Pith. sign in

REVIEW 5 major objections 6 minor 78 references

VLM-based Prompts as the Optimal Assistant for Unpaired Histopathology Virtual Staining

T0 review · 5 major / 6 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 This paper claims that text prompts from a pathology language model can guide unpaired virtual staining, beating prior methods on image quality and downstream detection.

desk verdict Plausible first integration of pathology VLM prompts with virtual staining, but the SOTA claim rests on test-set hyperparameter tuning and a one-patient test split. read the letter →

arxiv 2504.15545 v1 pith:MOOD3PA5 submitted 2025-04-22 eess.IV cs.CV

Zizhi Chen , Xinyu Zhang , Minghao Han , Yizhou Liu , Ziyun Qian , Weifeng Zhang , Xukun Zhang , Jingwei Wei
show 1 more author
Lihua Zhang
This is my paper · ORCID
classification eess.IVcs.CV
keywords virtualstaininghistopathologyvision-languagemodelprompttuningunpairedimage-to-imagetranslationglomerulardetectiongenerativeadversarialnetworkdiffusioninferenceenhancement
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 pathology vision-language model (VLM) can serve as an expert assistant for unpaired virtual staining, using text prompts as training constraints that preserve tissue structure while applying the correct stain colors. The authors build VPGAN, a CycleGAN augmented with three prompt-based losses: learnable contrastive prompts that describe the staining transition, a fixed tissue-invariance anchor, and stain-specific anchors that push the output toward the target stain. Adding a VLM-based multi-level calibration to DDIM-style inference enhancement yields HARBOR, which the paper reports avoids the staining-domain collapse of its baseline while improving structural metrics. Across H&E-to-MAS, H&E-to-PAS, and H&E-to-PASM kidney datasets, VPGAN reports the best zero-cost inference results on twelve metrics, HARBOR reports the best inference-enhanced results, and downstream glomerular detection and segmentation improve. If the claims hold, virtual staining can be guided by biomedical language knowledge rather than style statistics alone, making generated special stains more reliable for diagnosis.

What carries the argument

The load-bearing mechanism is the cosine-similarity geometry of a frozen pathology VLM's joint image-text embedding space, used as a supervisor from three directions. Contrastive Prompt Transfer (CPT) uses a binary classification loss over learnable positive and negative prompts to separate source-domain from target-domain images. Constant Concept Anchoring (CCA) computes cosine similarity of pre-stain and post-stain images to a fixed text anchor describing renal tissue and imposes an MSE loss so the 'kidney-ness' does not change during staining. Independent Concept Reinforcement (ICR) computes similarities to four stain-specific text anchors and applies a cross-entropy loss so the output is classified into the correct stain. In the inference-enhancement phase, the mechanism shifts to a trained zero-initialized noise prompt map Z added at every DDIM step, optimized by an SSIM structural constraint against the source path, an MSE stylization constraint against the VPGAN output, and a VLM calibration term that matches intermediate-layer features of the ResNet101 CLIP encoder to keep the style from collapsing.

What would settle it

Train VPGAN on H&E-to-PASM with the CONCH text anchors replaced by shuffled non-pathological sentences and by pure color-histogram descriptors, keeping all loss weights and training budgets fixed; if SSIM, FID, and glomerular detection scores stay within noise of the reported values, the semantic content of the prompts is not carrying the improvement. A cheaper check is to probe whether the trained generator's output embeddings move toward the target stain anchor because of texture and structure rather than hue distribution alone.

Watch

Extended reading notes

Core claim

The paper's central claim is that text embeddings from a pathology-specialized VLM can supervise virtual staining better than image-only adversarial or contrastive losses, and that the same VLM can stabilize diffusion-based inference enhancement. VPGAN adds three modules to a CycleGAN backbone: Contrastive Prompt Transfer trains a learnable positive/negative prompt pair so the VLM's text encoder describes the source-to-target stain transition; Constant Concept Anchoring measures cosine similarity between pre- and post-stain image embeddings and a fixed renal-tissue anchor, penalizing their difference with MSE; Independent Concept Reinforcement applies a softmax cross-entropy loss over stain-specific anchors so the generated image is classified into the intended stain. The authors report that VPGAN outperforms all zero-cost baselines on twelve metrics across the three kidney-stain tasks, and that among the VLMs compared, the pathology-specialized model CONCH gives the best average results while general CLIP degrades performance below the baseline. HARBOR then combines DDIM dual-path inversion with a multi-level calibration loss computed from intermediate ResNet101 CLIP features, which the paper says repairs staining-domain collapse and achieves the best inference-enhanced results. A final set of experiments claims that H&E-to-PASM virtual staining by VPGAN improves glomerular detection and segmentation relative to real H&E images, indicating clinical potential.

Load-bearing premise

The method rests on the premise that cosine similarity in the pathology VLM's image-text embedding space reliably tracks stain identity and tissue invariance; if those embeddings mostly encode color statistics rather than biological structure, VPGAN could improve style metrics while adding little pathological fidelity.

Editorial extensions

If this is right

  • VPGAN is presented as a drop-in optimization module, so any unpaired GAN-based virtual stainer could in principle be upgraded with the three prompt losses without changing its generator or discriminator architecture.
  • HARBOR's calibration repairs staining-domain collapse in DDIM-based inference enhancement, making diffusion-style high-resolution virtual staining usable where it previously failed.
  • The choice of embedding space matters: pathology-specialized VLMs outperform general CLIP, and an unsuitable VLM can make results worse than no VLM at all.
  • Virtual H&E-to-PASM staining can serve as data augmentation: glomerular detection and segmentation on generated PASM images exceed those on real H&E in the paper's experiments.
  • Normalizing PAS patches from different disease cohorts with VPGAN before training a patch-level segmentation model improves average segmentation accuracy over training on unnormalized merged data.

