4D Flow MRI: Validation of Advanced Flow Parameters
Julio Sotelo1,2,3
1School of Biomedical Engineering, Universidad de Valparaíso, Valparaíso, Chile, 2Millennium Institute for Intelligent Healthcare Engineering iHEALTH, Santiago, Chile, 3Biomedical Imaging Center, Pontificia Universidad Católica de Chile, Santiago, Chile

Synopsis

Keywords: Cardiovascular: Hemodynamics

In this educational talk, we will talk about the relationship between the velocity to noise ratio and acquisition related parameters. The preprocessing of 4D flow MRI is also important, because the acquisition could be affected by gradient-induced eddy currents, concomitant gradient fields, gradient non-linearity and phase wraps resulting in velocity aliasing, these problems need to be addressed to avoid additional errors in the quantification of hemodynamic parameters. Finally we will explore different methods used to validate 4D flow MRI advanced parameters as; wall shear stress, vorticity, helicity density, viscous energy loss, kinetic energy, circulation, pressure gradients, among others.

Introduction

In the last ten years, hemodynamic parameters quantified by the 4D Flow magnetic resonance technique have emerged as essential biomarkers in the early diagnosis of different cardiovascular diseases, providing new insights into complex flows [1]. Parameters such as wall shear stress, vorticity, helicity density, viscous energy loss, kinetic energy, circulation, pressure gradients, among others, have been extensively studied [2]. But, how much do we know about the validation of these hemodynamic parameters? In this work we explore that to have an adequate validation, we must consider that the 4D Flow MRI data must be correctly acquired and preprocessed. This work was divided in three different topics; a) relationship between the velocity to noise ratio and acquisition related parameters, b) the preprocessing steps necessary to avoid problems like eddy current effect, maxwell term, gradient field non-linearity and phase wraps and c) methods currently used for the validation of hemodynamics parameters.

4D Flow MRI and Velocity to Noise Ratio

The velocity-to-noise ratio (VNR) is a measure of the quality of 4D Flow MRI data. VNR is calculated by dividing the mean velocity (V) of the blood flow by the standard deviation of the velocity noise (N) in the image [3]. A higher VNR indicates a better-quality image, which is important for accurate measurement of blood flow velocity. Different parameters related to the acquisition can affect the VNR. Three of the most important are the velocity encoding value (Venc), spatial resolution, and temporal resolution. An overestimated velocity encoding value can generate a reduction in the VNR, and an underestimation can generate aliasing in the image. High spatial and temporal resolution increases the noise in the image which decreases the VNR. There are other acquisition related parameters that can affect the VNR, as field of view, high acceleration parameters used and intravoxel dephasing [1]. It is important to obtain a higher VNR, because this can lead to more accurate and reproducible measurements of blood flow velocity, which can improve the clinical utility of 4D Flow MRI.

4D Flow MRI and Preprocessing

Errors that cause bias in the measured velocity field in 4D Flow MRI are often related to gradient-induced eddy currents, concomitant gradient fields, and gradient non-linearity [4]. While clinical MRI systems can correct and calibrate the latter two sources of error with sufficient accuracy, eddy currents cause a background phase error that must be corrected. Post-processing methods, such as static-tissue interpolation offset correction, can be used to address this error with equivalent performance [5]. When the maximum blood flow velocity exceeds the chosen Venc, aliasing can be generated in velocity images. Phase unwrapping can be used to improve the accuracy of flow and velocity measurements in such cases, but it requires a visual inspection of the peak systolic and diastolic cardiac phases to ensure that the velocity images are free of uncorrected velocity aliasing [6]. In addition, an appropriate segmentation of the region of interest is an important aspect of the flow and velocity quantification process [7].

4D Flow MRI and Validation of Hemodynamics Parameters

Validating hemodynamic parameters in 4D flow MRI data is important for ensuring the accuracy and reliability of flow measurements. Hemodynamic parameters, such as wall shear stress, vorticity, helicity density, viscous energy loss, kinetic energy, circulation, pressure gradients, could be important indicators of cardiovascular disease [1,2]. In the literature most of the validation methods of these parameters involves comparing the MRI measurements to those obtained from in-silico models (Poiseuille equation [8], Womersley Equation [8] and Lamb-Ossen equation [9,10]), in-vitro models (Realistic Phantoms Studies [11,12]), and numerical simulations-based models [13,14]. This comparison can help to identify and correct any errors in the quantification of the hemodynamics parameters and to ensure that the measurements are consistent with expected values.

