Item

Abstract

Introduction Non-water suppressed metabolite cycled proton magnetic resonance spectroscopy (MC 1H-MRS) has been proven to enhance the frequency resolution and the signal to noise ratio (SNR) of the spectrum at 3 Tesla [1-2]. Previously the adiabatic inversion pulse for MC 1H-MRS was optimized to exploit these advantages for application in the human brain at 9.4T [3]. In this work, we examine the performance of STEAM [4] based MC 1H-MRS [3,5] compared to water suppressed 1H-MRS using a numerically optimized short water suppression (WS) sequence with respect to spectral resolution and signal-to-noise ratio (SNR) in the human brain at 9.4T. Methods All experiments were carried out using a 4 channel transceiver array coil [6] connected to a whole body 9.4Tesla Magnetom SIEMENS scanner. For 1H MRS localization a STEAM sequence (TE/TM/TR: 10/50/3000ms) was used. An optimized water suppression scheme consisting of 7 excitation pulses and orthogonal spoiler gradients was developed. The post processing of the data included in the given order: frequency alignment using the water reference [7], averaging, ECC [8], channel combination [9] and zero filling using factor of 2. Both sequences were applied on healthy volunteers (fig. 1-2) (NEX: 128-320, voxel size: 15x15x15mm3) placing a grey matter voxel in the occipital lobe. The optimized MC inversion pulse (IP) for the resulting B1+ of 22μT had a duration of 23ms and a frequency offset of 350Hz. Results The in vivo data (fig. 1) demonstrated that the simultaneously acquired water reference of MC data allowed for frequency and phase alignment of the different averages (fig. 4) leading to a line width of 25.9Hz and SNR of 38.2. In addition, the MC technique provided a WS factor of 99.8. On the other hand, the WS spectrum resulted in a line width of 28.9Hz (+3Hz difference), SNR of 33.3 (-5dB difference) and a WS factor of 99.7. The difference in SNR and linewidth is mainly produced by physiological motion (e.g. breathing, blood flow) as well as motion of the volunteer demonstrating the importance of the simultaneously acquired water reference spectrum both for the ECC and correct averaging of the acquisitions (fig. 2). Finally the MC data enabled the reconstruction of high frequency resolution spectra similar with other studies conducted on 9.4T [10]. Conclusions 1) MC 1H-MRS enables phase and frequency alignment of individual acquisitions as well as ECC of the spectrum at 9.4T 2) MC 1H-MRS performs better in terms of SNR and line width and thus effective spectral resolution compared to WS 1H-MRS and 3) MC 1H-MRS results in a free of gradient modulation sidebands and eddy current artefacts spectrum and excellent WS performance.