Pith. sign in

REVIEW 3 major objections 6 minor 1 cited by

The Role of Graph-based MIL and Interventional Training in the Generalization of WSI Classifiers

T0 review · 3 major / 6 minor · reviewed 2026-08-09 · deepseek-v4-flash

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

Pith's one-line read The paper claims that graph-based MIL models alone match or exceed the generalization gains that interventional training provides, and that adding the interventional module to graph models usually reduces performance.

desk verdict Useful graph-MIL comparison undermined by an overreaching central claim that Table 2 does not support; the more defensible result is that interventional training helps non-graph MIL but not graph models. read the letter →

arxiv 2501.19048 v1 pith:5MVWLRMR submitted 2025-01-31 eess.IV cs.CV

Rita Pereira , M. Rita Verdelho , Catarina Barata , Carlos Santiago This is my paper · ORCID
classification eess.IVcs.CV
keywords WholeSlideImagesMultipleInstanceLearningGraphNeuralNetworksInterventionalTrainingBackdoorAdjustmentDomainShiftCamelyon17BreastCancerMetastasisDetection
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

Whole-slide cancer images are classified from thousands of patches, and models are vulnerable to spurious visual biases such as staining and scanner differences. This paper tries to show that giving a multiple-instance model an explicit spatial graph of the slide, where nodes are patches and edges connect neighboring tissue, provides the domain-shift stability that interventional (causal) training was designed to add. Comparing graph construction methods, GNN backbones, and MIL pooling on Camelyon16 and Camelyon17, it finds that graph-based models alone match or beat their counterparts augmented with interventional training, and that added interventions usually lower graph-model performance and spread attention. If that holds, simpler graph-MIL pipelines are sufficient for cross-center generalization, and the extra confounder-modeling stage may be unnecessary for spatially structured models.

What carries the argument

The load-bearing object is the patch-graph: each patch of the WSI is a node carrying a 1024-dimensional ResNet-50 embedding, and edges connect spatially adjacent patches, defined by patch coordinates rather than feature similarity. On top of this graph sit GCN or GAT layers, a MIL pooling (ABMIL or DSMIL) that aggregates node embeddings into one bag vector, and, in GMIL-IT, an interventional stage that clusters bag embeddings with K-means into a confounder dictionary $C$ and applies the backdoor adjustment $P(Y|\mathrm{do}(X)) = \sum_i P(Y|X, h(X,c_i))P(c_i)$ with an attention function $h$ and fusion $B \oplus \sum_i \alpha_i c_i P(c_i)$. The graph is what carries the argument: it preserves spatial context, and the paper argues that its attention mechanism filters spurious features so thoroughly that the confounder dictionary ends up encoding class separation instead of adding robustness.

What would settle it

Compute the purity of the K-means confounder clusters against slide-level labels and against hospital or scanner provenance on both datasets, since the paper already reports class purity 1.0 for GAT-ABMIL; if clusters align with labels rather than scanners or stain groups, rerun interventional training with confounders defined by stain statistics or scanner identity, and if graph models still beat interventional training the claim stands, while if interventional training then improves, the claim fails.

Watch

Extended reading notes

Core claim

On the paper's own terms, the central discovery is that spatial structure, not the interventional module, is what buys robustness: PatchGAT-ABMIL without interventional training reaches AUC 0.923 on Camelyon16 and 0.837 on Camelyon17, while adding interventional training lowers those numbers to 0.910 and 0.823. The paper reports that for the GAT-based graph models interventional training decreases performance across nearly all evaluated metrics and datasets, with GCN results mixed, while the same intervention substantially helps the non-graph ABMIL baseline, raising its Camelyon16 AUC from 0.711 to 0.920. It interprets this as graph attention acting like an information bottleneck, filtering irrelevant visual features so that the K-means confounder clusters align with classes rather than adding useful causal adjustment.

Load-bearing premise

Interventional training is valid only if the K-means clusters of bag embeddings really capture visual biases like stain variation; if those clusters instead capture tumor class or other true signal, the backdoor adjustment is not a genuine intervention and the comparison says little about whether causal deconfounding helps graph models.

