King's College London

Research portal

Non-invasive local pulse wave velocity using 4D-flow MRI

Research output: Contribution to journalArticlepeer-review

Joaquín Mura, Julio Sotelo, Hernán Mella, James Wong, Tarique Hussain, Bram Ruijsink, Sergio Uribe

Original languageEnglish
Article number103259
JournalBiomedical Signal Processing and Control
Volume71
DOIs
PublishedJan 2022

Bibliographical note

Funding Information: This publication was funded by ANID - Millennium Science Initiative Program – NCN17_129. Also, has been supported by FONDECYT Postdoctorado #3170737, the ANID Ph.D. Scholarship #21170592, and ANID FONDECYT de Iniciación en Investigación #11200481, ANID FONDECYT #1181057, and Initiation-Reactivation USM project PI_LII_2021_86. Funding Information: This publication was funded by ANID - Millennium Science Initiative Program ? NCN17_129. Also, has been supported by FONDECYT Postdoctorado #3170737, the ANID Ph.D. Scholarship #21170592, and ANID FONDECYT de Iniciaci?n en Investigaci?n #11200481, ANID FONDECYT #1181057, and Initiation-Reactivation USM project PI_LII_2021_86. Publisher Copyright: © 2021 Elsevier Ltd

King's Authors

Abstract

Pulse Wave Velocity (PWV) corresponds to the velocity at which pressure waves, generated by the systolic contraction in the heart, propagate along the arterial tree. Due to the complex interplay between blood flow and the artery wall, PWV is related to inherent mechanical properties and arterial morphology. PWV has been widely accepted as a biomarker and early predictor to evaluate global arterial distensibility. Still, several local abnormalities often remain hidden or difficult to detect using non-invasive techniques. Here, we introduce a novel method to efficiently construct a local estimate of PWV along the aorta using 4D-Flow MRI data. A geodesic distance map was used to track advancing pulses for efficient flow calculations, based on the observation that the propagation of velocity wavefronts strongly depends on the arterial morphology. This procedure allows us a robust evaluation of the local transit time due to the pulse wave at each position in the aorta. Moreover, the estimation of the local PWV map did not require centerlines, and the final result is projected back to 3D using the same geodesic map. We evaluated PWV values in healthy young and adult volunteers and patients with univentricular physiology after a Fontan procedure. Our method is fast, semi-automatic, and depicts differences between young versus adult volunteers and young volunteers versus Fontan patients, showing consistent results compared to global methods. Remarkably, the technique could detect local differences of PWV on the aortic arch for all subjects, being consistent with previous findings of reduced PWV in the aortic arch.

View graph of relations

© 2020 King's College London | Strand | London WC2R 2LS | England | United Kingdom | Tel +44 (0)20 7836 5454