The DOI matches its Crossref record; 186 of the 186 references it cites were found in a registry or an index.
Towards a general-purpose foundation model for computational pathology
Open accessAbstract
Quantitative evaluation of tissue images is crucial for computational pathology (CPath) tasks, requiring the objective characterization of histopathological entities from whole-slide images (WSIs). The high resolution of WSIs and the variability of morphological features present significant challenges, complicating the large-scale annotation of data for high-performance applications. To address this challenge, current efforts have proposed the use of pretrained image encoders through transfer learning from natural image datasets or self-supervised learning on publicly available histopathology datasets, but have not been extensively developed and evaluated across diverse tissue types at scale. We introduce UNI, a general-purpose self-supervised model for pathology, pretrained using more than 100 million images from over 100,000 diagnostic H&E-stained WSIs (>77 TB of data) across 20 major tissue types. The model was evaluated on 34 representative CPath tasks of varying diagnostic difficulty. In addition to outperforming previous state-of-the-art models, we demonstrate new modeling capabilities in CPath such as resolution-agnostic tissue classification, slide classification using few-shot class prototypes, and disease subtyping generalization in classifying up to 108 cancer types in the OncoTree classification system. UNI advances unsupervised representation learning at scale in CPath in terms of both pretraining data and downstream evaluation, enabling data-efficient artificial intelligence models that can generalize and transfer to a wide range of diagnostically challenging tasks and clinical workflows in anatomic pathology. Pretrained using over 100,000 diagnostic histopathological slides across 20 major tissue types, a self-supervised model is shown to outperform existing baselines across various clinically relevant computational pathology tasks.
Overview
The claims and statistics of this paper haven’t been checked yet.
Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
The claims and statistics of this paper haven’t been checked yet.
Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.
Claims other papers make about this one
Each is checked against this paper’s own text: does it say what the citing paper says it does?
The source backs part of the claim, not all of it — or backs it for a different population, condition or comparison.
In the citing paper
“By increasing the patch size from the widely used 256 × 256 pixels, we effectively reduce the sequence length by four without impacting the representation quality due to higher resolution patch input, leveraging the robustness of the patch-level foundation models in generalizing to higher resolutions9,10,87.”
From A multimodal whole-slide foundation model for pathology (2025) · cites this paper as [9]
Experimental — accuracy not yet measuredThe source backs part of the claim, not all of it — or backs it for a different population, condition or comparison.
In the citing paper
“we present Virchow, the largest foundation model for computational pathology to date.”
From A foundation model for clinical-grade computational pathology and rare cancers detection (2024) · cites this paper as [41]
Experimental — accuracy not yet measuredThe source backs part of the claim, not all of it — or backs it for a different population, condition or comparison.
In the citing paper
“To overcome these limitations, several research groups have alternatively proposed the use of weakly-supervised deep learning models, which can be trained to perform various downstream computational pathology tasks such as tumor subtyping and prognosis32–37 without any region or pixel-level annotations.”
From Predicting the tumor microenvironment composition and immunotherapy response in non-small cell lung cancer from digital histopathology images (2024) · cites this paper as [32]
Experimental — accuracy not yet measuredThe source backs part of the claim, not all of it — or backs it for a different population, condition or comparison.
In the citing paper
“UNI [191], Prov-GigaPath [192], and Virchow [193] were pretrained solely on pathology image data to perform cancer classification, diagnosis, and prognosis.”
From Challenges in AI-driven Biomedical Multimodal Data Fusion and Analysis (2025) · cites this paper as [191]
Experimental — accuracy not yet measuredThe cited source says what the claim attributes to it.
In the citing paper
“Built on rich knowledge representations, these models have shown impressive performance across medical fields such as ophthalmology20, radiology21 and pathology22–25.”
From A Multimodal Vision Foundation Model for Clinical Dermatology (2025) · cites this paper as [23]
The cited source says what the claim attributes to it.
In the citing paper
“The vision module is based on the patch encoder CTransPath17 for the small and medium models, and UNI18 for the large model.”
From Generating dermatopathology reports from gigapixel whole slide images with HistoGPT (2025) · cites this paper as [18]
The cited source says what the claim attributes to it.
In the citing paper
“We found out that the UNI foundation model32, when paired with multi-task AB-MIL, achieved the best predictive performance for the various TME signature prediction tasks.”
From Predicting the tumor microenvironment composition and immunotherapy response in non-small cell lung cancer from digital histopathology images (2024) · cites this paper as [32]
In the citing paper
“In fact GMA is widely used to assess the performance of pathology foundation models23,25,26.”
From A clinical benchmark of public self-supervised pathology foundation models (2025) · cites this paper as [23]
In the citing paper
“Self-supervised learning has become a widely used approach for pretraining foundation models on thousands to tens of thousands of images without annotations or labels, marking an important paradigm shift in the field.20,21”
From Use of artificial intelligence for liver diseases: A survey from the EASL congress 2024 (2024) · cites this paper as [20]
In the citing paper
“An increasing number of published studies show the potential of foundation models20,24; however, they are currently underrepresented in the field of liver diseases.4”
From Use of artificial intelligence for liver diseases: A survey from the EASL congress 2024 (2024) · cites this paper as [20]
In the citing paper
“These works demonstrate that image features produced with self-supervised learning of pathology images outperform image features trained on natural images and that performance improves with scale.”
From A foundation model for clinical-grade computational pathology and rare cancers detection (2024) · cites this paper as [41]
In the citing paper
“Despite the challenges in collecting large-scale datasets in the pathology domain, recent pioneering works have utilized datasets ranging from 30,000 to 400,000 WSIs to train foundation models ranging in size from 28 million to 307 million parameters35–42”
From A foundation model for clinical-grade computational pathology and rare cancers detection (2024) · cites this paper as [41]
In the citing paper
“Specialty-specific foundation models20,21,23 typically focus on single imaging modalities”
From A Multimodal Vision Foundation Model for Clinical Dermatology (2025) · cites this paper as [23]
In the citing paper
“The scaling behavior observed in PanDerm’s performance aligns with recent foundation model trends20,22,23,59”
From A Multimodal Vision Foundation Model for Clinical Dermatology (2025) · cites this paper as [23]
Self-check
Percentages, sums and reported differences re-computed from the paper’s own numbers.
