Pith. sign in

REVIEW 5 major objections 5 minor 38 references

HIEGNet: A Heterogenous Graph Neural Network Including the Immune Environment in Glomeruli Classification

T0 review · 5 major / 5 minor · reviewed 2026-08-07 · deepseek-v4-flash

Pith's one-line read HIEGNet, a heterogeneous graph network that includes immune cells as nodes, classifies glomeruli and generalises between patients better than the CNN and random-forest baselines tested on the EXC dataset.

desk verdict Useful, honest computational pathology paper whose cross-patient immune-environment claim is weaker than it looks once you check the segmentation quality and the missing homogeneous GNN baseline. read the letter →

arxiv 2506.02542 v1 pith:S36KZLIJ submitted 2025-06-03 cs.LG cs.AIcs.CVq-bio.QM

classification cs.LGcs.AIcs.CVq-bio.QM
keywords graphneuralnetworkshistopathologyglomeruliclassificationimmuneenvironmentheterogeneousgraphswholeslideimageskidneytransplantglomerulosclerosis
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

The paper tries to establish that a heterogeneous graph representation of kidney tissue, where each glomerulus and the immune cells around it are nodes connected by distance-based edges, lets a graph neural network classify glomeruli as healthy, sclerotic, or dead, and that this representation transfers to unseen patients better than the tested convolutional baselines and a random forest. On the EXC dataset of six kidney-transplant patients with 2052 glomeruli, the proposed HIEGNet reaches a macro F1 of 0.73 between patients, against 0.60 for random forest, 0.59 for ResNet-18, 0.55 for EfficientNetV2, and 0.54 for U-Net. The paper further argues that the immune environment matters: deleting edges between glomeruli and immune cells drops the within-patient F1 by 0.07, the largest drop among the three edge groups. A sympathetic reader should care because, if true, it shows that explicitly modelling cell-level interactions with GNNs is a viable route to whole-slide histopathology classification that generalises across patients.

What carries the argument

The central object is the heterogeneous graph $G = (V, E, T, R)$ with node types $T = \{$glomeruli, macrophages, T-cells$\}$, built from whole-slide images by segmentation (manual glomeruli masks, Cellpose or contour detection for immune cells) and hand-crafted features (Local Binary Patterns plus shape descriptors). Edges are created by Euclidean distance, with edge-type-specific construction: k-NN plus $\epsilon$-neighbourhood ($\epsilon = 100\,\mu$m) among immune cells, $\epsilon = 277\,\mu$m between glomeruli and immune cells, and $\epsilon = 138.6\,\mu$m between glomeruli. The architecture uses RGCN-style aggregation with a different message passing function per edge-type group, namely GraphSAGE for glomerulus–glomerulus edges, GATv2 for glomerulus–immune edges, and SchNet's CFconv for immune–immune edges, sharing parameters only within the same edge type. That design is what lets the model exchange information between a glomerulus and its immune environment and across neighbouring glomeruli.

What would settle it

Run the between-patient protocol twice: once with the current pipeline's predicted immune-cell nodes and once with all macrophage and T-cell nodes removed entirely. If the macro F1 does not fall below 0.73 ± 0.01 when immune nodes are removed, the claim that the immune environment drives cross-patient generalisation is falsified.

Watch

Extended reading notes

Core claim

On its own terms, the paper reports that HIEGNet generalises best between patients among all models tested: macro F1 0.73 ± 0.01 on patients 004–006 after training on patients 001–003, versus 0.60 for Random Forest, 0.59 for ResNet-18, 0.55 for EfficientNetV2, and 0.54 for the U-Net. Within the same patients, the U-Net reaches 0.77 and HIEGNet 0.73; HIEGNet's between-patient drop is much smaller than the CNNs'. The paper also reports that removing glomerulus–immune-cell edges lowers within-patient F1 by 0.07, more than removing the other edge groups, which it reads as evidence that the immune environment carries classification signal.

Load-bearing premise

The claim depends on the automatically segmented immune cells around a glomerulus being accurate enough on new patients to provide the cross-patient signal attributed to the immune environment; the paper's own Appendix B reports macrophage segmentation AUC as low as 0.003 in between-patient settings, so those nodes may be mostly noise outside the training patients.

