2666

Ultra-high-field CAIPIRINHA modulated parallel transmit excitation for homogenous image reconstruction without RF shimming
Iulius Dragonu1, Craig Buckley1, Matthew D Robson2, and Aaron T Hess2

1Diagnostic Imaging, MR, Siemens Healthcare Ltd, Frimley, United Kingdom, 2Radcliffe Department of Medicine, University of Oxford, Oxford, United Kingdom

Synopsis

Ultra-high-field (UHF) provides higher SNR than conventional, clinically available field strengths. However, UHF suffers from heterogeneous transmit B1+ fields. At 7 T, the shortened transmit radio-frequency (RF) wavelengths have a similar value to the dimensions of the human head/thorax which may result in signal cancellation and local signal dropouts. In this paper, we propose a novel imaging scheme based on simultaneous excitation with all transmit channels. Controlled aliasing is used to encode each transmit channel independently which we term Tx-CAIPIRINHA.

Tx-CAIPIRINHA has been demonstrated in-vivo. The concept uses the linear superposition of B1+ fields via the excitation flip angle which only holds true in the low flip angle regime. When normalizing to transmit efficiency, Tx-CAIPIRINHA achieved a marginally higher SNR than B1+ shimming, demonstrating the constructive combination of transmit sensitives throughout the image.

Introduction

Ultra-high-field (UHF) provides higher SNR than conventional, clinically available field strengths. However, UHF suffers from heterogeneous transmit (B1+) fields [1]. At 7 T, the shortened transmit radio-frequency (RF) wavelengths have a similar value to the dimensions of the human head/thorax which may result in signal cancellation and local signal dropouts.

Radio-frequency shimming with multiple channel excitation of a parallel transmit (pTx) array is a well established method to increase the transverse magnetic field homogeneity at high magnetic field strength [2]. To perform RF shimming, relative transmit sensitivity maps are acquired. They are used to determine the combination of amplitudes and phases to either maximize the minimum B1+ or minimize the coefficient of variation throughout a region of interest. However, as the field of view increases, the excitation efficiency reduces to account for larger variation until that there is no satisfactory shim solution [3]. Subject motion between adjustment mapping and image acquisition means that the optimal B1+ solution can change [4]. Time interleaved acquisition of modes (TIAMO) [5] addresses the B1+ heterogeneity by combining two different transmit modes. In this paper, we propose a novel imaging scheme based on simultaneous excitation with all transmit channels. Controlled aliasing is used to encode each transmit channel independently which we term Tx-CAIPIRINHA

Methods

Cardiac measurements were performed in two healthy volunteers on a whole-body research 7Tesla MR-system (Siemens Healthcare GmbH, Erlangen, Germany). Healthy volunteer measurements were acquired according to our institutions ethical practices. Data acquisition was performed with a custom built eight-channel transmit/receive cardiac transverse electro-magnetic (TEM) coil.

Images were acquired using a modified non-product, gradient-echo (GRE) sequence allowing simultaneous transmit with controlled aliasing for each transmit channel. This was achieved by introducing a different phase modulation for each transmit channel resulting in a different shift within the FOV of excitation resulting from each transmitter. The increments in phase from one line to the next ranged from 0° to 315° in steps of 45° for each transmit channel. To calibrate the receive and transmit sensitives, low resolution, low flip angle, transmit maps were acquired by transmitting on one Tx channel at a time [6]. The acquisition parameters of the GRE sequence are: TR=9.64ms,TE =1.68 ms, acquisition matrix=352x352, FOV=350x350mm2. Single-slice images of 4mm thickness were acquired during four cardiac cycles in a single breath hold. For comparison purpose, an image of the same slice with identical orientation and RF pulse voltage was acquired using B1+ shimming.

Aliased images formed when transmitting on all channels were reconstructed using Tx channel–GRAPPA (Tx-GRAPPA) algorithm similar to slice-GRAPPA [7]. A 7×7 GRAPPA kernel is separately fitted in k-space to each transmit channel of the calibration data. In the case of the Tx-GRAPPA method, after reconstruction of separate images for each transmit and receive channel, the images were combined. First, images from all the receivers were combined with a root sum-of-squares (rSoS) reconstruction incorporating the noise covariance matrix. In the second step, rSoS reconstruction was performed along the transmit dimension.

For comparison a conjugate gradient iterative SENSE reconstruction was also employed [8] where the product of transmit and receive sensitivities was used to form an 8x8 or 64 virtual channel SENSE reconstruction. This method provides one combined image.

