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 901–1,000 of 1,003
The Neurosurgical Uncertainty Index: Self-Doubting AI for rare or unexpected surgical complicationsPreprint
Thiong’o, G. M., & Ogundokun, A. (2025) · medRxiv · doi:10.1101/2025.05.01.25326833
Preclinical Evaluation of an Interactive Image Search System of Oral Pathology
Herdiantoputri, R. R., Komura, D., Ochi, M., Fukawa, Y., Oba, K., Tsuchiya, M., Kikuchi, Y., Matsuyama, Y., Ushiku, T., Ikeda, T., et al. (2025) · Journal of Dental Research · doi:10.1177/00220345251329042 · Cited by 1
“State-of-the-art models trained on histopathological images, along with advanced vision–language foundation models for histology (Kang et al. 2023; Chen et al. 2024), show promise in enhancing retrieval accuracy and semantic understanding, potentially improving diagnostic precision and efficiency…”
Liu, A., & Zeng, P. (2025) · International Conference on Electronic Information Engineering and Data Processing (EIEDP) · doi:10.1117/12.3067471
A Foundation Model for Cell Segmentation from Immunohistochemistry: A Pilot Study
Li, W., Luo, X., Yang, Z., Jia, X., & Li, Z. (2025) · IEEE International Symposium on Biomedical Imaging (ISBI) · doi:10.1109/isbi60581.2025.10980902
“Then, we utilize the pre-trained feature extraction network UNI[2] to extract features, obtaining the feature vector for each patch as f n = g θ x in .”
Ensemble of Foundation Models for Pap Smear Cell Analysis
Jiang, H., Cai, J., Xu, M., Fang, Y., Zhang, L., & Wang, Q. (2025) · IEEE International Symposium on Biomedical Imaging (ISBI) · doi:10.1109/isbi60581.2025.10981009 · Cited by 1
“To overcome this limitation, we employ low-rank adaptation (LoRA) to fine-tune foundational pathological models [4, 5, 6] that have been pretrained from expansive, unlabeled medical histopathology datasets.”
Curd, A., Slaney, H., Brockmoeller, S. F., Miklosi, A.-G., Arora, D. S., Lewington, A. J. P., Felce, J. H., Nahhas, O. S. M. E., van Treeck, M., Kather, J. N., et al. (2025) · IEEE International Symposium on Biomedical Imaging (ISBI) · doi:10.1109/isbi60581.2025.10980913
Ding, R., Li, L., Soans, R. E., Shah, T., Krishnan, R., Sze, M. A., Lukyanov, S., Deshpande, Y., & Chen, A. (2025) · IEEE International Symposium on Biomedical Imaging (ISBI) · doi:10.1109/isbi60581.2025.10980783
“The first group is 1024 deep features extracted from a self-supervised model for digital pathology called UNI [11].”
Allen, K. E., Breen, J., Hall, G. D., Mappa, G., Zucker, K., Ravikumar, N., & Orsi, N. M. (2025) · Cancers · doi:10.3390/cancers17111789
“Patch features were extracted using the UNI foundation model, a self-supervised model that was pre-trained on over 100 million images from H&E-stained WSIs of 20 major tissue types [36].”
Zuo, Y., Wang, L., Liu, J., Feng, S., Wang, Y., Zhu, Q., Shao, W., & Zhang, D. (2025) · IEEE Transactions on Medical Imaging · doi:10.1109/tmi.2025.3575801
“V, VI, and VII, we replace the original feature learner with two feature extractors specialized for medical images, observing slight performance improvements with both CONCH [65] and UNI [66].”
Novel biomarkers for guiding treatment for prostate cancer
Yang, D. D., & Nguyen, P. L. (2025) · Seminars in Radiation Oncology · doi:10.1016/j.semradonc.2025.04.002 · Cited by 2
Yong, A. S. K., Husna, N., Tan, K. H., Manek, G., Sim, R. R. Y., Loi, R. M. L., Lee, O. Z. J., Tang, S. S., Soon, G. S. T., Chan, D. K. H., et al. (2025) · medRxiv · doi:10.1101/2025.06.11.25329397
“Three open-weights foundation models (FMs) for pathology, namely UNI2( 18 ), Phikon-v2( 19 ) and Virchow2( 20 ) were obtained and used to generate embeddings for FUSE images.”
