Quantitative analysis of PC MRI velocity maps: pulsatile flow in cylindrical vessels
Introduction
Atherosclerosis is the progressive condition that leads to a thickening of the arterial walls, corresponding to a reduction in the vessel’s elasticity. In severe cases, this process may eventually lead to the total occlusion of the artery, but in general any alterations in the innermost layer of the artery has a major influence on the disease of the vessel. Important indicators for the clinical diagnosis of atherosclerosis are the blood velocity and associated physical properties of the flow field [1], [2], [3]. In this particular application, magnetic resonance imaging has become a valuable non-invasive technique [4], [5].
Phase contrast (PC) velocity mapping has been used for both the qualitative and quantitative assessment of flow [4], [6], [7], [8]. The ability to obtain a temporal-spatial velocity map in this way permits a direct examination of blood velocities and flow rate throughout the cardiac cycle, which in turn are potentially useful in the diagnosis of vascular disease. Improvement on the basic experiment has been achieved through the introduction of novel methods of data acquisition [9], [10], [11], [12], [13] and the refinement of post-processing techniques that may be used to determine additional information about the velocity field [14], [15], [16], [17], [18].
Properties such as flow-rate [19], pressure [14], [15] and wall shear stress (WSS) [16], [17], [18] can be derived from the velocity data. The basic internal mechanical forces acting on a blood vessel are pressure and shear stress. These forces are related to the pulsatile flow of blood through the arteries. It is generally accepted that arterial WSS patterns influence the development of arterial disease [1], [2], [3]. The development of atherosclerosis, at least in its initial stages, may be related to the physiological response of arteries to the WSS on a localised basis. However, the inherent accuracy of specific PC MRI acquisitions, and post-processing techniques, has not been fully quantified for pulsatile flows.
Previous studies have demonstrated a close correlation between PC MRI and Doppler ultrasound, an established clinical technique, for both in vitro [19], [20] and in vivo [21] cases. However, caution is necessary when comparing different modalities such as these. The MRI technique provides an estimate of the mean velocity, while Doppler ultrasound measures the peak velocity within a voxel. The discrepancies between the quantities measured must be quantitatively addressed before inter-modality comparisons are feasible. Validation with an analytical model, although providing a reliable alternative have rarely been performed [22], [23]. Velocity maps for a pulsatile flow were obtained by Frayne et al [22], where the flow rate was measured directly from the MRI data, which was subsequently used to calculate the velocity with an analytical equation of motion. However, this procedure did not test the data for the presence of systematic errors in the velocity measurement, it only verified the analytical equation of motion. A more comprehensive examination would determine the flow rate from the flow simulator’s command waveform.
With the increased interest in information extracted from the flow field, the inherent accuracy of the measured velocities is of intrinsic importance. In particular, the WSS are extremely sensitive to errors in the velocity maps. As a result of the relatively limited spatial resolution of PC MRI data, the WSS is usually calculated from an empirical fit to the velocity map close to the vessel wall [16], [18], [24], [25]. However, this approach requires a phenomenological model that satisfactorily represents the velocity maps, and an accurate estimate for the position of the vessel wall. The calculation is subject to several potential errors, and therefore any new method must ideally be verified with an analytical model.
The aim of the present study was to investigate the agreement between an analytical calculation of the flow field, defined from a flow rate waveform, and a retrospectively gated PC MRI velocity map, where the flow was generated with a computer controlled flow simulator. A technique is developed in which the flow field at the measurement site is known for an arbitrary waveform, without the need for a separate flow rate calibration. Hence, the accuracy of a PC velocity measurement was established with an in vitro flow simulation in order to determine the inherent accuracy of this particular technique.
Section snippets
Theory
The flow of arterial blood is pulsatile, and thus may be characterised with a periodic function. An elegant method to analyse such a waveform is through a Fourier series, which defines the wave with a steady-state term and a series of sinusoidal components [26]. Calculation of the Fourier coefficients allows the decomposition of the original waveform into a series of simple harmonic terms, which may be individually studied. Subsequently, application of the equation of motion for the separate
Flow simulation
The flow phantom consisted of a reinforced PVC tube, with an internal radius of 3.15 mm. To maintain a straight rigid geometry, the tube was placed within a channel machined into a sheet of MDF (medium density fibreboard). The entrance length for the region of interest was approximately 1.2 m, which was twice the calculated inlet length required [27], and hence was sufficient to establish fully developed flow. Three metres of additional tubing were placed between the pump and the phantom, in
Fourier analysis
The period of the flow waveform was initially established, which corresponded to the period of the fundamental component of the Fourier series, was defined as the time between successive flow peaks. Addition of successive harmonic terms permitted a more accurate duplication of the command waveform. Reproduction of the waveform, from a given number of harmonic components, was qualitatively ascertained by inspection (Fig. 1). However, a more quantitative measure of the analysis was established
Discussion
The experiments described here were used to demonstrate the accuracy of the velocity measurement obtained with the triggered update implementation of retrospective gating, and the subsequent derivation of the wall shear stress. The Womersley oscillatory flow theory was shown to satisfactorily describe the PC MRI velocity data acquired for a sinusoidal flow rate waveform. The extension of this theory for a more general pulsatile waveform, through the use of a Fourier series, surprisingly
Acknowledgements
The financial support of the following organisations is gratefully acknowledged: the Engineering and Physical Science Research Council (MBR), and the European Union (UK). The study was performed at the SHEFC Brain Imaging Research Centre for Scotland.
