Pith. sign in

REVIEW 4 major objections 5 minor 41 references

Towards Accurate and Interpretable Neuroblastoma Diagnosis via Contrastive Multi-scale Pathological Image Analysis

T0 review · 4 major / 5 minor · reviewed 2026-08-16 · deepseek-v4-flash

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

Pith's one-line read A 1.32M-parameter, randomly initialized model is claimed to beat UNI, CONCH, and CTransPath on neuroblastoma patch classification, and to reach 100% WSI accuracy with soft voting.

desk verdict A plausible lightweight architecture with patch-level promise, but the WSI-level claims rest on an impossible comparison table and an under-specified, internally inconsistent voting protocol. read the letter →

arxiv 2504.13754 v3 pith:IX4CMGJL submitted 2025-04-18 cs.CV cs.AI

Zhu Zhu , Shuo Jiang , Jingyuan Zheng , Yawen Li , Yifei Chen , Manli Zhao , Weizhong Gu , Feiwei Qin
show 2 more authors
Jinhu Wang Gang Yu
This is my paper · ORCID
classification cs.CVcs.AI
keywords neuroblastomawhole-slideimageclassificationKolmogorov-ArnoldNetworkSwinTransformercontrastivelearningmulti-scalefeaturefusionsoftvotinghistopathology
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

This paper tries to establish that a lightweight model trained from scratch, without pretraining on massive pathology datasets, can outperform large pretrained pathology foundation models on neuroblastoma subtype classification. It argues that replacing the MLP layers in a Swin Transformer with Kolmogorov-Arnold Network (KAN) layers, combined with a contrastive multi-scale feature fusion module, improves both accuracy and interpretability. The paper also introduces a clinically guided soft voting mechanism intended to turn patch-level predictions into accurate whole-slide diagnoses. If these claims hold, small hospitals with limited data and computing resources could deploy accurate pediatric tumor classifiers without depending on large pretrained models.

What carries the argument

The central mechanism is a Swin Transformer variant in which every MLP is replaced by a Kolmogorov-Arnold Network (KAN) layer, whose activations are learnable B-spline functions. A contrastive-driven multi-scale aggregation module (CMSA), with its CDFA submodule, aligns feature maps from the first three transformer stages and uses cross-scale attention to weight local windows, while the KAN classification head fuses stage-4 residuals with the aggregated multi-scale features. For whole-slide diagnosis, a multi-class SVM trained on the model's features recognizes neuropil and Schwannian stroma, and a formula converts those tissue probabilities into patch voting weights, with hand-set coefficients α=1, β=8, and γ=1.

What would settle it

Re-run the WSI experiment with the tissue SVM trained on a held-out set of manually annotated neuropil and stroma patches, report the SVM's own accuracy, and freeze α, β, and γ before seeing the test WSIs; if the SVM is inaccurate or the frozen weights do not reproduce 100%, the central WSI claim fails.

Watch

Extended reading notes

Core claim

The paper claims that CMSwinKAN-micro, with only 1.32 million parameters and random initialization, reaches 93.58% accuracy on the five-class PpNTs patch classification task, beating UNI (93.35%), CONCH (92.17%), and CTransPath (92.22%), and reaches 87.79% accuracy on the public BreakHis dataset, again above the same baselines. With the proposed soft voting, CMSwinKAN-micro reaches 100% accuracy, balanced accuracy, and KAPPA on the private PpNTs whole-slide test set, while most baselines sit at 98.41% under hard or soft voting. The ablation studies attribute a 1.29 percentage point accuracy gain to the KAN-for-MLP replacement alone, and further gains to the contrastive multi-scale fusion components.

Load-bearing premise

The whole-slide 100% result rests on the soft-voting SVM that recognizes neuropil and Schwannian stroma, but the paper does not specify how that SVM is trained, on what annotations, or how its accuracy is validated; if the SVM is trained on the same model's features or the weights α=1, β=8, γ=1 are tuned on the test set, the perfect WSI accuracy would not be an independent result.