The self-check needs the paper’s full text.
Citations
Cites
Deep Residual Learning for Image Recognition
He, K., Zhang, X., Ren, S., & Sun, J. (2016) · 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr.2016.90 · Cited by 235,474
Scikit-learn: Machine Learning in Python
Pedregosa, F., Varoquaux, G., Gramfort, A., Michel, V., Thirion, B., Grisel, O., Blondel, M., Müller, A., Nothman, J., Louppe, G., et al. (2012) · arXiv · doi:10.48550/arxiv.1201.0490 · Cited by 71,016
ImageNet: A large-scale hierarchical image database
Deng, J., Dong, W., Socher, R., Li, L.-J., Li, K., & Fei-Fei, L. (2009) · 2009 IEEE Conference on Computer Vision and Pattern Recognition · doi:10.1109/cvpr.2009.5206848 · Cited by 65,333
ImageNet Large Scale Visual Recognition Challenge
Russakovsky, O., Deng, J., Su, H., Krause, J., Satheesh, S., Ma, S., Huang, Z., Karpathy, A., Khosla, A., Bernstein, M., et al. (2015) · International Journal of Computer Vision · doi:10.1007/s11263-015-0816-y · Cited by 39,380
Swin Transformer: Hierarchical Vision Transformer using Shifted Windows
Liu, Z., Lin, Y., Cao, Y., Hu, H., Wei, Y., Zhang, Z., Lin, S., & Guo, B. (2021) · 2021 IEEE/CVF International Conference on Computer Vision (ICCV) · doi:10.1109/iccv48922.2021.00986 · Cited by 35,534
PyTorch: An Imperative Style, High-Performance Deep Learning Library
Paszke, A., Gross, S., Massa, F., Lerer, A., Bradbury, J., Chanan, G., Killeen, T., Lin, Z., Gimelshein, N., Antiga, L., et al. (2019) · arXiv · doi:10.48550/arxiv.1912.01703 · Cited by 31,473
Language Models are Few-Shot Learners
Brown, T. B., Mann, B., Ryder, N., Subbiah, M., Kaplan, J., Dhariwal, P., Neelakantan, A., Shyam, P., Sastry, G., Askell, A., et al. (2020) · arXiv · doi:10.48550/arxiv.2005.14165 · Cited by 23,372
Advances in neural information processing systems 7
Cios, K. J., & Shields, M. E. (1997) · Neurocomputing · doi:10.1016/s0925-2312(97)90018-7 · Cited by 22,093
Least squares quantization in PCM
Lloyd, S. (1982) · IEEE Transactions on Information Theory · doi:10.1109/tit.1982.1056489 · Cited by 15,836
Goldberger, A. L., Amaral, L. A. N., Glass, L., Hausdorff, J. M., Ivanov, P. C., Mark, R. G., Mietus, J. E., Moody, G. B., Peng, C.-K., & Stanley, H. E. (2000) · Circulation · doi:10.1161/01.cir.101.23.e215 · Cited by 14,809
Research works printed but not linked: 13
- Chen, T., Kornblith, S., Norouzi, M. & Hinton, G. A simple framework for contrastive learning of visual representations. In International Conference on Machine Learning, 1597–1607 (PMLR, 2020).
- Chen, R. J. & Krishnan, R. G. Self-supervised vision transformers learn visual concepts in histopathology. In Learning Meaningful Representations of Life, NeurIPS 2021 (2022).
- Dosovitskiy, A. et al. An image is worth 16×16 words: transformers for image recognition at scale. In International Conference on Learning Representations (2021).
- Xiang, J. & Zhang, J. Exploring low-rank property in multiple instance learning for whole slide image classification. In The Eleventh International Conference on Learning Representations (2022).
- Wang, X. et al. SCL-WC: cross-slide contrastive learning for weakly-supervised whole-slide image classification. Advances in Neural Information Processing Systems 35, 18009–18021 (2022).
- Zhou, J. et al. iBOT: image BERT pre-training with online tokenizer. In International Conference on Learning Representations (2022).
- Devlin, J., Chang, M.-W., Lee, K. & Toutanova, K. BERT: pre-training of deep bidirectional transformers for language understanding. In Proceedings of the 2019 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies, Volume 1 (Long and Short Papers) (2018).
- Sablayrolles, A., Douze, M., Schmid, C. & Jégou, H. Spreading vectors for similarity search. In International Conference on Learning Representations (2019).
- Touvron, H., Vedaldi, A., Douze, M. & Jegou, H. Fixing the train–test resolution discrepancy. In Advances in Neural Information Processing Systems, Vol. 32 (eds Wallach, H. et al.) (Curran Associates, 2019).
- Dao, T., Fu, D. Y., Ermon, S., Rudra, A. & Ré, C. FlashAttention: fast and memory-efficient exact attention with IO-awareness. In Advances in Neural Information Processing Systems (2022).
- Loshchilov, I. & Hutter, F. Decoupled weight decay regularization. In International Conference on Learning Representations (2019).
- Sarıyıldız, M. B., Kalantidis, Y., Alahari, K. & Larlus, D. No reason for no supervision: improved generalization in supervised models. In The Eleventh International Conference on Learning Representations (2023).
- Yang, J., Chen, H., Yan, J., Chen, X. & Yao, J. Towards better understanding and better generalization of low-shot classification in histology images with contrastive learning. In International Conference on Learning Representations (2021).
Cited by
Sorted by citations
Items 601–700 of 1,003
Foundation models in computational pathology: methods, applications and clinical implications
Sali, R., Aschenaki, Y., Leveillee, R. J., Baba, F., Tessema, L., Dixon, C. M., Bachhamba, K., & Zhang, D. Y. (2026) · BMJ Oncology · doi:10.1136/bmjonc-2026-001102 · Cited by 4
Wang, Z., Fang, Y., & Li, Q. (2026) · Journal of Translational Medicine · doi:10.1186/s12967-026-08211-0 · Cited by 3
“UNI builds upon the DINOv2 paradigm, which combines contrastive learning with iBOT-based MLM [23].”