Applying Self-Supervised Tissue Patch Encoders to Enable Cellular Graph EncodingPreprint
Ding, Z., & Baras, A. S. (2025) · bioRxiv (Cold Spring Harbor Laboratory) · doi:10.1101/2025.06.10.658651
Zec, A., Declich, M. M., Sacco, M., Trovò, F., Lorenzini, D., Prelaj, A., & Miskovic, V. (2025) · IFIP advances in information and communication technology · doi:10.1007/978-3-031-96235-6_1
Digital Pathology Predicts Immunotherapy Response and Maps Spatial Biomarkers in Gastric CancerPreprint
Noh, M., Lee, J. H., Kim, T., Lee, S. H., Cha, S., Choi, Y. J., Choi, J., Choi, Y.-D., Kim, S., Choi, J.-I., et al. (2025) · Research Square · doi:10.21203/rs.3.rs-6772872/v1
Informatics at the Frontier of Cancer Research
Noller, K., Botsis, T., Cámara, P. G., Ciotti, L., Cooper, L., Goecks, J., Griffith, M., Haas, B. J., Ideker, T., Karchin, R., et al. (2025) · Cancer Research · doi:10.1158/0008-5472.can-24-2829 · Cited by 2
“Whereas the performance of tools for diagnostic, prognostic, and treatment predictive tasks have steadily increased, due in part to the development of pathology foundation models (58–60), clinical adoption of these tools remains low.”
Faure, G., Soglio, D. B. D., Patey, N., Oligny, L. L., Girard, S., & Séoud, L. (2025) · Computers in Biology and Medicine · doi:10.1016/j.compbiomed.2025.110634
Liu, F., Wu, X., Yu, S., Cheng, D., Pan, J., Wang, X., Yu, G., & Li, C. (2025) · Frontiers in Oncology · doi:10.3389/fonc.2025.1592044
Distilling Genomic Knowledge into Whole Slide Imaging for Glioma Molecular Classification
Chen, Q., Xu, H., Hao, X., Zhang, Q., Qin, H., Kärkkäinen, T., & Cong, F. (2025) · Proceedings - IEEE Symposium on Computer-Based Medical Systems · doi:10.1109/cbms65348.2025.00112
Brahim, I., Zeghlache, R., Frutoso, M., Rochais, M. L., Bandeira, M., Romão, V. C., Asselineau, J., Jaloustre, H., Hillion, S., Saraux, A., et al. (2025) · SSRN Electronic Journal · doi:10.2139/ssrn.5333602
PlasmoStage: A Hierarchical Deep Learning Framework for Plasmodium Parasite Staging in Malaria
Salem, M., Abouzahir, S., Ghedira, H., Saddik, A. E., & Yaqub, M. K. (2025) · IEEE International Symposium on Medical Measurements and Applications (MeMeA) · doi:10.1109/memea65319.2025.11067957
“In the context of medical image analysis, foundation models have shown great promise for tasks such as disease diagnosis, organ segmentation, and image retrieval [15].”
Clinical Stage Prompt Induced Multi-Modal Prognosis
Jin, T., Xie, X., Li, Q., Li, X., & Wang, Y. (2025) · IEEE Transactions on Medical Imaging · doi:10.1109/tmi.2025.3588836
“UNI [54] is a versatile self-supervised model for pathology, pre-trained on over 100 million images derived from more than 100,000 diagnostic H&E-stained whole slide images, and CTransPath [55] is pre-trained on massively unlabeled histopathological images, including around 15 million patches…”
Zhang, X., Wang, S., Gu, Q., Fu, Y., Li, H., Gan, J., Du, J., Chu, L., Li, X., Wang, C., et al. (2025) · JMIR Cancer · doi:10.2196/79227
Spatia: Multimodal Model for Prediction and Generation of Spatial Cell Phenotypes.Preprint
Kong, Z., Qiu, M., Boesen, J., Lin, X., Yun, S., Chen, T., Kellis, M., & Žitnik, M. (2025) · PubMed
AMKD: Adaptive Multi-modality Knowledge Distillation for Pathological Survival Analysis
Xu, Y., Cai, L., Wang, Y., Cheng, H., Liu, F., Wang, R., & Zhang, Y. (2025) · Lecture notes in computer science · doi:10.1007/978-981-95-0033-8_35
Sáez-Maldonado, F. J., García, L., Cooper, L., Goldstein, J. A., Molina, R., & Katsaggelos, A. K. (2025) · IEEE Access · doi:10.1109/access.2025.3593420 · Cited by 2
“Patch features are then extracted utilizing two different pre-trained models: Resnet50 with Barlow Twins ( BT ) self-supervised learning [49], using the weights provided in [50], and the general-purpose self-supervised foundation model for pathology UNI [51].”