Editorial extensions

If this is right

  • Graph-based MIL alone can give the cross-center generalization that motivated interventional training, so the confounder dictionary and backdoor module can be dropped for graph models without sacrificing AUC.
  • Patch-level graphs outperform region- and centroid-level graphs, so graph construction that keeps original patch information is preferable for WSI classification.
  • A domain-shift protocol that holds out entire medical centers, as in Camelyon17, exposes robustness differences that standard shuffled cross-validation on Camelyon16 hides.
  • Interventional training still helps non-graph MIL, especially ABMIL, showing that deconfounding is architecture-dependent rather than universally beneficial.
  • Attention heatmaps indicate interventions blur spatial focus, so causal adjustment may actively disrupt the spatial reasoning graph models use.

Reading between the lines

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

  • Because the reported cluster purity is 1.0, meaning confounder clusters equal classes, the K-means dictionary built from bag embeddings may be encoding the target label rather than stain or scanner confounders; if so, the negative result is about confounder misspecification, not about graph models being immune to all interventions.
  • A decisive extension would build the confounder dictionary from stain statistics, scanner identity, or hospital provenance instead of bag embeddings; if graph models still outperform interventional training under those true confounders, the claim that spatial structure substitutes for causal adjustment is much stronger.
  • Frontdoor adjustment, which does not require observing confounders, becomes an obvious next test for graph-MIL, since graph models may already learn invariant features and could avoid the misspecification that backdoor adjustment suffers here.
  • These results concern spatial artifacts such as staining and scanner differences in breast-cancer lymph-node slides; non-spatial biases like marker pen or air bubbles may still need interventional training even in graph models, because graph adjacency cannot filter what is not localized spatially.
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 / 6 minor

Summary. The paper studies whole-slide image classification with MIL, graph-based MIL, and interventional training via backdoor adjustment. It introduces GMIL-IT, a pipeline that combines feature extraction, graph construction (patch, region, centroid graphs), GNN backbones (GCN, GAT), MIL aggregators (ABMIL, DSMIL), and an IBMIL-style interventional training stage. Experiments on Camelyon16 and Camelyon17 compare these configurations under explicit domain shifts and report that patch graphs outperform other graph constructions, that graph-based models generally improve over plain MIL, and that interventional training harms graph-based models while helping plain MIL. The paper concludes that graph-based models alone achieve the generalization initially expected from interventional training, and that graph structure provides inherent robustness to confounders. Code is publicly available.

Significance. The question addressed is relevant: whether spatial graph structure in WSI classifiers can replace or reduce the need for causal interventional training is a useful empirical question for computational pathology. The paper contributes a broad comparison of graph construction methods and a reproducible implementation of IBMIL-style backdoor adjustment. The Camelyon17 evaluation is explicitly set up as a leave-one-center-out domain-shift test, which is a strength. Code release and the use of standard public datasets support reproducibility. However, the headline claim is not supported by the reported numbers: interventional training substantially improves plain ABMIL, and on Camelyon17 ABMIL with interventional training outperforms the best graph-only model. The central contribution therefore requires substantial re-framing rather than a small local fix.

major comments (3)
  1. [Abstract; Section 6; Table 2] The central claim that "graph-based models alone outperform models enhanced with interventional training" is contradicted by the paper's own Table 2. On Camelyon17, ABMIL with interventional training reaches AUC 0.852±0.03, higher than the best graph-only model, PatchGAT-ABMIL, at 0.837±0.04; on Camelyon16 the two are effectively tied (0.920±0.04 vs 0.923±0.06). The data support at most the weaker statement that interventional training did not improve graph-based models in these experiments. The abstract, contribution list, and Section 6 need to be rewritten to state that weaker claim.
  2. [Table 2; Section 5.3] No statistical significance testing is reported for the key comparisons. The differences between PatchGAT-ABMIL and PatchGAT-ABMIL w/IT are 0.013 on Camelyon16 and 0.014 on Camelyon17, with standard deviations between 0.04 and 0.07, and only 5 cross-validation folds are used. These differences are within one standard deviation and cannot support the conclusion of a "consistent decline" caused by interventional training. Paired tests, confidence intervals, or multi-seed results are needed before claiming that interventions harm graph-based models.
  3. [Section 3.3; Section 5.3; Figures 4/5] The confounder dictionary is built by K-means on the model's own bag embeddings, and the paper reports cluster purity 1.0 with centroids corresponding to normal and tumor classes. This indicates that the clusters encode class structure, not visual confounders such as stain variation, so the backdoor adjustment in Eqs. (7)–(9) may not implement a valid causal intervention. Without evidence that the clusters correspond to genuine confounders (for example, by comparing cluster membership against hospital or staining labels), the causal interpretation of the negative results for graph models is not established. Please validate the confounder clusters or explicitly limit the conclusions to the observed behavior of this particular IBMIL-style training procedure.
minor comments (6)
  1. [Eq. (6)] The first equality in Eq. (6) is not correct as written: P(Y|do(X)) = P(Y|X) Σ_i P(c_i) simplifies to P(Y|X), not to Σ_i P(Y|X,c_i)P(c_i). The correct backdoor adjustment is the second expression, and the text should be corrected to avoid a formal error in the central formula.
  2. [Abstract; Section 7] The text refers to "this dissertation" and "this thesis," but the manuscript is presented as a conference paper; please use consistent terminology such as "this paper."
  3. [Section 3.2; Section 3.3] There are typos: "explanability" should be "explainability," and "strataci" should be "strata."
  4. [Section 5.3] The statement that "standard deviations for MIL w\IT are generally larger" is ambiguous because Table 2 shows ABMIL w/IT has smaller standard deviations than ABMIL without IT (e.g., 0.04 vs 0.21 for Camelyon16 AUC); please specify the comparison group clearly.
  5. [Table 1] Several rows of Table 1 report identical Accuracy, Recall, and AUC values (e.g., the Region-GCN and Centroid-GCN rows). Please clarify whether this is a rounding artifact or a consequence of the evaluation setup, since identical values across three distinct metrics are otherwise surprising.
  6. [Figure 6] The caption of Figure 6 refers to GCN + ABMIL, while Section 5.4 and the surrounding text refer to GMIL and GAT-based models; please align the labels in the figure and text.

