@article{dayan2021federated,
author = {Dayan, Ittai and R. Roth, Holger and Zhong, Aoxiao and Harouni, Ahmed and Gentili, Amilcare and Z. Abidin, Anas and Liu, Andrew and Beardsworth Costa, Anthony and J. Wood, Bradford and Tsai, Chien-Sung and Wang, Chih-Hung and Hsu, Chun-Nan and K. Lee, C. and Ruan, Peiying and Xu, Daguang and Wu, Dufan and Huang, Eddie and Campos Kitamura, Felipe and Lacey, Griffin and César de Antônio Corradi, Gustavo and Nino, Gustavo and Shin, Hao-Hsin and Obinata, Hirofumi and Ren, Hui and C. Crane, Jason and Tetreault, Jesse and Guan, Jiahui and W. Garrett, John and D. Kaggie, Joshua and Gil Park, Jung and Dreyer, Keith and Juluru, Krishna and Kersten, Kristopher and Aloisio Bezerra Cavalcanti Rockenbach, Marcio and George Linguraru, Marius and A. Haider, Masoom and AbdelMaseeh, Meena and Rieke, Nicola and F. Damasceno, Pablo and Mario Cruz e Silva, Pedro and Wang, Pochuan and Xu, Sheng and Kawano, Shuichi and Sriswasdi, Sira and Young Park, Soo and M. Grist, Thomas and Buch, Varun and Jantarabenjakul, Watsamon and Wang, Weichung and Young Tak, Won and Li, Xiang and Lin, Xihong and Joon Kwon, Young and Quraini, Abood and Feng, Andrew and N. Priest, Andrew and Turkbey, Baris and Glicksberg, Benjamin and Bizzo, Bernardo and Seok Kim, Byung and Tor-Díez, Carlos and Lee, Chia-Cheng and Hsu, Chia-Jung and Lin, Chin and Lai, Chiu-Ling and P. Hess, Christopher and Compas, Colin and Bhatia, Deepeksha and K. Oermann, Eric and Leibovitz, Evan and Sasaki, Hisashi and Mori, Hitoshi and Yang, Isaac and Ho Sohn, Jae and Nand Keshava Murthy, Krishna and Fu, Li-Chen and Mendonça, Matheus R. F. and Fralick, Mike and Kyu Kang, Min and Adil, Mohammad and Gangai, Natalie and Vateekul, Peerapon and Elnajjar, Pierre and Hickman, Sarah and Majumdar, Sharmila and L. McLeod, Shelley and Reed, Sheridan and Gräf, Stefan and Harmon, Stephanie and Kodama, Tatsuya and Puthanakit, Thanyawee and Mazzulli, Tony and Lima de Lavor, Vitor and Rakvongthai, Yothin and Rim Lee, Yu and Wen, Yuhong and J. Gilbert, Fiona and G. Flores, Mona and Li, Quanzheng},
title = {Federated learning for predicting clinical outcomes in patients with COVID-19.},
year = {2021},
month = {September},
abstract = {Federated learning (FL) is a method used for training artificial intelligence models with data from multiple sources while maintaining data anonymity, thus removing many barriers to data sharing. Here we used data from 20 institutes across the globe to train a FL model, called EXAM (electronic medical record (EMR) chest X-ray AI model), that predicts the future oxygen requirements of symptomatic patients with COVID-19 using inputs of vital signs, laboratory data and chest X-rays. EXAM achieved an average area under the curve (AUC) >0.92 for predicting outcomes at 24 and 72 h from the time of initial presentation to the emergency room, and it provided 16% improvement in average AUC measured across all participating sites and an average increase in generalizability of 38% when compared with models trained at a single site using that site’s data. For prediction of mechanical ventilation treatment or death at 24 h at the largest independent test site, EXAM achieved a sensitivity of 0.950 and specificity of 0.882. In this study, FL facilitated rapid data science collaboration without data exchange and generated a model that generalized across heterogeneous, unharmonized datasets for prediction of clinical outcomes in patients with COVID-19, setting the stage for the broader use of FL in healthcare. Federated learning, a method for training artificial intelligence algorithms that protects data privacy, was used to predict future oxygen requirements of symptomatic patients with COVID-19 using data from 20 different institutes across the globe.},
url = {http://approjects.co.za/?big=en-us/research/publication/federated-learning-for-predicting-clinical-outcomes-in-patients-with-covid-19/},
pages = {1735-1743},
journal = {Nature Medicine},
volume = {27},
number = {10},
}