AI Models Close to your Chest: Robust Federated Learning Strategies for Multi-site CT

While it is well known that population differences from genetics, sex, race, and environmental factors contribute to disease, AI studies in medicine have largely focused on locoregional patient cohorts with less diverse data sources. Such limitation stems from barriers to large-scale data share and...

Full description

Saved in:
Bibliographic Details
Published in:arXiv.org 2023-04
Main Authors: Lee, Edward H, Kelly, Brendan, Altinmakas, Emre, Dogan, Hakan, Mohammadzadeh, Maryam, Colak, Errol, Fu, Steve, Choudhury, Olivia, Ratan, Ujjwal, Kitamura, Felipe, Chaves, Hernan, Zheng, Jimmy, Mourad, Said, Reis, Eduardo, Lim, Jaekwang, Yokoo, Patricia, Mitchell, Courtney, Houshmand, Golnaz, Ghassemi, Marzyeh, Killeen, Ronan, Qiu, Wendy, Hayden, Joel, Rafiee, Farnaz, Klochko, Chad, Bevins, Nicholas, Sazgara, Faeze, S Simon Wong, Moseley, Michael, Halabi, Safwan, Yeom, Kristen W
Format: Article
Language:English
Subjects:
Datasets
Federated learning
Heterogeneity
Hospitals
Sex
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:While it is well known that population differences from genetics, sex, race, and environmental factors contribute to disease, AI studies in medicine have largely focused on locoregional patient cohorts with less diverse data sources. Such limitation stems from barriers to large-scale data share and ethical concerns over data privacy. Federated learning (FL) is one potential pathway for AI development that enables learning across hospitals without data share. In this study, we show the results of various FL strategies on one of the largest and most diverse COVID-19 chest CT datasets: 21 participating hospitals across five continents that comprise >10,000 patients with >1 million images. We also propose an FL strategy that leverages synthetically generated data to overcome class and size imbalances. We also describe the sources of data heterogeneity in the context of FL, and show how even among the correctly labeled populations, disparities can arise due to these biases.
ISSN:2331-8422