Predicting Multiomics Histopathology Features with Surface Parameterizations
Huang, K., Tan, Z.-H., Yu, W., Wang, X., Ding, Y., Shi-hui, X., Chen, Z., Zhang, Y., Liu, Y., Lin, W., et al. (2025) · Engineering · doi:10.1016/j.eng.2025.07.026
Molecularly informed analysis of histopathology images using natural languagePreprint
Schaefer, M., Nonchev, K., Awasthi, A., Burton, J., Koelzer, V. H., Rätsch, G., & Bock, C. (2025) · bioRxiv (Cold Spring Harbor Laboratory) · doi:10.1101/2025.07.14.664402
“Representative examples of this approach include PLIP (Huang et al., 2023) and CONCH (Lu et al., 2024), which we use here as baseline methods.”
InfluentialMiroSCOPE: An AI-driven digital pathology platform for annotating functional tissue unitsPreprint
Fenner, M. R., Sevim, S., Wu, G., Beavers, D., Guo, P., Tang, Y., Eddy, C. Z., Ait-Ahmad, K., Rice-Stitt, T. L., THOMAS, G. V., et al. (2025) · bioRxiv (Cold Spring Harbor Laboratory) · doi:10.1101/2025.07.11.664228
BioX-CPath: Biologically-driven Explainable Diagnostics for Multistain IHC Computational Pathology
Gallagher-Syed, A., Senior, H., Alwazzan, O., Pontarini, E., Bombardieri, M., Pitzalis, C., Lewis, M., Barnes, M. R., Rossi, L., & Slabaugh, G. (2025) · IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings - IEEE Computer Society Conference on Computer Vision and Pattern Recognition/Proceedings · doi:10.1109/cvpr52734.2025.00970
Howard, F. M., Dolezal, J. M., Hieromnimon, H. M., Venters, S. J., Kochanny, S. E., Li, A., Borowsky, A. D., Symmans, W. F., Wolf, D. M., Swigart, L. B., et al. (2025) · medRxiv · doi:10.1101/2025.08.25.25334393 · Cited by 3
Dong, Z., McIntosh, C., & Schwartz, G. W. (2025) · bioRxiv (Cold Spring Harbor Laboratory) · doi:10.1101/2025.08.27.672636
“…process, recent advances in deep learning have lead to the development of high-performing foundational cell segmentors 7,8 and feature extractors 9–11 to segment cells and extract features from H&E images for downstream tasks such as diagnosis and survival prediction 12–14 , suggesting the…”
Noman, S. M., Henedak, M. T., & Elattar, M. (2025) · Machine Learning Science and Technology · doi:10.1088/2632-2153/ae0557
Dashtiahangar, F., & Yuan, J. S. (2025) · IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops (CVPRW) · doi:10.1109/cvprw67362.2025.00414
“These techniques have successfully been applied to predict gene expression, protein abundance, and other molecular signatures from his-045 tology, setting the stage for AI-driven virtual multi-omics profiling [8].”
Zhao, X., Zhou, J., Zhang, Y., Li, Y., Wang, L., Teng, G., Li, X., Yang, S. T., Liu, H., Li, F., et al. (2025) · Research Square · doi:10.21203/rs.3.rs-7440962/v1
Predicting Patient Self-reported Race From Skin Histological Images with Deep Learning
Chen, S., Verma, R., Clare, K., Jegminat, J., Alleva, E., Huang, K., Veremis, B. M., Fuchs, T. J., & Campanella, G. S. V. (2025) · Lecture notes in computer science · doi:10.1007/978-3-032-05870-6_11
“Tissue tiles were extracted at 20x magnification, and tile-level embeddings were generated using four pretrained FMs: SP22M [25], UNI [26], GigaPath [27], and Virchow [28], followed by an attention-based MIL (AB-MIL) model [18] for slide-level aggregation.”
