring 't := t #1 #1 substring "-" = "-" * t #2 global.max substring 't := while if t #1 #1 substring * t #2 gl...

Pith tools

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