Editorial extensions

If this is right

  • If HIEGNet's cross-patient F1 of 0.73 holds, GNNs that explicitly model immune cells become a viable alternative to CNNs for nephropathology, with the benefit of whole-slide context.
  • The linear-in-edges complexity and measured 1.21 s per epoch make the approach feasible on full WSIs, including graphs with up to roughly 4.1 million edges.
  • Hand-crafted, rotation- and stain-robust features (LBP and shape) appear to transfer across patients, as evidenced by Random Forest's small gap between the within- and between-patient settings.
  • The edge-removal ablation indicates that glomerulus–immune-cell edges carry more classification-relevant information than immune–immune or glomerulus–glomerulus edges.
  • The architecture's per-edge-type parameter sharing means additional cell types, such as B-cells, can be added without redesigning the network.

Reading between the lines

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

  • The paper does not show that the immune-cell nodes themselves drive the between-patient gain; given the reported macrophage segmentation AUC as low as 0.003 on unseen patients, the gain may be carried by the hand-crafted glomeruli features and glomerulus–glomerulus edges, and a direct ablation removing immune nodes would separate these.
  • The same graph construction and edge-type-group architecture could be applied to other histopathology tasks with different object types, provided the cells of interest can be stained and segmented, though the distance thresholds and feature sets would need re-derivation.
  • Because the paper's stain-augmentation experiments degrade HIEGNet much more than a pre-trained U-Net, the cross-patient advantage may depend on the training and test images sharing the same staining protocol; training on multiple sites is a testable extension the paper itself flags as future work.
  • The between-patient test set contains only three patients with heterogeneous class distributions, so the reported superiority should be read with the small sample size in mind.
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

5 major / 5 minor

Summary. The manuscript proposes HIEGNet, a heterogeneous graph neural network for classifying glomeruli in whole-slide kidney histopathology images. Nodes represent glomeruli, macrophages, and T-cells; edges are created from distance thresholds and k-nearest-neighbour rules; node features are hand-crafted LBP and shape descriptors. The authors evaluate HIEGNet on the EXC dataset of six kidney-transplant patients in two settings: within patients (patients 001-003, 85/15 split) and between patients (train on 001-003, test on 004-006). They report that HIEGNet reaches a between-patient macro F1 of 0.73 ± 0.01, outperforming Random Forest (0.60), ResNet-18 (0.59), EfficientNetV2 (0.55), and U-Net (0.54). The implementation is publicly available.

Significance. If substantiated, the between-patient generalisation result would be a useful step for graph-based histopathology, since it targets the practically important scenario of classifying glomeruli in patients not seen during training. Strengths of the paper include the public code, the detailed hyperparameter-search description, the reported computational cost of graph construction and training, and the explicit edge-importance ablation within patients. However, the evidence is currently thinner than the central claim: the between-patient comparison rests on a single train/test patient split, the segmentation of immune cells is reported to be very poor on some between-patient settings, and no homogeneous-GNN or glomeruli-only control is provided. The significance of the contribution is therefore conditional on closing these gaps.

