DeepSMILE: Contrastive self-supervised pre-training benefits MSI and HRD classification directly from H&E whole-slide images in colorectal and breast cancer
Open accessThe DOI matches its Crossref record; 24 of the 24 references it cites were found in a registry or an index.
Abstract
We propose a Deep learning-based weak label learning method for analyzing whole slide images (WSIs) of Hematoxylin and Eosin (H&E) stained tumor tissue not requiring pixel-level or tile-level annotations using Self-supervised pre-training and heterogeneity-aware deep Multiple Instance LEarning (DeepSMILE). We apply DeepSMILE to the task of Homologous recombination deficiency (HRD) and microsatellite instability (MSI) prediction. We utilize contrastive self-supervised learning to pre-train a feature extractor on histopathology tiles of cancer tissue. Additionally, we use variability-aware deep multiple instance learning to learn the tile feature aggregation function while modeling tumor heterogeneity. For MSI prediction in a tumor-annotated and color normalized subset of TCGA-CRC (n=360 patients), contrastive self-supervised learning improves the tile supervision baseline from 0.77 to 0.87 AUROC, on par with our proposed DeepSMILE method. On TCGA-BC (n=1041 patients) without any manual annotations, DeepSMILE improves HRD classification performance from 0.77 to 0.81 AUROC compared to tile supervision with either a self-supervised or ImageNet pre-trained feature extractor. Our proposed methods reach the baseline performance using only 40% of the labeled data on both datasets. These improvements suggest we can use standard self-supervised learning techniques combined with multiple instance learning in the histopathology domain to improve genomic label classification performance with fewer labeled data.
Overview
The claims and statistics of this paper haven’t been checked yet.
Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
The claims and statistics of this paper haven’t been checked yet.
Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.
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
An Image is Worth 16x16 Words: Transformers for Image Recognition at Scale
Dosovitskiy, A., Beyer, L., Kolesnikov, A., Weissenborn, D., Zhai, X., Unterthiner, T., Dehghani, M., Minderer, M., Heigold, G., Gelly, S., et al. (2020) · arXiv · doi:10.48550/arxiv.2010.11929 · Cited by 47,472
ShuffleNet V2: Practical Guidelines for Efficient CNN Architecture Design
Ma, N., Zhang, X., Zheng, H.-T., & Sun, J. (2018) · Lecture notes in computer science · doi:10.1007/978-3-030-01264-9_8 · Cited by 7,356 · cited 1×
“, 2016) and ShufflenetV2 (Ma et al., 2018) respectively.”
Representation Learning with Contrastive Predictive Coding
Oord, A. V. D., Li, Y., & Vinyals, O. (2018) · arXiv · doi:10.48550/arxiv.1807.03748 · Cited by 6,242
Bootstrap Your Own Latent: A New Approach to Self-Supervised Learning
Grill, J.-B., Strub, F., Altché, F., Tallec, C., Richemond, P. H., Buchatskaya, E., Doersch, C., Pires, B. A., Guo, Z. D., Azar, M. G., et al. (2020) · arXiv · doi:10.48550/arxiv.2006.07733 · Cited by 3,846 · cited 1×
“We conclude with the discussion and conclusion in section 6.”
BackgroundCoudray, N., Ocampo, P. S., Sakellaropoulos, T., Narula, N., Snuderl, M., Fenyö, D., Moreira, A. L., Razavian, N., & Tsirigos, A. (2018) · Nature Medicine · doi:10.1038/s41591-018-0177-5 · Cited by 2,930 · cited 9×
“This method has been applied to gene mutation classification in lung cancer (Coudray et al. (2018)), MSI status classification in colorectal and stomach cancer (Kather et al.”
BackgroundMethodInfluentialClinical-grade computational pathology using weakly supervised deep learning on whole slide images
Campanella, G., Hanna, M. G., Geneslaw, L., Miraflor, A., Werneck Krauss Silva, V., Busam, K. J., Brogi, E., Reuter, V. E., Klimstra, D. S., & Fuchs, T. J. (2019) · Nature Medicine · doi:10.1038/s41591-019-0508-1 · Cited by 2,799
Data-efficient and weakly supervised computational pathology on whole-slide images
Lu, M. Y., Williamson, D. F. K., Chen, T. Y., Chen, R. J., Barbieri, M., & Mahmood, F. (2021) · Nature Biomedical Engineering · doi:10.1038/s41551-020-00682-w · Cited by 2,282
Kather, J. N., Pearson, A. T., Halama, N., Jäger, D., Krause, J., Loosen, S. H., Marx, A., Boor, P., Tacke, F., Neumann, U. P., et al. (2019) · Nature Medicine · doi:10.1038/s41591-019-0462-y · Cited by 1,458 · cited 10×
“We obtain all FFPE WSIs for BC tissue from TCGA (referred to as TCGA-BC ), which is collected from multiple centers in the United States of America, and retrieve genomic DDR-related labels, including HRD Score, from Knijnenburg et al. (2018).”