Han, C., Li, D., Zhao, B., Zhang, X., Lin¹, J., Yan, X. Z., Mao, N., Lin, J., Deng, T., Huang, J., et al. (2026) · The Lancet Digital Health · doi:10.1016/j.landig.2025.100977 · Cited by 2
Self-citationSong, H.-J., Cho, Y. S., & Kim, Y. S. (2026) · Diagnostics · doi:10.3390/diagnostics16101570 · Cited by 3
“Recent pathology-specific foundation models such as CTransPath [11], UNI [12], Virchow [13], and Phikon-v2 [14] have substantially improved feature representations over ImageNet-pretrained encoders.”
Artificial Intelligence in Oncology: Clinical Applications, Challenges, and Opportunities
Parkinson, G., Patel, R. D., Bergstrom, C. P., Vadasz, B., Amgad, M., Cooper, L. A. D., Sparano, J. A., & Lu, J. (2026) · American Society of Clinical Oncology Educational Book · doi:10.1200/edbk-26-520716 · Cited by 4
Abdullah, A., Shahid, M., Ather, M. A., Fatima, Z., Sanchéz-Mejorada, C. G., Torres-Ruiz, M. J., Quintero, R., Mata, F., & Zagal-Flores, R. (2026) · Cancers · doi:10.3390/cancers18121952 · Cited by 2
“These include CONCH, a vision language foundation model for computational pathology [40]; UNI, a general-purpose pathology foundation model [41]; MCAT (Multimodal Co-Attention Transformer), a multimodal survival prediction model integrating genomic and pathology modalities [42]; and MultiMAE, a…”
Artificial general intelligence and the clinical laboratory: a paradigm shift toward Lab 2.0
Yuji, K., & Yuji, W. (2026) · Clinical Chemistry and Laboratory Medicine (CCLM) · doi:10.1515/cclm-2026-0789 · Cited by 2
“As foundation models in computational pathology, multi-omic analysis, and decision support become more common in laboratory work fl ows [19], this dual competency is likely to become a practical requirement for ensuring the safety, validity, and equity of diagnostic outputs rather than an optional…”
PathFound: An agentic multimodal model activating evidence-seeking pathological diagnosis
Hua, S., Wu, J., Shen, T., Hu, K., Huang, Z., Ni, S., Zhang, Z., Li, Y., Wang, Z., & Zhang, X. (2026) · Medical Image Analysis · doi:10.1016/j.media.2026.104200 · Cited by 4
Machine Learning for Radiomics in Oncology: Challenges, Limitations, and Future Directions
Missaoui, R., Saadaoui, W., Coco, M. D., Helali, A., Leo, M., & Carcagnì, P. (2026) · Sensors · doi:10.3390/s26144619 · Cited by 2
MerMED-FM: Multimodal, Multi-Disease Medical Imaging Foundation Model
Zhou, Y., Quek, C. W. N., Zhou, J., Wang, Y., Bai, Y., Gutiérrez, L., Ke, Y., Yao, J., Teo, Z. L., Ting, D. S. J., et al. (2026) · The Lancet Digital Health · doi:10.1016/j.landig.2026.101007 · Cited by 2
Beaufils, A., de Martino, J., Jiang, X. Y., Blanchard, T., Fraunhoffer, N. A., Mendes, D., Pignolet, C., Bourega, T. A., Meneghetti, A. R., Sainath, S., et al. (2026) · Journal of Clinical Oncology · doi:10.1200/jco-26-00327 · Cited by 2
“23 For patient-level prediction, tile features from all patient-associated slides were aggregated into bags and processed using attention-based multiple instance learning.”
Agentic systems in computational pathology: architectures, evidence, and translational challenges
Lu, X., Li, Q., Gao, Y., Wei, D., Lyu, M., Ma, S., Li, J., Wu, Y., Cai, L., Zhang, T., et al. (2026) · Journal of Translational Medicine · doi:10.1186/s12967-026-08829-0 · Cited by 2
“Staining, scanners, and local workflows can reduce generalization [103, 130].”
InfluentialBenchmarking biomedical foundation models
Sáez-Rodríguez, J., Schäfer, P. S. L., Kalavros, N., & Stolovitzky, G. A. (2026) · Nature Methods · doi:10.1038/s41592-026-03182-y · Cited by 6
“FMs are being rapidly developed and hold big promise across biomedical domains, from structural biology 3–5 to ‘virtual cells’ 6,7 , prediction of function from genetic sequence 8–10 , imaging 11 and electronic health records (EHRs) 12 .”
Akbar, A., Sajjad, U., Su, Z., Chowdhury, A. N., Tran, A. M., Xing, F., Ruiz, J., Parwani, A. V., Chen, W., Li, W., et al. (2026) · npj Digital Medicine · doi:10.1038/s41746-026-03264-3 · Cited by 12
Deep Learning-Based Automated Measurement of Murine Bone Length in Radiographs
Rong, R., Denton, K., Jin, K. W., Quan, P., Wen, Z., Kozlitina, J., Lyon, S., Wang, A., Wise, C. A., Beutler, B., et al. (2024) · Bioengineering · doi:10.3390/bioengineering11070670 · Cited by 2
Efficient annotation bootstrapping for cell identification in follicular lymphoma
Krawczyk, A., Osowska-Kurczab, A. M., Pakuło, S., Kotłowski, W., & Świderska-Chadaj, Ż. (2025) · Computer Methods and Programs in Biomedicine · doi:10.1016/j.cmpb.2025.108728 · Cited by 3
Yuan, H., Zhu, M., Yang, R., Liu, H., Li, I., & Hong, C. (2025) · Health care science · doi:10.1002/hcs2.70009 · Cited by 6
Nguyen, M.-T., Pham, N. D. H., Nguyen, P. L., & Do, M. N. (2025) · IEEE International Symposium on Biomedical Imaging (ISBI) · doi:10.1109/isbi60581.2025.10981240 · Cited by 2
“We evaluated SEMA against baseline architectures, including ResNet variants [9] and finetuned UNI [6], denoted as UNI-ft.”
