Federated Learning used for predicting outcomes in SARS-COV-2 patients.
Journal
Research square
Titre abrégé: Res Sq
Pays: United States
ID NLM: 101768035
Informations de publication
Date de publication:
08 Jan 2021
08 Jan 2021
Historique:
entrez:
14
1
2021
pubmed:
15
1
2021
medline:
15
1
2021
Statut:
epublish
Résumé
'Federated Learning' (FL) is a method to train Artificial Intelligence (AI) models with data from multiple sources while maintaining anonymity of the data thus removing many barriers to data sharing. During the SARS-COV-2 pandemic, 20 institutes collaborated on a healthcare FL study to predict future oxygen requirements of infected patients using inputs of vital signs, laboratory data, and chest x-rays, constituting the "EXAM" (EMR CXR AI Model) model. EXAM achieved an average Area Under the Curve (AUC) of over 0.92, an average improvement of 16%, and a 38% increase in generalisability over local models. The FL paradigm was successfully applied to facilitate a rapid data science collaboration without data exchange, resulting in a model that generalised across heterogeneous, unharmonized datasets. This provided the broader healthcare community with a validated model to respond to COVID-19 challenges, as well as set the stage for broader use of FL in healthcare.
Identifiants
pubmed: 33442676
doi: 10.21203/rs.3.rs-126892/v1
pmc: PMC7805458
pii:
doi:
Types de publication
Preprint
Langues
eng
Subventions
Organisme : Intramural NIH HHS
ID : ZIA CL040015
Pays : United States
Organisme : Intramural NIH HHS
ID : ZID BC011242
Pays : United States
Commentaires et corrections
Type : UpdateIn
Références
Nat Commun. 2020 Oct 12;11(1):5131
pubmed: 33046699
JMIR Med Inform. 2021 Jan 27;9(1):e24207
pubmed: 33400679
Methods Inf Med. 2015;54(1):65-74
pubmed: 25426730
NPJ Digit Med. 2019 Aug 19;2:78
pubmed: 31453373
Neuron. 2017 Dec 6;96(5):964-965
pubmed: 29216458
NPJ Digit Med. 2020 Sep 14;3:119
pubmed: 33015372
Med Image Anal. 2021 May;70:101993
pubmed: 33711739
J Infect. 2020 Aug;81(2):282-288
pubmed: 32479771
JAMA. 2020 Jun 9;323(22):2329-2330
pubmed: 32329799
Sci Rep. 2020 Jul 28;10(1):12598
pubmed: 32724046
J Am Coll Radiol. 2018 Mar;15(3 Pt B):504-508
pubmed: 29402533
Anaesthesia. 2020 Jun;75(6):785-799
pubmed: 32221970
Insights Imaging. 2019 Apr 4;10(1):44
pubmed: 30949865
JAMA Intern Med. 2020 Aug 1;180(8):1081-1089
pubmed: 32396163
Nat Med. 2020 Aug;26(8):1183-1192
pubmed: 32770165
PeerJ. 2018 Feb 13;6:e4375
pubmed: 29456894
Med Image Anal. 2021 May;70:101992
pubmed: 33601166
J Thromb Haemost. 2020 Jun;18(6):1324-1329
pubmed: 32306492
Health Aff (Millwood). 2014 Jul;33(7):1139-47
pubmed: 25006139
Nat Med. 2019 Jan;25(1):37-43
pubmed: 30617331
Nat Med. 2020 Jan;26(1):29-38
pubmed: 31932803
Bioinformatics. 2012 Jan 1;28(1):112-8
pubmed: 22039212
Int J Med Inform. 2018 Apr;112:59-67
pubmed: 29500022
BMJ. 2020 Apr 7;369:m1328
pubmed: 32265220
J Am Med Inform Assoc. 2020 Nov 1;27(11):1721-1726
pubmed: 32918447
Insights Imaging. 2018 Oct;9(5):745-753
pubmed: 30112675
BMC Anesthesiol. 2020 Jul 20;20(1):177
pubmed: 32689937
BMC Bioinformatics. 2011 Mar 17;12:77
pubmed: 21414208
Bull World Health Organ. 2020 Mar 1;98(3):150
pubmed: 32132744
Med Image Anal. 2020 Oct;65:101765
pubmed: 32679533
Acad Emerg Med. 2020 Oct;27(10):1039-1042
pubmed: 32853423
Nature. 2020 Mar;579(7798):193
pubmed: 32157233
J Am Coll Emerg Physicians Open. 2020 Apr 13;1(2):95-101
pubmed: 32427171
Infect Control Hosp Epidemiol. 2021 Apr;42(4):399-405
pubmed: 32928319
Cell Res. 2020 May;30(5):370-371
pubmed: 32350393
Nat Med. 2019 Jan;25(1):24-29
pubmed: 30617335