Zusammenfassung
Introduction:
In spite of significantly improved saturation pulses [1,2]
sufficient outer volume suppression (OVS) in MRSI could only be achieved in combination with PRESS localization [3]. PRESS causes immense chemical shift displacement at high field strength. Therefore metabolite ratios in
significantly large areas of the FOV are not reliable. Slice selective MRSI, which guarantees correct metabolite ratios, has been unfavorable due to high lipid contamination of the spectra. In this work flip angles of multiple, overlapping suppression bands were optimized considering crossing and progressing T1 relaxation during the OVS scheme. Thus sufficient T1 and B1 insensitive outer volume suppression in combination with slice selective
MRSI could be achieved. The optimization was performed for higher-order-phase pulses (HOPP) [4], which provide in comparison to VSS or QPP [1,2,3], an even larger bandwidth and hence a negligible chemical shift displacement.
Materials and Methods:
An iterative numerical optimization algorithm based on Bloch equations, neglecting properties of RF pulses and
spoiling gradients, has been implemented to calculate optimal flip angle combinations for each saturation band [5]. Due to progressing T1 relaxation the optimal flip angle depends on the time delay between saturation band
and the start of the MRSI sequence. To reach T1 and B1
independent OVS all saturation bands have been applied twice [3,6] (Figure 1). The optimized flip angle combinations have been verified using HOPP-pulses as well as
the original B1 and gradient settings by Bloch integration using the Runge-Kutta method (Figure 2). Finally the optimized OVS scheme was applied in Brain and prostate MRSI as well as SV spectroscopy of the spinal cord [7]. All experiments were performed on a Philips Achieva 3T scanner. The HOPP pulses were applied prior to the MRS or MRSI sequence.
Results and Discussion:
Highly selective T1 \& B1 independent suppression could be achieved (Figure 3). The verification of the OVS scheme in vivo shows consistently good fat suppression around the skull in slice selective brain MRSI and complete fat suppression using over-prescribed PRESS MRSI. But while PRESS MRSI data have distorted metabolite ratios, it
is possible to achieve entirely fat free spectra with the correct metabolite ratio directly adjacent to the skull combining the optimized OVS scheme and slice selective MRSI (Figure 4). The lowest residual Mz magnetization can be achieved using the smallest possible number of saturation bands to suit the anatomy and the shortest possible RF pulses.