Reading between the lines

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

  • A natural stress test is to replace the LLM-written, stain-specific anchors with color-histogram or photometric descriptors while keeping the same loss geometry; if the reported gains persist, the VLM's semantic knowledge is not the active ingredient, and if they vanish, language anchoring is doing the work.
  • The paper's own limitation section (Appendix D.1) notes that DDIM enhancement takes 5–10 minutes per 256×256 patch and that naive acceleration degrades quality, so a clinically usable HARBOR likely needs a distilled or consistency-model sampler, an open engineering problem.
  • The same three-loss recipe does not depend on the four kidney stains tested, so it should transfer to H&E-to-IHC virtual staining or label-free-to-stain synthesis; testing that would broaden the claim beyond renal pathology.
  • The calibration strength λ has a narrow optimal interval, with over-correction causing a secondary bright-pink collapse, suggesting VLM guidance is beneficial only with bounded authority; similar saturation effects may appear in other VLM-supervised generation tasks.
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

5 major / 6 minor

Summary. This manuscript introduces VPGAN and HARBOR, two methods for unpaired histopathology virtual staining guided by a pathology vision-language model. VPGAN extends CycleGAN with three auxiliary losses: a contrastive prompt-transfer loss based on learnable text prompts that classify source and target stain domains, a constant-concept anchor loss that penalizes changes in cosine similarity to a tissue-invariance text embedding, and an independent-concept reinforcement loss that encourages the generated image to match a stain-specific text embedding. HARBOR adds a DDIM-based inference-enhancement stage with a multi-level calibration loss computed in a ResNet101 CLIP embedding space. The methods are evaluated on H&E-to-MAS/PAS/PASM transfer on the ANHIR dataset, with downstream glomerular detection and segmentation on the GS and KPIs datasets. The authors report that VPGAN achieves the best metrics among zero-cost inference methods and that HARBOR improves over the DPI baseline, and they release code.

Significance. The idea of using a pathology-specific VLM as an auxiliary supervisor in virtual staining is genuinely novel relative to the cited GAN and diffusion baselines, and the paper includes useful ablations: Table 2 decomposes the contributions of the CPT, CCA, and ICR modules, and Table 3 compares four VLMs. If the results hold under a properly split validation protocol, the method would be of interest to the computational pathology community. The code release and the use of publicly available datasets are strengths. However, the current evidence is not yet sufficient for the paper's SOTA claims, because the main comparison appears to involve test-set hyperparameter selection and no uncertainty estimates; the paper also overstates what Table 1 shows. The central causal claim, that the gains come from the VLM's pathological knowledge rather than from additional style-matching losses, needs a stronger evaluation.

major comments (5)
  1. [Section 4.2 and Table 1] The statement that VPGAN 'achieved SOTA performance across 12 metrics' is not supported by the reported numbers. On H&E2MAS, VPGAN is below UMDST in PSNR (16.49 vs 17.16) and MS-SSIM (0.8526 vs 0.8571); on H&E2PAS it is below UNSB in FID (132.95 vs 112.46); and on H&E2PASM it is below UNSB in FID (125.28 vs 87.54). Please restate the claim with an exact metric count and identify which comparisons are actually won.
  2. [Appendix A, Section 4.4, and Table 7] The evaluation protocol allows test-set hyperparameter selection. There is no validation split: Patient 5 is the test set, Table 4 explicitly searches lambda on the H&E2MAS task using the same metrics reported in Table 1, and Table 7 reports per-dataset alpha, mu, and lambda. The baselines are not given the same per-task tuning, so the comparison is asymmetric and the headline numbers are an upper bound of a tuned protocol. Please introduce a held-out validation set, or nested cross-validation, and fix all hyperparameters before computing the final test metrics.
  3. [Section 4.1 and Tables 1, 5, 6] All quantitative claims are based on a single test patient without error bars or significance tests. The reported gains, for example H&E2PASM SSIM increasing from 0.5490 for CycleGAN to 0.6650 for VPGAN, could be within run-to-run or patient-to-patient variability. Please report means and standard deviations over multiple training runs or resamples, and state the number of independent test patches used for each metric.
  4. [Equation (13)] The printed formula for the SSIM-based structural constraint is not the standard SSIM and appears numerically inconsistent: it uses variances and covariance but omits the luminance terms, and the second factor in the denominator is sigma_Z * sigma_X + c2 rather than sigma_Z^2 + sigma_X^2 + c2. If this equation is the actual loss used in HARBOR, please provide a correct, reproducible definition; if it is a typographical error, please fix it and re-derive the corresponding gradient.
  5. [Section 3.2 and Tables 5-6] The causal claim that the VLM contributes pathological knowledge rather than style statistics is not fully established. All auxiliary losses are cosine-similarity terms in a frozen text-image embedding space, and the downstream validation has no error bars and no comparison against a non-VLM style-matching loss of comparable capacity. A concrete sanity check would be to show that the prompt losses improve stain classification accuracy or tissue-structure preservation on held-out annotations, and to compare against a color-matching loss such as histogram matching or an MMD loss with the same architecture.
minor comments (6)
  1. [Section 4.2] The claim that HARBOR achieves 'SOTA performance in all indicators among inference enhancement methods' should be phrased as 'compared with DPI', which is the only inference-enhancement baseline; DPI has no entries on H&E2PASM because of the reported staining-domain collapse.
  2. [Appendix C] The statement that 'most parameter settings have already been configured in the code' makes the paper not self-contained; please list all hyperparameters, including alpha, beta, gamma, mu, lambda, delta_l, and the prompt length N, in the paper.
  3. [Section 4.5] The dataset name 'ANINR' appears to be a typo for 'ANHIR'.
  4. [Equations (1)-(9)] The equations use e^{cos(...)} without stating a temperature or normalization; please clarify whether the exponential is intentional and, if so, specify the temperature.
  5. [Table 1 caption] The red/blue background legend should be repeated in the caption or replaced with explicit symbols, since the color coding may not survive printing or accessibility conversion.
  6. [Table 3] The VLM comparison reports point estimates only; without variance or a paired test, the conclusion that CONCH 'exhibited more balanced outcomes' is weak.

