Quantifying the effects of data augmentation and stain color normalization in convolutional neural n...
Quantifying the effects of data augmentation and stain color normalization in convolutional neural networks for computational pathology
About this item
Full title
Author / Creator
Publisher
Ithaca: Cornell University Library, arXiv.org
Journal title
Language
English
Formats
Publication information
Publisher
Ithaca: Cornell University Library, arXiv.org
Subjects
More information
Scope and Contents
Contents
Stain variation is a phenomenon observed when distinct pathology laboratories stain tissue slides that exhibit similar but not identical color appearance. Due to this color shift between laboratories, convolutional neural networks (CNNs) trained with images from one lab often underperform on unseen images from the other lab. Several techniques have...
Alternative Titles
Full title
Quantifying the effects of data augmentation and stain color normalization in convolutional neural networks for computational pathology
Authors, Artists and Contributors
Identifiers
Primary Identifiers
Record Identifier
TN_cdi_proquest_journals_2186334018
Permalink
https://devfeature-collection.sl.nsw.gov.au/record/TN_cdi_proquest_journals_2186334018
Other Identifiers
E-ISSN
2331-8422
DOI
10.48550/arxiv.1902.06543
