Elimination of oblique flow artifacts in magnetic resonance imaging.

We present an analysis of how flow oblique to the frequency-encoding direction generates displacement artifacts in MR imaging and show that for flow which has constant velocity between the start of the phase encoding and the center of the echo it is possible to eliminate these artifacts by gradient moment nulling in the phase-encoding direction. However, unlike the standard moment nulling calculations for flow compensating the frequency-encode and slice-selection gradients, the phase-encoding first moment must be nulled specifically with respect to the echo center. Limitations of this method imposed by finite gradient strengths are analyzed. In 3D volume acquisitions with two axes phase encoded it is possible to correct for oblique flow in all directions, and this is demonstrated in images of a human volunteer. Correction for oblique flow displacement artifacts may be particularly useful in quantitative flow and angiographic applications.     

Gradient moment smoothing: A new flow compensation technique for multi-shot echo-planar imaging

This work identifies an additional source of phase error across k_y in multi-shot echo-planar imaging resulting from flow or motion along the phase-encoding direction. A velocity-independent flow compensation technique, gradient moment smoothing, is presented that corrects this error by forcing the phase to have smooth quadratic behavior. The correction is implemented, without compromising scan time, by changing the first moment of a bipolar prephaser pulse on a shot-by-shot basis. In phantom and in vivo experiments, gradient moment smoothing effectively eliminates ghosting and signal loss due to phase-encoding flow. When used in conjunction with a “flyback” echo-planar readout, which compensates for flow in the frequency-encoding direction, gradient moment smoothing renders multi-shot echo-planar imaging relatively insensitive to in-plane flow. This can make multi-shot echo-planar imaging a viable technique for accurately imaging in-plane flow and may desensitize it to the otherwise serious problem of in-plane motion.     

Three-dimensional time-of-flight MR angiography with variable TE (VARIETE) for fat signal reduction

Uniform fat saturation over a large region of interest remains a problem in time-of-flight (TOF) magnetic resonance angiography applications. We demonstrate that a variable echo time with an opposed phase value at low spatial slice select frequencies can effectively reduce most of the fat signal in an otherwise standard three-dimensional TOF acquisition. We evaluated this method at 1.5 T using a short TE = 5.3 ms and a long TE = 6.75 ms for different values of the slice encoding gradient (i.e., different k₂ values). Shorter echo time (TE = 5.3 msec) was used at higher spatial slice select frequencies, but all echoes have the same gradient structures. By keeping the number of slice encoding steps with longer echoes to a minimum, field inhomogeneity effects on flow compensation remained small. A magnetization transfer saturation pulse was used to suppress signal of brain parenchyma. Overall, highly uniform and selective fat signal reduction was obtained while maintaining superior flow compensation in all volunteer studies.     

High-Field MR microscopy using fast spin-echoes

Fast spin-echo imaging has been investigated with attention to the requirements and opportunities for high-field MR microscopy. Two-and three-dimensional versions were implemented at 2.0 T, 7.1 T, and 9.4 T. At these fields, at least eight echoes were collectable with a 10 ms TE from fixed tissue specimens and living animals, giving an eightfold improvement in imaging efficiency. To reduce the phase-encoding gradient amplitude and its duty cycle, a modified pulse sequence with phase accumulation was developed. Images obtained using this pulse sequence exhibited comparable signal-to-noise (SNR) to those obtained from the conventional fast spin-echo pulse sequences. Signal losses due to imperfections in RF pulses and lack of phase rewinders were offset in this sequence by reduced diffusion losses incurred with the gradients required for MR microscopy. Image SNR, contrast, edge effects and spatial resolution for three k-space sampling schemes were studied experimentally and theoretically. One method of sampling k-space, 4-GROUP FSE, was found particularly useful in producing varied T₂ contrast at high field. Two-dimensional images of tissue specimens were obtained in a total acquisition time of 1 to 2 min with in-plane resolution between 30 to 70 μm, and 3D images with 256³ arrays were acquired from fixed rat brain tissue (isotropic voxel = 70 μm) and a living rat (isotropic voxel = 117 μm) in∼4.5 h.     