Results

A simulation of the efficiency (ratio to maximum B1+ possible) was performed for 100,000 random B1+ fields of an 8 channel system when using Tx-CAIPIRINHA and was compared to no B1+ shimming. The average attained flip angle is well behaved and it is approximately 1/√N-TxChannels i.e. 1/√8, compared to no B1+ shimming. The result of this simulation is summarized in fig.1.

Figure 2 shows the reconstructed images using Tx–CAIPIRINHA with a 7×7 GRAPPA kernel (a), iterative SENSE (b) and B1+ shimming (c). The efficiency of the applied B1+ shim was 67% compared to an expected 35% for Tx-CAIPIRINHA (Fig 1) a ratio of 1.9:1.

Both methods Tx-GRAPPA and iterative SENSE produced similar results with high signal homogeneity in the heart. This was achieved without any prior operator adjustment of the B1+ field of each transmit channel.

Discussion

Tx-CAPIRINHA has been demonstrated in-vivo. The concept uses the linear superposition of B1+ fields via the excitation flip angle which only holds true in the low flip angle regime. When normalizing to transmit efficiency, Tx-CAIPIRINHA achieved a marginally higher SNR than B1+ shimming, demonstrating the constructive combination of transmit sensitives throughout the image. This method has a number of further advantages over RF shimming, including no operator involvement to perform B1+ shimming and no limit to field of view size.

Acknowledgements

No acknowledgement found.

References

[1] Vaughan JT, Garwood M, Collins CM, Liu W, DelaBarre L, Adriany G, Andersen GP, Merkle H, Goebel R, Smith MB, Ugurbil K. 7T vs. 4T: RF power, homogeneity, and signal-to-noise comparison in head images. Magn Reson Med. 2001; 46:24–30

[2] Van de Moortele PF and Uburbil K. Very Fast Multi Channel B1 Calibration at High Field in the Small Flip Angle Regime” ISMRM 2009, pp 367

[3] Schmitter S, DelaBarre L, Wu X, Greiser A, Wang D, Auerbach EJ, Vaughan JT, Ugurbil K, Van de Moortele PF. Cardiac Imaging at 7T: Single- and Two-Spoke RF Pulse Design with 16-channel Parallel Excitation Magn Reson Med. 2001; 46:24-30

[4] Schmitter S, Wu X, Ugurbil K, Van de Moortele PF. Design of parallel transmission radiofrequency pulses robust against respiration in cardiac MRI at 7 Tesla. Magn Reson Med. 2015; 74:1291-1305

[5] Orzada S1, Maderwald S, Poser BA, Bitz AK, Quick HH, Ladd ME. RF excitation using time interleaved acquisition of modes (TIAMO) to address B1 inhomogeneity in high-field MRI. Magn Reson Med. 2010; 64:327-333

[6] Padormo F, Hess AT, Aljabar P, Malik SJ, Jezzard P, Robson MD, Hajnal JV, Koopmans PJ. Large dynamic range relative B1+ mapping. Magn Reson Med. 2016; 76:490-499

[7] Setsompop K, Gagoski BA, Polimeni JR, Witzel T, Wedeen VJ, Wald LL. Blipped-controlled aliasing in parallel imaging for simultaneous multislice echo planar imaging with reduced g-factor penalty. Magn Reson Med. 2012; 67:1210-1224

[8] Pruessmann KP, Weiger M, Börnert P, Boesiger P. Advances in sensitivity encoding with arbitrary k-space trajectories. Magn Reson Med. 2001; 46:638-651

Figures

Figure 1: Simulation of the efficiency (ratio to maximum B1+ possible) for 100,000 random B1+ fields of an eight-channel system when using Tx-CAIPIRINHA compared to no B1+ shimming. 1 represents a 100% efficient B1+ shim. It is shown that the average attained flip angle is well behaved and is approximately 1/√N-TxChannels i.e. 1/√8, compared to no B1+ shimming, the simple sum of fields which ranges from 0 to 1.

Figure 2: shows the reconstructed images obtained using Tx–CAIPIRINHA using a 7×7 GRAPPA kernel (a) for each transmit channel or iterative SENSE (b) and B1+ shimming (c). Both methods Tx-GRAPPA and iterative SENSE produced similar results with high signal homogeneity in the heart. This was achieved without any prior operator adjustment of the B1+ field of each transmit channel.

Proc. Intl. Soc. Mag. Reson. Med. 25 (2017)
2666