Kalweit, G., Klett, A., Silvestrini, P., Rahnfeld, J., Naouar, M., Vogt, Y., Infante, D., Berger, R., Duque‐Afonso, J., Hartmann, T. N., et al. (2025) · Frontiers in Oncology · doi:10.3389/fonc.2025.1480384 · Cited by 2
“Despite these advancements, training foundation models speci fi cally for medical applications often requires substantial computational resources (6, 7, 12), limiting accessibility for multiple iterations during model development.”
Simplified Artificial Intelligence Terminology for Pathologists
Zabihollahy, F., Mankaruos, M., Mohareb, M., Youssef, T., Soleymani, Y., & Yousef, G. M. (2025) · Diagnostics · doi:10.3390/diagnostics15131699 · Cited by 2
Cross-Disciplinary Applications of Generative AI in Oral Cancer Research and Early Detection
Vadisetty, R., & Polamarasetti, A. (2025) · Lecture notes in networks and systems · doi:10.1007/978-981-96-3728-7_11 · Cited by 2
Domain-incremental white blood cell classification with privacy-aware continual learning
Kumari, P., Bozorgpour, A., Reisenbüchler, D., Jost, E., Crysandt, M. M., Matek, C., & Merhof, D. (2025) · Scientific Reports · doi:10.1038/s41598-025-08024-z · Cited by 7
Benchmarking Histopathology Foundation Models in a Multi-center Dataset for Skin Cancer Subtyping
Meseguer, P., del Amor, R., & Naranjo, V. (2025) · Lecture notes in computer science · doi:10.1007/978-3-031-98688-8_2 · Cited by 2
Liu, H., Farahani, H., Li, X., Xie, Y., & Bashashati, A. (2025) · IEEE Journal of Biomedical and Health Informatics · doi:10.1109/jbhi.2025.3595101 · Cited by 2
“…First, back-bone networks have evolved from convolutional neural network (CNN)-based architectures (e.g., KimiaNet [25], Ciga [26], and REMEDIS [27]) to ViT-based ones (e.g., UNI [28], PLIP [29], CONCH [30], GigaPath [31], and Phikon [32]), as well as integrated structure (e.g., CTransPath [33]).”
Min, W., Xue, S., Zhu, F., Xu, T., Wang, C., & Xie, H.-B. (2025) · Tsinghua Science & Technology · doi:10.26599/tst.2025.9010087 · Cited by 2
Pathology Report Generation and Multimodal Representation Learning for Cutaneous Melanocytic Lesions
Lucassen, R. T., Moonemans, S. P. J., van de Luijtgaarden, T., Breimer, G. E., Blokx, W. A. M., & Veta, M. (2025) · Lecture notes in computer science · doi:10.1007/978-3-032-04978-0_48 · Cited by 2
Jiang, J., Liu, J., Wu, H., Wu, Y., Wang, Z., Bi, Q., & Zheng, Y. (2025) · ACM International Conference on Multimedia (ACM MM) · doi:10.1145/3746027.3754838 · Cited by 3
Foundation models in medical image analysis: A systematic review and quantitative analysis
Rajendran, P., Safari, M., He, W., Hu, M., Shah, K. D., Wang, S., Zhou, J., & Yang, X. (2026) · Medical Image Analysis · doi:10.1016/j.media.2026.104293 · Cited by 8
Vadori, V., Peruffo, A., Graïc, J., Finos, L., & Grisan, E. (2025) · Annual International Conference of the IEEE Engineering in Medicine and Biology Society (EMBC) · doi:10.1109/embc58623.2025.11253185 · Cited by 5
“…self-supervised learning approaches, these models learn image representations that generalize effectively across a range of downstream tasks, such as diagnosis, disease subtyping, biomarker quantification, treatment response estimation, and survival prediction [2], [17], [26], [27], [29], [30].”
InfluentialA Hierarchical Geometry-Guided Transformer for Histological Subtyping of Primary Liver Cancer
Lu, A., Liu, M., Jiao, Y., Xu, G., Gong, H., Cai, C., Chen, J., & Xu, J. (2025) · IEEE International Conference on Bioinformatics and Biomedicine (BIBM) · doi:10.1109/bibm66473.2025.11357200 · Cited by 2
“Finally, these two FoVs patch features were fed into the pathological foundation model UNI [7] independently, for capturing global region patterns (macro-organization feature, f macro ) and local tissue details (meso-pheotype feature, f meso ) simultaneously.”
Bursting the bubble on diagnostics: artificial intelligence in autoimmune bullous disease
Doolan, B. J., & Thomas, B. (2024) · British Journal of Dermatology · doi:10.1093/bjd/ljae197 · Cited by 2
Is Histopathology Deep Learning Artificial Intelligence the Future of Precision Oncology?
Wagner, V. M. (2024) · Journal of Clinical Oncology · doi:10.1200/jco-24-01271 · Cited by 2
Ayhan, M. S., Ong, A. Y., Ruffell, E., Wagner, S. K., Merle, D. A., & Keane, P. A. (2025) · medRxiv · doi:10.1101/2025.03.09.25323618 · Cited by 2
Laternser, S., Micheli, A. J. D., Brüningk, S. C., McDonough, E., Petrović, A., Rushing, E. J., Surrette, C., Bugajska, J., Eze, A., Kilburn, L. B., et al. (2025) · bioRxiv (Cold Spring Harbor Laboratory) · doi:10.1101/2025.03.23.644698 · Cited by 2
Trajanoski, Z., Plattner, C., Sturm, G., Kühl, A. A., Atreya, R. N., Carollo, S., Gronauer, R., Rieder, D., Günther, M., Ormanns, S., et al. (2025) · Research Square · doi:10.21203/rs.3.rs-6443303/v1 · Cited by 3
“…step where whole slide images (WSI) were tessellated into patches and color-normalized, followed by a feature extraction step leveraging four different foundation models: CHIEF 38 , UNI2 39 , Virchow2 40,41 and H-optimus-0 42 , which is the largest open-source AI foundation model for pathology.”
