Pith. sign in

REVIEW 3 major objections 7 minor 42 references

A hinge regularizer that borrows coarse anatomy masks from pretrained segmentation models steers medical classifiers toward the right tissue and raises accuracy across eight datasets.

Reviewed by Pith at T0; open to challenge. T0 means a machine referee read the full paper against a public rubric. the ladder, T0–T4 →

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

T0 review · grok-4.5

2026-07-14 08:46 UTC pith:V5ITZBIT

load-bearing objection Clean plug-and-play hinge regularizer that turns frozen medical SFMs into anatomy priors; consistent gains, real but bounded novelty. the 3 major comments →

arxiv 2607.10851 v1 pith:V5ITZBIT submitted 2026-07-12 cs.CV

Learning To Focus: Anatomy-Guided Attention Regularization for Medical Image Classification

Tonmoy Hossain , Atiqur Rahman , Farhana Hossain Swarnali , Miaomiao Zhang This is my paper
classification cs.CV
keywords medical image classificationattention regularizationanatomical shapessegmentation foundation modelsGrad-CAMhinge lossinterpretability
verification ladder T0 review T1 audit T2 compute T3 formal T4 reserved

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

Medical classifiers usually train on image-level labels alone, so they freely attend to background artifacts and scanner cues that do not generalize. This paper argues that you can supply the missing spatial common sense without any manual pixel masks or a second trained segmentation network. Frozen segmentation foundation models, prompted patch-wise with simple text, yield usable anatomical priors; a hinge loss then penalizes the classifier only when its attention energy on background exceeds that on the anatomy. The resulting framework, Locus, plugs into ordinary CNN and transformer backbones, needs no foundation-model fine-tuning, and produces both higher accuracy/F1 and more coherent Grad-CAM maps on dermoscopy, X-ray, histopathology, ultrasound and cardiac cine MRI. A sympathetic reader cares because it turns an expensive annotation problem into a free, off-the-shelf regularizer that also improves robustness to common imaging corruptions.

Core claim

Locus establishes that an adaptive hinge comparing aggregate attention inside versus outside a foundation-model anatomical mask is sufficient spatial supervision for medical image classification: it yields consistent gains in accuracy and F1-score together with more anatomically grounded attention maps across eight diverse datasets, without dedicated segmentation training or expert pixel annotations.

What carries the argument

The anatomy-centric hinge regularizer R(M,A)=max(0,A_bg−A_fg), which activates only when background attention dominates foreground attention; masks M are obtained by union of patch-wise text prompts to a frozen segmentation foundation model.

Load-bearing premise

The method assumes that the coarse, un-fine-tuned masks from patch-wise prompting of the foundation model correctly cover the diagnostically relevant anatomy on every dataset; if those masks systematically miss or misplace the true structures, the hinge supplies wrong spatial guidance.

What would settle it

On a held-out set where the foundation model systematically omits the true lesion or organ, measure whether Locus accuracy drops below plain cross-entropy while Grad-CAM still concentrates on the incorrect masked region; that outcome would refute the claim that the free masks supply useful supervision.

Watch this falsifier — get emailed when new claim-graph text bears on it.

If this is right

  • Medical classifiers can obtain anatomical focus and higher scores with zero extra pixel annotation cost.
  • The same plug-in regularizer works across CNN and transformer backbones and multiple saliency extractors.
  • Coarse foundation-model masks suffice when the loss only demands anatomy-dominant rather than pixel-perfect attention.
  • Robustness to brightness, contrast, blur and noise improves because background-correlated shortcuts are discouraged.
  • Temporal sequences such as cardiac cine MRI can be regularized with the same spatial hinge.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • As segmentation foundation models improve, free anatomical priors of this kind could become default training practice for any anatomy-dependent medical task.
  • The hinge idea may transfer outside medicine whenever a cheap off-the-shelf detector can supply a soft region prior.
  • Learning the patch grid end-to-end, as the authors note for future work, would likely close remaining cases of missed small or sparse structures.
  • Uncertainty-aware down-weighting of unreliable foundation masks would let the regularizer back off automatically on out-of-distribution anatomy.