Phase enhancement for time‐of‐flight and flow‐sensitive black‐blood MR angiography

A magnitude‐based MR angiography method of standard time‐of‐flight (TOF) employing a three‐dimensional gradient‐echo sequence with flow rephasing is widely used. A recently proposed flow‐sensitive black‐blood (FSBB) method combining three‐dimensional gradient‐echo sequence with a flow‐dephasing gradient and a hybrid technique, called hybrid of opposite‐contrast, allow depiction of smaller blood vessels than does standard TOF. To further enhance imaging of smaller vessels, a new enhancement technique combining phase with magnitude is proposed. Both TOF and FSBB pulse sequences were used with only 0th‐order gradient moment nulling, and suitable dephasing gradients were added to increase the phase shift introduced mainly by flow. Magnitude‐based vessel‐to‐background contrast‐to‐noise ratios in TOF and FSBB were further enhanced to increase the dynamic range between positive and negative signals through the use of cosine‐function‐based filters for white‐ and black‐blood imaging. The proposed phase‐enhancement processing both improved visualization of slow‐flow vessels in the brains of volunteer subjects with shorter echo time in TOF, FSBB, and hybrid of opposite‐contrast and reduced wraparound artifacts with smaller b values without sacrificing vessel‐to‐background contrast in FSBB. This method of enhancement processing has excellent potential to become clinically useful. Magn Reson Med, 2011. © 2011 Wiley‐Liss, Inc.     

Virtual frequency encoding: Achieving symmetry in k-space

Chemical shift artifacts and other off-resonance spatial shifts in 2DFT MRI arise from the linear time dependence in the k-space data in the readout direction. Introduction of a view-dependent time shift of the readout window adds a time dependence to the phase-encoding direction and results in a virtual frequency-encoding direction that is a linear combination of the phase-encode and readout axes. By this method, the readout and phase-encode directions can be made identical in their sensitivity to off-resonance effects and can be arbitrarily swapped with no change in chemical shift or inhomogeneity effects, improving previously reported methods that swap these axes for signal averaging or reduction of motion artifacts.     

Flow artifact reduction in MRI: a review of the roles of gradient moment nulling and spatial presaturation

In the past, flow artifacts and inconsistent depiction of vascular anatomy have represented significant problems in clinical MRI. These difficulties are now generally well addressed by the techniques of gradient moment nulling and spatial presaturation. Gradient moment nulling (GMN) is an effective method for eliminating flow artifacts in gradient echo images, while presaturation is more applicable to the same task in spin echo acquisitions. The GMN technique also has useful applications in spin echo imaging such as combating the effects of tissue and CSF motion in long TE sequences. In contrast to presaturation, however, GMN is not suitable for suppressing artifacts due to pulsatile blood flow in spin echo images.     

GRASE imaging at 3 Tesla with template interactive phase–encoding

A new method for ordering the phase-encoding gradient is proposed, and an application for short effective TE gradient-and spin-echo (GRASE) imaging is demonstrated. The proposed method calculates the phase-encoding order from the signal decay of a template scan (hence “template interactive phase-encoding” or TIPE). Computer simulations are used to compare the point spread functions of different phase-encoding orders giving short effective echo times (kb centric GRASE, centric GRASE, centric TIPE). The conventional centric phase-encoding order is also considered for GRASE. The conventional centric method is sensitive to both amplitude and phase modulation of the signal in K-space. The centric TIPE method gives the least amplitude modulation artifacts but is vulnerable to phase artifacts. The TIPE experiment was implemented on a 3 Tesla system. To the best of our knowledge, we present the first in vivo GRASE images at this field strength.     

3D Echo Planar Imaging: Application to the Human Head

In this work, the authors present 3D images acquired from the human head using echo planar encoding for two of the three dimensions of k-space. The third dimension of k-space is filled by selecting and phase encoding a slab of spins as in conventional 3D steady state (GRASS based) acquisition regimens. Using this approach, a 128 x 64 x 64 3D data matrix could be obtained in 3.4–4.7 sec using effective TE values of 24 and 34 ms, respectively. High quality 3D images could be acquired once phase ghosts present on 2D images were minimized through proper adjustments of scanner hardware.     

A Method to measure arbitrary k-space trajectories for rapid MR imaging

A method to measure arbitrary k-space trajectories was developed to compensate for nonideal gradient performance during rapid magnetic resonance (MR) imaging with actively or nonactively shielded gradients at a magnetic field strength of 4.1 T. Accurate MR image reconstruction requires knowledge of the k-trajectory produced by the gradient waveforms during k-space sampling. Even with shielded gradients, residual eddy currents and imperfections in gradient amplifier performance can cause the true k-space trajectory to deviate from the ideal trajectory. The k-space determination was used for spiral gradient-echo imaging of the human brain. While individual calibrations are needed for new pulse sequences, the method of k-space determination can be used for any sequence of preparation pulses and readout gradient waveforms and should prove useful for other trajectories, including the rastered lines of echo-planar imaging.     