Circularity Check

0 steps flagged · score 1.0 of 10

No significant circularity; the central claim is an empirical comparison and the only self-citation is not load-bearing.

full rationale

The paper's central assertion, that graph-based models alone achieve the generalization initially anticipated from interventional training, rests on the empirical comparisons in Table 2 rather than on any quantity that is defined in terms of itself. The interventional-training component follows IBMIL's published backdoor adjustment (Lin et al., 2023), an external method, and although the confounder dictionary is estimated from the model's own bag embeddings via PCA and K-means, the conclusion about graph-based models is not obtained by plugging those estimates back into the same formula; it is obtained by comparing separately trained configurations with and without the interventional module. No predicted value is a renamed fitted parameter, and no uniqueness or equivalence result is imported from the authors' prior work to force the choice of graph representation. The only same-group citation is the MMIST-ccRCC dataset (Mota et al., 2024), mentioned in the Limitations section as additional validation; removing it would not affect the Camelyon16/17 comparisons that carry the abstract and conclusion, so it is not load-bearing. The abstract's wording that graph-based models 'outperform' models with interventional training is stronger than Table 2 supports, since the key differences are within one standard deviation and no significance tests are reported—but that is an evidentiary weakness, not circularity. Similarly, the paper's admission that the causal graph is unknown and that clusters 'likely represent visual biases' concerns the validity of the intervention, not a circular reduction of the result to its inputs. No circular step can be exhibited from the paper's equations or citation chain.

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

No new physical or model entities are introduced; GMIL-IT is a combination of existing components. The central empirical claims rest on the free parameters and domain assumptions above, especially the confounder-dictionary assumption.

free parameters (4)
  • Confounder dictionary size K = not reported in main text
    Controls the number of confounder strata in the backdoor adjustment (Eqs. 6-9, Section 3.3); chosen by hand and not reported, yet it determines what the intervention can model.
  • Number of GNN layers L = 3 (main results, Table 1)
    Described as an adjustable hyperparameter; main results use L=3 with max pooling, and the choice affects receptive field and performance.
  • Number of centroid nodes k (Centroid-Graphs) = 9 (implied in Section 5.1)
    Centroid-graphs compress each WSI to k nodes; Section 5.1 notes the representation drops to 'only nine nodes', and k is chosen by hand without ablation.
  • PCA target dimension for confounder bag embeddings = not reported
    Bag embeddings are PCA-reduced before K-means in Section 3.3, but the target dimension is unspecified.
assumptions (4)
  • domain assumption Backdoor adjustment with an observable confounder set C is valid even though the causal graph is unknown
    Section 3.3 states 'the causal graph for our problem is unknown' yet applies Eqs. (6)-(9) following IBMIL; validity requires C to block all backdoor paths.
  • ad hoc to paper K-means clusters of the model's own bag embeddings represent visual confounders such as stain variation
    Section 3.3 states K-means produces clusters 'which likely represent visual biases in the images (e.g., variations in color staining)'. The 'likely' is untested, and clusters also separate classes (cluster purity 1.0 in Section 5.3).
  • domain assumption Frozen ImageNet-pretrained ResNet-50 patch features are adequate for all models under domain shift
    Section 3.1 fixes the feature extractor for every compared model; robustness conclusions are conditional on these features.
  • domain assumption A single-medical-center test fold in Camelyon17 appropriately measures out-of-distribution generalization
    Section 4.2 uses this protocol; it is a reasonable OOD proxy, but the paper overstates its novelty relative to the official Camelyon17 challenge design.