major comments (5)
  1. [Appendix B, Tables 2 and 3] The segmentation evaluation is internally inconsistent and does not cover the actual test patients. Table 3 reports between-patient AP-AUC values for macrophages as low as 0.003 and 0.000, and the column headers list test patients 001-004, whereas Section 5 defines the between-patient test set as patients 004-006 and Table 1 assigns only patient 004 patches to the between-patient test set. Consequently, the manuscript provides no segmentation-quality measurement on the patients used for the reported 0.73 between-patient F1 score. Since the immune-cell nodes in the graph are produced by this segmentation, the claim that the immune environment contributes to cross-patient performance is not supported without either improved segmentation metrics on patients 004-006 or a control that removes the immune nodes.
  2. [Section 5 and Appendix K] The only ablation that isolates the contribution of the immune environment is performed in the within-patient setting (Appendix K, Table 10), where segmentation quality is highest. No equivalent between-patient ablation is reported, and no homogeneous GNN or glomeruli-only graph control is provided. Because the Random Forest baseline already uses the same hand-crafted glomerulus features, the observed 0.73 versus 0.60 between-patient gap cannot be attributed to the immune-cell nodes or to the heterogeneous graph structure without such a control. The authors should add a between-patient ablation that removes immune-cell edge groups, and ideally a homogeneous GNN baseline on a glomeruli-only graph.
  3. [Section 5, 'between patients setting'] The between-patient result is based on a single partition of the six patients into training (001-003) and test (004-006) sets, with no significance testing across patients. The reported standard deviations come from 20 parameter initialisations and therefore quantify only initialisation variance, not the variance across possible patient splits. To support the claim that HIEGNet 'generalises best between patients among all baseline models', the authors should report patient-level cross-validation (or a permutation test) and, if possible, confidence intervals for the macro-F1 differences. Without this, the 0.13-point gap over Random Forest could be driven by the particular choice of patients.
  4. [Section 5, Baselines] The baseline set contains no standard GNN (e.g., GCN, GraphSAGE, or RGCN) applied to the same node features and graph structure. The comparison is therefore between HIEGNet, CNNs, and a Random Forest that ignores graph topology. This omission makes it impossible to isolate the contribution of the proposed heterogeneous message-passing architecture from the contribution of the graph representation itself. Adding a homogeneous GNN baseline on the same graph and features is necessary to support the architectural claims in Section 4.
  5. [Appendix B, threshold selection] The description of how the contour-detection thresholds were selected is contradictory and raises a possible leakage issue. The text states that 'the thresholds were optimised on the training set', but then says 'the final thresholds were determined on the test set within patients 001 - 003 with an intensity threshold value of 60 and an area threshold of 160 pixels'. If the threshold values were chosen using the within-patient test split, then the within-patient evaluation is not a clean evaluation, and the model selection performed on that setting may be compromised. The authors must clarify the protocol and, if thresholds were indeed selected on the test set, re-run the experiments with a proper training/validation split.
minor comments (5)
  1. [Section 4, display equation after Eq. (1)] The SAGE aggregation term writes 'sum_{u in N(v)} h_v / |N(v)|', which appears to be a typo: the summed quantity should be the neighbour embedding h_u, not h_v. The neighbouring set in the first term also reads 'N(v) ∪ {u}', which is unclear and should be written as N(v) or N(v) ∪ {v} depending on whether self-loops are intended.
  2. [Appendix G, Table 6] Precision and recall for ResNet-18 and EfficientNetV2 in the between-patients columns are reported as '0.– ± 0.–', which is not a usable result. The authors should report the actual values or explicitly state why they are unavailable (for example, due to undefined metrics from a missing class).
  3. [Appendix E, first paragraph] The text says 'all four patients are affected by glomerulosclerosis', but the dataset contains six patients and the following sentences refer to six patients. This is presumably a typo and should be corrected.
  4. [Appendix B, Tables 2 and 3] The column headers 'Between patients: 001 002 003 004' conflict with the main text, which uses patients 004-006 as the between-patient test set. The table headers should be relabelled to match the actual experimental setting.
  5. [Throughout] The manuscript contains several typographical or formatting issues, including 'Heterogenous' in the title, 'Gird Search Space' in Appendix H, and unusual spacing in 'F uture W ork'. These should be corrected during revision.

Circularity Check

0 steps flagged · score 1.0 of 10

No circular derivation is present: the central claim is an empirical held-out patient comparison, and the same-group citations are peripheral rather than load-bearing.

full rationale

The paper's central result is an empirical performance comparison (Section 5, Figure 4: between-patient macro F1 0.73 ± 0.01 for HIEGNet versus 0.60 for Random Forest), not a quantity derived from a fitted parameter or from a self-citation. Hyperparameters and message-passing functions are chosen by grid search with 4-fold cross-validation on training patients and then evaluated on held-out patients 004–006, which is standard model selection rather than circular reasoning. The same-group prior work used in the paper (Merveille et al., 2021 for the EXC dataset and the 277 µm immune-environment radius; Nisar and Lampert, 2024 for the U-Net baseline) provides inputs and baselines, but the between-patient F1 does not reduce to these citations by construction, and no uniqueness theorem is invoked to force the architecture. The weak macrophage segmentation results reported in Appendix B (contour-detection AUC as low as 0.003 between patients) are a serious evidential threat to the claim that the immune environment drives cross-patient performance, and the class definitions in Appendix E (healthy defined with 'presence of immune cells', dead with 'no immune cells') create a label-definition overlap with the immune-cell node features; however, these are correctness and confounding concerns, not circular derivations in the sense of an equation or fitted parameter being renamed as a prediction. Appendix B also contains an ambiguous statement that final thresholds were 'determined on the test set', which would be a leakage concern, but it does not by itself make the central comparison circular. The paper is therefore not circular in its derivation chain, though its central empirical claim carries substantial risk from segmentation quality and missing between-patient ablations.