Editorial extensions

If this is right

  • A randomly initialized 1.32M-parameter model can reach or exceed the patch-level accuracy of foundation models pretrained on over a million pathology images on a five-class pediatric tumor task.
  • KAN-based nonlinearity can serve as a drop-in replacement for MLPs in Swin-style blocks, with ablations attributing about 1.3 accuracy points and 1.2 KAPPA points to that replacement.
  • The clinical-prior soft voting scheme, if reproducible, lifts whole-slide accuracy to a perfect 100% on the private test set and also improves weaker models such as VGGNet.
  • The model's small size (0.22 GFLOPs) makes it feasible to run on edge devices, addressing a practical barrier to clinical deployment.
  • The ablation results indicate that each proposed component, KAN, CDFA, and CMSA, contributes positively, implying the full architecture is needed for the reported gains.

Reading between the lines

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

  • The perfect 100% WSI figure likely reflects a ceiling effect: nearly every baseline already reaches 98.41%, so the reported margin is concentrated in a few test slides, and a larger or harder test set would better separate methods.
  • Because the SVM that drives soft voting is trained on CMSwinKAN's own features, the WSI result is not an independent test of the patch classifier unless the SVM is trained on separate tissue annotations and its accuracy is reported.
  • The interpretability claim is architectural, resting on KAN splines, rather than demonstrated by a quantitative explanation metric; a follow-up examining which spline bases activate for neuropil versus stroma would test it.
  • The BreakHis WSI comparison uses CMSwinKAN-tiny while the patch experiments use CMSwinKAN-micro, so the cross-dataset gains mix model size with method; testing the same variant everywhere would isolate the contribution.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

4 major / 5 minor

Summary. The paper proposes CMSwinKAN, a lightweight Swin-Transformer variant in which MLP layers are replaced by Kolmogorov-Arnold Network (KAN) layers, augmented with a contrastive multi-scale feature aggregation module (CMSA/CDFA), for patch-level classification of H&E histopathology images. The authors evaluate it on a private pediatric peripheral neuroblastic tumor dataset (PpNTs) and on the public BreakHis dataset, and they introduce a soft voting mechanism with an SVM tissue-component classifier to aggregate patch predictions into whole-slide image (WSI) labels. The central claims are that randomly initialized CMSwinKAN-micro outperforms pathology foundation models UNI, CONCH, and CTransPath at the patch level, and that with the proposed soft voting it reaches 100% accuracy, balanced accuracy, and KAPPA on the PpNTs WSI test set (Table IV).

Significance. If the claims were fully substantiated, the paper would be noteworthy: a 1.32M-parameter model trained from random initialization outperforming large pathology foundation models would matter for resource-constrained deployment, and the interpretability angle through KAN layers is timely. The authors also state that source code is available and they report an ethics-approved private dataset, which are positive features. However, the load-bearing WSI-level result is not currently verifiable: the SVM-based soft voting module is underspecified, the reported weighting coefficients contradict the method text, and Table IV reports identical four-metric tuples across many heterogeneous models, which is implausible for independent classifiers. These issues directly affect the headline superiority claim at the WSI level, so the paper as submitted does not meet the standard for acceptance.

