Predicting survival from colorectal cancer histology slides using deep learning: A retrospective multicenter study

For virtually every patient with colorectal cancer (CRC), hematoxylin-eosin (HE)-stained tissue slides are available. These images contain quantitative information, which is not routinely used to objectively extract prognostic biomarkers. In the present study, we investigated whether deep convolutio...

Ausführliche Beschreibung

Gespeichert in:

Bibliographische Detailangaben
Veröffentlicht in:	PLoS medicine Jg. 16; H. 1; S. e1002730
Hauptverfasser:	Kather, Jakob Nikolas, Krisam, Johannes, Charoentong, Pornpimol, Luedde, Tom, Herpel, Esther, Weis, Cleo-Aron, Gaiser, Timo, Marx, Alexander, Valous, Nektarios A., Ferber, Dyke, Jansen, Lina, Reyes-Aldasoro, Constantino Carlos, Zörnig, Inka, Jäger, Dirk, Brenner, Hermann, Chang-Claude, Jenny, Hoffmeister, Michael, Halama, Niels
Format:	Journal Article
Sprache:	Englisch
Veröffentlicht:	United States Public Library of Science 24.01.2019 Public Library of Science (PLoS)
Schlagworte:	Aging Artificial intelligence Artificial neural networks Bioindicators Biological markers Biology and Life Sciences Biomarkers Cancer Cancer patients Cancer research Classification Colon - pathology Colorectal cancer Colorectal carcinoma Colorectal Neoplasms - diagnosis Colorectal Neoplasms - mortality Colorectal Neoplasms - pathology Coloring Agents Computer and Information Sciences Consortia Deep Learning Eosine Yellowish-(YS) Epidemiology Female Fibroblasts Funding Gastroenterology Gene expression Genes Genomes Genomics Hematoxylin Hepatology Histology Hospitals Humans Image Interpretation, Computer-Assisted - methods Immunotherapy Internal medicine Male Medical care Medical imaging equipment Medical prognosis Medical research Medicine Medicine and Health Sciences Neural networks Neurons Oncology Pathology Patient outcomes Pattern recognition People and Places Prognosis Rectum - pathology Research centers Retrospective Studies Software patches Statistical models Stroma Supervision Survival Systematic review Tissues Transfer learning Tumors Aachen Germany Germany
ISSN:	1549-1676, 1549-1277, 1549-1676
Online-Zugang:	Volltext
Tags:	Tag hinzufügen Keine Tags, Fügen Sie den ersten Tag hinzu!

Beschreibung
Zusammenfassung:	For virtually every patient with colorectal cancer (CRC), hematoxylin-eosin (HE)-stained tissue slides are available. These images contain quantitative information, which is not routinely used to objectively extract prognostic biomarkers. In the present study, we investigated whether deep convolutional neural networks (CNNs) can extract prognosticators directly from these widely available images. We hand-delineated single-tissue regions in 86 CRC tissue slides, yielding more than 100,000 HE image patches, and used these to train a CNN by transfer learning, reaching a nine-class accuracy of >94% in an independent data set of 7,180 images from 25 CRC patients. With this tool, we performed automated tissue decomposition of representative multitissue HE images from 862 HE slides in 500 stage I-IV CRC patients in the The Cancer Genome Atlas (TCGA) cohort, a large international multicenter collection of CRC tissue. Based on the output neuron activations in the CNN, we calculated a "deep stroma score," which was an independent prognostic factor for overall survival (OS) in a multivariable Cox proportional hazard model (hazard ratio [HR] with 95% confidence interval [CI]: 1.99 [1.27-3.12], p = 0.0028), while in the same cohort, manual quantification of stromal areas and a gene expression signature of cancer-associated fibroblasts (CAFs) were only prognostic in specific tumor stages. We validated these findings in an independent cohort of 409 stage I-IV CRC patients from the "Darmkrebs: Chancen der Verhütung durch Screening" (DACHS) study who were recruited between 2003 and 2007 in multiple institutions in Germany. Again, the score was an independent prognostic factor for OS (HR 1.63 [1.14-2.33], p = 0.008), CRC-specific OS (HR 2.29 [1.5-3.48], p = 0.0004), and relapse-free survival (RFS; HR 1.92 [1.34-2.76], p = 0.0004). A prospective validation is required before this biomarker can be implemented in clinical workflows. In our retrospective study, we show that a CNN can assess the human tumor microenvironment and predict prognosis directly from histopathological images.
Bibliographie:	ObjectType-Article-1 SourceType-Scholarly Journals-1 ObjectType-Feature-2 content type line 14 content type line 23 The authors have declared that no competing interests exist.
ISSN:	1549-1676 1549-1277 1549-1676
DOI:	10.1371/journal.pmed.1002730