MR angiography using velocity-selective preparation pulses and segmented gradient-echo acquisition

We describe a cardiac-gated MR angiographic imaging method that employs velocity-selective preparation (VSP) pulses in conjunction with segmented gradient-echo acquisitioin and subtraction to produce images that, ideally, contain no signal from stationary tissues and display vessels with a signal intensity that is dependent on the velocity of the blood in the vessels. The novel features of this method are a) it acquires several phase-encoding valueslapplication of a single VSP pulse, b) it uses subtraction to eliminate signal that is not sufficiently suppressed by the VSP pulses, and c) it uses VSP pulses that are synchronized with the cardiac cycle so it can be used to produce ghost-free images of pulsatile blood. An advantage of this sequence is that it detects a signal that, after preparation, is relatively unaffected by changes in blood velocity. This leads to a large signal-to-noise ratio for all the phase-encoding values, a reduction of ghosting artifacts, and the ability to visualize blood that is in motion for only a short time during the cardiac cycle. Because the signal is prepared during peak flow, venous signal can be suppressed by making the sequence sensitive to high velocities. An additional advantage of this sequence is that it permits sampling with a short TE because the velocity-encoding gradient can be applied in a preparatory interval. Signal loss that results from dephasing during the longer TE preparation interval can be reduced or eliminated by allowing the dephased spins to flow out of the region of complex flow, and perhaps out of the field-of-view, by introducing a delay between the finish of the VSP pulse and the beginning of data acquisition.     

Reducing flow artifacts in echo-planar imaging

Echo-planar imaging (EPI) is very susceptible to flow artifacts. Two ways to improve its flow properties are presented. First, “partial flyback” is proposed to reduce artifacts arising from flow in the readout direction. Near the center of k-space, only the even echoes of the EPI echo-train are used. Partial flyback is shown to improve the readout-flow properties at the expense of a slight worsening of the phase-encode flow and off-resonance properties. We recommend that the flyback region acquire 95% of the energy in k-space. Second, “inside-out” EPI is used to reduce artifacts arising from flow in the phase-encode direction. Data collection begins at the center of k-space, with separate interleaves to acquire the top and bottom halves of k-space. Partial flyback is combined with partial-Fourier EPI and inside-out EPI. Partial-flyback inside-out EPI has worse off-resonance properties than partial-flyback partial-Fourier EPI but demonstrates better flow properties and does not require partial k-space reconstruction.     

A Nyquist Modulated Echo-to-View Mapping Scheme for Fast Spin-Echo Imaging

A new scheme for assignment of echoes to views in fast spin-echo imaging was developed. The scheme places early and late echoes in alternating lines of the periphery of k-space; continued alternation of echoes arising closer to the middle of the echo train encodes the central portion of k-space. The scheme has two effects: a) The echo-to-echo signal decay that usually gives rise to multiple faint ghosts, now yields a single Nyquist ghost that is removed by phase over-sampling. b) Mapping earlier echoes (and thus more energy) to the periphery of k-space yields sharper edges.     

Comparison of k-space sampling schemes for multidimensional MR spectroscopic imaging

For clinical ³¹P MR spectroscopic imaging (MRSI) studies, where signal averaging is necessary, some improvement of sensitivity and spatial response function may be achieved by acquiring data over a spherical k-space volume and varying the number of averages acquired in proportion to the desired spatial filter. Eight different k-space sampling schemes are compared through simulations that provide graphs of the spatial response functions (SRF), and tabulations of voxel volumes, relative signal-to-noise ratios (SNR), and relative data collection efficiencies (SNR per unit volume over the same time). All schemes were based on practical experiments, each of which could be implemented in the same length of time. The results show that in comparison with cubic k-space sampling with the same number of signal averages at each point, spherical and acquisition-weighted k-space sampling can be used to achieve reduced Gibbs ringing along the principal axes directions, and thus reduced contamination from adjacent tissue in these directions, without degradation of voxel volume or SNR.     