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

3 major / 7 minor

Summary. The paper proposes Locus, an anatomy-guided attention regularization framework for medical image classification. Pretrained segmentation foundation models (BiomedParse, Medical SAM3) supply anatomical masks via a patch-wise text-prompting strategy; a hinge regularizer R = max(0, A_bg − A_fg) (Eq. 3) then penalizes the classifier only when Grad-CAM background attention exceeds foreground attention. The regularizer is added to standard cross-entropy with a warm-up schedule (Eq. 4). The method is claimed to be plug-and-play across backbones and saliency extractors, requiring neither manual pixel annotations nor fine-tuning of the foundation model. Empirical validation spans multiple public datasets (dermoscopy, X-ray, histopathology, ultrasound) plus an in-house cardiac cine MRI set, with reported accuracy/F1 gains over CE-only, masked-input, and MGA baselines, plus qualitative improvements in attention localization and robustness under test-time perturbations.

Significance. If the reported gains hold under stronger statistical scrutiny and more complete reporting, Locus is a practically useful contribution: it converts frozen, off-the-shelf medical SFMs into spatial supervision for classifiers without the annotation cost of mask-guided training or the optimization friction of joint segmentation–classification. The hinge design is deliberately tolerant of coarse masks, the method is architecture- and saliency-agnostic, and the multi-modality evaluation (including temporal cine MRI) is broader than typical single-domain studies. Code is promised (anonymous repository). These are genuine engineering strengths for MICCAI-style venues. The work does not introduce a new theoretical principle; its value is empirical and methodological.

major comments (3)
  1. Section 2.2 and Fig. 3: The central claim rests on the assumption that un-fine-tuned SFM masks obtained by patch-wise text prompting are sufficiently complete and correctly localized for the diagnostically relevant anatomy on every dataset. Only qualitative examples are shown; there is no quantitative mask-quality evaluation (e.g., coverage of lesion/organ, false-positive background rate, or failure-case rate) on the eight evaluation sets. Because R (Eq. 3) uses these masks as the sole spatial prior, systematic mislocalization would supply incorrect supervision. A short quantitative audit of mask quality (or explicit failure analysis) is needed to bound this risk.
  2. Table 1: Accuracy and F1 gains are reported as single point estimates with no standard deviations across seeds/runs and no statistical significance tests. Several improvements are modest (e.g., PneumoniaMNIST ResNet Acc 89.90 → 92.15; Cardiac DenseNet Acc 78.32 → 82.02). Without multi-seed statistics it is difficult to judge whether the “consistent gains” claim is reliable. At minimum, report mean ± std over ≥3 seeds and a paired test or confidence interval for the main comparisons.
  3. Abstract and §3 claim validation on eight datasets, yet Table 1 reports only six (HAM10000, PneumoniaMNIST, BTXRD, PanNuke, ChestX-ray8, Cardiac). BUSI and CheXpert appear in figures and text but lack the same quantitative baseline comparison. Either complete the main table for all eight datasets or revise the claim to match the reported numbers.
minor comments (7)
  1. Table 2 (λ ablation): performance is non-monotonic (dips at λ = 0.3 and 0.7). A brief discussion of why intermediate strengths can over-constrain the classifier would help readers choose λ.
  2. §3.1 Implementation: number of patches P is “selected empirically per dataset” with no sensitivity study or default rule. A short ablation or heuristic would improve reproducibility.
  3. Fig. 1 caption and several figure panels use “SOTA classifiers” without naming which models are shown; label the baselines explicitly.
  4. Leftover running head “Title Suppressed Due to Excessive Length” appears on multiple pages; remove before camera-ready.
  5. Medical SAM3 is cited as arXiv:2601.10880 (2026); confirm the citation is stable and publicly available at submission time.
  6. §3 Datasets: exclusion of “No Finding” / normal classes is reasonable but should be stated once with a clear criterion so readers can judge selection bias.
  7. Clarify whether Grad-CAM is computed only on the predicted class or on the ground-truth class during training of R; this affects the gradient path.