major comments (4)
  1. [§IV.D.1, Table IV] The identical tuple (ACC=98.41, BACC=98.33, KAPPA=98.01, F1=98.39) is reported for numerous diverse method variants, including ResNet-50, DenseNet, ViT, Swin, UNI, CTransPath, CONCH, FMDNN, StoHisNet, HiFuse, and CMSwinKAN-micro under hard voting, with hard and soft voting giving the same numbers for most models. With only 63 test WSIs, identical four-metric tuples across such heterogeneous classifiers cannot plausibly arise from independent evaluations, so the table as printed cannot support the WSI-level comparison. The authors must report per-model confusion matrices or exact test-set predictions, and explain how these identical values were produced.
  2. [§III.E.2 vs. §IV.C] The method text states that α > β, 'emphasizing the significance of neuropil,' but Section IV.C sets α = 1 for neuropil and β = 8 for Schwannian stroma, so the implemented weighting is β > α. This contradiction makes the clinical-prior weighting ill-defined, and Eq. (11) as configured would give Schwannian stroma eight times the weight of neuropil, opposite to the stated rationale. The authors must reconcile the text with the implementation and provide a sensitivity analysis over α and β, ideally with the weights chosen on a validation set rather than the test set.
  3. [§III.D and §III.E.2] The WSI soft voting depends on a multi-class SVM that recognizes neuropil, Schwannian stroma, and other tissue components, but the manuscript does not describe the SVM input features, the tissue-component labels, the annotation protocol, the number of annotated patches, the train/test split, or the SVM's own classification accuracy. Section III.D explicitly says 'Further details on the voting mechanism and SVM training will follow,' yet no such details appear anywhere in the paper. Without this information, the 100% soft-vote result for CMSwinKAN-micro in Table IV cannot be reproduced or independently checked, and the WSI-level claim is not established.
  4. [§IV.C, Table III] The patch-level improvements over UNI, CONCH, and CTransPath are reported only as point estimates, with no confidence intervals, significance tests, or per-patient statistics. Given the small test sets, differences of 0.2–0.3% in ACC and BACC are within plausible sampling noise, so the claim that a 1.32M-parameter randomly initialized model beats large pretrained pathology foundation models needs statistical substantiation.
minor comments (5)
  1. [§III.D] The sentence 'Further details on the voting mechanism and SVM training will follow' is inappropriate in a final manuscript; the details should either be fully integrated into Section III.E.2 or clearly stated as future work rather than deferred within the same paper.
  2. [§III.C] The term 'contrastive learning' appears in the model name and contribution list, but no contrastive loss is defined; the CDFA module compares attention maps across scales, which is not standard contrastive learning. Please clarify the terminology or specify the actual contrastive objective.
  3. [Eqs. (8)–(11)] The notation is not fully connected: w_{i,c} in Eq. (9) is not explicitly linked to w_patch in Eq. (11), and P1 and P2 are introduced without explaining how the SVM's multiclass probabilities are obtained or normalized. Please unify the notation and define all variables.
  4. [Fig. 5] The circle-size encoding of parameter count is hard to read because many points overlap; a table listing parameter counts and FLOPs for all compared models would be much clearer.
  5. [Table V] The text says soft voting improves generalization on BreakHis, but Table V reports hard-voting results only; please either add soft-vote results for BreakHis or adjust the wording.

Circularity Check

1 steps flagged · score 6.0 of 10

WSI-level soft-vote result is constructed from arbitrary hand-set weights and an unspecified SVM trained on the same features, so the 100% accuracy is not an independent prediction; the patch-level comparisons are self-contained.

  1. fitted input called prediction [Section III.E.2 (Training and Application of the SVM Model), Eq. (11); Section IV.C (Experimental Setup); Table IV]
    "Leveraging the features extracted by the patch classification model CMSwinKAN, we train a multi-class SVM model to determine whether a patch contains neuropil, Schwannian stroma, or other tissue components. ... wpatch = α·P1 + β·P2 ... α > β emphasizing the significance of neuropil. ... We set α = 1 for neuropil and β = 8 for Schwannian stroma, with γ = 1 assigned to non-key tissue regions."

    Eq. (11) defines wpatch from hand-set α, β, and γ, and Eq. (10) then sets LabelWSI = argmax of the weighted sum. The coefficients are never fitted or validated on independent data, and the paper contradicts its own α > β rule by setting α=1 and β=8. The SVM supplying P1 and P2 is trained on the same CMSwinKAN features, with no described tissue-label protocol, train/test split, or accuracy check; Section III.D only promises details 'will follow'. Hence the reported 100% soft-vote WSI row is not an independent prediction from the architecture; it is an output of arbitrary weights and an unvalidated SVM, i.e., a constructed/fitted input presented as a WSI-level prediction.

