A validation of a flow quantification by MR phase mapping software

https://doi.org/10.1016/S0720-048X(97)00105-8Get rights and content

Abstract

Aim: We evaluated a Siemens software of flow quantification (FQ) by MR phase mapping, in the framework of a common practical use. Methods: Experiments with a laminar flow phantom and in vivo pulsatile flow were performed. In particular, FQ in ascending aorta was investigated in healthy volunteers. Results and conclusion: Flow phantom experiments reveal that the FQ slightly underestimates (8% on the average) actual velocities (mean velocities over a vessel area), and also that velocity uncertainties are related to the encoding velocity value, whatever the measured velocity. Furthermore, using well characterized working criteria, we found low intraobserver variability and negligible interobserver variability in ascending aorta FQs. The role played by the choice of reference area in FQ accuracy is emphasized. When recording several cardiac cycles during the same acquisition, it is shown that the FQ software may provide erroneous results. Several comments for FQ software use in the ascending aorta are added.

Introduction

Techniques of flow quantification (FQ) by MR phase mapping have recently been validated 1, 2, 3, 4. The present study is a complementary contribution to these previous validations concerning a commercial Siemens software. Although this work involved a particular software/hardware (derived from a major manufacturer), we expect the results to be useful for the radiologic community. Current use of FQ software necessitates knowing and understanding the uncertainties of the computed values. We have evaluated these uncertainties in the various conditions that we describe in detail.

We first evaluated phantom measurement uncertainties in a very simple case, an experimental permanent laminar flow model. Then we performed measurements of pulsatile flow in healthy volunteers. Various comparisons were made between FQ performed in ascending aorta, FQ performed in right and left pulmonary arteries (sum) and also left ventricular stroke volumes (end diastolic and end systolic volume difference: EDV-ESV) that were measured from left ventricular outlines. The major part of the work deals with FQ measurements in the ascending aorta for assessment of intraobserver, interobserver and interexamination variability. Choice of FQs in the ascending aorta was made because of highest velocities and hence, the maximal variations of velocity are found there. This choice was made also because of the large size of this vessel and its constant quasicylindrical shape over the cardiac cycle.

Section snippets

Flow quantification software

The FQ software used is a Siemens product (version 3b31a) that was developed on a field even echo rephasing (FEER) concept, flow encoding and flow compensation being achieved alternatively. Use of a proper encoding velocity avoids aliasing and the software thus allows quantification of mean flow over a region of interest (ROI) sketched in images that are acquired every 30 ms. The number of successive images available cannot exceed 64. Both a magnitude image and a phase image are shown on the

Results

Results of experiments using a laminar flow model are shown in Table 1. A comparison between expected velocities and velocities measured by means of three encoding velocities is presented. All measured velocities (n=15), except one, were lower than expected velocities. The results also show that uncertainty for a given encoding velocity did not significantly vary with the value of the measured velocity. For 150, 75 and 40 cm/s encoding velocities averaged uncertainties were 3, 1 and 0.8 cm/s,

Laminar flow phantom experiments

Nine different velocity measurements performed with 150 cm/s encoding velocity showed that the FQ software by phase mapping underestimated expected (actual) velocities by an average of 8%. This was close to Burkart's results [4], who found a 6% underestimation with flow phantom experiments (encoding velocity was not precisely reported in this particular experiment). We suggest this underestimation could be correlated with intravoxel phase dispersion due to the velocity gradient that is found at

Acknowledgements

This work was made possible by anonymous volunteers. S. Dupond's flow-phantom work was very useful for the present experiments.

References (9)

  • D.N. Firmin et al.

    In vivo validation of MR velocity imaging

    J Comput Assisted Tomogr

    (1987)
  • C. Kondo et al.

    Right and left ventricular stroke volume measurements with velovity-encoded cine MR imaging: In vitro and in vivo validation

    A.J.R

    (1991)
  • Buonocore MH, Bogren H. A technique to estimate the accuracy and precision of phase contrast flow measurement. Ninth...
  • D.J. Burkart et al.

    Volumetric flow rates in the portal venous system: measurement with Cine phase-contrast MR imaging

    A.J.R

    (1993)
There are more references available in the full text version of this article.

Cited by (30)

  • Optimized super-selective Arterial Spin Labeling for quantitative flow territory mapping

    2018, Magnetic Resonance Imaging
    Citation Excerpt :

    Using the optimized implementation of super-selective ASL, measurements of the labeling efficiency were subsequently performed using flow information from quantitative phase-contrast angiography (PCA) [9]. PCA was used as technique of choice given that it is an established technique with reported error rates below 5% and a negligible intra- and interobserver variability [10, 11]. In this study, quantitative PCA was performed separately prior to each super-selective scan.

  • The effect of end-range cervical rotation on vertebral and internal carotid arterial blood flow and cerebral inflow: A sub analysis of an MRI study

    2015, Manual Therapy
    Citation Excerpt :

    MRA offers a more complete picture of blood flow to the brain in response to movement, through simultaneous consideration of all four arteries and subsequently TCI, than the isolated study of individual arteries. Additionally, MRA offers more detailed imaging and evaluation of blood flow characteristics than ultrasound (Laffon et al., 1998; Paciaroni et al., 2005). Future studies could repeat this methodology in patients with an absent cervical or cerebral vessel or those with ischaemic symptoms; practically however this may present some ethical challenges.

  • In vitro validation of flow measurements in an aortic nitinol stent graft by velocity-encoded MRI

    2011, European Journal of Radiology
    Citation Excerpt :

    This implies that changes of haemodynamics after EVR as well as during a longer follow-up period could be accurately detected by VEC MRI. Previous studies presented the accuracy of VEC MRI for measuring aortic flow [26–28]. In the present study we observed a tendency of VEC MRI to slightly underestimate aortic flow which had been indicated by previous studies [26,28].

View all citing articles on Scopus
View full text