how reviews work

0 comments
Cite this review

Pith. "Pith review of The Role of Graph-based MIL and Interventional Training in the Generalization of WSI Classifiers." pith.science (2026). https://pith.science/paper/5MVWLRMR

@misc{pith2026250119048,
  author       = {Pith},
  title        = {Pith review of: The Role of Graph-based MIL and Interventional Training in the Generalization of WSI Classifiers},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/5MVWLRMR}},
  note         = {Machine review of arXiv:2501.19048}
}
read the original abstract

Whole Slide Imaging (WSI), which involves high-resolution digital scans of pathology slides, has become the gold standard for cancer diagnosis, but its gigapixel resolution and the scarcity of annotated datasets present challenges for deep learning models. Multiple Instance Learning (MIL), a widely-used weakly supervised approach, bypasses the need for patch-level annotations. However, conventional MIL methods overlook the spatial relationships between patches, which are crucial for tasks such as cancer grading and diagnosis. To address this, graph-based approaches have gained prominence by incorporating spatial information through node connections. Despite their potential, both MIL and graph-based models are vulnerable to learning spurious associations, like color variations in WSIs, affecting their robustness. In this dissertation, we conduct an extensive comparison of multiple graph construction techniques, MIL models, graph-MIL approaches, and interventional training, introducing a new framework, Graph-based Multiple Instance Learning with Interventional Training (GMIL-IT), for WSI classification. We evaluate their impact on model generalization through domain shift analysis and demonstrate that graph-based models alone achieve the generalization initially anticipated from interventional training. Our code is available here: github.com/ritamartinspereira/GMIL-IT

Figures

Figures reproduced from arXiv: 2501.19048 by the authors.

Figure 1
Figure 1. General Pipeline of GMIL-IT. The process begins with a feature extractor and graph construction module 3.1. A GNN model is then used to generate spatially aware instance representations, with a MIL aggregator creating the bag embedding and a classifier determining the label 3.2. Bag￾level clustering is performed to build a confounder dictionary, which is concatenated to the bag embeddings during the Interventional T… view at source ↗
Figure 2
Figure 2. Performance Comparison of Baseline and GAT-based MIL Models Across Camelyon16 (left) and Camelyon17 (right) Datasets. recall obtained by Local Region-GCN and a 20% im￾provement over the best accuracy and AUC achieved by Centroid-GCN. These findings suggest a strong correlation between graph representation compactness and performance. As the representation becomes more compact, per￾formance tends to decrease across a… view at source ↗
Figure 3
Figure 3. Performance Comparison of Baseline and GCN-based MIL Models Across Camelyon16 (left) and Camelyon17 (right) Datasets [PITH_FULL_IMAGE:figures/full_fig_p014_3.png] view at source ↗
Figures from the paper (3 more)
Figure 4
Figure 4. Figure 4: t-SNE visualization of bag embeddings from Camelyon16, comparing ABMIL (left) and GAT￾ABMIL (right) for confounder dictionary construction [PITH_FULL_IMAGE:figures/full_fig_p015_4.png]
Figure 5
Figure 5. Figure 5: t-SNE visualization of bag embeddings from Camelyon17, comparing GAT-ABMIL fold 0 (left) and GAT-ABMIL fold 2 (right) for confounder dictionary construction. 15 [PITH_FULL_IMAGE:figures/full_fig_p015_5.png]
Figure 6
Figure 6. Figure 6: Heatmap visualizations for a tumor slides of Camelyon16 and Camelyon17. The blue lines outline tumor regions. Red color indicates patches with high attention score, while blue indicate patches with lowest attention score. 16 [PITH_FULL_IMAGE:figures/full_fig_p016_6.png]

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. Integrating Implicit and Explicit Relational Biases through Graph-Based Multiple Instance Learning: A Case Study in Skin Lesion Diagnosis

    cs.AI 2026-08 conditional novelty 4.0 of 10

    Grid-based graph attention over masked-autoencoder patch embeddings improves balanced accuracy on ISIC-2018 and ISIC-2019 skin lesion classification over image-level and attention-pooling baselines.