Lin, Y., Lin, W., Guo, C., Yang, X., Meng, H., & Wang, L. (2025) · Lecture notes in computer science · doi:10.1007/978-3-032-04978-0_63
How Fair are Foundation Models? Exploring the Role of Covariate Bias in Histopathology
Shafique, A., Dy, A., Qin, X., Alshamlan, N., Done, S. J., Androutsos, D., & Khademi, A. (2025) · Lecture notes in computer science · doi:10.1007/978-3-032-05870-6_4 · Cited by 1
Explainable Classifier for Malignant Lymphoma Subtyping via Cell Graph and Image Fusion
Nishiyama, D., Miyoshi, H., Hashimoto, N., Ohshima, K., Hontani, H., Takeuchi, I., & Sakuma, J. (2025) · Lecture notes in computer science · doi:10.1007/978-3-032-05162-2_31
Yu, L., Ma, T., Zhao, W., Liang, Z., & Yu, L. (2025) · Lecture notes in computer science · doi:10.1007/978-3-032-05162-2_47 · Cited by 1
Graph Laplacian Transformer with Progressive Sampling for Prostate Cancer Grading
Junayed, M. S., Vessem, J. D. V., Wan, Q., Nam, G., & Nabavi, S. (2025) · Lecture notes in computer science · doi:10.1007/978-3-032-05162-2_35
“The IRM operates in two stages: (1) local feature extraction using ResNet50 [11] and (2) context-aware scoring using a frozen FM, specifically the UNI model [7].”
Sahai, S., Ramos-Guerra, A. D., Almagro-Pérez, C., Jaume, G., Zhang, A., Rennke, H. G., Weins, A., Ortuño, J. E., Ledesma‐Carbayo, M. J., & Mahmood, F. (2025) · Lecture notes in computer science · doi:10.1007/978-3-032-04984-1_9
Self-citationKoga, S., Guda, A., Wang, Y., Sahni, A., Wu, J. Z., Rosen, A., Nield, J., Nandish, N., Patel, K., Goldman, H. M., et al. (2025) · medRxiv · doi:10.1101/2025.09.16.25335833
A Flexible Deep Learning Framework for Survival Analysis with Medical Data
Campanella, G. S. V., Häggström, I., Kook, L., Hothorn, T., & Fuchs, T. J. (2025) · Lecture notes in computer science · doi:10.1007/978-3-032-05182-0_1
Hierarchical Self-supervised Adversarial Training for Robust Vision Models in Histopathology
Malik, H. S., Kunhimon, S., Naseer, M., Khan, F. S., & Khan, S. H. (2025) · Lecture notes in computer science · doi:10.1007/978-3-032-04927-8_23
MeDi: Metadata-Guided Diffusion Models for Mitigating Biases in Tumor Classification
Drexlin, D. J., Dippel, J., Hense, J., Prenißl, N., Montavon, G., Klauschen, F., & Müller, K. R. (2025) · Lecture notes in computer science · doi:10.1007/978-3-032-05185-1_37
Chen, J., Liu, P., Chen, C., Su, Y., Wang, J., Chen, C., Ai, X., & Lv, X. (2025) · Interdisciplinary Sciences Computational Life Sciences · doi:10.1007/s12539-025-00744-0
“Finally, the histopathologically pre-trained visual coder UNI [31], which was pre-trained on WSI's private dataset, was used to perform feature extraction on a large number of patches.”
Self-citationYadav, A., Alvarez, K., Yip, K. Y.-L., Ruppin, E., Maher, J. C. Y., Lobo-Gomez, V., Kumsta, C., & Sinha, S. (2025) · bioRxiv (Cold Spring Harbor Laboratory) · doi:10.1101/2025.09.29.679316
FUSED: Cross-Domain Integration of Foundation Models for Cancer Drug Response PredictionPreprint
Rössner, T., Balke, J., & Tang, M. (2025) · bioRxiv (Cold Spring Harbor Laboratory) · doi:10.1101/2025.09.30.679434
“In the biomedical domain, pre-trained FMs for medical images [1, 2], drug molecules [3, 4], single-cells [5, 6] and many other modalities are increasingly applied and systematically explored.”
Automated Computational Pathology to Assess Degenerative Disc Histology
Bell, R. D., Reddick, E. C., Lillyman, D. J., Lee, F. S., & Wachs, R. A. (2025) · JOR Spine · doi:10.1002/jsp2.70119
“Further, histopathology foundation models, or models that have been trained on very diverse datasets, have recently been developed and show great promise for computational pathology tasks [51, 52].”
