Privacy-first health research with federated learning

Abstract Privacy protection is paramount in conducting health research. However, studies often rely on data stored in a centralized repository, where analysis is done with full access to the sensitive underlying content. Recent advances in federated learning enable building complex machine-learned m...

Descripción completa

Guardado en:

Detalles Bibliográficos
Autores principales:	Adam Sadilek, Luyang Liu, Dung Nguyen, Methun Kamruzzaman, Stylianos Serghiou, Benjamin Rader, Alex Ingerman, Stefan Mellem, Peter Kairouz, Elaine O. Nsoesie, Jamie MacFarlane, Anil Vullikanti, Madhav Marathe, Paul Eastham, John S. Brownstein, Blaise Aguera y. Arcas, Michael D. Howell, John Hernandez
Formato:	article
Lenguaje:	EN
Publicado:	Nature Portfolio 2021
Materias:	Computer applications to medicine. Medical informatics R858-859.7
Acceso en línea:	https://doaj.org/article/decac40d76f04fc18e73402b619ffedc
Etiquetas:	Agregar Etiqueta Sin Etiquetas, Sea el primero en etiquetar este registro!

id	oai:doaj.org-article:decac40d76f04fc18e73402b619ffedc
record_format	dspace
spelling	oai:doaj.org-article:decac40d76f04fc18e73402b619ffedc2021-12-02T17:19:10ZPrivacy-first health research with federated learning10.1038/s41746-021-00489-22398-6352https://doaj.org/article/decac40d76f04fc18e73402b619ffedc2021-09-01T00:00:00Zhttps://doi.org/10.1038/s41746-021-00489-2https://doaj.org/toc/2398-6352Abstract Privacy protection is paramount in conducting health research. However, studies often rely on data stored in a centralized repository, where analysis is done with full access to the sensitive underlying content. Recent advances in federated learning enable building complex machine-learned models that are trained in a distributed fashion. These techniques facilitate the calculation of research study endpoints such that private data never leaves a given device or healthcare system. We show—on a diverse set of single and multi-site health studies—that federated models can achieve similar accuracy, precision, and generalizability, and lead to the same interpretation as standard centralized statistical models while achieving considerably stronger privacy protections and without significantly raising computational costs. This work is the first to apply modern and general federated learning methods that explicitly incorporate differential privacy to clinical and epidemiological research—across a spectrum of units of federation, model architectures, complexity of learning tasks and diseases. As a result, it enables health research participants to remain in control of their data and still contribute to advancing science—aspects that used to be at odds with each other.Adam SadilekLuyang LiuDung NguyenMethun KamruzzamanStylianos SerghiouBenjamin RaderAlex IngermanStefan MellemPeter KairouzElaine O. NsoesieJamie MacFarlaneAnil VullikantiMadhav MarathePaul EasthamJohn S. BrownsteinBlaise Aguera y. ArcasMichael D. HowellJohn HernandezNature PortfolioarticleComputer applications to medicine. Medical informaticsR858-859.7ENnpj Digital Medicine, Vol 4, Iss 1, Pp 1-8 (2021)
institution	DOAJ
collection	DOAJ
language	EN
topic	Computer applications to medicine. Medical informatics R858-859.7
spellingShingle	Computer applications to medicine. Medical informatics R858-859.7 Adam Sadilek Luyang Liu Dung Nguyen Methun Kamruzzaman Stylianos Serghiou Benjamin Rader Alex Ingerman Stefan Mellem Peter Kairouz Elaine O. Nsoesie Jamie MacFarlane Anil Vullikanti Madhav Marathe Paul Eastham John S. Brownstein Blaise Aguera y. Arcas Michael D. Howell John Hernandez Privacy-first health research with federated learning
description	Abstract Privacy protection is paramount in conducting health research. However, studies often rely on data stored in a centralized repository, where analysis is done with full access to the sensitive underlying content. Recent advances in federated learning enable building complex machine-learned models that are trained in a distributed fashion. These techniques facilitate the calculation of research study endpoints such that private data never leaves a given device or healthcare system. We show—on a diverse set of single and multi-site health studies—that federated models can achieve similar accuracy, precision, and generalizability, and lead to the same interpretation as standard centralized statistical models while achieving considerably stronger privacy protections and without significantly raising computational costs. This work is the first to apply modern and general federated learning methods that explicitly incorporate differential privacy to clinical and epidemiological research—across a spectrum of units of federation, model architectures, complexity of learning tasks and diseases. As a result, it enables health research participants to remain in control of their data and still contribute to advancing science—aspects that used to be at odds with each other.
format	article
author	Adam Sadilek Luyang Liu Dung Nguyen Methun Kamruzzaman Stylianos Serghiou Benjamin Rader Alex Ingerman Stefan Mellem Peter Kairouz Elaine O. Nsoesie Jamie MacFarlane Anil Vullikanti Madhav Marathe Paul Eastham John S. Brownstein Blaise Aguera y. Arcas Michael D. Howell John Hernandez
author_facet	Adam Sadilek Luyang Liu Dung Nguyen Methun Kamruzzaman Stylianos Serghiou Benjamin Rader Alex Ingerman Stefan Mellem Peter Kairouz Elaine O. Nsoesie Jamie MacFarlane Anil Vullikanti Madhav Marathe Paul Eastham John S. Brownstein Blaise Aguera y. Arcas Michael D. Howell John Hernandez
author_sort	Adam Sadilek
title	Privacy-first health research with federated learning
title_short	Privacy-first health research with federated learning
title_full	Privacy-first health research with federated learning
title_fullStr	Privacy-first health research with federated learning
title_full_unstemmed	Privacy-first health research with federated learning
title_sort	privacy-first health research with federated learning
publisher	Nature Portfolio
publishDate	2021
url	https://doaj.org/article/decac40d76f04fc18e73402b619ffedc
work_keys_str_mv	AT adamsadilek privacyfirsthealthresearchwithfederatedlearning AT luyangliu privacyfirsthealthresearchwithfederatedlearning AT dungnguyen privacyfirsthealthresearchwithfederatedlearning AT methunkamruzzaman privacyfirsthealthresearchwithfederatedlearning AT stylianosserghiou privacyfirsthealthresearchwithfederatedlearning AT benjaminrader privacyfirsthealthresearchwithfederatedlearning AT alexingerman privacyfirsthealthresearchwithfederatedlearning AT stefanmellem privacyfirsthealthresearchwithfederatedlearning AT peterkairouz privacyfirsthealthresearchwithfederatedlearning AT elaineonsoesie privacyfirsthealthresearchwithfederatedlearning AT jamiemacfarlane privacyfirsthealthresearchwithfederatedlearning AT anilvullikanti privacyfirsthealthresearchwithfederatedlearning AT madhavmarathe privacyfirsthealthresearchwithfederatedlearning AT pauleastham privacyfirsthealthresearchwithfederatedlearning AT johnsbrownstein privacyfirsthealthresearchwithfederatedlearning AT blaiseaguerayarcas privacyfirsthealthresearchwithfederatedlearning AT michaeldhowell privacyfirsthealthresearchwithfederatedlearning AT johnhernandez privacyfirsthealthresearchwithfederatedlearning
_version_	1718381114859454464

Privacy-first health research with federated learning

Ejemplares similares