Reference graph

Works this paper leans on

40 extracted references · 35 canonical work pages · cited by 1 Pith paper

  1. [1]

    Learning whole-slide segmentation from inexact and incomplete labels using tissue graphs

    Valentin Anklin, Pushpak Pati, Guillaume Jaume, Behzad Bozorgtabar, Antonio Foncubierta-Rodriguez, Jean-Philippe Thiran, Mathilde Sibony, Maria Gabrani, and Orcun Goksel. Learning whole-slide segmentation from inexact and incomplete labels using tissue graphs. Medical Image Computing and Computer Assisted Intervention – MICCAI 2021, page 636–646, 2021. do...

    work page doi:10.1007/978-3-030-87196-3_59 2021

  2. [2]

    Van Der Laak, Meyke Hermsen, Quirine F

    Babak Ehteshami Bejnordi, Mitko Veta, Paul Johannes Van Diest, Bram Van Ginneken, Nico Karssemeijer, Geert Litjens, Jeroen A.W.M. Van Der Laak, Meyke Hermsen, Quirine F. Manson, Maschenka Balkenhol, Oscar Geessink, Nikolaos Stathonikos, Marcory C.R.F. Van Dijk, Peter Bult, Francisco Beca, Andrew H. Beck, Dayong Wang, Aditya Khosla, Rishab Gargeya, Humayun...

    work page arXiv 2017

  3. [3]

    Enhancing pfi prediction with gds-mil: A graph-based dual stream mil approach

    Gianpaolo Bontempo, Nicola Bartolini, Marta Lovino, Federico Bolelli, Anni Virtanen, and Elisa Ficarra. Enhancing pfi prediction with gds-mil: A graph-based dual stream mil approach. In Gian Luca Foresti, Andrea Fusiello, and Edwin Hancock, editors, Image Analysis and Processing -- ICIAP 2023, pages 550--562, Cham, 2023 a . Springer Nature Switzerland. IS...

    work page 2023

  4. [4]

    Das-mil: Distilling across scales for mil classification of histological wsis

    Gianpaolo Bontempo, Angelo Porrello, Federico Bolelli, Simone Calderara, and Elisa Ficarra. Das-mil: Distilling across scales for mil classification of histological wsis. In Hayit Greenspan, Anant Madabhushi, Parvin Mousavi, Septimiu Salcudean, James Duncan, Tanveer Syeda-Mahmood, and Russell Taylor, editors, Medical Image Computing and Computer Assisted ...

    work page 2023

  5. [5]

    Whole slide image quality in digital pathology: Review and perspectives

    Romain Brixtel, Sébastien Bougleux, Olivier Lézoray, Yann Caillot, Benoît Lemoine, Mathieu Fontaine, Dalal Nebati, and Arnaud Renouf. Whole slide image quality in digital pathology: Review and perspectives. IEEE Access, 10: 0 131005--131035, 2022. doi:10.1109/ACCESS.2022.3227437

    work page arXiv 2022

  6. [6]

    Péter Bándi, Oscar Geessink, Quirine Manson, Marcory Van Dijk, Maschenka Balkenhol, Meyke Hermsen, Babak Ehteshami Bejnordi, Byungjae Lee, Kyunghyun Paeng, Aoxiao Zhong, Quanzheng Li, Farhad Ghazvinian Zanjani, Svitlana Zinger, Keisuke Fukuta, Daisuke Komura, Vlado Ovtcharov, Shenghua Cheng, Shaoqun Zeng, Jeppe Thagaard, Anders B. Dahl, Huangjing Lin, Hao...

    work page arXiv 2019

  7. [7]

    Camil: Causal multiple instance learning for whole slide image classification

    Kaitao Chen, Shiliang Sun, and Jing Zhao. Camil: Causal multiple instance learning for whole slide image classification. Proceedings of the AAAI Conference on Artificial Intelligence, 38 0 (2): 0 1120--1128, Mar. 2024. doi:10.1609/aaai.v38i2.27873. URL https://ojs.aaai.org/index.php/AAAI/article/view/27873

    work page doi:10.1609/aaai.v38i2.27873 2024

  8. [8]

