4226
On the use of frequency modulated pulses in sat-UTE1Université de Strasbourg, CNRS, ICube, FMTS, Strasbourg, France, 2NMR laboratory, Institute of Myology, Paris, France, 3NMR laboratory, CEA/DRF/IBFJ/MIRCen, Paris, France, 4Institut Mines Télécom Atlantique, INSERM, LaTIM, Brest, France
Synopsis
Conventional slice selection in 2D-UTE sequences is challenging as eddy currents and gradient non-idealities make difficult to achieve an appropriate slice selection and a minimum TE. The sat-UTE sequence proposes a simplification that separates slice selection from excitation, ensuring an easily implementable 2D-UTE sequence. The selection part was originally proposed with a short gaussian pulse, constraining slice selectivity to a gaussian profile and demanding exceedingly high RF power for ensuring efficient saturation. In this work, we propose an alternative implementation using low-peak-amplitude and selective quadratic phase pulses for slice saturation, providing an efficient and sharp short-T2 slice selection in sat-UTE.
Introduction
Achieving ultra-short echo times in conventional 2D-UTE sequences is challenging since it necessitates the use of specific reshaped half-pulses to avoid the need for any rephasing moment inherent to the slice select gradients1,2. Eddy currents also constitute a source of artifacts at short echo times. In sat-UTE3 (see Figure 1), magnetization excitation is performed using a short rectangular pulse, and the slice selectivity is ensured by a slice-saturation preparation module. Two k-spaces acquired with and without (αsat = 0°) the slice saturation are subtracted prior to image reconstruction. Hence, the saturation pulse also must ensure an efficient short-T2 saturation. In the original sat-UTE, this was accomplished using a short duration and extremely high amplitude gaussian RF-pulse. In the present work, we propose an alternative to the use of short conventional saturation pulses in sat-UTE by taking advantage of low-amplitude, frequency modulated and selective pulses based on the Shinnar-Le Roux (SLR) framework4. The advantages of the method are demonstrated in simulations and in vitro.Method
The framework proposed by Balchandani et al.4 for generating quadratic-phase SLR (QPSLR) pulses was used with the following parameters: N = 1023, k = 1000, BW = 8 kHz, FTW = 0.1, fs = 1 MHz. This resulted in a 2.902 ms pulse with an excitation bandwidth of 16 kHz. Its peak amplitude (20.18 μT) was determined from a slice profile simulation study, and defined as the required amplitude for applying a 90° tilt on a short-T2 component (T2 = 0.5 ms).Equivalently, the pulse tilts a long-T2 component of T2 = 100 ms by 97°. The resulting QPSLR was therefore compared to a gaussian and two linear-phase SLR (LPSLR) pulses. B1 peak amplitudes were set to 20.18 μT, and the durations adjusted to ensure a 97° flip angle on a long-T2 component. The respective pulse durations and TBW are given in Table 1.
Slice profiles were simulated for a 0.3-ms T2-species, with a 1-mm slice thickness (ST). Out-of-slice excitation is expected in the short-T2 component due to its broad frequency linewidth.Hence, the experimental in-plane resulting signal to noise ratio (SNR) may be biased by the out-of-slice undesired excitation. We propose to quantify this bias by assessing the ratio of the out-of-slice magnetization to the total magnetization generated in sat-UTE:
$$\Gamma=\frac{\int_{-\beta}^{-ST/2}M_{xy}(z)dz+\int_{ST/2}^{\beta}M_{xy}(z)dz}{\int_{-\beta}^{\beta}M_{xy}(z)dz}$$
where $$$M_0$$$ is the thermal equilibrium and β symmetrical distance from the slice center to the position where the short-T2 transverse magnetization reaches 1% of the maximum magnetization amplitude. Additionally, the full widths at half-maximum (F) of the slice profiles were estimated.
Experiments were conducted on a 7T preclinical scanner (Bruker BioSpec, Ettlingen, Germany), using a 86-mm Tx volume coil and mouse head surface Rx coil. A piece of Lego brick (T2* $$$\approx$$$ 300 μs5) was scanned using the four pulses from simulations. In a first experiment, one of the readout directions was tilted into the slice direction in order to assess the slice selectivity. In a second experiment, ROI-based SNR on axial views were evaluated along the saturation pulses.
Relevant parameters were: TR/TE/TS = 500/0.01/2.1 ms, α = 90°, τ = 0.07 ms, matrix size = 128x128, slice thickness = 1 mm, receiver bandwidth = 200 kHz, number of trajectories = 1604 with in-plane voxels size = 100x100 μm2 in the first experiment (4 accumulations, Tacq/scan = 1h47min), and voxels size = 150x150 μm2 in the second one (2 accumulations, Tacq/scan = 53min).
Results
The respective simulated slice profiles are shown in Figure 2, and the corresponding evaluated scores in Table 1. The QPSLR pulse yielded the sharpest slice profile (β = 1.33 mm) with the minimum out-of-slice contamination ($$$\Gamma$$$ = 10.7%) and the highest short-T2 saturation efficiency. Gaussian and LPSLR pulses show flattened profiles yielding theoretical $$$\Gamma$$$ ratios above 27%, and β above 2.25 mm.
Figure 3 shows the slice projections and their respective extracted profiles, as well as the actual acquired slices in the Lego brick along the four pulses. In-vitro results confirm the superior performance of the QPSLR pulse concerning short-T2 saturation and slice sharpness. Differences between in vitro and simulations results can be attributed to a much more complex T2 and T1 distribution of the Lego brick material. On axial views, the SNR ratios are comparable between the SLR-designed pulses, however mostly likely resulting of an out-of-slice accumulation effects for the LPSLR pulses.
Conclusion
Low amplitude and selective QPSLR pulses can be preferably used as saturation pulses in sat-UTE sequences, improving slice sharpness and short-T2 saturation efficiency while minimizing the B1-peak amplitude.Acknowledgements
No acknowledgement found.References
1. Conolly S, Nishimura D, Macovski A, Glover G. Variable-rate selective excitation. Journal of Magnetic Resonance (1969) 1988;78:440–458.
2. Pauly J, Conolly SM. Slice-selective excitation for very short T2 species. in Proceedings of the 8th Annual Meeting of SMRM, Amsterdam, The Netherlands, 1989. p. 28.
3. Harkins KD, Horch RA, Does MD. Simple and robust saturation-based slice selection for ultrashort echo time MRI. Magnetic Resonance in Medicine 2015;73:2204–2211.
4. Balchandani P, Pauly J, Spielman D. Designing adiabatic radio frequency pulses using the Shinnar-Le Roux algorithm. Magnetic Resonance in Medicine 2010;64:843–851.
5. Codd S, Mallett M, Halse M, Strange J, Vennart W, Doorn T. A Three-Dimensional NMR Imaging Scheme Utilizing Doubly Resonant Gradient Coils. Journal of MagneticResonance, Series B 1996;113:214–221.
Figures