Circularity Check

0 steps flagged

No circularity: empirical regularization method whose hinge and SFM masks are independent of the classification objective and whose gains are measured on external benchmarks.

full rationale

Locus is a purely empirical method paper. The anatomy-centric regularizer is defined as R(M,A)=max(0,A_bg-A_fg) (Eq. 3) from Grad-CAM attention aggregates over an externally supplied, frozen SFM mask M; this expression does not contain the class labels y, is not fitted to any classification metric, and is not claimed to be a first-principles derivation of anything. The network loss (Eq. 4) simply adds a weighted hinge to ordinary cross-entropy; the reported accuracy/F1 gains (Table 1) and attention maps (Figs. 4–7) are ordinary empirical comparisons against CE, Masked and MGA baselines on eight held-out datasets. Patch-wise prompting is a fixed, non-learned heuristic. Author self-citations ([7],[8]) appear only as background and are not load-bearing for uniqueness, ansatz or any claimed prediction. Nothing reduces by construction to its own inputs.

Axiom & Free-Parameter Ledger

3 free parameters · 3 axioms · 2 invented entities

The central claim rests on a small set of free hyperparameters, standard domain assumptions about Grad-CAM and foundation-model localization, and two invented algorithmic constructs (the hinge regularizer and the patch-wise prompting procedure). No new physical entities are postulated.

free parameters (3)
  • λ_max (regularization strength) = 0.1
    Set to 0.1 after warm-up; ablation in Table 2 shows performance varies with λ, so the reported gains depend on this hand-chosen value.
  • number of patches P = dataset-specific
    Chosen empirically per dataset according to expected anatomy scale; not learned and not reported as a fixed global constant.
  • T_warmup / T_ramp = 5 / 5
    Fixed at 5 epochs each; the schedule controls when the regularizer becomes active.
axioms (3)
  • domain assumption Grad-CAM (or other listed saliency maps) is a faithful enough proxy for the spatial regions the classifier uses for its decision.
    Invoked throughout Section 2.1 and all visualization experiments; the regularizer acts only on these maps.
  • domain assumption Pretrained medical segmentation foundation models localize the diagnostically relevant anatomy sufficiently well across the eight modalities without any fine-tuning.
    Core premise of Sections 1–2; if false the hinge supplies wrong spatial targets.
  • ad hoc to paper Penalizing only the case A_bg > A_fg is sufficient; no pixel-wise alignment or fixed foreground/background ratio is required.
    Design choice of the hinge in Eq. (3); justified by robustness arguments but not derived from first principles.
invented entities (2)
  • Locus hinge regularizer R = max(0, A_bg − A_fg) no independent evidence
    purpose: Provides soft anatomical attention supervision without requiring exact mask alignment or extra learnable parameters.
    Defined in Eq. (3); the paper’s performance claims rest on this specific functional form.
  • Patch-wise anatomical prompting (PAP) no independent evidence
    purpose: Recovers multiple or small anatomical instances that full-image prompting of the foundation model misses.
    Introduced in Section 2.2 and Fig. 3; without it the masks used by R would be incomplete on several datasets.

pith-pipeline@v1.1.0-grok45 · 16188 in / 2656 out tokens · 48772 ms · 2026-07-14T08:46:32.937568+00:00 · methodology

0 comments
read the original abstract

Medical image classification models are ideally expected to identify diagnostically relevant regions while making predictions, yet standard classification losses rarely provide spatial supervision. Explicit supervision via anatomical shape information, such as segmentation masks of task-relevant anatomy, has been shown to guide the network toward regions relevant to the target prediction. However, obtaining such masks incurs substantial manual annotation effort and computational overhead. With the advent of segmentation foundation models that exhibit strong localization of anatomical structures across diverse imaging modalities, we leverage this capability to extract anatomical shape priors without the burden of training a dedicated segmentation model. In this paper, we propose a new framework, Locus, an anatomical attention regularization framework that leverages pretrained segmentation foundation models to guide a classifier's attention toward diagnostically meaningful anatomical structures across diverse imaging modalities. Instead of enforcing pixel-wise alignment with the foundation-model-derived mask, we introduce a regularization term that adaptively balances attention between anatomical (foreground) and background regions, penalizing the classifier when background attention dominates. We validate Locus on eight diverse medical imaging datasets spanning dermoscopy, X-ray, histopathology, and cardiac MRI, showing consistent gains in classification performance alongside improved anatomically grounded attention.