Circularity Check

1 steps flagged · score 2.0 of 10

No derivation-level circularity; the only fitted-input issue is per-task hyperparameter selection on the same single-patient test set used for the SOTA claims.

  1. fitted input called prediction [Section 4.4 (Table 4) and Appendix C (Table 7), with test split defined in Appendix A]
    "This intriguing phenomenon necessitates the search for the optimal hyperparameter λ in the correction loss function on the H&E2MAS dataset, guiding the reasoning enhancement to the optimal staining range. ... We use four sets (Patient 1, Patient 2, Patient 3, and Patient 4) as the training set and one set (Patient 5) as the testing set. ... Some hyperparameters in our method may have different numerical values depending on the dataset, as shown in the table 7 below."

    The ANHIR evaluation uses a single test patient with no validation split. Section 4.4 explicitly searches for the optimal λ on the H&E2MAS test metrics in Table 4, and Table 7 reports task-specific α, μ, and λ values. The same test set then produces the Table 1 headline results ('VPGAN achieved SOTA performance across 12 metrics', 'HARBOR achieved the SOTA performance in all indicators'). Thus the reported numbers are partly the result of fitting hyperparameters to the evaluation set, while baselines are not given the same per-task test-set tuning. This is not an equation-level derivation equivalence, but it is a fitted-input selection presented as a fixed-protocol prediction and is the paper's main circularity-adjacent weakness.

full rationale

The three VLM-based losses (Eqs. 3, 6, and 9) are auxiliary training constraints imposed through frozen CONCH embeddings; none of them equals any reported evaluation metric by construction. SSIM, CSS, MS-SSIM, PSNR, and FID are computed against real target images or real-image feature distributions, and the downstream glomerular detection and segmentation tasks use independent annotations, so the central claim is externally testable rather than self-validating. The paper's self-citations (e.g., MSCPT) appear only in related-work context and are not load-bearing. The choice of CONCH over CLIP, PLIP, and MUSK is empirical (Table 3), not stipulated. The only concern is procedural: Section 4.4 tunes λ on the H&E2MAS test set, and Table 7 gives per-dataset α, μ, and λ values while Appendix A defines no validation split, so the SOTA comparisons in Table 1 include test-set selection that baselines did not receive. This can inflate the reported advantage, but it does not make the derivation equivalent to its inputs; hence a low circularity score of 2.

Assumptions & free parameters 7 free parameters · 4 assumptions · 0 invented entities

The method's central claim depends on the availability of a matching pathology VLM, on the validity of text-image similarity as a staining supervisor, and on a registered test set for pixel-level metrics. Hyperparameters are numerous and partly tuned on the test set, which increases the ledger cost.

free parameters (7)
  • alpha (contrastive loss weight) = 30 (H&E2MAS, H&E2PASM), 50 (H&E2PAS)
    Table 7; selected per dataset, not justified by a validation protocol; controls the contribution of L_contrastive in Equation 10.
  • beta (constant concept loss weight) = 0.1 (all tasks)
    Table 7; hyperparameter for Equation 10, no sensitivity analysis is provided.
  • gamma (stain concept loss weight) = 0.1 (all tasks)
    Table 7; hyperparameter for Equation 10, no sensitivity analysis is provided.
  • mu (structure/style balance) = 0.05, 0.55, 0.8 per task
    Table 7; balances L_struct and L_style in Equation 18; chosen per task.
  • lambda (calibration weight) = 0.001 (H&E2MAS, H&E2PAS), 0.05 (H&E2PASM)
    Table 7; explicitly searched on H&E2MAS test metrics in Section 4.4, Table 4; selection is done on test data.
  • learnable contrastive prompts T_p and T_n = learned vectors of length N x 512 (N unspecified)
    Section 3.1; optimized with Equations 1 and 2 on source and target training images, then used in the CPT loss Equation 3.
  • delta_l (ResNet layer calibration weights) = not reported in main text
    Equation 17; weights for five CLIP/ResNet layers, referenced to the supplementary materials; not enough detail to reproduce without code.
assumptions (4)
  • domain assumption CONCH text embeddings of LLM-generated stain descriptions are semantically accurate for staining domains.
    All auxiliary losses in Equations 4 to 9 rely on this; the paper manually removes LLM hallucinations but provides no systematic validation that the embeddings align with stain biology.
  • domain assumption ANHIR serial sections can be treated as approximately registered ground truth for computing SSIM, PSNR, and CSS across domains.
    Evaluation in Table 1 uses pixel-level similarity between generated and real target patches; no registration-error analysis is reported.
  • domain assumption A single held-out patient (Patient 5) is representative for stain-transfer generalization.
    Appendix A; all stain-transfer results come from one test patient after excluding H&E Patient 1 from training.
  • domain assumption DDIM inversion with a pretrained diffusion model provides a valid trajectory for inference enhancement.
    Section 3.3; the method relies on the DPI baseline and gives no independent derivation of validity for pathology patches.