InfluentialGevaert, O., Bareja, R., Carrillo‐Pérez, F., Zheng, Y., Pizurica, M., Nandi, T., Shen, J., & Madduri, R. (2025) · Research Square · doi:10.21203/rs.3.rs-6823810/v1 · Cited by 3
“Many pathology foundation models rely on self-supervised pre-training using datasets like TCGA 24,33,44 or large proprietary datasets from a single institution 25,26 , which can inadvertently result in domain-specic over�tting rather than true generalization across diverse clinical settings.”
Wong, B., & Yi, M. Y. (2025) · SSRN Electronic Journal · doi:10.2139/ssrn.5362908 · Cited by 2
SAME: Topology-flexible transforms enable robust integration of multimodal spatial omicsPreprint
Pratapa, A., Mansouri, S., Nikulina, N., Matuck, B. F., Schneider, M. A., Byrd, K. M., Savai, R., Tata, P. R., & Singh, R. (2025) · bioRxiv (Cold Spring Harbor Laboratory) · doi:10.1101/2025.07.12.664419 · Cited by 2
“Introduced for metabolite integration, our use of H&E foundation models[19, 20] to extract phenotypic features from morphology offers a general strategy for modalities lacking direct cell-type information.”
Zhang, N., Long, Y., Xia, S., Jia, P., Jin, Y., Chen, S., Chen, J., & Li, M. (2025) · bioRxiv (Cold Spring Harbor Laboratory) · doi:10.1101/2025.08.07.669202 · Cited by 2
Nie, G., Chen, X., Wang, Y., Chen, J., Shi, Y., Zhong, J., Fang, X., Shi, J., Liu, C.-F., Huang, B., et al. (2025) · medRxiv · doi:10.1101/2025.09.06.25335216 · Cited by 2
“Such models, which leverage large-scale pretraining and generalize across diverse downstream tasks, have shown considerable promise in the medical domain, in-83 cluding computational pathology 24,25 , echocardiography 26 , and physiological signal analysis 27–31 .”
Integrative spatial multi-omics reveals prognostic tumor niches in female genital tumorsPreprint
Tian, X., Lin, S., Wu, Y., He, Y., Zhang, Z., Zhao, Y., Zhang, P., Zhao, L., Yang, W., Ye, M., et al. (2025) · bioRxiv (Cold Spring Harbor Laboratory) · doi:10.1101/2025.11.24.690313 · Cited by 2
scPRINT-2: Towards the next-generation of cell foundation models and benchmarksPreprint
Kalfon, J., Gabriel Peyré, A., & Cantini, L. (2025) · bioRxiv (Cold Spring Harbor Laboratory) · doi:10.64898/2025.12.11.693702 · Cited by 2
Rao, V. M., Sadanandappa, M. K., Black, C. C., Palisoul, S. M., Workman, A. A., MacKenzie, T. A., Liu, X., Chamberlin, M. D., Vaickus, L. J., Zanazzi, G. J., et al. (2026) · bioRxiv (Cold Spring Harbor Laboratory) · doi:10.64898/2026.02.09.704587 · Cited by 2
A foundation AI model enhances electron microscopy image analysisPreprint
Du, M., Wang, Y., Xie, L., Deng, G., Guo, J., Han, B., Chen, Z., Rui, C., Han, J., Chen, Y., et al. (2026) · bioRxiv (Cold Spring Harbor Laboratory) · doi:10.64898/2026.02.28.708664 · Cited by 2
Deep learning representations of human Immune Health for precision immunologyPreprint
Lee, M., Kim, J., Ionita, M., McKeague, M. L., Lee, J., Nam, Y., Jeong, C.-U., Nair, A., Moyo, E. T., Khavin, I., et al. (2026) · bioRxiv (Cold Spring Harbor Laboratory) · doi:10.64898/2026.07.15.724605 · Cited by 2
Buzzanca, G., Pala, C., He, J., Hofstraat, R., Tammaro, A., van Midden, D., Buelow, R. D., Hölscher, D. L., Muehlfeld, A., Koeller, M. C., et al. (2026) · medRxiv · doi:10.64898/2026.08.28.26360029 · Cited by 2
Myles, C., Um, I. H., Harrison, D. J., & Harris‐Birtill, D. (2024) · Lecture notes in computer science · doi:10.1007/978-3-031-66955-2_23 · Cited by 2
AutoEncoder-Based Feature Transformation with Multiple Foundation Models in Computational Pathology
Chung, W., Park, Y., & Nam, Y. (2024) · Lecture notes in computer science · doi:10.1007/978-3-031-73471-7_5 · Cited by 1
Computer vision methods under rapid evolution for pathology image tasks
Maher, N. G., Scolyer, R. A., & Liu, S. (2024) · Histopathology · doi:10.1111/his.15352 · Cited by 1
“This is a game-changing leap forward, as these pathology domain-specific, task-agnostic models can be more readily adapted for task-specific functions due to their reduced need for task-specific training examples, 7 owing to their ability to extract meaningful features from histopathology images.”
Chen, Y., Chen, G., Shi, G., Zhou, Y., Bai, J., Corredor, G., Lu, C. Y., & Lei, X. (2024) · Big Data Mining and Analytics · doi:10.26599/bdma.2024.9020057 · Cited by 1
Ovarian cancer subtype classification based on deep learning models on WSI and TMA images
Gu, Y. (2024) · International Symposium on Artificial Intelligence for Medicine Sciences (ISAIMS) · doi:10.1145/3706890.3706985 · Cited by 1
Foundation Models for Slide-Level Cancer Subtyping in Digital Pathology
Meseguer, P., del Amor, R., Colomer, A., & Naranjo, V. (2025) · Lecture notes in computer science · doi:10.1007/978-3-031-78241-1_18 · Cited by 2
Pruitt, K. T., Pasupuleti, H., Yu, J. J. Q., DeAtley, W., & Fei, B. (2025) · SPIE Medical Imaging: Computer-Aided Diagnosis · doi:10.1117/12.3048802 · Cited by 1
Mehrabian, H., Mahadevan, S. K., Sharpnack, M. F., Moon, C., & Diehl, L. J. (2025) · SPIE Medical Imaging: Digital and Computational Pathology · doi:10.1117/12.3047319 · Cited by 1
“The UNI pathology foundation model [8] was adapted for 3-class classification of the disease marker signal by replacing its original classification head with a two-layer classification network (Figure 2).”
