ausblenden:
Schlagwörter:
-
Zusammenfassung:
Introduction:
Measuring metabolic profiles at ultra-high fields usually re
quires short echo times to capture also low concentrated or fast relaxing metabolites. One sequence beside STEAM that was proposed in this regard is SPECIAL [1]. Due to B 1
limitations of volume coils at ultra-high fields surface coils are often preferred. A SPECIAL semi-LASER (SPECIAL-sLASER) sequence is presented that is less sensitive to the strong B 1 inhomogeneities than regular SPECIAL. Furthermore, slight adjustment to the pre-localization pulse were made to diminish lipid artifacts that occur due to magnetization transfer (MT) effects [2].
Materials Methods:
All measurements were performed using a Philips
7T Achieva Scanner (Philips Medical Systems, Cleveland) using a RAPID quadrature surface coil. A voxel of 25x18x20mm was placed in the occipital lope. The regular SPECIAL refocusing pulse was replaced by a pair of trapezoidal adiabatic inversion pulses. 2nd order shimming and volume based power optimization [3] were used. The minimum echo time was TE=16ms with an TR=7000ms. Two outer-volume suppression pulses were interleaved with VAPOR [4]. Additional care was taken to minimize lipid artifacts due
to magnetization transfer or transient NOE effects. Instead of turning the pre-inversion pulse off every other scan the frequency offset of the pulse was changed such that the inverted slice was shifted outside the VOI. The data
were reconstructed offline and fitted (Table 1) in LCModel [5]. Real pulse simulations in GAMMA [6] were carried out to obtain suitable basis sets. 2D sequence profiles incorporating in-vivo B 1 distribution data were simulated
in Matlab to show the improved insensitivity to B 1
variations of SPECIAL-sLASER. Furthermore, phantom measurements were performed to validate the suitability of the MT compensation scheme to eliminate lipid artifacts.
Results Discussion:
Representative good quality in-vivo data acquired
with SPECIAL-sLASER is shown in Figure 1. In the phantom measurements (Fig. 2) it could be shown how the lipid contamination in the phantom data is influenced by the RF of the pre-localization pulse and how this problem
could be tackled. Furthermore, Figures 3a 3b show comparisons of B1 sensitive slice profiles for regular SPECIAL and SPECIAL-sLASER that demonstrate the potential improvement due to the adiabatic behavior.