Figures

Figures reproduced from arXiv: 2607.10851 by Atiqur Rahman, Farhana Hossain Swarnali, Miaomiao Zhang, Tonmoy Hossain.

Figure 1
Figure 1. Figure 1: Visualization of overlaid attention maps from SOTA classifiers vs. our [PITH_FULL_IMAGE:figures/full_fig_p002_1.png] view at source ↗
Figure 2
Figure 2. Figure 2: Overview of the proposed framework. The image encoder [PITH_FULL_IMAGE:figures/full_fig_p004_2.png] view at source ↗
Figure 3
Figure 3. Figure 3: Comparison of anatomical masks obtained from the pretrained Biomed [PITH_FULL_IMAGE:figures/full_fig_p006_3.png] view at source ↗
Figure 4
Figure 4. Figure 4: GradCAM visualizations across four imaging modalities and cardiac cine [PITH_FULL_IMAGE:figures/full_fig_p010_4.png] view at source ↗
Figure 5
Figure 5. Figure 5: Comparison of GradCAM visualizations between network trained with [PITH_FULL_IMAGE:figures/full_fig_p011_5.png] view at source ↗
Figure 6
Figure 6. Figure 6: Robustness of our model on ChestX-ray under varying image perturba [PITH_FULL_IMAGE:figures/full_fig_p012_6.png] view at source ↗
Figure 7
Figure 7. Figure 7: Generalizability of our approach across various attention extraction meth [PITH_FULL_IMAGE:figures/full_fig_p013_7.png] view at source ↗

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Reference graph

Works this paper leans on

42 extracted references · 2 linked inside Pith

  1. [1]

    Data in brief28, 104863 (2020)

    Al-Dhabyani, W., Gomaa, M., Khaled, H., Fahmy, A.: Dataset of breast ultrasound images. Data in brief28, 104863 (2020)

    2020

  2. [2]

    In: 2018 IEEE winter conference on applications of computer vision (WACV)

    Chattopadhay, A., Sarkar, A., Howlader, P., Balasubramanian, V.N.: Grad- cam++: Generalized gradient-based visual explanations for deep convolutional networks. In: 2018 IEEE winter conference on applications of computer vision (WACV). pp. 839–847. IEEE (2018)

    2018

  3. [3]

    Nature Machine Intelligence3(7), 610–619 (2021)

    DeGrave, A.J., Janizek, J.D., Lee, S.I.: Ai for radiographic covid-19 detection se- lects shortcuts over signal. Nature Machine Intelligence3(7), 610–619 (2021)

    2021

  4. [4]

    In: European congress on digital pathology

    Gamper, J., Alemi Koohbanani, N., Benet, K., Khuram, A., Rajpoot, N.: Pannuke: an open pan-cancer histology dataset for nuclei instance segmentation and classi- fication. In: European congress on digital pathology. pp. 11–19. Springer (2019)

    2019

  5. [5]

    IEEE Transactions on Image Processing (2025)

    Gao, H., Sheng, R., Su, Y., Chen, Z., Xu, S., Gao, L.: Multiscale segmentation- guided fusion network for hyperspectral image classification. IEEE Transactions on Image Processing (2025)

    2025

  6. [6]

    Nature Machine In- telligence2(11), 665–673 (2020)

    Geirhos, R., Jacobsen, J.H., Michaelis, C., Zemel, R., Brendel, W., Bethge, M., Wichmann, F.A.: Shortcut learning in deep neural networks. Nature Machine In- telligence2(11), 665–673 (2020)

    2020

  7. [7]

    arXiv preprint arXiv:2411.12201 (2024)

    Hossain, T., Ma, J., Li, J., Zhang, M.: Invariant shape representation learning for image classification. arXiv preprint arXiv:2411.12201 (2024)

    Pith/arXiv arXiv 2024

  8. [8]

    In: 2019 1st international conference on advances in science, engineering and robotics technology (ICASERT)

    Hossain, T., Shishir, F.S., Ashraf, M., Al Nasim, M.A., Shah, F.M.: Brain tumor detection using convolutional neural network. In: 2019 1st international conference on advances in science, engineering and robotics technology (ICASERT). pp. 1–6. IEEE (2019)

    2019

  9. [9]