    Chen, Ming Y

    Richard J. Chen, Ming Y. Lu, Muhammad Shaban, Chengkuan Chen, Tiffany Y. Chen, Drew F. K. Williamson, and Faisal Mahmood. Whole slide images are 2d point clouds: Context-aware survival prediction using patch-based graph convolutional networks. In Marleen de Bruijne, Philippe C. Cattin, St \'e phane Cotin, Nicolas Padoy, Stefanie Speidel, Yefeng Zheng, and...

    work page 2021

Show all 40 references
  1. [9]

    Causal inference meets machine learning

    Peng Cui, Zheyan Shen, Sheng Li, Liuyi Yao, Yaliang Li, Zhixuan Chu, and Jing Gao. Causal inference meets machine learning. In Proceedings of the 26th ACM SIGKDD International Conference on Knowledge Discovery & Data Mining, KDD '20, page 3527–3528, New York, NY, USA, 2020. As...

    2020

  2. [10]

    Zuckerman, Tairan Liu, Anthony E

    Kevin de Haan, Yijie Zhang, Jonathan E. Zuckerman, Tairan Liu, Anthony E. Sisk, Miguel F.P. Diaz, Kuang Yu Jen, Alexander Nobori, Sofia Liou, Sarah Zhang, Rana Riahi, Yair Rivenson, W. Dean Wallace, and Aydogan Ozcan. Deep learning-based transformation of h&e stained tissues i...

    2021 doi

  3. [11]

    Combining graph neural network and mamba to capture local and global tissue spatial relationships in whole slide images, 2024

    Ruiwen Ding, Kha-Dinh Luong, Erika Rodriguez, Ana Cristina Araujo Lemos da Silva, and William Hsu. Combining graph neural network and mamba to capture local and global tissue spatial relationships in whole slide images, 2024. URL https://arxiv.org/abs/2406.04377

    2024 arXiv

  4. [12]

    A comprehensive review on multiple instance learning

    Samman Fatima, Sikandar Ali, and Hee-Cheol Kim. A comprehensive review on multiple instance learning. Electronics, 12 0 (20), 2023. ISSN 2079-9292. doi:10.3390/electronics12204323. URL https://www.mdpi.com/2079-9292/12/20/4323

    2023 doi

  5. [13]

    Node-aligned graph convolutional network for whole-slide image representation and classification

    Yonghang Guan, Jun Zhang, Kuan Tian, Sen Yang, Pei Dong, Jinxi Xiang, Wei Yang, Junzhou Huang, Yuyao Zhang, and Xiao Han. Node-aligned graph convolutional network for whole-slide image representation and classification. In Proceedings of the IEEE/CVF Conference on Computer Vis...

    2022

  6. [14]

    Investigating out-of-distribution generalization of gnns: An architecture perspective

    Kai Guo, Hongzhi Wen, Wei Jin, Yaming Guo, Jiliang Tang, and Yi Chang. Investigating out-of-distribution generalization of gnns: An architecture perspective. In Proceedings of the 30th ACM SIGKDD Conference on Knowledge Discovery and Data Mining, KDD '24, page 932–943, New Yor...

    2024

  7. [15]

    Ziaul Hoque, Anja Keskinarkaus, Pia Nyberg, Taneli Mattila, and Tapio Seppänen

    Md. Ziaul Hoque, Anja Keskinarkaus, Pia Nyberg, Taneli Mattila, and Tapio Seppänen. Whole slide image registration via multi-stained feature matching. Computers in Biology and Medicine, 144: 0 105301, 2022. ISSN 0010-4825. doi:https://doi.org/10.1016/j.compbiomed.2022.105301. ...

    2022

  8. [16]

    Tomczak, and Max Welling

    Maximilian Ilse, Jakub M. Tomczak, and Max Welling. Attention-based deep multiple instance learning. In 35th International Conference on Machine Learning, ICML 2018, volume 5, 2018

    2018

  9. [17]

    Guido W. Imbens. Causal inference in the social sciences. Annual Review of Statistics and Its Application, 11 0 (Volume 11, 2024): 0 123--152, 2024. ISSN 2326-831X. doi:https://doi.org/10.1146/annurev-statistics-033121-114601. URL https://www.annualreviews.org/content/journals...