how reviews work

0 comments
Cite this review

Pith. "Pith review of VLM-based Prompts as the Optimal Assistant for Unpaired Histopathology Virtual Staining." pith.science (2026). https://pith.science/paper/MOOD3PA5

@misc{pith2026250415545,
  author       = {Pith},
  title        = {Pith review of: VLM-based Prompts as the Optimal Assistant for Unpaired Histopathology Virtual Staining},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/MOOD3PA5}},
  note         = {Machine review of arXiv:2504.15545}
}
read the original abstract

In histopathology, tissue sections are typically stained using common H&E staining or special stains (MAS, PAS, PASM, etc.) to clearly visualize specific tissue structures. The rapid advancement of deep learning offers an effective solution for generating virtually stained images, significantly reducing the time and labor costs associated with traditional histochemical staining. However, a new challenge arises in separating the fundamental visual characteristics of tissue sections from the visual differences induced by staining agents. Additionally, virtual staining often overlooks essential pathological knowledge and the physical properties of staining, resulting in only style-level transfer. To address these issues, we introduce, for the first time in virtual staining tasks, a pathological vision-language large model (VLM) as an auxiliary tool. We integrate contrastive learnable prompts, foundational concept anchors for tissue sections, and staining-specific concept anchors to leverage the extensive knowledge of the pathological VLM. This approach is designed to describe, frame, and enhance the direction of virtual staining. Furthermore, we have developed a data augmentation method based on the constraints of the VLM. This method utilizes the VLM's powerful image interpretation capabilities to further integrate image style and structural information, proving beneficial in high-precision pathological diagnostics. Extensive evaluations on publicly available multi-domain unpaired staining datasets demonstrate that our method can generate highly realistic images and enhance the accuracy of downstream tasks, such as glomerular detection and segmentation. Our code is available at: https://github.com/CZZZZZZZZZZZZZZZZZ/VPGAN-HARBOR

Figures

Figures reproduced from arXiv: 2504.15545 by the authors.

Figure 1
Figure 1. Three auxiliary methods for virtual staining tasks [PITH_FULL_IMAGE:figures/full_fig_p001_1.png] view at source ↗
Figure 2
Figure 2. Overview of the proposed VPGAN and HARBOR. In the prompt generation phase, we employed prompt tuning method [PITH_FULL_IMAGE:figures/full_fig_p003_2.png] view at source ↗
Figure 3
Figure 3. We demonstrate a fine-grained verification process [PITH_FULL_IMAGE:figures/full_fig_p005_3.png] view at source ↗
Figures from the paper (3 more)
Figure 4
Figure 4. Figure 4: The performance comparison of various existing methods and our proposed method for multiple stain transfer of the [PITH_FULL_IMAGE:figures/full_fig_p006_4.png]
Figure 5
Figure 5. Figure 5: Overview of Downstream Task Datasets enhancing the performance of glomerular object detection at the patch level. Please refer to the supplementary materials for all specific details of dataset division and preprocessing. Evaluation Metrics. In this experiment, we empl…
Figure 6
Figure 6. Figure 6: Unlike H&E2PASM, H&E2MAS correction shows [PITH_FULL_IMAGE:figures/full_fig_p008_6.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

78 extracted references · 43 canonical work pages

  1. [1]

    Josh Achiam, Steven Adler, Sandhini Agarwal, Lama Ahmad, Ilge Akkaya, Floren- cia Leoni Aleman, Diogo Almeida, Janko Altenschmidt, Sam Altman, Shyamal Anadkat, et al. 2023. Gpt-4 technical report. arXiv preprint arXiv:2303.08774 (2023)

    arXiv 2023

  2. [2]

    Jiří Borovec, Jan Kybic, Ignacio Arganda-Carreras, Dmitry V Sorokin, Gloria Bueno, Alexander V Khvostikov, Spyridon Bakas, Eric I-Chao Chang, Stefan Heldmann, Kimmo Kartasalo, et al. 2020. ANHIR: automatic non-rigid histological image registration challenge. IEEE transactions on medical imaging 39, 10 (2020), 3042–3052

    work page 2020

  3. [3]

    Gloria Bueno, Lucia Gonzalez-Lopez, Marcial Garcia-Rojo, Arvydas Laurinavicius, and Oscar Deniz. 2020. Data for glomeruli characterization in histopathological images. Data in brief 29 (2020), 105314

    work page 2020

  4. [4]

    Fuqiang Chen, Ranran Zhang, Boyun Zheng, Yiwen Sun, Jiahui He, and Wenjian Qin. 2024. Pathological semantics-preserving learning for H&E-to-IHC virtual staining. In International Conference on Medical Image Computing and Computer- Assisted Intervention. Springer, 384–394

    2024

  5. [5]

