Coupling synthetic and real-world data for a deep learning-based segmentation process of 4D flow MRI

Simone Garzia; Martino Andrea Scarpolini; Marilena Mazzoli; Katia Capellini; Angelo Monteleone; Filippo Cademartiri; Vincenzo Positano; Simona Celi

doi:10.1016/j.cmpb.2023.107790

Coupling synthetic and real-world data for a deep learning-based segmentation process of 4D flow MRI

Comput Methods Programs Biomed. 2023 Dec:242:107790. doi: 10.1016/j.cmpb.2023.107790. Epub 2023 Sep 6.

Authors

Simone Garzia¹, Martino Andrea Scarpolini², Marilena Mazzoli¹, Katia Capellini³, Angelo Monteleone⁴, Filippo Cademartiri⁴, Vincenzo Positano³, Simona Celi⁵

Affiliations

¹ BioCardioLab, UOC Bioingegneria, Fondazione Toscana G Monasterio, Via Aurelia Sud, Massa, 54100, Italy; Department of Information Engineering, University of Pisa, Via Caruso, Pisa, 56122, Italy.
² BioCardioLab, UOC Bioingegneria, Fondazione Toscana G Monasterio, Via Aurelia Sud, Massa, 54100, Italy; Department of Industrial Engineering, University of Rome "Tor Vergata", Via del Politecnico, Roma, 00133, Italy.
³ BioCardioLab, UOC Bioingegneria, Fondazione Toscana G Monasterio, Via Aurelia Sud, Massa, 54100, Italy.
⁴ Department of Radiology, Fondazione Toscana G Monasterio, Via Moruzzi, Pisa, 56122, Italy.
⁵ BioCardioLab, UOC Bioingegneria, Fondazione Toscana G Monasterio, Via Aurelia Sud, Massa, 54100, Italy. Electronic address: s.celi@ftgm.it.

PMID: 37708583
DOI: 10.1016/j.cmpb.2023.107790

Abstract

Background and objective: Phase contrast magnetic resonance imaging (4D flow MRI) is an imaging technique able to provide blood velocity in vivo and morphological information. This capability has been used to study mainly the hemodynamics of large vessels, such as the thoracic aorta. However, the segmentation of 4D flow MRI data is a complex and time-consuming task. In recent years, neural networks have shown great accuracy in segmentation tasks if large datasets are provided. Unfortunately, in the context of 4D flow MRI, the availability of these data is limited due to its recent adoption in clinical settings. In this study, we propose a pipeline for generating synthetic thoracic aorta phase contrast magnetic resonance angiography (PC_MRA) to expand the limited dataset of patient-specific PC_MRA images, ultimately improving the accuracy of the neural network segmentation even with a small real dataset.

Methods: The pipeline involves several steps. First, a statistical shape model is used to synthesize new artificial geometries to improve data numerosity and variability. Secondly, computational fluid dynamics simulations are employed to simulate the velocity fields and, finally, after a downsampling and a signal-to-noise and velocity limit adjustment in both frequency and spatial domains, volumes are obtained using the PC_MRA formula. These synthesized volumes are used in combination with real-world data to train a 3D U-Net neural network. Different settings of real and synthetic data are tested.

Results: Incorporating synthetic data into the training set significantly improved the segmentation performance compared to using only real data. The experiments with synthetic data achieved a DICE score (DS) value of 0.83 and a better target reconstruction with respect to the case with only real data (DS = 0.65).

Conclusion: The proposed pipeline demonstrated the ability to increase the dataset in terms of numerosity and variability and to improve the segmentation accuracy for the thoracic aorta using PC_MRA.

Keywords: 4D flow MRI; Aorta; CFD; Deep learning; Synthetic data.

MeSH terms

Blood Flow Velocity
Deep Learning*
Humans
Magnetic Resonance Angiography / methods
Magnetic Resonance Imaging / methods
Neural Networks, Computer