Distortions during excitation and acquisition in zoomed DWI combined with parallel transmission
Denis Kokorin1, Jürgen Hennig1, and Maxim Zaitsev1

1Medical Physics, University Medical Center Freiburg, Freiburg, Germany

Synopsis

In this work, we discuss off-resonance effects observed in zoomed DWI combined with 2D pulses based on EPI trajectories. Our main focus is placed on the use of parallel excitation (PEX) for shortening 2D pulses and further minimizing the distortion. Experimental data were obtained using a 3T MRI system with an 8-channel TxArray extension and analyzed in detail. We conclude that the use of PEX improves the matching between excitation and acquisition in zoomed EPI applications.

Introduction

Excitation of slice profiles with a limited FOV was shown to be advantageous for distortion reduction in zoomed EPI applications such as DWI [1-4]. Profile selection is accomplished normally by 2D pulses based on EPI excitation trajectories [3,4]. However, due to long durations, 2D pulses are sensitive to B0 deviations, manifesting themselves as distortions in the excited profile [5]. In this study, we analyze off-resonance effects observed in SE-EPI combined with 2D pulses. The main focus is placed on new opportunities brought by parallel excitation (PEX, [5-7]) for shortening 2D pulses and further minimizing the distortion.

Theory

The thickness of a designed slice is usually encoded either along the phase encoding (PE) or frequency encoding (FE) directions of the excitation trajectory (Fig. 1). In PEX, the trajectories are undersampled by skipping lines and the resulting excitation replicates are suppressed by using the B1 sensitivities during pulse design [6,7]. When FE is along the slice selection direction (Fig. 1A), the PE gradients of excitation and acquisition are applied in the same directions in SE-EPI (Fig. 2A). Therefore, excitation and acquisition distortions are observed along the same spatial dimension and combined with each other in the final image. In another encoding scheme (Fig. 1B), the excitation and acquisition PE gradients are in orthogonal dimensions (Fig. 2B). Thus, strong field deviations might shift the profile outside the refocusing position during excitation, resulting in an SNR decrease during the EPI readout. PEX allows shortening of the 2D pulses and thus reduces these effects in both techniques in Fig. 2.

Materials and Methods

Experiments were carried out on a 3T MR scanner (Siemens Magnetom Trio) with an 8-channel TxArray extension. The scanned object was a container filled with water and plastic tubes. 2D PEX was exploited for both techniques in Fig. 1 and 2. The RF pulses were designed using the small-tip-angle algorithm combined with conjugate gradient optimization to incorporate B1 data [8]. The excitation trajectories with slice selection in the PE direction were defined over an FOE of 38.4×24 cm2 with a grid size of 192×16. The trajectories with FE along the slice selection direction had an FOE of 38.4×9.6 cm2 and 32×32 samples. Both trajectories were undersampled by skipping every second k-space line.

The selected profiles were imaged using GRE, which reflected the excitation distortions. Afterwards, SE-EPI methods (Fig. 2) were tested for varying shim settings. For the sequence in Fig. 2A, imaging was in a sagittal orientation for PE along the "anterior-posterior" and "posterior-anterior" directions, which corresponded to opposite polarities of the PE acquisition gradients. The imaging in Fig. 2B was in a coronal orientation.

Results

For default shim settings (Fig. 3A), the profiles exhibited no visible distortions, when FE of the pulses was along the slice selection direction. After the shim currents were modified and scanning was repeated, stronger B0 deviations led to shearing of the profiles (Fig. 3B, C). The distortion during excitation for the accelerated pulse was weaker compared to the fully-sampled counterpart as shown in the GRE images. The profiles underwent further distortions in EPI and appeared to be sheared dramatically when PE was "anterior-posterior". For PE along the "posterior-anterior" direction, the observed distortion was minimal.

For excitation of profiles with slice selection in the PE direction and default shim settings, the observed distortions were also minimal (Fig. 4A). After the resonance frequency was changed by 100 Hz, the selected slices were shifted from the original position, resulting in a signal decrease in EPI (Fig. 4B,C). However, the signal was partially restored for the accelerated pulse due to its shorter duration.

Conclusions

PEX allows for a more robust selection of limited profiles against off-resonance effects. For the imaging scheme with slice selection along the FE dimension, PEX offers additional degrees of freedom, in order to better confine the FOV to the targeted organ in SE-EPI and DWI. For slice selection in the PE direction, PEX makes DWI more robust again signal loss.

Acknowledgements

The authors thank Dr. Kelvin Layton for helpful discussions.

References

1. P. Mansfield, J. Phys C 10:L5-L58, 1977

2. J. Pauly, J Magn Reson, 82, p. 571, 1989

3. J. Finsterbusch, JMRI, 29, p. 987, 2009

4. EU Saritas et al., MRM, 60, p. 468, 2008

5. D. Kokorin et al., Concepts in MR (B), 45(4), p. 153, 2015

6. U. Katscher et al., MRM, 49, p. 144, 2003

7. Y. Zhu, MRM, 51, p. 775, 2004

8. HP Fautz et al., ISMRM 2008, #1247

Figures

Figure 1. Illustration of encoding slice thickness along the FE (A) or PE (B) dimensions of an EPI excitation trajectories. The slices extend in the third spatial non-encoded direction across the whole object.


Figure 2. Sequence diagrams for SE-EPI (e.g. DWI) of limited profiles for schemes from Figures 1A and B, respectively. The slices are limited in the PE dimension of the acquisition block.


Figure 3. Off-resonance effects for the case when slice thickness was encoded in the FE dimension of the excitation trajectory. (A) Profiles excited by a fully-sampled pulse for default shim settings (B,C) Profiles excited by fully-sampled and undersampled pulses for shim currents modified by -40 μT/m in the Z-direction.


Figure 4. Off-resonance effects for good (A) and bad shim settings (B,C) when the slice thickness was encoded in the FE dimension of the excitation. (A) and (B) show the performance of the fully-sampled pulse and (C) corresponds to the undersampled excitation.




Proc. Intl. Soc. Mag. Reson. Med. 24 (2016)
4279