    2024 doi

  10. [18]

    The devil is in the details: Whole slide image acquisition and processing for artifacts detection, color variation, and data augmentation: A review

    Neel Kanwal, Fernando Pérez-Bueno, Arne Schmidt, Kjersti Engan, and Rafael Molina. The devil is in the details: Whole slide image acquisition and processing for artifacts detection, color variation, and data augmentation: A review. IEEE Access, 10: 0 58821--58844, 2022. doi:10...

    2022

  11. [19]

    Lubanski, and Peter M

    Yongnam Kim, Stanley A. Lubanski, and Peter M. Steiner. Matching Strategies for Causal Inference with Observational Data in Education, pages 173--191. Springer International Publishing, Cham, 2018. ISBN 978-3-319-93539-3. doi:10.1007/978-3-319-93539-3_9. URL https://doi.org/10...

    2018 doi

  12. [20]

    Kipf and Max Welling

    Thomas N. Kipf and Max Welling. Semi-supervised classification with graph convolutional networks. In International Conference on Learning Representations, 2017. URL https://openreview.net/forum?id=SJU4ayYgl

    2017

  13. [21]

    Machine learning methods for histopathological image analysis

    Daisuke Komura and Shumpei Ishikawa. Machine learning methods for histopathological image analysis. Computational and Structural Biotechnology Journal, 16: 0 34--42, 2018. ISSN 2001-0370. doi:https://doi.org/10.1016/j.csbj.2018.01.001. URL https://www.sciencedirect.com/science...

    2018 doi

  14. [22]

    Eliceiri

    Bin Li, Yin Li, and Kevin W. Eliceiri. Dual-stream multiple instance learning network for whole slide image classification with self-supervised contrastive learning. In Proceedings of the IEEE Computer Society Conference on Computer Vision and Pattern Recognition, 2021. doi:10...

    2021

  15. [23]

    Graph cnn for survival analysis on whole slide pathological images

    Ruoyu Li, Jiawen Yao, Xinliang Zhu, Yeqing Li, and Junzhou Huang. Graph cnn for survival analysis on whole slide pathological images. In Alejandro F. Frangi, Julia A. Schnabel, Christos Davatzikos, Carlos Alberola-L \'o pez, and Gabor Fichtinger, editors, Medical Image Computi...

    2018

  16. [24]

    A comprehensive review of computer-aided whole-slide image analysis: from datasets to feature extraction, segmentation, classification and detection approaches

    Xintong Li, Chen Li, Md Mamunur Rahaman, Hongzan Sun, Xiaoqi Li, Jian Wu, Yudong Yao, and Marcin Grzegorzek. A comprehensive review of computer-aided whole-slide image analysis: from datasets to feature extraction, segmentation, classification and detection approaches. Artific...

    2022 doi

  17. [25]

    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 Computer Society Conference on Computer Vision and Pattern Recognition, volume 2023-June, 2023. doi:10.110...

    2023

  18. [26]

    1399 H&E-stained sentinel lymph node sections of breast cancer patients: the CAMELYON dataset

    Geert Litjens, Peter Bandi, Babak Ehteshami Bejnordi, Oscar Geessink, Maschenka Balkenhol, Peter Bult, Altuna Halilovic, Meyke Hermsen, Rob van de Loo, Rob Vogels, Quirine F Manson, Nikolas Stathonikos, Alexi Baidoshvili, Paul van Diest, Carla Wauters, Marcory van Dijk, and Je...

    2018 doi

  19. [27]

    Lu, Drew F.K

    Ming Y. Lu, Drew F.K. 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, 2021. ISSN 2157846X. doi:10.1038/s41551-020-00682-w

    2021 doi

  20. [28]

    Rita Verdelho, Diogo J

    Tiago Mota, M. Rita Verdelho, Diogo J. Ara\'ujo, Alceu Bissoto, Carlos Santiago, and Catarina Barata. Mmist-ccrcc: A real world medical dataset for the development of multi-modal systems. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVP...

    2024

  21. [29]

    Graph attention multi-instance learning for accurate colorectal cancer staging

    Ashwin Raju, Jiawen Yao, Mohammad MinHazul Haq, Jitendra Jonnagaddala, and Junzhou Huang. Graph attention multi-instance learning for accurate colorectal cancer staging. In Anne L. Martel, Purang Abolmaesumi, Danail Stoyanov, Diana Mateus, Maria A. Zuluaga, S. Kevin Zhou, Dani...