Yang, X., Ruan, J., & Yue, J. (2025) · International Conference on Computer Vision, Application, and Algorithm (CVAA) · doi:10.1109/cvaa66438.2025.11193255
“Specifically, given an input WSI, it is first divided into multiple image patches, and instance features are extracted using a pretrained encoder UNI [16].”
KI in der Diagnostik – aus Sicht der Pathologie
Schulz, S., Jesinghaus, M., & Foersch, S. (2025) · Die Onkologie · doi:10.1007/s00761-025-01849-3
Blondel, A., Krucker, C., Harter, V., Silva, M. D., Groeneveld, C. S., de Reyniès, A., Karimi, M., Benhamou, S., Bernard‐Pierrot, I., Pfister, C., et al. (2025) · bioRxiv (Cold Spring Harbor Laboratory) · doi:10.1101/2025.10.23.684013
Web engine for tumor pathology image retrievals on massive scalesPreprint
Zhong, Z., Huang, J., Li, X., Song, L., Gong, L., Ru, B., Paul, R., Levine, J., Zhang, Y., Aldape, K., et al. (2025) · bioRxiv (Cold Spring Harbor Laboratory) · doi:10.1101/2025.10.25.684566
Pathomic Signature for Stratification of Metastasis Risk in Colon Cancer
Gil, D., Arenos, N., & Musulén, E. (2025) · Communications in computer and information science · doi:10.1007/978-3-032-10209-6_16
Toward Linear Representations of Foundation Models for Histopathology Image Retrieval
Santos, M., Borja, I., Lima, D., de Oliveira, G. F., Duarte, A. A., dos‐Santos, W. L. C., & de Oliveira, L. R. (2025) · SIBGRAPI Conference on Graphics, Patterns and Images (SIBGRAPI) · doi:10.1109/sibgrapi67909.2025.11223348
“…the DINOv2 [18] self-supervised protocol and is pre-trained on MASS-100K, a corpus of aproximately 100,000 H&E WSIs; its larger sibling UNI2-h [14] uses a ViT-h/14-reg8 [19] backbone and an even broader slide collection, but we note that both variants underperform on glomerular tissue, which…”
Quo Vadis Pathology? Advancing Glomerular Lesion Classification with Foundation Models
Lima, D., de Oliveira, G. F., Duarte, A. A., dos‐Santos, W. L. C., & de Oliveira, L. R. (2025) · SIBGRAPI Conference on Graphics, Patterns and Images (SIBGRAPI) · doi:10.1109/sibgrapi67909.2025.11223393
“Particularly in our work, the FMs selected – UNI [11], UNI2 [11], Phikon [12], and PhikonV2 [13] – represent cutting-edge efforts in developing large-scale and generalizable AI for histopathology, and based primarily on ViTs.”
InfluentialReliable Cervical Cancer Screening Using Cytology Data Through a Histopathology Foundation Model
Tchokponhoue, G. A. D., & Idri, A. (2025) · Lecture notes in networks and systems · doi:10.1007/978-3-031-97119-8_50
Contrasting low- and high-resolution features for HER2 scoring using deep learning
Chauhan, E., Sharma, A., Sharma, A., Nishadham, V., Padariya, K. V., Ghughtyal, A., Kumar, A., Gupta, G., Mehta, A., Jawahar, C. V., et al. (2025) · Journal of Pathology Informatics · doi:10.1016/j.jpi.2025.100529
“Low-resolution images may fail to capture fi ne details necessary for precise classi fi cation, potentially degrading performance.”
Kates-Harbeck, D., Kreipe, H. H., Gluz, O., Christgen, M., Küemmel, S., Graeser, M., Nitz, U. A., Mahner, S., Mayr, D. C., Wuerstlein, R., et al. (2025) · JCO Clinical Cancer Informatics · doi:10.1200/cci-24-00287
Jiang, Z., Lin, H., Zhao, Z., Bai, X., Su, C., Sun, K., Zhang, Z., Fu, W., & Zhou, X. (2025) · Journal of Gastroenterology · doi:10.1007/s00535-025-02326-9
Yuan, Y., Yue, Q., Wang, Y., Du, Z., Wang, X., Yan, F., Cheng, H., Chen, K., Wu, Z., Song, J., et al. (2025) · Nature Communications · doi:10.1038/s41467-025-66853-y · Cited by 1
“Unlike prior studies demonstrating their validity in FFPE-based tasks 23 – 26 , we found that directly transferring a pathological foundation model to cryo-section images yielded suboptimal performance (Fig.”