Zheng, W., Huang, A., Yuan, J., Zhao, H., Zhao, Z., Xu, Y., & Géraud, T. (2025) · IEEE International Conference on Acoustics Speech and Signal Processing · doi:10.1109/icassp49660.2025.10890868 · Cited by 1
Ács, B., Fend, F., Guettier, C., L’Imperio, V., Montezuma, D., Zerbe, N., & Zlobec, I. (2025) · Archiv für Pathologische Anatomie und Physiologie und für Klinische Medicin · doi:10.1007/s00428-025-04084-8 · Cited by 2
“The vision of foundation models as a comprehensive AI tool in pathology, e.g., “Virchow model” capable of pan-cancer detection, or molecular status prediction, represents a significant leap forward [3–5].”
Pruitt, K. T., DeAtley, W., Rathgeb, A. P., Johnson, B. A., Gahan, J. C., & Fei, B. (2025) · SPIE Medical Imaging: Image-Guided Procedures, Robotic Interventions, and Modeling · doi:10.1117/12.3048439 · Cited by 1
“…medical imaging tasks, such as Mass-100K, The Cancer Genome Atlas, and Providence, all of which contain thousands of H&E stained histopathology slides that have been used in the training and evaluation of whole-slide digital pathology models including UNI [11], CONCH [12], and Prov-GigaPath [13].”
Kasireddy, H. R., Lucarelli, N., Yun, D., Moon, K. C., Rosa, P. S. L., Tomaszewski, J. E., Han, S. S., Shickel, B., Naglah, A., & Sarder, P. (2025) · SPIE Medical Imaging: Digital and Computational Pathology · doi:10.1117/12.3047908 · Cited by 2
Foundation Modelle in der Pathologie
Klauschen, F., Dippel, J., & Müller, K. R. (2025) · Die Pathologie · doi:10.1007/s00292-025-01429-7 · Cited by 1
Kläger, J. P., Koeller, M. C., & Compérat, É. (2025) · Diagnostic histopathology · doi:10.1016/j.mpdhp.2025.04.005 · Cited by 1
Bui, D. C., Pham, H. L., Le, V. T. D., Hai, V. T., Tran, V. D., Nguyen, K. T. T. M., & Nakashima, Y. (2025) · IEEE Access · doi:10.1109/access.2025.3580470 · Cited by 1
Mehr, H. D., Cong, C., Noorani, I., Ieva, A. D., & Liu, S. (2025) · Lecture notes in computer science · doi:10.1007/978-3-031-95838-0_9 · Cited by 1
Should Healthcare Resist AI?: The Transformative Impact of LLMs in Digital Pathology
Mobilio, G., Orlandi, E., Riccio, M. D., Ugolini, F., Grassi, S., Simoncini, A., Massi, D., & Mucchi, L. (2025) · International Symposium on Medical Information and Communication Technology · doi:10.1109/ismict64722.2025.11059396 · Cited by 1
Bitto, V., Jiang, X., Baumann, M. A., Kather, J. N., & Kurth, I. (2025) · Modern Pathology · doi:10.1016/j.modpat.2025.100845 · Cited by 2
Akebli, H., Mea, V. D., & Roitero, K. (2025) · IEEE International Conference on Healthcare Informatics (ICHI) · doi:10.1109/ichi64645.2025.00099 · Cited by 1
An adaptive multi-graph fusion for tumor grading in pathology images
Alzoubi, I., Xin, B., Bjerkvig, R., Wang, J., & Wang, X. (2025) · Pattern Recognition · doi:10.1016/j.patcog.2025.112214 · Cited by 2
Self-citationHaile, G., Zhu, Y., Maqsood, M. H., Yin, X., & Liew, A. W. (2025) · Computers in Biology and Medicine · doi:10.1016/j.compbiomed.2025.110734 · Cited by 1
Li, H., Chen, W., Pan, Y., Wu, L., & Qiu, J. (2025) · Biomedical Signal Processing and Control · doi:10.1016/j.bspc.2025.108402 · Cited by 1
The value of morphology: osteoclast-like cells in soft tissue tumours
Khader, A. A., Seghetti, C., Oumlil, F., Tollit, A., Tirabosco, R., Amary, F. M., O’Donnell, P., & Flanagan, A. M. (2025) · Pathology & Oncology Research · doi:10.3389/pore.2025.1612175 · Cited by 1
PatHME: Hierarchical Multi-Expert Knowledge Distillation for Whole Slide Image Analysis
Tan, J., Kim, G., & Jeong, W. (2025) · IEEE Access · doi:10.1109/access.2025.3610117 · Cited by 1
“Single-scale multi-expert model: GPFM [7] is trained using multiple pre-trained expert models (UNI [6], Phikon [28], and CONCH [13]) for knowledge distillation on single-scale patches.”
