Please use this identifier to cite or link to this item:
http://hdl.handle.net/2445/214020
Title: | Domain generalization in deep learning for contrast-enhanced imaging |
Author: | Sendra-Balcells, Carla Campello Román, Víctor Manuel Martín-Isla, Carlos Viladés Medel, David Descalzo, Martín Luis Guala, Andrea Rodríguez-Palomares, José Fernando Lekadir, Karim, 1977- |
Keywords: | Aprenentatge automàtic Imatges mèdiques Imatges per ressonància magnètica Diagnòstic per la imatge Machine learning Imaging systems in medicine Magnetic resonance imaging Diagnostic imaging |
Issue Date: | 1-Oct-2022 |
Publisher: | Elsevier Ltd |
Abstract: | Background: The domain generalization problem has been widely investigated in deep learning for noncontrast imaging over the last years, but it received limited attention for contrast-enhanced imaging. However, there are marked differences in contrast imaging protocols across clinical centers, in particular in the time between contrast injection and image acquisition, while access to multi-center contrast-enhanced image data is limited compared to available datasets for non-contrast imaging. This calls for new tools for generalizing single-domain, single-center deep learning models across new unseen domains and clinical centers in contrast-enhanced imaging. Methods: In this paper, we present an exhaustive evaluation of deep learning techniques to achieve generalizability to unseen clinical centers for contrast-enhanced image segmentation. To this end, several techniques are investigated, optimized and systematically evaluated, including data augmentation, domain mixing, transfer learning and domain adaptation. To demonstrate the potential of domain generalization for contrast-enhanced imaging, the methods are evaluated for ventricular segmentation in contrast-enhanced cardiac magnetic resonance imaging (MRI). Results: The results are obtained based on a multi-center cardiac contrast-enhanced MRI dataset acquired in four hospitals located in three countries (France, Spain and China). They show that the combination of data augmentation and transfer learning can lead to single-center models that generalize well to new clinical centers not included during training. Conclusions: Single-domain neural networks enriched with suitable generalization procedures can reach and even surpass the performance of multi-center, multi-vendor models in contrast-enhanced imaging, hence eliminating the need for comprehensive multi-center datasets to train generalizable models. |
Note: | Reproducció del document publicat a: https://doi.org/10.1016/j.compbiomed.2022.106052 |
It is part of: | Computers in Biology and Medicine, 2022, vol. 149 |
URI: | http://hdl.handle.net/2445/214020 |
Related resource: | https://doi.org/10.1016/j.compbiomed.2022.106052 |
ISSN: | 0010-4825 |
Appears in Collections: | Articles publicats en revistes (Matemàtiques i Informàtica) |
Files in This Item:
File | Description | Size | Format | |
---|---|---|---|---|
824745.pdf | 4.05 MB | Adobe PDF | View/Open |
This item is licensed under a
Creative Commons License