Assumptions & free parameters 6 free parameters · 5 assumptions · 0 invented entities

The central claim rests on many hand-tuned parameters (segmentation thresholds, graph radii, architecture hyperparameters) and domain assumptions (segmentation quality, biological relevance of immune proximity, label accuracy). No entirely new entities are introduced; HIEGNet is a model architecture, not a new physical object.

free parameters (6)
  • Macrophage contour detection thresholds = Intensity threshold 60, area threshold 160 pixels
    Optimized on the training set of patients 001-003 (Appendix B); used to generate all macrophage nodes.
  • T-cell segmentation model hyperparameters = Cellpose fine-tuned with lr 0.001, weight decay 0.01, 100 epochs
    Selected for T-cell segmentation (Appendix B).
  • Glomeruli-glomeruli edge radius = epsilon = 550 pixels (138.6 um)
    Chosen from 4-fold CV on the training set as best-performing (Appendix C); main text Section 3.3.
  • Immune-immune edge radius and k = epsilon = 100 um, k = 5
    Chosen by design for immune-immune edges (Section 3.3).
  • HIEGNet hyperparameters = dropout 0.2, hidden dim 64, layers 2, FC layers 2, message passing SAGE/GATv2/CFconv
    Grid search with 4-fold CV on the within-patients setting (Appendix H).
  • Node feature set = LBP (8,1), area, perimeter, eccentricity, circularity, aspect ratio, inside/outside binary
    Hand-selected to be rotation-invariant and stain-robust (Section 3.2).
assumptions (5)
  • domain assumption Glomerulus segmentation masks are available for every WSI
    Section 3.1 states glomeruli segmentations are assumed available through manual annotation or a learned model; the graph pipeline depends on this.
  • domain assumption Macrophages and T-cells near glomeruli carry information about glomerular fibrosis state
    Motivated by cited biology (Wynn and Barron, 2010; Xu et al., 2023; Braun et al., 2021); it is the biological premise of the immune-environment graph.
  • domain assumption Euclidean distance between cells is a valid proxy for biological interaction likelihood
    Section 3.3 builds edges by distance thresholds and kNN; no functional validation of this proxy is provided.
  • domain assumption The EXC dataset labels (healthy/sclerotic/dead) are accurate
    Labels were obtained by consensus-based manual annotation by pathologists (Appendix E); no external validation or inter-rater agreement is reported.
  • domain assumption Stain deconvolution and segmentation methods isolate CD3 and CD68 positive cells
    Appendix B shows the evaluated segmentation methods have low AUC (e.g., macrophage contour detection AUC 0.265 within patients), so this assumption is partially contradicted by the paper's own evaluation.

how reviews work

0 comments
Cite this review

Pith. "Pith review of HIEGNet: A Heterogenous Graph Neural Network Including the Immune Environment in Glomeruli Classification." pith.science (2026). https://pith.science/paper/S36KZLIJ

@misc{pith2026250602542,
  author       = {Pith},
  title        = {Pith review of: HIEGNet: A Heterogenous Graph Neural Network Including the Immune Environment in Glomeruli Classification},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/S36KZLIJ}},
  note         = {Machine review of arXiv:2506.02542}
}
read the original abstract

Graph Neural Networks (GNNs) have recently been found to excel in histopathology. However, an important histopathological task, where GNNs have not been extensively explored, is the classification of glomeruli health as an important indicator in nephropathology. This task presents unique difficulties, particularly for the graph construction, i.e., the identification of nodes, edges, and informative features. In this work, we propose a pipeline composed of different traditional and machine learning-based computer vision techniques to identify nodes, edges, and their corresponding features to form a heterogeneous graph. We then proceed to propose a novel heterogeneous GNN architecture for glomeruli classification, called HIEGNet, that integrates both glomeruli and their surrounding immune cells. Hence, HIEGNet is able to consider the immune environment of each glomerulus in its classification. Our HIEGNet was trained and tested on a dataset of Whole Slide Images from kidney transplant patients. Experimental results demonstrate that HIEGNet outperforms several baseline models and generalises best between patients among all baseline models. Our implementation is publicly available at https://github.com/nklsKrmnn/HIEGNet.git.

Figures

Figures reproduced from arXiv: 2506.02542 by the authors.

