doi:10.1016/S1090-7807(03)00044-2
Copyright © 2003 Elsevier Science (USA). All rights reserved.
Flow effects in long-range dipolar field MRI
References and further reading may be available for this article. To view references and further reading you must
purchase this article.
Paulo Loureiro de Sousa
,
, Daniel Gounot and Daniel Grucker
Institut de Physique Biologique, Université Louis Pasteur, UMR 7004-CNRS, 4, rue Kirschléger, Strasbourg Cedex 67085, France
Received 16 October 2002;
revised 23 January 2003.
Available online 29 April 2003.
Abstract
Incoherent spin motion, such as diffusion, can lead to significant signal loss in multiple spin echoes (MSE) experiments, sometimes to its complete extinction. Coherent spin motion, such as laminar flow, can also modify the magnetization in MSE imaging and yield additional contrast. Our experimental results indicate that MSE is flow-sensitive. Our theoretical analysis and experimental results show how the effect of the distant dipolar field can be annihilated by flow. This effect can be quantified by directly solving the nonlinear Bloch equation, taking into account the deformation of the dipolar field by motion. Unexpected results have been observed, such as a recovery of the dipolar interaction due to flow in the “magic angle” condition.
Author Keywords: MSE; Distant dipolar field; CRAZED; Flow
Fig. 1. Sequences used in MSE: (a) the basic MSE sequence consisting of two r.f. pulses (angle 90° and β°) and a steady gradient field; (b) the nth-order MSE sequence (or CRAZED sequence), consisting of two r.f. pulses (angle 90° and β°) and the two gradient pulses of duration δ and strength G and nG, respectively. If t1=τ then t2=nτ. (c) The MSE–MRI sequence with a slice-selective refocusing (180°) pulse. Only the first pulse is phase-cycled. The gradient g can be chosen to point to any direction (cf. text for greater details).
Fig. 2. Simulation of the main parameters of the distant dipolar field in a laminar flow. (a) Calculated variation of θ, the angle between the vector
k and , as a function of the normalized radial position ρ/ρ
0 to a Poiseuille flow where ρ
0 is the center of the tube. Each line corresponds to distinct values of the adimensional parameter
vmaxt/ρ
0 from Eq. (
29): (–) 0, (– –) 0.5, (

) 1, (–·–) 1.5, (–··–) 2. (b) Calculated variation of Λ=(3cos
2θ−1)/2, using θ values obtained in (a).
Fig. 3. Comparison of conventional (SE) images and multiple spin echoes (MSE) images. (a) SE and (b) MSE images of stationary fluid using the sequence shown in
Fig. 1c. Signal-to-noise ratios were 158 and 64, respectively. (c) SE and (d) MSE images of laminar fluid flow (
vmax=5 cm/s), using the same sequence as (a) and (b). MSE images were obtained by setting
G2=−2×
G1=145 mT/m, while in SE images
G2=−
G1=145 mT/m, keeping the other parameters constant (
t1,
t2, δ, TR, FOV, and matrix size). Relevant pulse sequence parameters for MSE–MRI are described in
Section 3. Images have been normalized to their maximum intensity.
Fig. 4. Comparison of experimental (symbols) and simulated data (lines) from radial profiles obtained to (a) MSE and (b) SE images of laminar fluid flow, to five distinct
vmax values (in cm/s). simulation data: (–) 0, (– –) 5, (

) 8, (–·–) 10.8, and (–··–) 14; experimental data: (□) 0, (•) 5, (

) 8, (

) 10.8, and (

) 14. Simulation data were obtained from Eqs. (
(26) and
(29)). Absolute signal intensities |
M+(ρ,
v)| were normalized by the mean intensity of the respective images without flow .
Fig. 5. MSE image in the “magic angle” condition (Θ=54.7°) of a stationary fluid (a) and a Poiseuille fluid flow (b). Simulation of MSE image of the flow in the “magic angle” condition is displayed on (c). In both (experimental and simulated) images
vmax was set to 15 cm/s. Images have been normalized to their maximum intensity.
Fig. 6. Experimental (symbols) data from diametrical profiles obtained from the images displayed in
Fig. 5: (○) stationary fluid and (▪) laminar flow. The line is a profile taken from the simulated image in
Fig. 5.
Corresponding author