full rationale

The patch-level claims (Tables III and VI) are not circular: CMSwinKAN is trained from random initialization and compared against other models on fixed patch test sets, so those results stand independently of the WSI machinery. The circularity is concentrated in the WSI-level soft-voting claim. The decisive WSI label is computed as LabelWSI = argmax of a weighted sum whose weights come from Eq. (11), with coefficients set by hand (α=1, β=8, γ=1) rather than derived or cross-validated, and the SVM producing P1/P2 is explicitly trained on the same CMSwinKAN features. The paper itself admits the voting/SVM details are missing ('Further details on the voting mechanism and SVM training will follow'). Additionally, Table IV reports identical four-metric tuples (98.41/98.33/98.01/98.39) for 13 heterogeneous models under both hard and soft voting, which is not consistent with independently computed per-model predictions and reinforces that the WSI comparison is constructed rather than derived. The 100% soft-vote row for CMSwinKAN is therefore not supported by a reproducible derivation chain: it is produced by arbitrary weighting and an unspecified classifier trained on the same features. This raises the circularity score to 6 because the central WSI-level claim reduces to fitted choices, while the patch-level core retains independent content.

Assumptions & free parameters 4 free parameters · 3 assumptions · 0 invented entities

The central empirical claims rely on hand-set voting weights, undisclosed SVM training details, and clinical priors taken as given. No new physical or theoretical entities are introduced; the CDFA, CMSA, and voting components are software modules.

free parameters (4)
  • Voting weight alpha = 1
    Hand-set in Section IV.C; weights neuropil probability in Eq. (11).
  • Voting weight beta = 8
    Hand-set in Section IV.C; contradicts stated design alpha > beta.
  • Voting weight gamma = 1
    Hand-set baseline weight for non-key tissue patches in Eq. (11).
  • CDFA window size K
    Required by Eq. (2)-(3) but its value is not reported in the paper.
assumptions (3)
  • standard math Kolmogorov-Arnold representation theorem justifies replacing MLPs with KAN layers.
    Invoked in Section III.B and III.D via the KAN design, following Liu et al. [33].
  • domain assumption Clinical priors that neuropil indicates potential malignancy and Schwannian stroma >= 50% suggests GNBi or GN.
    Stated in Section III.E.1 and used to design the soft voting weights.
  • domain assumption The private dataset's pathologist labels are correct and the WSI split prevents patient-level leakage.
    Section IV.A describes data collection and splitting but does not state patient-level stratification or inter-observer agreement.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Towards Accurate and Interpretable Neuroblastoma Diagnosis via Contrastive Multi-scale Pathological Image Analysis." pith.science (2026). https://pith.science/paper/IX4CMGJL

@misc{pith2026250413754,
  author       = {Pith},
  title        = {Pith review of: Towards Accurate and Interpretable Neuroblastoma Diagnosis via Contrastive Multi-scale Pathological Image Analysis},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/IX4CMGJL}},
  note         = {Machine review of arXiv:2504.13754}
}
read the original abstract

Neuroblastoma, adrenal-derived, is among the most common pediatric solid malignancies, characterized by significant clinical heterogeneity. Timely and accurate pathological diagnosis from hematoxylin and eosin-stained whole-slide images is critical for patient prognosis. However, current diagnostic practices primarily rely on subjective manual examination by pathologists, leading to inconsistent accuracy. Existing automated whole-slide image classification methods encounter challenges such as poor interpretability, limited feature extraction capabilities, and high computational costs, restricting their practical clinical deployment. To overcome these limitations, we propose CMSwinKAN, a contrastive-learning-based multi-scale feature fusion model tailored for pathological image classification, which enhances the Swin Transformer architecture by integrating a Kernel Activation Network within its multilayer perceptron and classification head modules, significantly improving both interpretability and accuracy. By fusing multi-scale features and leveraging contrastive learning strategies, CMSwinKAN mimics clinicians' comprehensive approach, effectively capturing global and local tissue characteristics. Additionally, we introduce a heuristic soft voting mechanism guided by clinical insights to bridge patch-level predictions to whole-slide image-level classifications seamlessly. We verified the CMSwinKAN on the publicly available BreakHis dataset and the PpNTs dataset, which was established by our hospital. Results demonstrate that CMSwinKAN performs better than existing state-of-the-art pathology-specific models pre-trained on large datasets. Our source code is available at https://github.com/JSLiam94/CMSwinKAN.