Figure 1
Figure 1. Images (a) and (b) show a healthy and dead glomerulus, respectively. Image (c) shows renal tissue with multiple glomeruli coloured by class membership (green: “healthy”, yellow: “sclerotic”, red: “dead”). In the field of medical diagnosis, histopathology is crucial to detect alterations in tissue and understand biological phenomena on the microscopic level (Welsch and Deller, 2006; Brussee et al., 2025). Glomerulosc… view at source ↗
Figure 2
Figure 2. Illustration of a graph with three glomeruli vg,i (orange) and their surrounding macrophages vm,j (green) and T-cells vt,k (blue). The different line styles illus￾trate different message passing methods and their colour different edge types. Edges of these edge types were constructed using a k-nearest neighbour graph construc￾tion with k = 5 combined with a radius-based ϵ-neighbourhood graph construction to limit ed… view at source ↗
Figure 3
Figure 3. An illustration of all model variations. The regular HIEGNet uses only hand￾crafted node features. In a hybrid model, the CNN output replaces the glomeruli node features. Jumping knowledge concatenates node embeddings from all layers. We also explored the usage of Jumping-Knowledge and a modified U-Net pre-trained on WSIs (Nisar and Lampert, 2024) for feature extraction. These model variations are illustrated in [P… view at source ↗
Figures from the paper (4 more)
Figure 4
Figure 4. Figure 4: F1-Scores macro-averaged of all models evaluated on the test sets. [PITH_FULL_IMAGE:figures/full_fig_p008_4.png]
Figure 5
Figure 5. Figure 5: illustrates the graph construction with the pipeline proposed in Section 3 [PITH_FULL_IMAGE:figures/full_fig_p013_5.png]
Figure 6
Figure 6. Figure 6: F1-Scores macro-averaged of all models including HIEGNet-JK and Hybrid￾HIEGNet evaluated on the test sets. 17 [PITH_FULL_IMAGE:figures/full_fig_p017_6.png]
Figure 7
Figure 7. Figure 7 [PITH_FULL_IMAGE:figures/full_fig_p019_7.png]

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

38 extracted references · 36 canonical work pages

  1. [1]

    Principles of cell circuits for tissue repair and fibrosis

    Miri Adler, Avi Mayo, Xu Zhou, Ruth Franklin, Matthew Meizlish, Ruslan Medzhitov, Stefan Kallenberger, and Uri Alon. Principles of cell circuits for tissue repair and fibrosis. iScience, 2019

  2. [2]

    A survey on graph-based deep learning for computational histopathology

    David Ahmedt-Aristizabal, Ali Armin, Simon Denman, Clinton Fookes, and Lars Petersson. A survey on graph-based deep learning for computational histopathology. Computerized Medical Imaging and Graphics, 2021

  3. [3]

    Performance and limitations of a supervised deep learning approach for the histopathological O xford classification of glomeruli with I g A nephropathy

    Nicola Altini et al. Performance and limitations of a supervised deep learning approach for the histopathological O xford classification of glomeruli with I g A nephropathy. Computer Methods and Programs in Biomedicine, 242, 2023

  4. [4]

    Harnessing AI for kidney glomeruli classification

    Meghna Ayyar, Puneet Mathur, Rajiv Ratn Shah, and Shree G Sharma. Harnessing AI for kidney glomeruli classification. In IEEE International Symposium on Multimedia, 2018

  5. [5]

    Deep learning in histopathology: A review

    Sugata Banerji and Sushmita Mitra. Deep learning in histopathology: A review. WIREs Data Mining and Knowledge Discovery, 2022

  6. [6]

    Immune-mediated entities of (primary) focal segmental glomerulosclerosis

    Fabian Braun, Inka Homeyer, Nada Alachkar, and Tobias Huber. Immune-mediated entities of (primary) focal segmental glomerulosclerosis. Cell and Tissue Research, 2021

  7. [7]

    How attentive are graph attention networks? In International Conference on Learning Representations, 2022

    Shaked Brody, Uri Alon, and Eran Yahav. How attentive are graph attention networks? In International Conference on Learning Representations, 2022

  8. [8]

    Graph neural networks in histopathology: Emerging trends and future directions

    Siemen Brussee, Giorgio Buzzanca, Anne Schrader, and Jesper Kers. Graph neural networks in histopathology: Emerging trends and future directions. Medical Image Analysis, 2025