    Richard J Chen, Tong Ding, Ming Y Lu, Drew FK Williamson, Guillaume Jaume, Bowen Chen, Andrew Zhang, Daniel Shao, Andrew H Song, Muhammad Shaban, et al. 2024. Towards a General-Purpose Foundation Model for Computational Pathology. Nature Medicine (2024)

    work page 2024

  6. [6]

    Ting Chen, Simon Kornblith, Mohammad Norouzi, and Geoffrey Hinton. 2020. A simple framework for contrastive learning of visual representations. In Interna- tional conference on machine learning . PmLR, 1597–1607

    2020

  7. [7]

    Cong Cong, Sidong Liu, Antonio Di Ieva, Maurice Pagnucco, Shlomo Berkovsky, and Yang Song. 2022. Colour adaptive generative networks for stain normalisation of histopathology images. Medical Image Analysis 82 (2022), 102580

    work page 2022

  8. [8]

    Kevin de Haan, Yijie Zhang, Jonathan E Zuckerman, Tairan Liu, Anthony E Sisk, Miguel FP Diaz, Kuang-Yu Jen, Alexander Nobori, Sofia Liou, Sarah Zhang, et al

Show all 78 references
  1. [9]

    Ruining Deng, Tianyuan Yao, Yucheng Tang, Junlin Guo, Siqi Lu, Juming Xiong, Lining Yu, Quan Huu Cap, Pengzhou Cai, Libin Lan, et al . 2025. KPIs 2024 Challenge: Advancing Glomerular Segmentation from Patch-to Slide-Level.arXiv preprint arXiv:2502.07288 (2025)

    2025 arXiv

  2. [10]

    Kexue Fu, Linhao Qu, Shuo Wang, Ying Xiong, Ilias Maglogiannis, Longxiang Gao, Manning Wang, et al. 2024. Fast: A dual-tier few-shot learning paradigm for whole slide image classification. Advances in Neural Information Processing Systems 37 (2024), 105090–105113

    2024

  3. [11]

    Ian J Goodfellow, Jean Pouget-Abadie, Mehdi Mirza, Bing Xu, David Warde-Farley, Sherjil Ozair, Aaron Courville, and Yoshua Bengio. 2014. Generative adversarial nets. Advances in neural information processing systems 27 (2014)

    2014

  4. [12]

    Jiaxiang Gou, Luping Ji, Pei Liu, and Mao Ye. 2024. Queryable Prototype Multiple Instance Learning with Vision-Language Models for Incremental Whole Slide Image Classification. arXiv preprint arXiv:2410.10573 (2024)

    2024 arXiv

  5. [13]

    Xianchao Guan, Yifeng Wang, Yiyang Lin, Xi Li, and Yongbing Zhang. 2024. Unsupervised multi-domain progressive stain transfer guided by style encoding dictionary. IEEE Transactions on Image Processing 33 (2024), 767–779

    2024

  6. [14]

    Daya Guo, Dejian Yang, Haowei Zhang, Junxiao Song, Ruoyu Zhang, Runxin Xu, Qihao Zhu, Shirong Ma, Peiyi Wang, Xiao Bi, et al . 2025. Deepseek-r1: Incentivizing reasoning capability in llms via reinforcement learning. arXiv preprint arXiv:2501.12948 (2025)

    2025 arXiv

  7. [15]

    Zhengrui Guo, Conghao Xiong, Jiabo Ma, Qichen Sun, Lishuang Feng, Jinzhuo Wang, and Hao Chen. 2024. FOCUS: Knowledge-enhanced Adaptive Visual Compression for Few-shot Whole Slide Image Classification. arXiv preprint arXiv:2411.14743 (2024)

    2024 arXiv

  8. [16]

    Minghao Han, Linhao Qu, Dingkang Yang, Xukun Zhang, Xiaoying Wang, and Lihua Zhang. 2024. MSCPT: Few-shot Whole Slide Image Classification with Multi-scale and Context-focused Prompt Tuning. arXiv preprint arXiv:2408.11505 (2024)

    2024 arXiv

  9. [17]

    Kaiming He, Haoqi Fan, Yuxin Wu, Saining Xie, and Ross Girshick. 2020. Mo- mentum contrast for unsupervised visual representation learning. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition . 9729–9738

    2020

  10. [18]

    Kaiming He, Georgia Gkioxari, Piotr Dollár, and Ross Girshick. 2017. Mask r-cnn. In Proceedings of the IEEE international conference on computer vision . 2961–2969

    2017

  11. [19]

    Kaiming He, Xiangyu Zhang, Shaoqing Ren, and Jian Sun. 2016. Deep residual learning for image recognition. In Proceedings of the IEEE conference on computer vision and pattern recognition . 770–778

    2016

  12. [20]

    Martin J Hetz, Tabea-Clara Bucher, and Titus J Brinker. 2024. Multi-domain stain normalization for digital pathology: A cycle-consistent adversarial network for whole slide images. Medical Image Analysis 94 (2024), 103149

    2024

  13. [21]

    Jonathan Ho, Ajay Jain, and Pieter Abbeel. 2020. Denoising diffusion probabilistic models. Advances in neural information processing systems 33 (2020), 6840–6851

    2020

  14. [22]

    Ming-Yang Ho, Che-Ming Wu, Min-Sheng Wu, and Yufeng Jane Tseng. 2024. Every pixel has its moments: Ultra-high-resolution unpaired image-to-image translation via dense normalization. In European Conference on Computer Vision . Springer, 312–328

    2024

  15. [23]