Deep Learning for Clinical Ultrasound Imaging: From Supervised Approaches to Foundation Models
He, K., Wei, D., & Ranger, B. J. (2025) · IEEE Journal of Biomedical and Health Informatics · doi:10.1109/jbhi.2025.3636830
“Vision-language foundation models have demonstrated promising results in domains such as computational pathology [175], [176].”
FigATree: a novel framework for histological subtyping and grading of lung adenocarcinoma
Huang, Q., Zhang, J., He, Q., Wang, S., Zhang, Q., Lin, L., Yu, X., Wang, Y., He, Y., Chen, X., et al. (2025) · Archiv für Pathologische Anatomie und Physiologie und für Klinische Medicin · doi:10.1007/s00428-025-04355-4
Chen, Y., Chan, T. H., Chen, J., Li, L., Yin, G., & Yu, L. (2025) · IEEE Transactions on Medical Imaging · doi:10.1109/tmi.2025.3638243
“For a fair and complete comparison of di ff erent models, we use the same pretrained image encoder for feature extraction on all slides and evaluate all models on three widely used feature extractors, i.e., UNI2-H [45], KimiaNet [33] and CONCH [41].”
Jessen, A., Blattgerste, C., Legnar, M., Popović, Z. V., Porubský, Š., & Weis, C. (2025) · Archiv für Pathologische Anatomie und Physiologie und für Klinische Medicin · doi:10.1007/s00428-025-04351-8
Belete, M. K., Thakkar, N., Khatri, I., Guan, M., Sun, Y., Muthuswamy, A., Kolder, I. C. R. M., Si, H., Soong, D., Higgs, B. W., et al. (2026) · Translational Oncology · doi:10.1016/j.tranon.2026.102822
“We evaluated the following feature extractors: PLIP [30], UNI2 [31], H-optimus-0 [32], and Virchow2 [33] to determine which patch-level feature extraction embeddings were most informative in predicting PFS and stratifying risk.”
Liu, H., Xu, H., Hao, X., Zhang, M., Lu, C., Pölönen, I., & Cong, F. (2025) · Annual International Conference of the IEEE Engineering in Medicine and Biology Society (EMBC) · doi:10.1109/embc58623.2025.11254327
“UNI-HoverNet enhances Hover-Net by integraing a UNI foundation model [12] into the encoder and integrating Squeeze-and-Excitation (SE) modules [13] with multi-scale encoder features into the decoder.”
MuCoSA: Multi-contextual similarity assessment for histopathology image search
Kim, G. Y., Jeon, Y., Jeong, H., Lee, S., Kim, H., Lee, B., Jung, K.-H., Ho, D. J., & Choi, Y. (2025) · Journal of Pathology Informatics · doi:10.1016/j.jpi.2025.100533
MPT-MIL: Multimodal Aware Prompt Tuning for Prediction of Cancer Survival
Zhang, R., Fang, M., Liu, S., Wang, Z., Xie, X., Ma, X., Tian, J., & Dong, D. (2025) · IEEE Journal of Biomedical and Health Informatics · doi:10.1109/jbhi.2025.3640398
“Many foundational models based on self-supervised pre-training, either in pathology or natural images, are already in use as pre-trained feature extractors [8], [9], [10], [11].”
Deep Bayesian Blind Color Deconvolution Using a Gaussian Scale Mixture Prior
Castro-Macías, F. M., Miguel Vega, B., Pérez-Bueno, F., Molina, R., & Katsaggelos, A. K. (2025) · European Signal Processing Conference (EUSIPCO) · doi:10.23919/eusipco63237.2025.11226814
“Today, BCD techniques are essential in large-scale applications, such as developing foundation models for pathology and automating image diagnosis [4], [5].”
Balan, J., & Larson, N. B. (2025) · Scientific Reports · doi:10.1038/s41598-025-30947-w
“The median [IQR] MMD for the H-optimus-0 between sample related samples was 0.013 [0.004, 0.031], and the unrelated sample pairs was 0.024 [0.014, 0.038].”