Show all 38 references
  1. [9]

    Cattran et al

    Daniel C. Cattran et al. The O xford classification of I g A nephropathy: Rationale, clinicopathological correlations, and classification. Kidney International, 2009

  2. [10]

    Schoenholz, Patrick F

    Justin Gilmer, Samuel S. Schoenholz, Patrick F. Riley, Oriol Vinyals, and George E. Dahl. Neural message passing for quantum chemistry. In International Conference on Machine Learning, 2017

  3. [11]

    Bootstrap your own latent: A new approach to self-supervised learning

    Jean-Bastien Grill, Florian Strub, Florent Altché, Corentin Tallec, Pierre Richemond, Elena Buchatskaya, Carl Doersch, Bernardo Pires, Zhaohan Guo, Mohammad Azar, Bilal Piot, Koray Kavukcuoglu, Remi Munos, and Michal Valko. Bootstrap your own latent: A new approach to self-sup...

  4. [12]

    Hamilton, Zhitao Ying, and Jure Leskovec

    William L. Hamilton, Zhitao Ying, and Jure Leskovec. Inductive representation learning on large graphs. In Neural Information Processing Systems, 2017

  5. [13]

    Zhang, Shaoqing Ren, and Jian Sun

    Kaiming He, X. Zhang, Shaoqing Ren, and Jian Sun. Deep residual learning for image recognition. In IEEE Conference on Computer Vision and Pattern Recognition, 2015

  6. [14]

    Global attention based GNN with B ayesian collaborative learning for glomerular lesion recognition

    Qiming He, Shuang Ge, Siqi Zeng, Yanxia Wang, Jing Ye, Yonghong He, Jing Li, Zhe Wang, and Tian Guan. Global attention based GNN with B ayesian collaborative learning for glomerular lesion recognition. Computers in Biology and Medicine, 2024

  7. [15]

    Strategies for pre-training graph neural networks

    Weihua Hu, Bowen Liu, Joseph Gomes, Marinka Zitnik, Percy Liang, Vijay Pande, and Jure Leskovec. Strategies for pre-training graph neural networks. In International Conference on Learning Representations, 2020

  8. [16]

    Kipf and Max Welling

    Thomas N. Kipf and Max Welling. Semi-supervised classification with graph convolutional networks. In International Conference on Learning Representations, 2017

  9. [17]

    Epidemiology of chronic kidney disease: An update 2022

    Csaba Kovesdy. Epidemiology of chronic kidney disease: An update 2022. Kidney international supplements, 2022

  10. [18]

    Imagenet classification with deep convolutional neural networks

    Alex Krizhevsky, Ilya Sutskever, and Geoffrey E Hinton. Imagenet classification with deep convolutional neural networks. In Neural Information Processing Systems. Curran Associates, Inc., 2012

  11. [19]

    Renal fibrosis: New insights into the pathogenesis and therapeutics

    Youhua Liu. Renal fibrosis: New insights into the pathogenesis and therapeutics. Kidney International, 2006

  12. [20]

    An automatic framework for fusing information from differently stained consecutive digital whole slide images: A case study in renal histology

    Odyssee Merveille, Thomas Lampert, Jessica Schmitz, Germain Forestier, Friedrich Feuerhake, and Cédric Wemmert. An automatic framework for fusing information from differently stained consecutive digital whole slide images: A case study in renal histology. Computer Methods and ...

  13. [21]

    Maximising histopathology segmentation using minimal labels via self-supervision

    Zeeshan Nisar and Thomas Lampert. Maximising histopathology segmentation using minimal labels via self-supervision. arXiv:2412.15389, 2024

  14. [22]

    Ojala, M

    T. Ojala, M. Pietikainen, and D. Harwood. Performance evaluation of texture measures with classification based on K ullback discrimination of distributions. In International Conference on Pattern Recognition, 1994

  15. [23]

    U- N et: Convolutional networks for biomedical image segmentation

    Olaf Ronneberger, Philipp Fischer, and Thomas Brox. U- N et: Convolutional networks for biomedical image segmentation. In Medical Image Computing and Computer-Assisted Intervention, 2015

  16. [24]

    Sauls, Cassidy McCausland, and Bryce N

    Ryan S. Sauls, Cassidy McCausland, and Bryce N. Taylor. Histology, T-Cell Lymphocyte. StatPearls Publishing, Treasure Island (FL), 2023