Conclusions

Validating hemodynamic parameters in 4D flow MRI data is particularly important for clinical applications, where the data is used to diagnose and monitor cardiovascular disease. Accurate and reliable measurements of hemodynamic parameters can help clinicians to make better informed decisions about patient care and treatment and can ultimately lead to improved patient outcomes. Most of the validation method used in the literature are in base of in-silico, in-vitro and simulations-based models.

Acknowledgements

Thanks to ANID – Millennium Science Initiative Program – ICN2021_004 and ANID – FONDECYT de iniciación en investigación 11200481.

References

1.- Soulat G, McCarthy P, Markl M. 4D Flow with MRI. Annu Rev Biomed Eng. 2020 Jun 4;22:103-126. doi: 10.1146/annurev-bioeng-100219-110055.

2.- Zhuang B, Sirajuddin A, Zhao S, et al. The role of 4D flow MRI for clinical applications in cardiovascular disease: current status and future perspectives. Quant Imaging Med Surg. 2021 Sep;11(9):4193-4210. doi: 10.21037/qims-20-1234.

3.- Nilsson A, Markenroth Bloch K, Carlsson M, et al. Variable velocity encoding in a three-dimensional, three-directional phase contrast sequence: Evaluation in phantom and volunteers. J Magn Reson Imaging. 2012 Dec;36(6):1450-9. doi: 10.1002/jmri.23778.

4.- Dyverfeldt P, Bissell M, Barker AJ, et al. 4D flow cardiovascular magnetic resonance consensus statement. J Cardiovasc Magn Reson. 2015 Aug 10;17(1):72. doi: 10.1186/s12968-015-0174-5.

5.- Minderhoud SCS, van der Velde N, Wentzel JJ, et al. The clinical impact of phase offset errors and different correction methods in cardiovascular magnetic resonance phase contrast imaging: a multi-scanner study. J Cardiovasc Magn Reson. 2020 Sep 17;22(1):68. doi: 10.1186/s12968-020-00659-3.

6.- Loecher M, Schrauben E, Johnson KM, et al. Phase unwrapping in 4D MR flow with a 4D single-step laplacian algorithm. J Magn Reson Imaging. 2016 Apr;43(4):833-42. doi: 10.1002/jmri.25045.

7.- Casciaro ME, Pascaner AF, Guilenea FN, et al. 4D flow MRI: impact of region of interest size, angulation and spatial resolution on aortic flow assessment. Physiol Meas. 2021 Apr 6;42(3). doi: 10.1088/1361-6579/abe525.

8.- Sotelo J, Urbina J, Valverde I, et al. Quantification of wall shear stress using a finite-element method in multidimensional phase-contrast MR data of the thoracic aorta. J Biomech. 2015 Jul 16;48(10):1817-27. doi: 10.1016/j.jbiomech.2015.04.038.

9.- Sotelo J, Bissell MM, Jiang Y, et al. Three-dimensional quantification of circulation using finite-element methods in four-dimensional flow MR data of the thoracic aorta. Magn Reson Med. 2022 Feb;87(2):1036-1045. doi: 10.1002/mrm.29004.

10.- Sotelo J, Urbina J, Valverde I, et al. Three-dimensional quantification of vorticity and helicity from 3D cine PC-MRI using finite-element interpolations. Magn Reson Med. 2018 Jan;79(1):541-553. doi: 10.1002/mrm.26687.

11.- Urbina J, Sotelo JA, Springmüller D, et al. Realistic aortic phantom to study hemodynamics using MRI and cardiac catheterization in normal and aortic coarctation conditions. J Magn Reson Imaging. 2016 Sep;44(3):683-97. doi: 10.1002/jmri.25208.

12.- Montalba C, Urbina J, Sotelo J, et al. Variability of 4D flow parameters when subjected to changes in MRI acquisition parameters using a realistic thoracic aortic phantom. Magn Reson Med. 2018 Apr;79(4):1882-1892. doi: 10.1002/mrm.26834.

13.- Sotelo J, Urbina J, Valverde I, et al. 3D Quantification of Wall Shear Stress and Oscillatory Shear Index Using a Finite-Element Method in 3D CINE PC-MRI Data of the Thoracic Aorta. IEEE Trans Med Imaging. 2016 Jun;35(6):1475-87. doi: 10.1109/TMI.2016.2517406.

14.- Marlevi D, Ruijsink B, Balmus M, et al. Estimation of Cardiovascular Relative Pressure Using Virtual Work-Energy. Sci Rep. 2019 Feb 4;9(1):1375. doi: 10.1038/s41598-018-37714-0.

Proc. Intl. Soc. Mag. Reson. Med. 31 (2023)