    Ming-Yang Ho, Min-Sheng Wu, and Che-Ming Wu. 2022. Ultra-high-resolution unpaired stain transformation via kernelized instance normalization. InEuropean Conference on Computer Vision . Springer, 490–505

    2022

  16. [24]

    Shengyi Hua, Fang Yan, Tianle Shen, Lei Ma, and Xiaofan Zhang. 2024. Pathoduet: Foundation models for pathological slide analysis of H&E and IHC stains.Medical Image Analysis 97 (2024), 103289

    2024

  17. [25]

    Zhi Huang, Federico Bianchi, Mert Yuksekgonul, Thomas J Montine, and James Zou. 2023. A visual–language foundation model for pathology image analysis using medical twitter. Nature medicine 29, 9 (2023), 2307–2316

    2023

  18. [26]

    Wisdom Ikezogwo, Saygin Seyfioglu, Fatemeh Ghezloo, Dylan Geva, Fatwir Sheikh Mohammed, Pavan Kumar Anand, Ranjay Krishna, and Linda Shapiro

  19. [27]

    Maximilian Ilse, Jakub Tomczak, and Max Welling. 2018. Attention-based deep multiple instance learning. InInternational conference on machine learning. PMLR, 2127–2136

    2018

  20. [28]

    Chanyong Jung, Gihyun Kwon, and Jong Chul Ye. 2024. Patch-wise graph con- trastive learning for image translation. In Proceedings of the AAAI Conference on Artificial Intelligence, Vol. 38. 13013–13021

    2024

  21. [29]

    Beomsu Kim, Gihyun Kwon, Kwanyoung Kim, and Jong Chul Ye. [n. d.]. Un- paired Image-to-Image Translation via Neural Schrödinger Bridge. InThe Twelfth International Conference on Learning Representations

  22. [30]

    Junho Kim, Minjae Kim, Hyeonwoo Kang, and Kwang Hee Lee. 2020. U-GAT-IT: Unsupervised Generative Attentional Networks with Adaptive Layer-Instance Normalization for Image-to-Image Translation. In International Conference on Learning Representations

    2020

  23. [31]

    Murli Krishna. 2013. Role of special stains in diagnostic liver pathology. Clinical liver disease 2, S1 (2013), S8–S10. Preprint, 2025 , Zizhi Chen, Xinyu Zhang and Minghao Han, et al

    2013

  24. [32]

    Thomas Lampert, Odyssée Merveille, Jessica Schmitz, Germain Forestier, Friedrich Feuerhake, and Cédric Wemmert. 2019. Strategies for training stain in- variant CNNs. In 2019 IEEE 16th International Symposium on Biomedical Imaging (ISBI 2019). IEEE, 905–909

    2019

  25. [33]

    Fangda Li, Zhiqiang Hu, Wen Chen, and Avinash Kak. 2023. Adaptive super- vised patchnce loss for learning h&e-to-ihc stain translation with inconsistent groundtruth image pairs. InInternational Conference on Medical Image Computing and Computer-Assisted Intervention. Springer, 632–641

    2023

  26. [34]

    Hao Li, Ying Chen, Yifei Chen, Rongshan Yu, Wenxian Yang, Liansheng Wang, Bowen Ding, and Yuchen Han. 2024. Generalizable whole slide image classifica- tion with fine-grained visual-semantic interaction. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern ...

    2024

  27. [35]

    Jiahan Li, Jiuyang Dong, Shenjin Huang, Xi Li, Junjun Jiang, Xiaopeng Fan, and Yongbing Zhang. 2024. Virtual immunohistochemistry staining for histological images assisted by weakly-supervised learning. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern R...

    2024

  28. [36]

    Jingxiong Li, Sunyi Zheng, Chenglu Zhu, Yuxuan Sun, Pingyi Chen, Zhongyi Shui, Yunlong Zhang, Honglin Li, and Lin Yang. 2024. PathUp: Patch-wise Timestep Tracking for Multi-class Large Pathology Image Synthesising Diffusion Model. In Proceedings of the 32nd ACM International C...

    2024

  29. [37]

    Zhexin Liang, Chongyi Li, Shangchen Zhou, Ruicheng Feng, and Chen Change Loy. 2023. Iterative prompt learning for unsupervised backlit image enhancement. In Proceedings of the IEEE/CVF International Conference on Computer Vision . 8094– 8103

    2023

  30. [38]

    Yiyang Lin, Yifeng Wang, Zijie Fang, Zexin Li, Xianchao Guan, Danling Jiang, and Yongbing Zhang. 2024. A Multi-Perspective Self-Supervised Generative Adversarial Network for FS to FFPE Stain Transfer. IEEE Transactions on Medical Imaging (2024)

    2024

  31. [39]

    Yiyang Lin, Bowei Zeng, Yifeng Wang, Yang Chen, Zijie Fang, Jian Zhang, Xi- angyang Ji, Haoqian Wang, and Yongbing Zhang. 2022. Unpaired multi-domain stain transfer for kidney histopathological images. In Proceedings of the AAAI Conference on Artificial Intelligence, Vol. 36. ...