InfluentialMorphological Landscape Mapping Decodes Pathological Heterogeneity and Proteomic Programs in HCCPreprint
Wang, X., Cheng, Y., Peng, X., Li, J., Han, J., Yang, Z., Ding, G., Zhao, H., Qu, P., Sun, Q., et al. (2025) · medRxiv · doi:10.64898/2025.11.27.25341163
Self-citationFoundation models in oncology win benchmarks but miss the clinic
Duan, Z., Duan, Q., Wild, B., & Eils, R. (2025) · Nature Cancer · doi:10.1038/s43018-025-01071-5
KI in der Diagnostik – aus Sicht der Pathologie
Schulz, S., Jesinghaus, M., & Foersch, S. (2025) · Die Pathologie · doi:10.1007/s00292-025-01524-9
Xu, C., Yang, R., Li, W., Fu, X., Fang, L., Feng, Z., Zhang, X., Wang, C., Song, M., & Chen, W. (2025) · IEEE Transactions on Visualization and Computer Graphics · doi:10.1109/tvcg.2025.3634833
“With advances in artificial intelligence, numerous diagnostic algorithms and models for pathology image analysis have been developed and validated, demonstrating successful applications across multiple cancer types [7, 27].”
Deep Learning as a Diagnostic Discovery Engine in Solid Organ Transplant Pathology
Wu, J., Jiang, C., Fang, Z., Shang, G., Yang, C., & Zhu, D. (2025) · Organ medicine. · doi:10.1002/orm2.70024
Practical Trade-Offs in 3D Cancer Imaging
Dequiedt, L., & Kiemen, A. L. (2025) · Cancer Research · doi:10.1158/0008-5472.can-25-3336
CAMIL-A Class Activation Mapping Integrated with MIL-Attention
Córdoba-Vargas, D., Cano, F., Revelo-Fuelagán, J., & Romero, E. (2025) · International Symposium on Medical Information Processing and Analysis (SIPAIM) · doi:10.1109/sipaim67325.2025.11283315
“…extraction: Each patch was processed us-ing state-of-the-art models pretrained with ImageNet [20], including ConvNeXt-Tiny [21], EfficientNet [22], and Vision Transformer (ViT) [20], as well as the foundational model UNI2-H [23] pretrained with histopathological images from the TCGA repository.”
IHGAMP: Pan-cancer HRD prediction from routine H&E whole-slide images using foundation modelsPreprint
Zafar, S. A., Qin, W., Chengliang, L., Khan, A. A., Nazir, A., Batool, H., Khalid, F., & Faisal, M. S. (2025) · medRxiv · doi:10.64898/2025.12.19.25342650
Valieris, R., Martins, L. V. C., Defelicibus, A., Bueno, A. P., de Toledo Osorio, C. A. B., Carraro, D. M., Dias-Neto, E., Rosales, R., de Figueiredo, J. M. B., & da Silva, I. T. (2024) · Figshare · doi:10.6084/m9.figshare.26785956
Golfe, A., Meseguer, P., Naranjo, V., & Colomer, A. (2025) · Knowledge-Based Systems · doi:10.1016/j.knosys.2025.114423
Lin, W., Zhu, R., Hou, W., Wang, J., Lin, Y., Chen, R., Ta, N., & Wang, L. (2025) · Lecture notes in computer science · doi:10.1007/978-3-032-05185-1_16
Self-citationVasiljević, J., Veiga, I. B., Hahn, K., Schwalie, P. C., & Valdeolivas, A. (2025) · Lecture notes in computer science · doi:10.1007/978-3-032-06118-8_27
Rao, V. R., Workman, A. A., Lu, L., Liu, X., & Sukhadia, S. (2025) · bioRxiv (Cold Spring Harbor Laboratory) · doi:10.1101/2025.09.17.674367
A perspective on foundation models in intensive care medicine
Harris, C., Schmidgall, S., Rapuri, S., Hwang, K., Moor, M., & Stevens, R. D. (2026) · Artificial Intelligence in Medicine · doi:10.1016/j.artmed.2026.103515 · Cited by 1
Papagoras, K. A., Lund, O., & Barra, C. (2025) · bioRxiv (Cold Spring Harbor Laboratory) · doi:10.64898/2025.12.19.695372
“Following earlier work on this dataset, 14 we developed a simpler, and more resource-efficient pipeline, by utilizing UNI-2, 15 a pathology foundation model as a feature extractor and later used those representations to develop models in a reduced feature space.”