    2020

  22. [30]

    D. Sculley. Web-scale k-means clustering. In Proceedings of the 19th International Conference on World Wide Web, WWW '10, page 1177–1178, New York, NY, USA, 2010. Association for Computing Machinery. ISBN 9781605587998. doi:10.1145/1772690.1772862. URL https://doi.org/10.1145/...

    2010

  23. [31]

    A structure-aware hierarchical graph-based multiple instance learning framework for pt staging in histopathological image

    Jiangbo Shi, Lufei Tang, Yang Li, Xianli Zhang, Zeyu Gao, Yefeng Zheng, Chunbao Wang, Tieliang Gong, and Chen Li. A structure-aware hierarchical graph-based multiple instance learning framework for pt staging in histopathological image. IEEE Transactions on Medical Imaging, 42...

    2023

  24. [32]

    Attention-based deep multiple instance learning with adaptive instance sampling

    Aliasghar Tarkhan, Trung Kien Nguyen, Noah Simon, Thomas Bengtsson, Paolo Ocampo, and Jian Dai. Attention-based deep multiple instance learning with adaptive instance sampling. In 2022 IEEE 19th International Symposium on Biomedical Imaging (ISBI), pages 1--5, March 2022. doi:...

    2022

  25. [33]

    Hal R. Varian. Causal inference in economics and marketing. Proceedings of the National Academy of Sciences, 113 0 (27): 0 7310--7315, 2016. doi:10.1073/pnas.1510479113. URL https://www.pnas.org/doi/abs/10.1073/pnas.1510479113

    2016 doi

  26. [34]

    Graph attention networks

    Petar Veličković, Guillem Cucurull, Arantxa Casanova, Adriana Romero, Pietro Liò, and Yoshua Bengio. Graph attention networks. In International Conference on Learning Representations, 2018. doi:10.1007/978-3-031-01587-8_7. URL https://openreview.net/forum?id=rJXMpikCZ

    2018 doi

  27. [35]

    Dual-stream multi-dependency graph neural network enables precise cancer survival analysis

    Zhikang Wang, Jiani Ma, Qian Gao, Chris Bain, Seiya Imoto, Pietro Liò, Hongmin Cai, Hao Chen, and Jiangning Song. Dual-stream multi-dependency graph neural network enables precise cancer survival analysis. Medical Image Analysis, 97: 0 103252, 2024. ISSN 1361-8415. doi:https:/...

    2024

  28. [36]

    Show, attend and tell: Neural image caption generation with visual attention

    Kelvin Xu, Jimmy Ba, Ryan Kiros, Kyunghyun Cho, Aaron Courville, Ruslan Salakhudinov, Rich Zemel, and Yoshua Bengio. Show, attend and tell: Neural image caption generation with visual attention. In Francis Bach and David Blei, editors, Proceedings of the 32nd International Con...

    2015

  29. [37]

    Individual and structural graph information bottlenecks for out-of-distribution generalization

    Ling Yang, Jiayi Zheng, Heyuan Wang, Zhongyi Liu, Zhilin Huang, Shenda Hong, Wentao Zhang, and Bin Cui. Individual and structural graph information bottlenecks for out-of-distribution generalization. IEEE Transactions on Knowledge and Data Engineering, 36 0 (2): 0 682--693, 20...

    2024

  30. [38]

    Gigapixel whole-slide images classification using locally supervised learning

    Jingwei Zhang, Xin Zhang, Ke Ma, Rajarsi Gupta, Joel Saltz, Maria Vakalopoulou, and Dimitris Samaras. Gigapixel whole-slide images classification using locally supervised learning. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligen...

    2022 doi

  31. [39]

    Predicting lymph node metastasis using histopathological images based on multiple instance learning with deep graph convolution

    Yu Zhao, Fan Yang, Yuqi Fang, Hailing Liu, Niyun Zhou, Jun Zhang, Jiarui Sun, Sen Yang, Bjoern Menze, Xinjuan Fan, and Jianhua Yao. Predicting lymph node metastasis using histopathological images based on multiple instance learning with deep graph convolution. In Proceedings o...

    2020

  32. [40]

    Yanning Zhou, Simon Graham, Navid Alemi Koohbanani, Muhammad Shaban, Pheng-Ann Heng, and Nasir M. Rajpoot. Cgc-net: Cell graph convolutional network for grading of colorectal cancer histology images. 2019 IEEE/CVF International Conference on Computer Vision Workshop (ICCVW), p...

    2019

Pith tools

Reviewed August 9, 2026 · model on record in the stance chip above.