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

41 extracted references · 36 canonical work pages

  1. [1]

    Neu- roblastoma, version 2.2024, nccn clinical practice guidelines in oncology,

    R. Bagatell, J. R. Park, S. Acharya, J. Aldrink, J. Allison, E. Alva, C. Arndt, D. Benedetti, E. Brown, S. Cho et al., “Neu- roblastoma, version 2.2024, nccn clinical practice guidelines in oncology,” Journal of the National Comprehensive Cancer Network, vol. 22, no. 6, pp. 413–433, 2024

    work page 2024

  2. [2]

    Single-cell transcriptomics and epigenomics unravel the role of monocytes in neuroblastoma bone marrow metastasis,

    I. S. Fetahu, W. Esser-Skala, R. Dnyansagar, S. Sindelar, F. Ri- fatbegovic, A. Bileck, L. Skos, E. Bozsaky, D. Lazic, L. Shaw et al., “Single-cell transcriptomics and epigenomics unravel the role of monocytes in neuroblastoma bone marrow metastasis,” Nature communications, vol. 14, no. 1, p. 3620, 2023

    work page 2023

  3. [3]

    Pathological prognosis classification of patients with neuroblastoma using computational pathology analysis,

    Y . Liu, Y . Jia, C. Hou, N. Li, N. Zhang, X. Yan, L. Yang, Y . Guo, H. Chen, J. Li et al. , “Pathological prognosis classification of patients with neuroblastoma using computational pathology analysis,” Computers in Biology and Medicine , vol. 149, p. 105980, 2022

    work page 2022

  4. [4]

    Sd-mil: Multiple instance learning with dual perception of scale and distance information fusion for whole slide image classification,

    Y . Xie, Z. Liu, J. Chen, W. Zhang, J. Zhao, and J. Ma, “Sd-mil: Multiple instance learning with dual perception of scale and distance information fusion for whole slide image classification,” Expert Systems with Applications , p. 126831, 2025

    work page 2025

  5. [5]

    Pseudo-bag mixup augmen- tation for multiple instance learning-based whole slide image classification,

    P. Liu, L. Ji, X. Zhang, and F. Ye, “Pseudo-bag mixup augmen- tation for multiple instance learning-based whole slide image classification,” IEEE Transactions on Medical Imaging , vol. 43, no. 5, pp. 1841–1852, 2024

    work page 2024

  6. [6]

    Vision transformers for computational histopathol- ogy,

    H. Xu, Q. Xu, F. Cong, J. Kang, C. Han, Z. Liu, A. Madabhushi, and C. Lu, “Vision transformers for computational histopathol- ogy,” IEEE Reviews in Biomedical Engineering , vol. 17, pp. 63–79, 2023

    work page 2023

  7. [7]

    Deep learning to assess microsatellite instability directly from histopathological whole slide images in endometrial cancer,

    C.-W. Wang, H. Muzakky, N. P. Firdi, T.-C. Liu, P.-J. Lai, Y .- C. Wang, M.-H. Yu, and T.-K. Chao, “Deep learning to assess microsatellite instability directly from histopathological whole slide images in endometrial cancer,” NPJ Digital Medicine , vol. 7, no. 1, p. 143, 2024

    work page 2024

  8. [8]

    Multi- scale convolutional attention frequency-enhanced transformer network for medical image segmentation,

    S. Yan, B. Yang, A. Chen, X. Zhao, and S. Zhang, “Multi- scale convolutional attention frequency-enhanced transformer network for medical image segmentation,” Information Fusion , p. 103019, 2025

    work page 2025

Show all 41 references
  1. [9]

    Scunet++: Swin-unet and cnn bottleneck hybrid architecture with multi-fusion dense skip connection for pul- monary embolism ct image segmentation*,

    Y . Chen, B. Zou, Z. Guo, Y . Huang, Y . Huang, F. Qin, Q. Li, and C. Wang, “Scunet++: Swin-unet and cnn bottleneck hybrid architecture with multi-fusion dense skip connection for pul- monary embolism ct image segmentation*,” in 2024 IEEE/CVF Winter Conference on Applications ...

