Time

Prog #

Anke Henning¹, Alexander Fuchs¹, James B. Murdoch², David Foxall², Peter Boesiger¹

¹University and ETH Zurich, Zurich, Switzerland; ²Philips Medical Systems, Cleveland, USA

For MRSI acquisition at 7T direct acquisition of the Free Induction Decay Localized by Outer Volume Suppression (FIDLOVS) is proposed to minimize SNR loss due to short T2 relaxation times. To that, a broadband frequency-modulated excitation pulse for slice-selection and a numerically optimized outer-volume-suppression scheme, which considers T1 relaxation of skull lipid, B1 inhomogeneity, saturation band crossing and is based on broadband PPR-pulses, for in-plane localization are applied. The resulting signal-to-noise ratio (SNR) and spectral resolution enable mapping of 15 metabolites. The high SNR is also the base for highly spatially resolved metabolite mapping accelerated by sensitivity encoding.

Fa-Hsuan Lin¹, ², Shang-Yueh Tsai², Yi-Ru Lin³, Ricardo Otazo⁴, Graham Wiggins¹, Lawrence Wald¹, John Belliveau¹, Stefan Posse⁴

¹Massachusetts General Hospital, Charlestown, Massachusetts, USA; ²National Taiwan University, Taipei, Taiwan; ³National Taiwan University of Science and Technology, Taipei, Taiwan; ⁴Univesity of New Mexico, Albuquerque, New Mexico, USA

A proton spectroscopic inverse imaging (SInI) acquisition and reconstruction protocol is introduced to measure 2-dimensional spectroscopic maps in a single excitation with high spectral resolution. Combining with echo-planar readout for spectral and 1D spatial information, we use the solution of the minimum-norm estimate (MNE) inverse problem along the other spatial dimension from all channels of a coil array. Feasibility of short TE proton MRSI was demonstrated on 3 T scanners equipped with a 32-channel array. This method enables flexible tradeoff between phase encoding and parallel imaging to maximize spectral width for applications in dynamic single-shot MRSI.

Yoshitaka Bito¹, Koji Hirata¹, Toru Shirai¹, Yukari Yamamoto¹, Yoshihisa Soutome¹, Toshihiko Ebisu², Masahiro Umeda³, Yuko Kawai³, Toshihiro Higuchi³, Chuzo Tanaka³

¹Hitachi, Ltd., Kokubunji-shi, Japan; ²Nantan General Hospital, Japan; ³Meiji University of Oriental Medicine, Japan

A fast lactate-discriminating echo-planar spectroscopic imaging (EPSI) technique suitable for 7-T MRI is developed. The technique uses just a single TE measurement for acquiring lactate-discriminated images, in which the TE is chosen to shift the echo peak of a lactate signal away from that of an overlapping lipid signal. The optimum TE at 7 T is calculated so that the lipid signal is sufficiently suppressed while the lactate signal is not attenuated so much. Acquisition of discriminated lactate images is demonstrated by applying this technique to a phantom and a rat.

Nicolas Kunz¹, Cristina Cudalbu¹, Vladimir Mlynarik¹, Stéphane V. Sizonenko, Rolf Gruetter¹, ²

¹Ecole Polytechnique Fédérale de Lausanne, Lausanne, Switzerland; ²University of Lausanne and Geneva, Switzerland

Quantification of the neurochemical profile by 1H-MRS critically depends on the estimation of the contribution of macromolecule resonances to the spectrum. As the linewidth of macromolecules does not increase with B0, the resonances attributed to macromolecules are increasingly difficult to distinguish from those of coupled spin systems. Due to heterogeneity of T1, in the metabolite-nulled spectra residuals attributed to metabolites are still observed. The study shows that  an IR 1H-MRS sequence combined with diffusion weighting allows near-complete removal of metabolites signals and excellent definition of the spectral contribution of macromolecules due to one order of magnitude smaller ADC of macromolecules.