    Huang, Y., Yang, X., Liu, L., Zhou, H., Chang, A., Zhou, X., Chen, R., Yu, J., Chen, J., Chen, C., et al.: Segment anything model for medical images? Medical Image Analysis92, 103061 (2024)

    2024

  10. [10]

    Biomedical Signal Processing and Control75, 103553 (2022)

    Inan, M.S.K., Alam, F.I., Hasan, R.: Deep integrated pipeline of segmentation guided classification of breast cancer from ultrasound images. Biomedical Signal Processing and Control75, 103553 (2022)

    2022

  11. [11]

    In: Proceedings of the AAAI conference on artificial intelligence

    Irvin, J., Rajpurkar, P., Ko, M., Yu, Y., Ciurea-Ilcus, S., Chute, C., Marklund, H., Haghgoo, B., Ball, R., Shpanskaya, K., et al.: Chexpert: A large chest radiograph dataset with uncertainty labels and expert comparison. In: Proceedings of the AAAI conference on artificial intelligence. vol. 33, pp. 590–597 (2019)

    2019

  12. [12]

    arXiv preprint arXiv:2601.10880 (2026)

    Jiang, C., Ding, T., Song, C., Tu, J., Yan, Z., Shao, Y., Wang, Z., Shang, Y., Han, T., Tian, Y.: Medical sam3: A foundation model for universal prompt-driven medical image segmentation. arXiv preprint arXiv:2601.10880 (2026)

    arXiv 2026

  13. [13]

    In: International Conference on Medical Image Computing and Computer-Assisted Intervention

    Jiang, H., Miao, C.: Anatomy-aware gating network for explainable alzheimer’s disease diagnosis. In: International Conference on Medical Image Computing and Computer-Assisted Intervention. pp. 90–100. Springer (2024)

    2024

  14. [14]

    In: 2020 42nd Annual International Conference of the IEEE Engineering in Medicine & Biology Society (EMBC)

    Jiang, H., Xu, J., Shi, R., Yang, K., Zhang, D., Gao, M., Ma, H., Qian, W.: A multi-label deep learning model with interpretable grad-cam for diabetic retinopa- thy classification. In: 2020 42nd Annual International Conference of the IEEE Engineering in Medicine & Biology Society (EMBC). pp. 1560–1563. IEEE (2020) Title Suppressed Due to Excessive Length 15

    2020

  15. [15]

    cell172(5), 1122–1131 (2018)

    Kermany, D.S., Goldbaum, M., Cai, W., Valentim, C.C., Liang, H., Baxter, S.L., McKeown, A., Yang, G., Wu, X., Yan, F., et al.: Identifying medical diagnoses and treatable diseases by image-based deep learning. cell172(5), 1122–1131 (2018)

    2018

  16. [16]

    arXiv preprint arXiv:1412.6980 (2014)

    Kingma, D.P.: Adam: A method for stochastic optimization. arXiv preprint arXiv:1412.6980 (2014)

    Pith/arXiv arXiv 2014

  17. [17]

    In: Proceedings of the IEEE/CVF international conference on computer vision

    Kirillov, A., Mintun, E., Ravi, N., Mao, H., Rolland, C., Gustafson, L., Xiao, T., Whitehead, S., Berg, A.C., Lo, W.Y., et al.: Segment anything. In: Proceedings of the IEEE/CVF international conference on computer vision. pp. 4015–4026 (2023)

    2023

  18. [18]

    In: Proceedings of the AAAI Conference on Artificial Intelligence

    Li, W., Ju, L., Tang, F., Xia, P., Xiong, X., Hu, M., Zhu, L., Ge, Z.: Towards realistic semi-supervised medical image classification. In: Proceedings of the AAAI Conference on Artificial Intelligence. vol. 39, pp. 4968–4976 (2025)

    2025

  19. [19]

    Medical Image Analysis p

    Lin, M., Holste, G., Wang, S., Zhou, Y., Wei, Y., Banerjee, I., Chen, P., Dai, T., Du, Y., Dvornek, N.C., et al.: Cxr-lt 2024: A miccai challenge on long-tailed, multi- label, and zero-shot disease classification from chest x-ray. Medical Image Analysis p. 103739 (2025)