InfluentialHao, X., Xu, H., Wang, X., Wang, T., Hämäläinen, T., & Cong, F. (2025) · Engineering Applications of Artificial Intelligence · doi:10.1016/j.engappai.2025.112302 · Cited by 1
Bridging imaging and genomics: Domain knowledge guided spatial transcriptomics analysis
Zhang, W., Liu, X., Chen, T., Xu, W., Sakal, C., Nie, X., Wang, L., & Li, X. (2025) · Information Fusion · doi:10.1016/j.inffus.2025.103746 · Cited by 1
ConVLM: Context-guided vision-language model for fine-grained histopathology image classification
Sohail, A., Ganapathi, I. I., Alawode, B., Javed, S., Bennamoun, M., & Mahmood, A. (2025) · Information Fusion · doi:10.1016/j.inffus.2025.103737 · Cited by 2
Robust Sensitivity Control in Digital Pathology via Tile Score Distribution Matching
Pignet, A., Klein, J., Robin, G., & Olivier, A. (2025) · Lecture notes in computer science · doi:10.1007/978-3-032-04978-0_54 · Cited by 1
Xu, S., Hou, J., & Chen, H. (2025) · Lecture notes in computer science · doi:10.1007/978-3-032-04978-0_21 · Cited by 2
Revisiting Automatic Data Curation for Vision Foundation Models in Digital Pathology
Chen, B., Vincent-Cuaz, C., Schoenpflug, L. A., Madeira, M., Fournier, L. M., Subramanian, V., Andani, S., Ruipérez-Campillo, S., Vogt, J. E., Luisier, R., et al. (2025) · Lecture notes in computer science · doi:10.1007/978-3-032-04978-0_53 · Cited by 1
Self-citationBio2Vol: Adapting 2D Biomedical Foundation Models for Volumetric Medical Image Segmentation
Zhuang, J., Wu, L., Ni, X., Wang, X., Wang, L., & Chen, H. (2025) · Lecture notes in computer science · doi:10.1007/978-3-032-04978-0_3 · Cited by 1
Probabilistic Integration of Renal Cancer Radiology and Pathology Using Graph Neural Networks
Gao, S., Gao, S., Machado, I. P., & Crispin‐Ortuzar, M. (2025) · Lecture notes in computer science · doi:10.1007/978-3-032-05162-2_54 · Cited by 1
HiLa: Hierarchical Vision-Language Collaboration for Cancer Survival Prediction
Cui, J., Wen, L., Fei, Y., Лю, Б., Zhou, L., Shen, D. G., & Wang, Y. (2025) · Lecture notes in computer science · doi:10.1007/978-3-032-04971-1_23 · Cited by 3
PathoCellBench: A Comprehensive Benchmark for Cell Phenotyping
Lüscher, J., Koreuber, N., Franzen, J., Reith, F. H., Winklmayr, C., Baumann, E., Schürch, C. M., Kainmueller, D., & Rumberger, J. L. (2025) · Lecture notes in computer science · doi:10.1007/978-3-032-04981-0_39 · Cited by 1
Xue, S., Wang, C., Fan, X., & Min, W. (2025) · Lecture notes in computer science · doi:10.1007/978-3-032-05169-1_29 · Cited by 1
Le, H., Coudray, N., Yeaton, A. H., Ramaswami, S., Karimkhan, A., Cheng, W., Cai, J. J., Yang, T.-H. O., Punekar, S. R., Chiriboga, L. A., et al. (2025) · Scientific Reports · doi:10.1038/s41598-025-17117-8 · Cited by 1
Pathological classification of non-ischaemic dilated cardiomyopathy based on deep learning
Jia, H., Wang, Y., Lv, Z., Zhao, Y., Zhang, N., Zhang, X., Wang, W., Feng, Y., Wang, W., Cui, H., et al. (2025) · European Heart Journal - Digital Health · doi:10.1093/ehjdh/ztaf113 · Cited by 1
“12 To address variability in slide staining, we applied data augmentation techniques, including random colour jitter and contrast jitter, to simulate commonly observed staining variations.”
Unbiased self supervised learning of kidney histology reveals phenotypic and prognostic insights
Pandit, K., Coudray, N., Quiros, A. C., Surapaneni, A., Upadhyay, D., Vanguri, R., Hirohama, D., Mohandes, S., Schlosser, P., Thiessen‐Philbrook, H., et al. (2025) · Scientific Reports · doi:10.1038/s41598-025-19193-2 · Cited by 1
“To assess the current SSL algorithm against off-the-shelf publicly available models, a comparative analysis was conducted against pre-trained ResNet-50 and UNI 24–26 .”
Age-informed, attention-based weakly supervised learning for neuropathological image assessment
Li, S., Malamut, M., McKee, A. C., Cherry, J. D., & Tian, L. (2025) · Brain Informatics · doi:10.1186/s40708-025-00276-9 · Cited by 1
Super resolution of pathology images with hierarchical feature integration and local image patterns
Xu, F., Li, L., Wang, S., Ling, R., Zhang, X., Deng, X., Zhou, M., Ling, J., & Gao, C. (2025) · The Journal of Pathology · doi:10.1002/path.6482 · Cited by 1
“We applied the UNI model pretrained on the Mass-100 K dataset [32] as it has been reported to generalize robustly across diverse tissue types.”
Deep learning finds convergent melanocytic morphology despite noisy archival slides
Tada, M., Gaskins, G., Ghandian, S., Mew, N. C., Keiser, M. J., & Keiser, E. (2025) · Cell Reports Methods · doi:10.1016/j.crmeth.2025.101201 · Cited by 1
Kaur, M., Kumar, A., Ceccarelli, M., Mall, R., & Gupta, S. (2025) · International Workshop on Multimedia Computing for Health and Medicine (MCHM) · doi:10.1145/3728424.3760768 · Cited by 1
The use of artificial intelligence methods in pathology
Damian, V., Palichenka, U., Koblížek, M., Škapa, P., & Zámečník, J. (2025) · Česko-slovenská pediatrie · doi:10.55095/cspediatrie2025/043 · Cited by 1
Valesano, A. L., Skala, S. L., & Yousif, M. Q. (2025) · Journal of Pathology Informatics · doi:10.1016/j.jpi.2025.100522 · Cited by 1
“17 We trained aMIL models based on these feature bags for binary classi fi cation of benign or STIC/STIL with a PyTorch backend.”
