EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

Shape-driven deep neural networks for fast acquisition of aortic 3D pressure and velocity flow fields

Endrit Pajaziti iD, Javier Montalt-Tordera, Claudio Capelli, Raphaël Sivera, Emilie Sauvage, Michael Quail, Silvia Schievano, Vivek Muthurangu

PLOS Computational Biology · 24.04.2023 · 19 (4) · e1011055

DOI: 10.1371/journal.pcbi.1011055 ↗

Kurz gefasst

Fachartikel in der Zeitschrift PLOS Computational Biology, online erschienen am 24.04.2023. Verfasst von Endrit Pajaziti und 7 weiteren Autor:innen. Beteiligte Einrichtung: University College London, Institution of Cardiovascular Science, London, United Kingdom. Hauptthemen laut MEDLINE: Models, Cardiovascular, Hemodynamics. Die Arbeit wurde bisher 34-mal zitiert (Crossref) und nennt 47 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: British Heart Foundation, Medical Research Council, UK Research and Innovation.

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Computational fluid dynamics (CFD) can be used to simulate vascular haemodynamics and analyse potential treatment options. CFD has shown to be beneficial in improving patient outcomes. However, the implementation of CFD for routine clinical use is yet to be realised. Barriers for CFD include high computational resources, specialist experience needed for designing simulation set-ups, and long processing times. The aim of this study was to explore the use of machine learning (ML) to replicate conventional aortic CFD with automatic and fast regression models. Data used to train/test the model consisted of 3,000 CFD simulations performed on synthetically generated 3D aortic shapes. These subjects were generated from a statistical shape model (SSM) built on real patient-specific aortas (N = 67). Inference performed on 200 test shapes resulted in average errors of 6.01% ±3.12 SD and 3.99% ±0.93 SD for pressure and velocity, respectively. Our ML-based models performed CFD in ∼0.075 seconds (4,000x faster than the solver). This proof-of-concept study shows that results from conventional vascular CFD can be reproduced using ML at a much faster rate, in an automatic process, and with reasonable accuracy.

Schlagwörter

MeSH-Schlagwörter (MEDLINE)

Models, CardiovascularHemodynamicsHumansBlood Flow VelocityComputer SimulationNeural Networks, ComputerHydrodynamics

Autor:innen und Institutionen

  • University College London, Institution of Cardiovascular Science, London, United Kingdom.
  • Great Ormond Street Hospital, Cardiac Unit, London, United Kingdom.

ORCID-Profile: Endrit Pajaziti

Förderung

  • British Heart Foundation
  • Medical Research Council
  • UK Research and Innovation
  • Dorothy M. Davis Heart and Lung Research Institute
  • Engineering and Physical Sciences Research Council
  • Action medical research
  • European Research Council

Zitieren

APA

Pajaziti, E., Montalt-Tordera, J., Capelli, C., Sivera, R., Sauvage, E., Quail, M., Schievano, S., & Muthurangu, V. (2023). Shape-driven deep neural networks for fast acquisition of aortic 3D pressure and velocity flow fields. PLOS Computational Biology, 19(4), e1011055. https://doi.org/10.1371/journal.pcbi.1011055

Vancouver

Pajaziti E, Montalt-Tordera J, Capelli C, Sivera R, Sauvage E, Quail M, et al. Shape-driven deep neural networks for fast acquisition of aortic 3D pressure and velocity flow fields. PLOS Computational Biology. 2023;19(4):e1011055. doi:10.1371/journal.pcbi.1011055

BibTeX
@article{Pajaziti2023,
  author = {Pajaziti, Endrit and Montalt-Tordera, Javier and Capelli, Claudio and Sivera, Raphaël and Sauvage, Emilie and Quail, Michael and Schievano, Silvia and Muthurangu, Vivek},
  title = {Shape-driven deep neural networks for fast acquisition of aortic 3D pressure and velocity flow fields},
  journal = {PLOS Computational Biology},
  year = {2023},
  volume = {19},
  number = {4},
  pages = {e1011055},
  doi = {10.1371/journal.pcbi.1011055}
}
Kostenlos starten →