    2024

  20. [20]

    IEEE Transactions on Medical Imaging43(2), 674–685 (2023)

    Ling, Y., Wang, Y., Dai, W., Yu, J., Liang, P., Kong, D.: Mtanet: Multi-task attention network for automatic medical image segmentation and classification. IEEE Transactions on Medical Imaging43(2), 674–685 (2023)

    2023

  21. [21]

    Nature Communications15(1), 654 (2024)

    Ma, J., He, Y., Li, F., Han, L., You, C., Wang, B.: Segment anything in medical images. Nature Communications15(1), 654 (2024)

    2024

  22. [22]

    In: International Conference on Medical Image Computing and Computer-Assisted Intervention

    Madan, C., Satia, A., Basu, S., Gupta, P., Dutta, U., Arora, C.: Focus on texture: rethinking pre-training in masked autoencoders for medical image classification. In: International Conference on Medical Image Computing and Computer-Assisted Intervention. pp. 207–217. Springer (2025)

    2025

  23. [23]

    Com- puter Methods and Programs in Biomedicine197, 105725 (2020)

    Mahbod, A., Tschandl, P., Langs, G., Ecker, R., Ellinger, I.: The effects of skin le- sion segmentation on the performance of dermatoscopic image classification. Com- puter Methods and Programs in Biomedicine197, 105725 (2020)

    2020

  24. [24]

    In: European conference on computer vision

    Müller, P., Kaissis, G., Zou, C., Rueckert, D.: Joint learning of localized represen- tations from medical images and reports. In: European conference on computer vision. pp. 685–701. Springer (2022)

    2022

  25. [25]

    In: International Conference on Medical Image Computing and Computer-Assisted Intervention

    Müller, P., Meissen, F., Brandt, J., Kaissis, G., Rueckert, D.: Anatomy-driven pathology detection on chest x-rays. In: International Conference on Medical Image Computing and Computer-Assisted Intervention. pp. 57–66. Springer (2023)

    2023

  26. [26]

    IEEE trans- actions on medical imaging39(12), 4011–4022 (2020)

    Noothout, J.M., De Vos, B.D., Wolterink, J.M., Postma, E.M., Smeets, P.A., Takx, R.A., Leiner, T., Viergever, M.A., Išgum, I.: Deep learning-based regression and classification for automatic landmark localization in medical images. IEEE trans- actions on medical imaging39(12), 4011–4022 (2020)

    2020

  27. [27]

    ACM Computing Surveys56(4), 1–41 (2023)

    Patrício,C.,Neves,J.C.,Teixeira,L.F.:Explainabledeeplearningmethodsinmed- ical image classification: A survey. ACM Computing Surveys56(4), 1–41 (2023)

    2023

  28. [28]

    PLoS medicine15(11), e1002686 (2018)

    Rajpurkar, P., Irvin, J., Ball, R.L., Zhu, K., Yang, B., Mehta, H., Duan, T., Ding, D., Bagul, A., Langlotz, C.P., et al.: Deep learning for chest radiograph diagnosis: A retrospective comparison of the chexnext algorithm to practicing radiologists. PLoS medicine15(11), e1002686 (2018)

    2018

  29. [29]

    In: Proceedings of the IEEE international conference on computer vision

    Selvaraju, R.R., Cogswell, M., Das, A., Vedantam, R., Parikh, D., Batra, D.: Grad- cam: Visual explanations from deep networks via gradient-based localization. In: Proceedings of the IEEE international conference on computer vision. pp. 618–626 (2017)

    2017

  30. [30]

    In- ternational journal of computer vision128(2), 336–359 (2020) 16 Hossain et al

    Selvaraju, R.R., Cogswell, M., Das, A., Vedantam, R., Parikh, D., Batra, D.: Grad- cam: visual explanations from deep networks via gradient-based localization. In- ternational journal of computer vision128(2), 336–359 (2020) 16 Hossain et al

    2020

  31. [31]

    Scientific data5(1), 180161 (2018)

    Tschandl, P., Rosendahl, C., Kittler, H.: The ham10000 dataset, a large collection of multi-source dermatoscopic images of common pigmented skin lesions. Scientific data5(1), 180161 (2018)