Ricardo Otazo¹, Fa-Hsuan Lin², Graham Wiggins², Ramiro Jordan¹, Stefan Posse¹

¹University of New Mexico, Albuquerque, New Mexico, USA; ²Massachusetts General Hospital, Boston, Massachusetts, USA

Standard parallel imaging may produce artifacts when applied to MR spectroscopic imaging (MRSI) due to coil sensitivity variation within the typical large voxels in low spatial resolution MRSI. In this work, a novel approach for accelerating the spatial encoding process of MRSI known as Superresolution SENSE (SURE-SENSE) is presented. Acceleration is performed by acquiring the low spatial resolution representation of the object being imaged and intra-voxel reconstruction is performed using coil sensitivity maps acquired with higher target spatial resolution. The method is particularly suitable for array coils with a large number of small elements that present stronger sensitivity variation. We show feasibility of the method for human brain MRSI using Proton Echo Planar Spectroscopic Imaging (PEPSI) and a 32-channel receiver array coil.

Borjan Aleksandar Gagoski¹, Kawin Setsompop¹, Vijay Alagappan², Franz Schmitt³, Ulrich Fontius³, Andreas Potthast⁴, Lawrence Wald², ⁵, Elfar Adalsteinsson¹, ⁵

¹Massachusetts Institute of Technology, Cambridge, Massachusetts, USA; ²MGH, Charlestown, Massachusetts, USA; ³Siemens Medical Solutions, Erlangen, Germany; ⁴Siemens Medical Solutions, Charlestown, Massachusetts, USA; ⁵Harvard-MIT Division of Health Sciences and Technology, Cambridge, Massachusetts, USA

In this work we combine spiral CSI readouts with parallel RF transmission to mitigate B1+ inhomogeneities. We limit this initial demonstration to the low flip-angle domain where excitation k-space analysis holds, and apply “spokes”-based slice selective RF design to an eight channel transmit system at 7T. The 8 transmit channels enable reduced-duration, slice-selective RF pulses that implement excellent wide band B1+ mitigation over a 600 Hz bandwidth. The goal of this work is to demonstrate efficient spiral CSI encoding with B1+-mitigated spatial-spectral excitation over a spatial FOV and frequencies of interest for 1H brain spectroscopic imaging.

Meng Gu¹, Chunlei Liu¹, Daniel Spielman¹

¹Stanford University, Stanford, USA

Parallel reconstruction has been successfully applied to magnetic resonance spectroscopic imaging to reduce scan times. To reconstruct undersampled MRSI data with arbitrary k-Space trajectories, image-domain based iterative-SENSE algorithm and k-Space based PARS algorithm have been proposed at costs of long computing times.  In this abstract, a new k-Space-domain based parallel spectroscopic imaging reconstruction with arbitrary trajectories using sparse matrices is applied to MRSI with spiral trajectories. It achieves MRSI reconstruction with reduced computing times. The results are demonstrated in an in-vivo study. Results very similar to that reconstructed with fully sampled spiral k-Space data are obtained with different reduction factors.

Richard Young¹, Hacene Serrai¹

¹National Research Council Institute for Biodiagnostics, Winnipeg, Canada

This work describes the extension of the wavelet encoding method in MRSI from 2 dimensions to 3 dimensions. The obtained in-vivo results demonstrate the usefulness of the wavelet encoding method in reducing the acquisition time as compared to CSI. In addition, wavelet encoding appears to preserve the spatial distribution better than the CSI (phantom results not shown). As expected a reduction in signal-to-noise ratio is noticed in wavelet encoding; however, this reduction is minimum.

Lazar Fleysher¹, Roman Fleysher¹, Songtao Liu¹, Oded Gonen¹

¹NYU School of Medicine, New York, New York, USA

While spatial Fourier gradient phase-encoding and spatial Hadamard radio-frequency encoding are two established MR localization techniques, the absence of voxel-shift and interpolation post-processing algorithms for the latter has always placed it at a discouraging disadvantage. In this paper we present a method for voxel-shift and interpolation of Hadamard-encoded data.

Julien Valette¹, Jang-Yeon Park¹, Olli Gröhn¹, Kamil Ugurbil¹, Michael Garwood¹, Pierre-Gilles Henry¹

¹CMRR, University of Minnesota, Minneapolis, Minnesota, USA

In spectroscopy, volume selection can be advantageously achieved using adiabatic π pulses, which enable high bandwidth and B1 insensitivity. In order to avoid the generation of non-linear phase profiles, these pulses are usually used in pairs. However, in the context of spectroscopic imaging, a high enough spatial resolution may limit phase dispersion within each pixel when using only one pulse per selected spatial dimension, yielding a reduced TE and reduced power deposition. In this work, the feasibility of this approach is explored theoretically and experimentally, using a new adiabatic sequence named Pseudo-LASER in the rat brain at 9.4 T.