BackgroundMethodInfluentialLord, C. J., & Ashworth, A. (2016) · Nature Reviews Cancer · doi:10.1038/nrc.2015.21 · Cited by 1,257
State-of-the-art strategies for targeting the DNA damage response in cancer
Pilié, P. G., Tang, C., Mills, G. B., & Yap, T. A. (2019) · Nature Reviews Clinical Oncology · doi:10.1038/s41571-018-0114-z · Cited by 1,189
HRDetect is a predictor of BRCA1 and BRCA2 deficiency based on mutational signatures
Davies, H., Glodzik, D., Morganella, S., Yates, L. R., Staaf, J., Zou, X., Ramakrishna, M., Martin, S., Boyault, S., Sieuwerts, A. M., et al. (2017) · Nature Medicine · doi:10.1038/nm.4292 · Cited by 1,086 · cited 4×
“Additionally, the molecular features that indicate functional deficiencies in the homologous recombination pathway, thus indicating targeted therapy sensitivity, are inconclusive and debated (Davies et al., 2017).”
BackgroundMethodInfluentialMlecnik, B., Bindea, G., Angell, H. K., Maby, P., Angelova, M., Tougeron, D., Church, S. E., Lafontaine, L., Fischer, M., Fredriksen, T., et al. (2016) · Immunity · doi:10.1016/j.immuni.2016.02.025 · Cited by 1,034 · cited 1×
“…improvements may reduce the need for expensive genome sequencing techniques, provide personalized therapy recommendations based on widely available H&E WSIs, and improve patient care with quicker treatment decisions - also in medical centers without access to genome sequencing resources.”
BackgroundBlank, C. U., Haanen, J. B., Ribas, A., & Schumacher, T. N. (2016) · Science · doi:10.1126/science.aaf2834 · Cited by 748
AI-based pathology predicts origins for cancers of unknown primary
Lu, M. Y., Chen, T. Y., Williamson, D. F. K., Zhao, M., Shady, M., Lipkova, J., & Mahmood, F. (2021) · Nature · doi:10.1038/s41586-021-03512-4 · Cited by 636
Using Self-Supervised Learning Can Improve Model Robustness and Uncertainty
Hendrycks, D., Mazeika, M., Kadavath, S., & Song, D. (2019) · arXiv · doi:10.48550/arxiv.1906.12340 · Cited by 634
Therapeutic opportunities within the DNA damage response
Pearl, L. H., Schierz, A. C., Ward, S. E., Al-Lazikani, B., & Pearl, F. M. G. (2015) · Nature Reviews Cancer · doi:10.1038/nrc3891 · Cited by 587 · cited 1×
“The main contributions of this work can be summarized as follows: 1) We outperform existing tile-supervised WSI-label learning methods for MSI classification for a tumor-annotated and color-normalized colorectal cancer dataset using a feature extractor pre-trained with SimCLR.”
Deep learning-based classification of mesothelioma improves prediction of patient outcome
Courtiol, P., Maussion, C., Moarii, M., Pronier, E., Pilcer, S., Sefta, M., Manceron, P., Toldo, S., Zaslavskiy, M., Le Stang, N., et al. (2019) · Nature Medicine · doi:10.1038/s41591-019-0583-3 · Cited by 544 · cited 2×
“Courtiol et al. (2019) extend the work by Durand et al. (2016) to tissue segmentation and WSI-level label classification using an MLP on the top and bottom instance scores.”