    2024

  2. [10]

    Sckansformer: Fine- grained classification of bone marrow cells via kansformer 10 backbone and hierarchical attention mechanisms,

    Y . Chen, Z. Zhu, S. Zhu, L. Qiu, B. Zou, F. Jia, Y . Zhu, C. Zhang, Z. Fang, F. Qin et al. , “Sckansformer: Fine- grained classification of bone marrow cells via kansformer 10 backbone and hierarchical attention mechanisms,” IEEE Journal of Biomedical and Health Informatics , 2024

    2024

  3. [11]

    Explainable ai-driven iomt fusion: Unravelling techniques, opportunities, and challenges with explainable ai in healthcare,

    N. A. Wani, R. Kumar, J. Bedi, I. Rida et al. , “Explainable ai-driven iomt fusion: Unravelling techniques, opportunities, and challenges with explainable ai in healthcare,” Information Fusion, p. 102472, 2024

    2024

  4. [12]

    Bayesian collaborative learning for whole-slide image classification,

    J.-G. Yu, Z. Wu, Y . Ming, S. Deng, Q. Wu, Z. Xiong, T. Yu, G.-S. Xia, Q. Jiang, and Y . Li, “Bayesian collaborative learning for whole-slide image classification,” IEEE Transactions on Medical Imaging , vol. 42, no. 6, pp. 1809–1821, 2023

    2023

  5. [13]

    Patch-slide discrim- inative joint learning for weakly-supervised whole slide image representation and classification,

    J. Yu, X. Wang, T. Ma, X. Li, and Y . Xu, “Patch-slide discrim- inative joint learning for weakly-supervised whole slide image representation and classification,” in International Conference on Medical Image Computing and Computer-Assisted Interven- tion. Springer, 2024, pp. 713–722

    2024

  6. [14]

    Computer aided classification of neuroblastoma histological images using scale invariant feature transform with feature encoding,

    S. Gheisari, D. R. Catchpoole, A. Charlton, Z. Melegh, E. Grad- hand, and P. J. Kennedy, “Computer aided classification of neuroblastoma histological images using scale invariant feature transform with feature encoding,” Diagnostics, vol. 8, no. 3, p. 56, 2018

    2018

  7. [15]

    Convolutional deep belief network with feature encoding for classification of neuroblastoma histological images,

    S. Gheisari, D. R. Catchpoole, A. Charlton, and P. J. Kennedy, “Convolutional deep belief network with feature encoding for classification of neuroblastoma histological images,” Journal of Pathology Informatics, vol. 9, no. 1, p. 17, 2018

    2018

  8. [16]

    Patch-based convolutional neural network for whole slide tissue image classification,

    L. Hou, D. Samaras, T. M. Kurc, Y . Gao, J. E. Davis, and J. H. Saltz, “Patch-based convolutional neural network for whole slide tissue image classification,” in 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR) , 2016, pp. 2424–2433

    2016

  9. [17]

    Task-specific fine-tuning via variational informa- tion bottleneck for weakly-supervised pathology whole slide im- age classification,

    H. Li, C. Zhu, Y . Zhang, Y . Sun, Z. Shui, W. Kuang, S. Zheng, and L. Yang, “Task-specific fine-tuning via variational informa- tion bottleneck for weakly-supervised pathology whole slide im- age classification,” in 2023 IEEE/CVF Conference on Computer Vision and Pattern Reco...

    2023

  10. [18]

    Transformer-based unsupervised con- trastive learning for histopathological image classification,

    X. Wang, S. Yang, J. Zhang, M. Wang, J. Zhang, W. Yang, J. Huang, and X. Han, “Transformer-based unsupervised con- trastive learning for histopathological image classification,” Medical Image Analysis , vol. 81, p. 102559, 2022

    2022

  11. [19]

    To- wards a general-purpose foundation model for computational pathology,

    R. J. Chen, T. Ding, M. Y . Lu, D. F. Williamson, G. Jaume, A. H. Song, B. Chen, A. Zhang, D. Shao, M. Shaban et al., “To- wards a general-purpose foundation model for computational pathology,” Nature Medicine, vol. 30, no. 3, pp. 850–862, 2024

    2024

  12. [20]