Chu, H., Chang, Z., Shao, Z., Deng, P., Wang, Y., Guo, X., Wang, Y., Yang, J., Guan, T. M., & Li, X. (2025) · Biomedical Signal Processing and Control · doi:10.1016/j.bspc.2025.108699 · Cited by 1
Predicting Immunotherapy Outcomes in NSCLC Using RNA and Pathology from Multicenter Clinical Trials
Wang, Z., Fang, Y., Huang, X., Ma, G., Mao, Q., Lu, X., Rong, G., Yu, Y., Wang, Y., Huang, Z., et al. (2025) · Advanced Science · doi:10.1002/advs.202502037 · Cited by 1
Self-citationMCS-Stain: Boosting FFPE-to-HE Virtual Staining With Multiple Cell Semantics
Hu, Y., Du, Z., Lin, W., Yang, S., Yu, L., Zhang, G., & Wang, L. (2025) · IEEE Transactions on Medical Imaging · doi:10.1109/tmi.2025.3628174 · Cited by 1
“For instance, metrics designed in the feature space of pathology-specific foundation models, such as PLIP [54], UNI [55], and CONCH [56], may provide more reliable and domain-relevant assessments than that computed from natural image–trained networks (e.g., VGG, Inception-Net).”
Medina, S., Romero, E., Cruz-Roa, Á., & Gonzalez, F. A. (2025) · PLoS ONE · doi:10.1371/journal.pone.0335826 · Cited by 1
“More remarkably, when compared to foundation models like UNI [33] that achieve 𝜅 = 0 .”
Universal and transferable attacks on pathology foundation models using microscopic perturbations
Wang, Y., Yang, X., Shen, C., Dong, S., Pillar, N., & Özcan, A. (2026) · Light Science & Applications · doi:10.1038/s41377-026-02347-w · Cited by 2
“We primarily used a foundation model pool containing the following seven pathology foundation models directly accessed from HuggingFace: [ MahmoodLab/UNI2-h, prov-gigapath/prov-gigapath, paige-ai/Virchow2, paige-ai/ Virchow, bioptimus/H-optimus-0, bioptimus/H0-mini, bioptimus/H-optimus-1 ] 29 – 35 .”
InfluentialPathViT Model for Automated Disease Classification from Skeletal Muscle Histopathology
Akan, T., Alp, S., Aishwarya, R., Xing, D. G., Dicharry, D., Bhuiyan, M. S., & Bhuiyan, M. A. N. (2025) · American Journal Of Pathology · doi:10.1016/j.ajpath.2025.10.009 · Cited by 1
“14,15 The use of deep-learning algorithms in digital pathology has shown promising results.”
Yan, C., Li, L., Qian, X., Ou, Y., Huang, Z., Ruan, Z., Xiang, W., Liu, H., & Liu, J. (2025) · Advanced Science · doi:10.1002/advs.202510307 · Cited by 1
Chinese expert opinion on AI-based identification of lymph node metastases in solid tumors
Wang, X., Tang, H., Chen, Y., Zhang, X., Liu, X., Zhu, M., Ni, C.-R., & Yu, G. (2025) · Journal of Histotechnology · doi:10.1080/01478885.2025.2580157 · Cited by 1
“Models like UNI [13], CONCH [14], and PathChat [15] have already demonstrated potential in complex pathology image analysis and interactive diagnostic support.”
Kockwelp, J., Kliesch, S., Gromoll, J., Wistuba, J., & Risse, B. (2025) · Computers in Biology and Medicine · doi:10.1016/j.compbiomed.2025.111270 · Cited by 1
Foundation Models for Hepatocellular Carcinoma: Challenges in Generalization under Data Scarcity
Corso, G., Lovino, M., Akpinar, R., Tommaso, L. D., Ficarra, E., & Ranzini, M. B. M. (2025) · International Joint Conference on Neural Networks (IJCNN) · doi:10.1109/ijcnn64981.2025.11228716 · Cited by 1
“Foundation models have found broad use as feature extractors in histopathology [1], [4], [5], however, adapting the feature aggregation for the medical task at hand (MIL component) is not always straightforward.”
Influential
Metadata
- Authors
- Chen, R. J., Ding, T., Lu, M. Y., Williamson, D. F. K., Jaume, G., Song, A. H., Chen, B., Zhang, A., Shao, D., Shaban, M., Williams, M., Oldenburg, L., Weishaupt, L. L., Wang, J. J., Vaidya, A., Le, L. P., Gerber, G., Sahai, S., Williams, W., & Mahmood, F.
- Affiliations
- Broad Institute (US)
- Brigham and Women's Hospital (US)
- Harvard University (US)
- Corresponding author
- Mahmood, Faisal
- ORCID
- Venue
- Nature Medicine
- Published
- 2024-03
- Volume
- 30
- Issue
- 3
- Pages
- 850-862
- ISSN
- 1078-8956 (ISSN-L)
- 1546-170X
- PMC
- PMC11403354
- OpenAlex
- W4392947521
- Semantic Scholar
- cac496ddebbff6bbba78242780cde01bd962bfc9
- Cited by
- 1,560
- References
- 170
- Citations by year
- 2026: 789 · 2025: 613 · 2024: 154
- FWCI
- 229.15
- Type
- journal article
- Language
- en
- Open access
- green
- www.ncbi.nlm.nih.gov
- Licence
- Locations
- Nature Medicine (publishedVersion)
- PubMed (publishedVersion)
- PubMed Central (acceptedVersion)
- Funders
- National Human Genome Research Institute (T32HG002295)
- National Cancer Institute (T32CA251062)
- National Institute of General Medical Sciences (R35GM149270)
- Topics
- Physical Sciences › Computer Science › Artificial Intelligence › AI in cancer detection
- Health Sciences › Medicine › Radiology, Nuclear Medicine and Imaging › Radiomics and Machine Learning in Medical Imaging
- Physical Sciences › Computer Science › Artificial Intelligence › Domain Adaptation and Few-Shot Learning
- Keywords
- computational pathology, whole slide images, self-supervised learning, foundation models, unsupervised feature learning, tissue classification, few-shot learning, cancer subtyping, histopathology, transfer learning
- MeSH terms
- Artificial Intelligence, Workflow
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