  17. [25]

    Kipf, Peter Bloem, Rianne van den Berg, Ivan Titov, and Max Welling

    Michael Schlichtkrull, Thomas N. Kipf, Peter Bloem, Rianne van den Berg, Ivan Titov, and Max Welling. Modeling relational data with graph convolutional networks. In The Semantic Web, 2018

  18. [26]

    Sauceda, Stefan Chmiela, Alexandre Tkatchenko, and Klaus-Robert Müller

    Kristof Schütt, Pieter-Jan Kindermans, Huziel E. Sauceda, Stefan Chmiela, Alexandre Tkatchenko, and Klaus-Robert Müller. Schnet: A continuous-filter convolutional neural network for modeling quantum interactions. In Neural Information Processing Systems, 2017

  19. [27]

    Cellpose documentation, 2020

    Carsen Stringer and Marius Pachitariu. Cellpose documentation, 2020. URL https://cellpose.readthedocs.io. Accessed: 2024-11-06

  20. [28]

    Cellpose 2.0: H ow to train your own model

    Carsen Stringer and Marius Pachitariu. Cellpose 2.0: H ow to train your own model. Nat Methods, 2022

  21. [29]

    Cellpose: A generalist algorithm for cellular segmentation

    Carsen Stringer, Michalis Michaelos, and Marius Pachitariu. Cellpose: A generalist algorithm for cellular segmentation. Nat Methods, 2020

  22. [30]

    Topological structural analysis of digitized binary images by border following

    Satoshi Suzuki and Keiichi Abe. Topological structural analysis of digitized binary images by border following. Computer Vision, Graphics, and Image Processing, 1985

  23. [31]

    Mingxing Tan and Quoc V. Le. Efficient N et V 2: Smaller models and faster training. In International Conference on Machine Learning, 2021

  24. [32]

    Whole-slide mitosis detection in H & E breast histology using PHH 3 as a reference to train distilled stain-invariant convolutional networks

    David Tellez, Maschenka Balkenhol, Irene Otte-Höller, Rob van de Loo, Rob Vogels, Peter Bult, Carla Wauters, Willem Vreuls, Suzanne Mol, Nico Karssemeijer, Geert Litjens, Jeroen van der Laak, and Francesco Ciompi. Whole-slide mitosis detection in H & E breast histology using P...

  25. [33]

    Towards histopathological stain invariance by unsupervised domain augmentation using generative adversarial networks

    Jelica Vasiljevic, Friedrich Feuerhake, Cédric Wemmert, and Thomas Lampert. Towards histopathological stain invariance by unsupervised domain augmentation using generative adversarial networks. Neurocomputing, 2021

  26. [34]

    Histostargan: A unified approach to stain normalisation, stain transfer and stain invariant segmentation in renal histopathology

    Jelica Vasiljevic, Friedrich Feuerhake, Cédric Wemmert, and Thomas Lampert. Histostargan: A unified approach to stain normalisation, stain transfer and stain invariant segmentation in renal histopathology. Knowledge-Based Systems, 2022

  27. [35]

    Sabotta Lehrbuch Histologie

    Ulrich Welsch and Thomas Deller. Sabotta Lehrbuch Histologie. Elsevier, 2nd edition, 2006

  28. [36]

    Macrophages: M aster regulators of inflammation and fibrosis

    Thomas Wynn and Luke Barron. Macrophages: M aster regulators of inflammation and fibrosis. Seminar in Liver Disease, 2010

  29. [37]

    Representation learning on graphs with jumping knowledge networks

    Keyulu Xu, Chengtao Li, Yonglong Tian, Tomohiro Sonobe, Ken-ichi Kawarabayashi, and Stefanie Jegelka. Representation learning on graphs with jumping knowledge networks. In International Conference on Machine Learning, 2018

  30. [38]

    Cd8 t cell‐derived exosomal mir‐186‐5p elicits renal inflammation via activating tubular tlr7/8 signal axis

    Xiaodong Xu, Shuang Qu, Changming Zhang, Mingchao Zhang, Weisong Qin, Guisheng Ren, Hao Bao, Limin Li, Ke Zen, and Hehe Liu. Cd8 t cell‐derived exosomal mir‐186‐5p elicits renal inflammation via activating tubular tlr7/8 signal axis. Advanced Science, 2023

Pith tools

Reviewed August 7, 2026 · model on record in the stance chip above.