InfluentialPathQC: Determining Molecular and Structural Integrity of Tissues from Histopathological Slides
Sinha, R. K., Yadav, A., & Sinha, S. (2025) · Bioengineering · doi:10.3390/bioengineering13010005 · Cited by 1
“Patch-Level Feature Extraction: For deep-learning feature extraction, we process these normalized patches through the UNI encoder [10], a vision transformer architecture pre-trained on over 100 million tissue patches across multiple cancer types and normal tissues.”
Low-Resource Vision–Language Learning for Brain Organoid Mitotic Classification
Ulhaq, A., & Vallejo, J. R. (2025) · International Conference on Digital Image Computing: Techniques and Applications (DICTA) · doi:10.1109/dicta68720.2025.11302452
“The performance comparison shows that the proposed approach is competitive with SOTA foundation models, including CLIP [14], BiomedCLIP [22], and UNI [4].”
Panaccione, F. P., Sgaravatti, C., & Pinoli, P. (2026) · Lecture notes in computer science · doi:10.1007/978-3-032-11317-7_33
“We use UNI [3] as an Image Encoder, which offers a pretrained Vision Trans-former for histopathology tiles.”
WeakSupCon: Weakly Supervised Contrastive Learning for Encoder Pre-training
Zhang, B., Manoochehri, H., Li, X., Knudsen, B. S., & Taşdizen, T. (2026) · Lecture notes in computer science · doi:10.1007/978-3-032-13961-0_31
UniDis: Universal Distillation for Efficient and Personalized Pathology Diagnosis
Ma, R., Lu, M., & Xia, Y. (2026) · Lecture notes in computer science · doi:10.1007/978-3-032-09513-8_20
Self-citationEvaluating Foundation Models with Pathological Concept Learning for Kidney Cancer
Gao, S., Wang, S., Gao, Y., Wang, B., Zhuang, X., Warren, A. Y., Stewart, G., Jones, J. O., & Crispin‐Ortuzar, M. (2026) · Lecture notes in computer science · doi:10.1007/978-3-032-09569-5_4
Harness Behavioural Analysis for Unpacking the Bio-Interpretability of Pathology Foundation ModelsPreprint
Hu, Y., Batchkala, G., Gaitskell, K. L., Domingo, E., Li, B., Zhang, T., Li, Z., Friedrich, M., Woodcock, D. J., Verrill, C. L., et al. (2026) · Research Square · doi:10.21203/rs.3.rs-8090378/v1 · Cited by 1
A Local-Global Multi-Scale Feature Fusion Learning Network for Whole Slide Images Segmentation
Sun, J., Fei, H., Zhu, Z., Sun, Y. L., He, W., & Cai, H. (2025) · IEEE International Conference on Industrial Informatics (INDIN) · doi:10.1109/indin64977.2025.11279496
“Patch-based processing has become a widely adopted strategy in large-scale pathology foundation models, such as UNI, CONCH, CHIEF, and TITAN [10,11,12,13], where gigapixel WSIs are divided into smaller tiles for feature extraction and disease classification [14].”
Witte, A., Fuhlert, P., Lennartz, M., Baumbach, J., Huber, T. B., Bengtsson, E. W., Sauter, G., Zimmermann, M., & Bonn, S. (2026) · medRxiv · doi:10.64898/2026.01.06.25342815 · Cited by 3
Unambiguous signatures of malignancies extracted from images of growing cellsPreprint
Kalweit, G., Kalweit, M., Checinska, W., Saric, M., Berger, R., Bodurova-Spassova, E., Rawluk, J. K., Talvard-Balland, N., Klett, A., Follo, M., et al. (2026) · medRxiv · doi:10.64898/2026.01.10.26343803
Akbar, A., Levya, A., Esnakula, A. K., Hasanov, E., Noonan, A. M., Manne, U., Sahai, V., Meng, L., Tsai, S., Parwani, A. V., et al. (2026) · Research Square · doi:10.21203/rs.3.rs-8545050/v1
Wen, H., Wu, Y., Huang, D., & Liu, C. (2026) · Scientific Reports · doi:10.1038/s41598-026-36256-0
Wu, Y., Xie, X., Zhao, C., Song, P., Yin, F., Wu, G., & Yu, J. (2026) · IEEE Transactions on Image Processing · doi:10.1109/tip.2026.3654402
“We demonstrated the experimental results of the proposed method in the three datasets and compared them with SOTA methods that include standard DSMIL [9], TransMIL [16], DTFD [1], CLAM [2], UNI [42], RankMix [5], PMIL [43], PseMix [6] and Prov-GigaPath [44].”
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
