The impact of site-specific digital histology signatures on deep learning model accuracy and bias

Deep learning models have been trained on The Cancer Genome Atlas to predict numerous features directly from histology, including survival, gene expression patterns, and driver mutations. Here, the authors demonstrate that site-specific histologic signatures can lead to biased estimates of accuracy...

Descripción completa

Guardado en:

Detalles Bibliográficos
Autores principales:	Frederick M. Howard, James Dolezal, Sara Kochanny, Jefree Schulte, Heather Chen, Lara Heij, Dezheng Huo, Rita Nanda, Olufunmilayo I. Olopade, Jakob N. Kather, Nicole Cipriani, Robert L. Grossman, Alexander T. Pearson
Formato:	article
Lenguaje:	EN
Publicado:	Nature Portfolio 2021
Materias:	Science Q
Acceso en línea:	https://doaj.org/article/63d9acc99eb6472c87b682f300c925c8
Etiquetas:	Agregar Etiqueta Sin Etiquetas, Sea el primero en etiquetar este registro!

id	oai:doaj.org-article:63d9acc99eb6472c87b682f300c925c8
record_format	dspace
spelling	oai:doaj.org-article:63d9acc99eb6472c87b682f300c925c82021-12-02T17:57:04ZThe impact of site-specific digital histology signatures on deep learning model accuracy and bias10.1038/s41467-021-24698-12041-1723https://doaj.org/article/63d9acc99eb6472c87b682f300c925c82021-07-01T00:00:00Zhttps://doi.org/10.1038/s41467-021-24698-1https://doaj.org/toc/2041-1723Deep learning models have been trained on The Cancer Genome Atlas to predict numerous features directly from histology, including survival, gene expression patterns, and driver mutations. Here, the authors demonstrate that site-specific histologic signatures can lead to biased estimates of accuracy for such models, and propose a method to minimize such bias.Frederick M. HowardJames DolezalSara KochannyJefree SchulteHeather ChenLara HeijDezheng HuoRita NandaOlufunmilayo I. OlopadeJakob N. KatherNicole CiprianiRobert L. GrossmanAlexander T. PearsonNature PortfolioarticleScienceQENNature Communications, Vol 12, Iss 1, Pp 1-13 (2021)
institution	DOAJ
collection	DOAJ
language	EN
topic	Science Q
spellingShingle	Science Q Frederick M. Howard James Dolezal Sara Kochanny Jefree Schulte Heather Chen Lara Heij Dezheng Huo Rita Nanda Olufunmilayo I. Olopade Jakob N. Kather Nicole Cipriani Robert L. Grossman Alexander T. Pearson The impact of site-specific digital histology signatures on deep learning model accuracy and bias
description	Deep learning models have been trained on The Cancer Genome Atlas to predict numerous features directly from histology, including survival, gene expression patterns, and driver mutations. Here, the authors demonstrate that site-specific histologic signatures can lead to biased estimates of accuracy for such models, and propose a method to minimize such bias.
format	article
author	Frederick M. Howard James Dolezal Sara Kochanny Jefree Schulte Heather Chen Lara Heij Dezheng Huo Rita Nanda Olufunmilayo I. Olopade Jakob N. Kather Nicole Cipriani Robert L. Grossman Alexander T. Pearson
author_facet	Frederick M. Howard James Dolezal Sara Kochanny Jefree Schulte Heather Chen Lara Heij Dezheng Huo Rita Nanda Olufunmilayo I. Olopade Jakob N. Kather Nicole Cipriani Robert L. Grossman Alexander T. Pearson
author_sort	Frederick M. Howard
title	The impact of site-specific digital histology signatures on deep learning model accuracy and bias
title_short	The impact of site-specific digital histology signatures on deep learning model accuracy and bias
title_full	The impact of site-specific digital histology signatures on deep learning model accuracy and bias
title_fullStr	The impact of site-specific digital histology signatures on deep learning model accuracy and bias
title_full_unstemmed	The impact of site-specific digital histology signatures on deep learning model accuracy and bias
title_sort	impact of site-specific digital histology signatures on deep learning model accuracy and bias
publisher	Nature Portfolio
publishDate	2021
url	https://doaj.org/article/63d9acc99eb6472c87b682f300c925c8
work_keys_str_mv	AT frederickmhoward theimpactofsitespecificdigitalhistologysignaturesondeeplearningmodelaccuracyandbias AT jamesdolezal theimpactofsitespecificdigitalhistologysignaturesondeeplearningmodelaccuracyandbias AT sarakochanny theimpactofsitespecificdigitalhistologysignaturesondeeplearningmodelaccuracyandbias AT jefreeschulte theimpactofsitespecificdigitalhistologysignaturesondeeplearningmodelaccuracyandbias AT heatherchen theimpactofsitespecificdigitalhistologysignaturesondeeplearningmodelaccuracyandbias AT laraheij theimpactofsitespecificdigitalhistologysignaturesondeeplearningmodelaccuracyandbias AT dezhenghuo theimpactofsitespecificdigitalhistologysignaturesondeeplearningmodelaccuracyandbias AT ritananda theimpactofsitespecificdigitalhistologysignaturesondeeplearningmodelaccuracyandbias AT olufunmilayoiolopade theimpactofsitespecificdigitalhistologysignaturesondeeplearningmodelaccuracyandbias AT jakobnkather theimpactofsitespecificdigitalhistologysignaturesondeeplearningmodelaccuracyandbias AT nicolecipriani theimpactofsitespecificdigitalhistologysignaturesondeeplearningmodelaccuracyandbias AT robertlgrossman theimpactofsitespecificdigitalhistologysignaturesondeeplearningmodelaccuracyandbias AT alexandertpearson theimpactofsitespecificdigitalhistologysignaturesondeeplearningmodelaccuracyandbias AT frederickmhoward impactofsitespecificdigitalhistologysignaturesondeeplearningmodelaccuracyandbias AT jamesdolezal impactofsitespecificdigitalhistologysignaturesondeeplearningmodelaccuracyandbias AT sarakochanny impactofsitespecificdigitalhistologysignaturesondeeplearningmodelaccuracyandbias AT jefreeschulte impactofsitespecificdigitalhistologysignaturesondeeplearningmodelaccuracyandbias AT heatherchen impactofsitespecificdigitalhistologysignaturesondeeplearningmodelaccuracyandbias AT laraheij impactofsitespecificdigitalhistologysignaturesondeeplearningmodelaccuracyandbias AT dezhenghuo impactofsitespecificdigitalhistologysignaturesondeeplearningmodelaccuracyandbias AT ritananda impactofsitespecificdigitalhistologysignaturesondeeplearningmodelaccuracyandbias AT olufunmilayoiolopade impactofsitespecificdigitalhistologysignaturesondeeplearningmodelaccuracyandbias AT jakobnkather impactofsitespecificdigitalhistologysignaturesondeeplearningmodelaccuracyandbias AT nicolecipriani impactofsitespecificdigitalhistologysignaturesondeeplearningmodelaccuracyandbias AT robertlgrossman impactofsitespecificdigitalhistologysignaturesondeeplearningmodelaccuracyandbias AT alexandertpearson impactofsitespecificdigitalhistologysignaturesondeeplearningmodelaccuracyandbias
_version_	1718379048549220352

The impact of site-specific digital histology signatures on deep learning model accuracy and bias

Ejemplares similares