Increased flexibility in GRASE imaging by k space-banded phase encoding

GRASE (GRadient and Spin Echo) is an echo train imaging technique that combines gradient and RF refocusing. Although overall signal decay is with T₂ and field inhomogeneity phase errors do not accumulate, the small residual phase errors are periodic with echo number. The echo order described previously eliminates the phase error periodicity in k space but instead creates periodicity in the T₂ modulation function that can also cause artifacts. In addition, with this order, the effective TE must be half the echo train time, and asymmetric Fourier sampling is difficult to implement. A new method is described that greatly reduces artifacts due to T₂ decay, permits greater control of T₂ contrast, and lends itself to asymmetric Fourier sampling. Different time segments of the echo train are encoded with different bands of spatial frequency in k space (hence “k banding”). Both computer simulations and experimental results demonstrate improvements in GRASE images acquired by this method.     

Excitation of arbitrary shapes by gradient optimized random walk in discrete k-space

A new technique for the excitation of arbitrary shapes is proposed. It is based on a parallel sequence of small tip angle RF pulses and gradient pulses. The small tip angle rotations co-add yielding a 90° excitation pulse within the selected excitation profile while outside the profile, the rotations cancel each other. A full theory of the completely arbitrary regional volume excitation (CARVE) method is presented and experimentally verified. In CARVE, k-space is discrete because the RF is applied in pulses. The discrete character of k-space permits an arbitrary trajectory for the k-space walk. The optimal random trajectory is found by minimizing the gradient load using simulated annealing. It is shown, both theoretically and experimentally, that such a trajectory is much better than any other systematic or random trajectory in k-space.     

Interleaved asymmetric echo-planar imaging

A version of interleaved echo-planar imaging (EPI) is presented in which only one polarity of the readout gradient is used for signal acquisition to avoid ghosting artifacts. Two possible forms of the phase encoding gradient, blipped and constant, are discussed. With the constant phase encoding, interleaving of partial trajectories in the Fourier domain (k-space) is controlled automatically by the echo train delay. The constant phase encoding gradient introduces a shear distortion of the k-space grid. A modification of the reconstruction procedure is given which corrects for this effect. The method provides a 128 × 128 image in 1 s on a clinical system with standard gradients.     

Frequency resolved single-shot MR imaging using stochastic k-space trajectories

A new single-shot stochastic imaging technique with a random k-space path that provides very selective filtering with respect to chemical shift or off-resonance signals of the investigated tissue is proposed. It is demonstrated that in stochastic imaging only on-resonance compartments are visible whereas frequency shifted compartments cancel to noise that is distributed over the whole image. This method can be used as a single-shot chemical shift selective imaging technique and allows to calculate frequency resolved spectra for each spatial position of the image based on a single signal aquisition. The single-shot stochastic imaging sequence makes high demands on the gradient system and the theoretical k-space trajectory is distorted by imperfect gradient performance. Therefore an additional k-space guided imaging technique that uses the true, measured k-space trajectory to correct artifacts generated by eddy currents and delay times of the rapid switched gradients is presented. In vitro and in vivo measurements demonstrate the successful implementation of single-shot stochastic imaging on a conventional MR scanner with unshielded gradient systems.     

Gradient moment nulling in fast spin echo

The fast spin echo sequence combines data from many echo signals in a Carr-Purcell-Meiboom-Gill echo train to form a single image. Much of the signal in the second and later echoes results from the coherent addition of stimulated echo signal components back to the spin echo signal. Because stimulated echoes experience no dephasing effects during the time that they are stored as M, magnetization, they experience a different gradient first moment than does the spin echo. This leads to flow-related phase differences between different echo components and results in flow voids and ghosting, even when the first moment is nulled for the spin echo signal. A method of gradient moment nulling that correctly compensates both spin echo and stimulated echo components has been developed. The simplest solution involves nulling the first gradient moment at least at the RF pulses and preferably at both the RF pulses and the echoes. Phantom and volunteer studies demonstrate good suppression of flowrelated artifacts.     

Artifacts and signal loss due to flow in the presence of Bo inhomogeneity

An in vitro study was performed to investigate the effects of B_o inhomogeneity on magnetic resonance images of flow. Controlled inhomogeneity gradients (G₁) were applied and the magnitude of the artifacts produced was quantified for different echo delay times (TE). Both steady and pulsatile flows were examined. In the presence of an inhomogeneity gradient, signal loss is apparent if the flow is pulsatile and/or if the slice thickness is large. The signal loss increases with increasing TE and G₁. With pulsatile flow, ghosting artifacts are also generated. These increase in intensity with increasing TE and G₁. In vivo, field inhomogeneity due to susceptibility variations is large enough to produce these effects. Representative time-of-flight images obtained of a normal volunteer with two different TEs demonstrate the effect in vivo. Flow-related signal loss and artifacts, therefore, increase with increasing TE independent of the moments of the applied gradients.