    2018

  32. [32]

    In: MICCAI Workshop on Data Engineering in Medical Imaging

    Uddin, I.I., Wang, L., Santosh, K.: Expert-guided explainable few-shot learning for medical image diagnosis. In: MICCAI Workshop on Data Engineering in Medical Imaging. pp. 95–104. Springer (2025)

    2025

  33. [33]

    Cardiovascular Imaging15(3), 413–427 (2022)

    Wang, S., Patel, H., Miller, T., Ameyaw, K., Narang, A., Chauhan, D., Anand, S., Anyanwu, E., Besser, S.A., Kawaji, K., et al.: Ai based cmr assessment of biven- tricular function: clinical significance of intervendor variability and measurement errors. Cardiovascular Imaging15(3), 413–427 (2022)

    2022

  34. [34]

    In: Proceedings of the IEEE conference on computer vision and pattern recognition

    Wang, X., Peng, Y., Lu, L., Lu, Z., Bagheri, M., Summers, R.M.: Chestx-ray8: Hospital-scale chest x-ray database and benchmarks on weakly-supervised classi- fication and localization of common thorax diseases. In: Proceedings of the IEEE conference on computer vision and pattern recognition. pp. 2097–2106 (2017)

    2097

  35. [35]

    Medical Image Analysis98, 103304 (2024)

    Xie, Y., Gu, L., Harada, T., Zhang, J., Xia, Y., Wu, Q.: Rethinking masked image modelling for medical image representation. Medical Image Analysis98, 103304 (2024)

    2024

  36. [36]

    Scientific Data10(1), 41 (2023)

    Yang,J.,Shi,R.,Wei,D.,Liu,Z.,Zhao,L.,Ke,B.,Pfister,H.,Ni,B.:Medmnistv2- a large-scale lightweight benchmark for 2d and 3d biomedical image classification. Scientific Data10(1), 41 (2023)

    2023

  37. [37]

    IEEE Transactions on Medical Imaging44(5), 2244–2255 (2025)

    Yang, Y., Fu, H., Aviles-Rivero, A.I., Xing, Z., Zhu, L.: Diffmic-v2: Medical im- age classification via improved diffusion network. IEEE Transactions on Medical Imaging44(5), 2244–2255 (2025)

    2025

  38. [38]

    Scientific Data12(1), 88 (2025)

    Yao, S., Huang, Y., Wang, X., Zhang, Y., Paixao, I.C., Wang, Z., Chai, C.L., Wang, H., Lu, D., Webb, G.I., et al.: A radiograph dataset for the classification, localization, and segmentation of primary bone tumors. Scientific Data12(1), 88 (2025)

    2025

  39. [39]

    PLoS medicine15(11), e1002683 (2018)

    Zech, J.R., Badgeley, M.A., Liu, M., Costa, A.B., Titano, J.J., Oermann, E.K.: Variable generalization performance of a deep learning model to detect pneumo- nia in chest radiographs: a cross-sectional study. PLoS medicine15(11), e1002683 (2018)

    2018

  40. [40]

    IEEE transactions on medical imaging40(6), 1618–1631 (2021)

    Zhang, Y., Li, H., Du, J., Qin, J., Wang, T., Chen, Y., Liu, B., Gao, W., Ma, G., Lei, B.: 3d multi-attention guided multi-task learning network for automatic gastric tumor segmentation and lymph node classification. IEEE transactions on medical imaging40(6), 1618–1631 (2021)

    2021

  41. [41]

    Nature methods22(1), 166–176 (2025)

    Zhao, T., Gu, Y., Yang, J., Usuyama, N., Lee, H.H., Kiblawi, S., Naumann, T., Gao, J., Crabtree, A., Abel, J., et al.: A foundation model for joint segmentation, detection and recognition of biomedical objects across nine modalities. Nature methods22(1), 166–176 (2025)

    2025

  42. [42]

    In: Proceedings of the IEEE conference on computer vision and pattern recognition

    Zhou, B., Khosla, A., Lapedriza, A., Oliva, A., Torralba, A.: Learning deep features for discriminative localization. In: Proceedings of the IEEE conference on computer vision and pattern recognition. pp. 2921–2929 (2016)

    2016