    2022

  32. [40]

    Pei Liu, Luping Ji, Jiaxiang Gou, Bo Fu, and Mao Ye. 2025. Interpretable Vision- Language Survival Analysis with Ordinal Inductive Bias for Computational Pathology. In The Thirteenth International Conference on Learning Representations . https://openreview.net/forum?id=trj2Jq8riA

    2025

  33. [41]

    Shengjie Liu, Chuang Zhu, Feng Xu, Xinyu Jia, Zhongyue Shi, and Mulan Jin. 2022. Bci: Breast cancer immunohistochemical image generation through pyramid pix2pix. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition. 1815–1824

    2022

  34. [42]

    Ying-Chih Lo, I-Fang Chung, Shin-Ning Guo, Mei-Chin Wen, and Chia-Feng Juang. 2021. Cycle-consistent GAN-based stain translation of renal pathology images with glomerulus detection application. Applied Soft Computing 98 (2021), 106822

    2021

  35. [43]

    Wei Lou, Guanbin Li, Xiang Wan, and Haofeng Li. 2024. Multi-modal Denoising Diffusion Pre-training for Whole-Slide Image Classification. In Proceedings of the 32nd ACM International Conference on Multimedia . 10804–10813

    2024

  36. [44]

    Ming Y Lu, Bowen Chen, Drew FK Williamson, Richard J Chen, Ivy Liang, Tong Ding, Guillaume Jaume, Igor Odintsov, Long Phi Le, Georg Gerber, et al. 2024. A visual-language foundation model for computational pathology. Nature Medicine 30, 3 (2024), 863–874

    2024

  37. [45]

    Ming Y Lu, Drew FK Williamson, Tiffany Y Chen, Richard J Chen, Matteo Barbieri, and Faisal Mahmood. 2021. Data-efficient and weakly supervised computational pathology on whole-slide images. Nature Biomedical Engineering 5, 6 (2021), 555–570

    2021

  38. [46]

    Ziwei Luo, Fredrik K Gustafsson, Zheng Zhao, Jens Sjölund, and Thomas B Schön

  39. [47]

    Jeffrey H Miner. 2012. The glomerular basement membrane. Experimental cell research 318, 9 (2012), 973–978

    2012

  40. [48]

    Ron Mokady, Amir Hertz, Kfir Aberman, Yael Pritch, and Daniel Cohen-Or. 2023. Null-text inversion for editing real images using guided diffusion models. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition . 6038–6047

    2023

  41. [49]

    arXiv preprint arXiv:2310.01018 (2023)

    Controlling Vision-Language Models for Universal Image Restoration. arXiv preprint arXiv:2310.01018 (2023)

    2023 arXiv

  42. [50]

    Qiong Peng, Weiping Lin, Yihuang Hu, Ailisi Bao, Chenyu Lian, Weiwei Wei, Meng Yue, Jingxin Liu, Lequan Yu, and Liansheng Wang. 2024. Advancing H&E- to-IHC Virtual Staining with Task-Specific Domain Knowledge for HER2 Scoring. In International Conference on Medical Image Compu...

    2024

  43. [51]

    Linhao Qu, Kexue Fu, Manning Wang, Zhijian Song, et al. 2023. The rise of ai language pathologists: Exploring two-level prompt learning for few-shot weakly- supervised whole slide image classification. Advances in Neural Information Processing Systems 36 (2023), 67551–67564

    2023

  44. [52]

    Taesung Park, Alexei A Efros, Richard Zhang, and Jun-Yan Zhu. 2020. Contrastive learning for unpaired image-to-image translation. InComputer Vision–ECCV 2020: 16th European Conference, Glasgow, UK, August 23–28, 2020, Proceedings, Part IX

    2020

  45. [53]

    Olaf Ronneberger, Philipp Fischer, and Thomas Brox. 2015. U-net: Convolu- tional networks for biomedical image segmentation. In Medical image computing and computer-assisted intervention–MICCAI 2015: 18th international conference, Munich, Germany, October 5-9, 2015, proceeding...

    2015

  46. [54]

    Jiangbo Shi, Chen Li, Tieliang Gong, Chunbao Wang, and Huazhu Fu. 2024. CoD- MIL: Chain-of-Diagnosis Prompting Multiple Instance Learning for Whole Slide Image Classification. IEEE Transactions on Medical Imaging (2024)

    2024

  47. [55]

    Alec Radford, Jong Wook Kim, Chris Hallacy, Aditya Ramesh, Gabriel Goh, Sandhini Agarwal, Girish Sastry, Amanda Askell, Pamela Mishkin, Jack Clark, et al. 2021. Learning transferable visual models from natural language supervision. In International conference on machine learni...

    2021

  48. [56]

    Yang Song, Jascha Sohl-Dickstein, Diederik P Kingma, Abhishek Kumar, Ste- fano Ermon, and Ben Poole. [n. d.]. Score-Based Generative Modeling through Stochastic Differential Equations. In International Conference on Learning Repre- sentations

  49. [57]

    Yuxuan Sun, Yixuan Si, Chenglu Zhu, Xuan Gong, Kai Zhang, Pingyi Chen, Ye Zhang, Zhongyi Shui, Tao Lin, and Lin Yang. 2024. CPath-Omni: A Unified Multimodal Foundation Model for Patch and Whole Slide Image Analysis in Computational Pathology. arXiv preprint arXiv:2412.12077 (2024)

    2024 arXiv

  50. [58]

    Jiangbo Shi, Chen Li, Tieliang Gong, Yefeng Zheng, and Huazhu Fu. 2024. ViLa- MIL: Dual-scale Vision-Language Multiple Instance Learning for Whole Slide Image Classification. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition . 11248–11258

    2024

  51. [59]

    Yuxuan Sun, Chenglu Zhu, Sunyi Zheng, Kai Zhang, Lin Sun, Zhongyi Shui, Yunlong Zhang, Honglin Li, and Lin Yang. 2024. Pathasst: A generative foundation ai assistant towards artificial general intelligence of pathology. In Proceedings of the AAAI Conference on Artificial Intel...