References (44)
- et al.
Flow properties of fast three-dimensional sequences for MR angiography
Magn Reson Imaging
(1999) - et al.
Multiphase segmented k-space velocity mapping in pulsatile flow waveforms
Magn Reson Imaging
(1998) - et al.
In vivo wall shear stress measurement by magnetic resonance velocity mapping in the normal human abdominal aorta
Eur J Vasc Endovasc Surg
(1997) - et al.
The accuracy of magnetic resonance velocity measurements in stenotic flow
J Biomechanics
(1996) - et al.
A non-invasive method to estimate wall shear rate using ultrasound
Ultrasound in Med Biol
(1995) - et al.
Accuracy and precision of time-averaged flow as measured by nontriggered 2D phase-contrast MR angiography, a phantom evaluation
Magn Reson Imaging
(1995) - et al.
Measuring blood flow by nontriggered 2D phase-contrast MR angiography
Magn Reson Imaging
(1996) - et al.
Pulsatile flow and atherosclerosis in human carotid bifurcationpositive correlation between plaque location and low and oscillating shear stress
Arteriosclerosis
(1985) - et al.
Haemodynamics and atherosclerosis—insights and perspectives gained from studies of human arteries
Arch Path Lab Med
(1988) - et al.
Phase contrast magnetic resonance imaging
Magn Reson Q
Quantitative NMR imaging of flow
Concepts in Magn Reson
A flow velocity zeugmatographic interlace for NMR imaging in humans
Magn Reson Imaging
The application of phase shifts in NMR for Flow Measurement
Magn Reson Med
Evaluation of magnetic resonance velocimetry for steady flow
J Biomech Eng
Simultaneous acquisition of phase-contrast angiograms and stationary-tissue images with Hadamard encoding of flow-induced phase shifts
J Magn Reson Imaging
Phase-contrast MR angiography with reduced acquisition timenew concepts in sequence design
J Magn Reson Imaging
Magnetic resonance imaging using a fast spiral phase contrast sequence
Magn Reson Med
Three-dimensional time resolved (4D) relative pressure mapping using magnetic resonance imaging
J Magn Reson Imaging
Nonsubtractive spiral phase contrast velocity imaging
Magn Reson Med
Cited by (29)
Assessment of Pulmonary Hypertension. What CT and MRI Can Provide
2011, Academic RadiologyCitation Excerpt :Isovolumetric relaxation time was reported to have a good correlation with RV mass and RV afterload and, furthermore, better response to medication to decrease RV afterload (142). Phase-contrast MRI techniques may provide reliable quantitative measurements of blood flow: velocities and flow volumes passing through blood vessels and cardiac valves, such as SV, CO, EF; valvular regurgitant fractions; and the extent of cardiac shunts, which correlate well with ex vivo measurements and in vivo measurements derived from RHC and Doppler echocardiography (108,121,122,143,144). They may also produce ventricular diastolic filling patterns, PVR, acceleration time (AT, defined as time from onset to the peak velocity of flow in the main PA), and acceleration volume by assessing the time-velocity curve.
Quality-by-Design (QbD): An integrated approach for evaluation of powder blending process kinetics and determination of powder blending end-point
2009, Journal of Pharmaceutical SciencesCitation Excerpt :Knowledge of their strengths and limitations are essential for establishing an integrated QbD approach for process monitoring and controlling. This method has been widely used in the areas of imaging and signal processing.19–21 Previously we18 reported the application of this method for Terahertz spectral data analysis.
Process analytical technology (PAT): Quantification approaches in terahertz spectroscopy for pharmaceutical application
2008, Journal of Pharmaceutical SciencesCitation Excerpt :For pharmaceutical materials studies, the method is relative easy as we only deal with medium molecular weight substances. This method has been finding widespread applications in the areas of imaging and signal processing.15–17 However, to the best of our knowledge, this method has not been used for THz spectral data analysis since THz spectroscopy is an emerging technology.
MRI evaluation of right heart functions in children with mild cystic fibrosis
2023, Cardiology in the YoungMeasuring stroke volume: Impedance cardiography vs phasecontrast magnetic resonance imaging
2017, American Journal of Critical CareEmpirical retinal venous pulse wave velocity using modified photoplethysmography
2023, BMC Research Notes
- 1
Present address: Department of Physics and Astronomy, The University of Edinburgh, Mayfield Road, Edinburgh EH9 3JZ UK.