Federated Learning for Multi-Center Imaging Diagnostics: A Study in Cardiovascular Disease

TL;DR

This study demonstrates that federated learning can effectively train cardiovascular MRI diagnostic models across multiple centers, preserving privacy and achieving results comparable to centralized methods, while enhancing robustness and domain shift sensitivity.

Contribution

First federated learning application to multi-center cardiovascular MRI diagnosis, exploring model adaptation, shape priors, and data augmentation impacts.

Findings

Federated learning achieves competitive accuracy with centralized training.

Models trained federatively show increased robustness.

Federated models are more sensitive to domain shifts.

Abstract

Deep learning models can enable accurate and efficient disease diagnosis, but have thus far been hampered by the data scarcity present in the medical world. Automated diagnosis studies have been constrained by underpowered single-center datasets, and although some results have shown promise, their generalizability to other institutions remains questionable as the data heterogeneity between institutions is not taken into account. By allowing models to be trained in a distributed manner that preserves patients' privacy, federated learning promises to alleviate these issues, by enabling diligent multi-center studies. We present the first federated learning study on the modality of cardiovascular magnetic resonance (CMR) and use four centers derived from subsets of the M\&M and ACDC datasets, focusing on the diagnosis of hypertrophic cardiomyopathy (HCM). We adapt a 3D-CNN network pretrained on action recognition and explore two different ways of incorporating shape prior information to the model, and four different data augmentation set-ups, systematically analyzing their impact on the different collaborative learning choices. We show that despite the small size of data (180 subjects derived from four centers), the privacy preserving federated learning achieves promising results that are competitive with traditional centralized learning. We further find that federatively trained models exhibit increased robustness and are more sensitive to domain shift effects.