    2024

  52. [60]

    Cheng-Chang Tsai, Yuan-Chih Chen, and Chun-Shien Lu. 2024. Test-Time Stain Adaptation with Diffusion Models for Histopathology Image Classification. In European Conference on Computer Vision . Springer, 257–275

    2024

  53. [61]

    Yuxuan Sun, Yunlong Zhang, Yixuan Si, Chenglu Zhu, Kai Zhang, Zhongyi Shui, Jingxiong Li, Xuan Gong, XINHENG LYU, Tao Lin, et al. [n. d.]. PathGen-1.6 M: 1.6 Million Pathology Image-text Pairs Generation through Multi-agent Collabo- ration. In The Thirteenth International Conf...

  54. [62]

    Patrick D Walker, Tito Cavallo, and Stephen M Bonsib. 2004. Practice guidelines for the renal biopsy. Modern Pathology 17, 12 (2004), 1555–1563

    2004

  55. [63]

    Yuanbo Wen, Tao Gao, and Ting Chen. 2024. Unpaired Photo-realistic Image Deraining with Energy-informed Diffusion Model. In Proceedings of the 32nd ACM International Conference on Multimedia . 360–369

    2024

  56. [64]

    Jelica Vasiljević, Zeeshan Nisar, Friedrich Feuerhake, Cédric Wemmert, and Thomas Lampert. 2022. CycleGAN for virtual stain transfer: Is seeing really believing? Artificial Intelligence in Medicine 133 (2022), 102420

    2022

  57. [65]

    Bing Xiong, Yue Peng, RanRan Zhang, Fuqiang Chen, JiaYe He, and Wenjian Qin

  58. [66]

    Renao Yan, Qiming He, Yiqing Liu, Peng Ye, Lianghui Zhu, Shanshan Shi, Jizhou Gou, Yonghong He, Tian Guan, and Guangde Zhou. 2023. Unpaired virtual histological staining using prior-guided generative adversarial networks. Com- puterized Medical Imaging and Graphics 105 (2023), 102185

    2023

  59. [67]

    Jinxi Xiang, Xiyue Wang, Xiaoming Zhang, Yinghua Xi, Feyisope Eweje, Yijiang Chen, Yuchen Li, Colin Bergstrom, Matthew Gopaulchan, Ted Kim, et al. 2025. A vision–language foundation model for precision oncology. Nature (2025), 1–10

    2025

  60. [68]

    Yijie Zhang, Kevin de Haan, Yair Rivenson, Jingxi Li, Apostolos Delis, and Aydo- gan Ozcan. 2020. Digital synthesis of histological stains using micro-structured and multiplexed virtual staining of label-free tissue. Light: Science & Applications 9, 1 (2020), 78

    2020

  61. [69]

    Zhanjie Zhang, Quanwei Zhang, Wei Xing, Guangyuan Li, Lei Zhao, Jiakai Sun, Zehua Lan, Junsheng Luan, Yiling Huang, and Huaizhong Lin. 2024. Artbank: Artistic style transfer with pre-trained diffusion model and implicit style prompt bank. In Proceedings of the AAAI Conference ...

    2024

  62. [70]

    Min Zhao, Fan Bao, Chongxuan Li, and Jun Zhu. 2022. Egsde: Unpaired image-to- image translation via energy-guided stochastic differential equations. Advances in Neural Information Processing Systems 35 (2022), 3609–3623

    2022

  63. [71]

    Wei Zhang, Tik Ho Hui, Pui Ying Tse, Fraser Hill, Condon Lau, and Xinyue Li

  64. [72]

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

    High-Resolution Medical Image Translation via Patch Alignment-Based Bidirectional Contrastive Learning. In International Conference on Medical Image Computing and Computer-Assisted Intervention. Springer, 178–188

  65. [73]

    Jun-Yan Zhu, Taesung Park, Phillip Isola, and Alexei A Efros. 2017. Unpaired image-to-image translation using cycle-consistent adversarial networks. In Pro- ceedings of the IEEE international conference on computer vision . 2223–2232. VLM-based Prompts as the Optimal Assistant...

    2017

  66. [76]

    Kaiyang Zhou, Jingkang Yang, Chen Change Loy, and Ziwei Liu. 2022. Conditional prompt learning for vision-language models. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition . 16816–16825

    2022

  67. [77]

    Kaiyang Zhou, Jingkang Yang, Chen Change Loy, and Ziwei Liu. 2022. Learning to prompt for vision-language models. International Journal of Computer Vision 130, 9 (2022), 2337–2348

    2022

  68. [2021]

    Nature communications 12, 1 (2021), 4884

    Deep learning-based transformation of H&E stained tissues into special stains. Nature communications 12, 1 (2021), 4884

    2021

  69. [2023]

    Advances in neural information processing systems 36 (2023), 37995–38017

    Quilt-1m: One million image-text pairs for histopathology. Advances in neural information processing systems 36 (2023), 37995–38017

    2023

  70. [2024]

    arXiv preprint arXiv:2412.11106 (2024)

    Unpaired Multi-Domain Histopathology Virtual Staining using Dual Path Prompted Inversion. arXiv preprint arXiv:2412.11106 (2024)

    2024 arXiv

Pith tools

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