    A visual-language foundation model for computational pathology,

    M. Y . Lu, B. Chen, D. F. Williamson, R. J. Chen, I. Liang, T. Ding, G. Jaume, I. Odintsov, L. P. Le, G. Gerber et al. , “A visual-language foundation model for computational pathology,” Nature Medicine, vol. 30, no. 3, pp. 863–874, 2024

    2024

  13. [21]

    Robust and data-efficient generalization of self-supervised machine learning for diagnostic imaging,

    S. Azizi, L. Culp, J. Freyberg, B. Mustafa, S. Baur, S. Kornblith, T. Chen, N. Tomasev, J. Mitrovi´c, P. Strachan et al., “Robust and data-efficient generalization of self-supervised machine learning for diagnostic imaging,” Nature Biomedical Engineering, vol. 7, no. 6, pp. 75...

    2023

  14. [22]

    Superhistopath: A deep learning pipeline for mapping tumor heterogeneity on low-resolution whole-slide digital histopathology images,

    K. Zormpas-Petridis, R. Noguera, D. K. Ivankovic, I. Roxanis, Y . Jamin, and Y . Yuan, “Superhistopath: A deep learning pipeline for mapping tumor heterogeneity on low-resolution whole-slide digital histopathology images,” Frontiers in oncol- ogy, vol. 10, p. 586292, 2021

    2021

  15. [23]

    Fmdnn: A fuzzy-guided multigranular deep neural network for histopathological image classification,

    W. Ding, T. Zhou, J. Huang, S. Jiang, T. Hou, and C.-T. Lin, “Fmdnn: A fuzzy-guided multigranular deep neural network for histopathological image classification,” IEEE Transactions on Fuzzy Systems , 2024

    2024

  16. [24]

    Scorenet: Learning non-uniform attention and augmentation for transformer-based histopathological image classification,

    T. Stegm ¨uller, B. Bozorgtabar, A. Spahr, and J.-P. Thiran, “Scorenet: Learning non-uniform attention and augmentation for transformer-based histopathological image classification,” in 2023 IEEE/CVF Winter Conference on Applications of Com- puter Vision (WACV) , 2023, pp. 6170–6179

    2023

  17. [25]

    Detexnet: Accurately diagnosing frequent and challenging pediatric malignant tumors,

    Y . Liu, M. Yin, and S. Sun, “Detexnet: Accurately diagnosing frequent and challenging pediatric malignant tumors,” IEEE Transactions on Medical Imaging , vol. 40, no. 1, pp. 395–404, 2020

    2020

  18. [26]

    Crossvit: Cross-attention multi-scale vision transformer for image classification,

    C.-F. R. Chen, Q. Fan, and R. Panda, “Crossvit: Cross-attention multi-scale vision transformer for image classification,” in 2021 IEEE/CVF International Conference on Computer Vision (ICCV), 2021, pp. 357–366

    2021

  19. [27]

    Mvitv2: Improved multiscale vi- sion transformers for classification and detection,

    Y . Li, C.-Y . Wu, H. Fan, K. Mangalam, B. Xiong, J. Ma- lik, and C. Feichtenhofer, “Mvitv2: Improved multiscale vi- sion transformers for classification and detection,” in 2022 IEEE/CVF Conference on Computer Vision and Pattern Recog- nition (CVPR) , 2022, pp. 4804–4814

    2022

  20. [28]

    Shunted self- attention via multi-scale token aggregation,

    S. Ren, D. Zhou, S. He, J. Feng, and X. Wang, “Shunted self- attention via multi-scale token aggregation,” in 2022 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2022, pp. 10 853–10 862

    2022

  21. [29]

    Dilateformer: Multi-scale dilated transformer for visual recognition,

    J. Jiao, Y .-M. Tang, K.-Y . Lin, Y . Gao, A. J. Ma, Y . Wang, and W.-S. Zheng, “Dilateformer: Multi-scale dilated transformer for visual recognition,” IEEE Transactions on Multimedia , vol. 25, pp. 8906–8919, 2023

    2023

  22. [30]

    Hifuse: Hierarchical multi-scale feature fusion network for medical image classification,