MethodClinical-Grade Detection of Microsatellite Instability in Colorectal Tumors by Deep Learning
Echle, A., Grabsch, H. I., Quirke, P., van den Brandt, P. A., West, N. P., Hutchins, G. G. A., Heij, L. R., Tan, X., Richman, S. D., Krause, J., et al. (2020) · Gastroenterology · doi:10.1053/j.gastro.2020.06.021 · Cited by 376
Bilal, M., Raza, S. E. A., Azam, A., Graham, S., Ilyas, M., Cree, I. A., Snead, D., Minhas, F., & Rajpoot, N. M. (2021) · The Lancet Digital Health · doi:10.1016/s2589-7500(21)00180-1 · Cited by 299
Self supervised contrastive learning for digital histopathology
Ciga, O., Xu, T., & Martel, A. L. (2022) · Machine Learning with Applications · doi:10.1016/j.mlwa.2021.100198 · Cited by 294
Pathologic Predictors of Microsatellite Instability in Colorectal Cancer
Greenson, J. K., Huang, S.-C., Herron, C., Moreno, V., Bonner, J. D., Tomsho, L. P., Ben-Izhak, O., Cohen, H. I., Trougouboff, P., Bejhar, J., et al. (2009) · American Journal of Surgical Pathology · doi:10.1097/pas.0b013e31817ec2b1 · Cited by 286 · cited 1×
“We believe that this is due to the tumor annotations done in TCGA-CRCk, strengthening the supervisory signal for the MSI classification task as most tiles contain tumor tissue that reflects information about the cancer genotype.”
MethodMuti, H. S., Heij, L. R., Keller, G., Kohlruss, M., Langer, R., Dislich, B., Cheong, J.-H., Kim, Y.-W., Kim, H., Kook, M.-C., et al. (2021) · The Lancet Digital Health · doi:10.1016/s2589-7500(21)00133-3 · Cited by 126 · cited 1×
“For the MSI classification task, our SimCLR-DeepMIL model reproduces the results achieved by Saillard et al. (2021) when using DeepMIL, showing that the use of MOCOv2 or SimCLR is likely equivalent in the histopathology domain.”
MethodA Subset of Colorectal Cancers with Cross-Sensitivity to Olaparib and Oxaliplatin
Arena, S., Corti, G., Durinikova, E., Montone, M., Reilly, N. M., Russo, M., Lorenzato, A., Arcella, P., Lazzari, L., Rospo, G., et al. (2020) · Clinical Cancer Research · doi:10.1158/1078-0432.ccr-19-2409 · Cited by 115 · cited 5×
“(2020); Arena et al. (2020)) patients. Currently, the most common techniques for determining HRD or MSI are next-generation whole-genome and exome sequencing methods (Davies et al. (2017); Zhu et al. (2018)), targeted DNA sequencing methods using Polymerase Chain Reaction (Boland et al. (1998)), and immunohistochemist…”
BackgroundMethodInfluential
Cited by
Towards a general-purpose foundation model for computational pathology
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., et al. (2024) · Nature Medicine · doi:10.1038/s41591-024-02857-3 · Cited by 1,560
Wagner, S. J., Reisenbüchler, D., West, N. P., Niehues, J. M., Zhu, J., Foersch, S., Veldhuizen, G. P., Quirke, P., Grabsch, H. I., van den Brandt, P. A., et al. (2023) · Cancer Cell · doi:10.1016/j.ccell.2023.08.002 · Cited by 275
Scaling Self-Supervised Learning for Histopathology with Masked Image Modeling
Filiot, A., Ghermi, R., Olivier, A., Jacob, P., Fidon, L., Camara, A., Mac Kain, A., Saillard, C., & Schiratti, J.-B. (2023) · doi:10.1101/2023.07.21.23292757 · Cited by 148
Bergstrom, E. N., Abbasi, A., Díaz-Gay, M., Galland, L., Ladoire, S., Lippman, S. M., & Alexandrov, L. B. (2024) · Journal of Clinical Oncology · doi:10.1200/jco.23.02641 · Cited by 83
One label is all you need: Interpretable AI-enhanced histopathology for oncology
Tavolara, T. E., Su, Z., Gurcan, M. N., & Niazi, M. K. K. (2023) · Seminars in Cancer Biology · doi:10.1016/j.semcancer.2023.09.006 · Cited by 28
Metadata
- Authors
- Schirris, Y., Gavves, E., Nederlof, I., Horlings, H. M., & Teuwen, J.
- Venue
- Medical Image Analysis
- Publisher
- Elsevier BV
- Published
- 2022-07
- Volume
- 79
- Pages
- 102464
- ISSN
- 1361-8415 (ISSN-L)
- OpenAlex
- W4225136345
- Cited by
- 143
- References
- 24
- FWCI
- 9.13
- Type
- journal article
- Language
- en
- Open access
- green
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
Data sources
Bibliographic data: Semantic Scholar, OpenAlex, Crossref.
Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).
Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.