    X. Huo, G. Sun, S. Tian, Y . Wang, L. Yu, J. Long, W. Zhang, and A. Li, “Hifuse: Hierarchical multi-scale feature fusion network for medical image classification,” Biomedical Signal Processing and Control, vol. 87, p. 105534, 2024

    2024

  23. [31]

    Accurate leukocyte detection based on deformable-detr and multi-level feature fusion for aiding diagnosis of blood diseases,

    Y . Chen, C. Zhang, B. Chen, Y . Huang, Y . Sun, C. Wang, X. Fu, Y . Dai, F. Qin, Y . Peng et al. , “Accurate leukocyte detection based on deformable-detr and multi-level feature fusion for aiding diagnosis of blood diseases,” Computers in Biology and Medicine, vol. 170, p. 10...

    2024

  24. [32]

    Condseg: A general med- ical image segmentation framework via contrast-driven feature enhancement,

    M. Lei, H. Wu, X. Lv, and X. Wang, “Condseg: A general med- ical image segmentation framework via contrast-driven feature enhancement,” arXiv preprint arXiv:2412.08345 , 2024

    2024 arXiv

  25. [33]

    KAN: Kolmogorov–arnold net- works,

    Z. Liu, Y . Wang, S. Vaidya, F. Ruehle, J. Halverson, M. Soljacic, T. Y . Hou, and M. Tegmark, “KAN: Kolmogorov–arnold net- works,” in The Thirteenth International Conference on Learning Representations, 2025

    2025

  26. [34]

    Delving deep into rectifiers: Surpassing human-level performance on imagenet classification,

    K. He, X. Zhang, S. Ren, and J. Sun, “Delving deep into rectifiers: Surpassing human-level performance on imagenet classification,” IEEE Computer Society , 2015

    2015

  27. [35]

    Very deep convolutional networks for large-scale image recognition,

    K. Simonyan and A. Zisserman, “Very deep convolutional networks for large-scale image recognition,” arXiv preprint arXiv:1409.1556, 2014

    2014 arXiv

  28. [36]

    Deep residual learning for image recognition,

    K. He, X. Zhang, S. Ren, and J. Sun, “Deep residual learning for image recognition,” in 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR) , 2016, pp. 770–778

    2016

  29. [37]

    Densely connected convolutional networks,

    G. Huang, Z. Liu, L. Van Der Maaten, and K. Q. Wein- berger, “Densely connected convolutional networks,” in 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 2017, pp. 4700–4708

    2017

  30. [38]

    An image is worth 16x16 words: Transformers for image recognition at scale,

    A. Dosovitskiy, L. Beyer, A. Kolesnikov, D. Weissenborn, X. Zhai, T. Unterthiner, M. Dehghani, M. Minderer, G. Heigold, S. Gelly, J. Uszkoreit, and N. Houlsby, “An image is worth 16x16 words: Transformers for image recognition at scale,” ICLR, 2021

    2021

  31. [39]

    Swin transformer: Hierarchical vision transformer using shifted windows,

    Z. Liu, Y . Lin, Y . Cao, H. Hu, Y . Wei, Z. Zhang, S. Lin, and B. Guo, “Swin transformer: Hierarchical vision transformer using shifted windows,” in 2021 IEEE/CVF International Con- ference on Computer Vision (ICCV) , 2021, pp. 10 012–10 022

    2021

  32. [40]

    An enhanced vision transformer with wavelet position embedding for histopathological image classification,

    M. Ding, A. Qu, H. Zhong, Z. Lai, S. Xiao, and P. He, “An enhanced vision transformer with wavelet position embedding for histopathological image classification,” Pattern Recognition, vol. 140, p. 109532, 2023

    2023

  33. [41]

    Stohisnet: A hybrid multi-classification model with cnn and transformer for gastric pathology images,

    B. Fu, M. Zhang, J. He, Y . Cao, Y . Guo, and R. Wang, “Stohisnet: A hybrid multi-classification model with cnn and transformer for gastric pathology images,” Computer Methods and Programs in Biomedicine , vol. 221, p. 106924, 2022

    2022

Pith tools

Reviewed August 16, 2026 · model on record in the stance chip above.