Datensatz

Zusammenfassung

Purpose/Introduction: Continuous ASL with labeling coils placed at the subject´s neck (DC-CASL, [1]) has a high sensitivity [2], does not suffer from magnetization transfer effects and leads to lower power deposition than any other CASL technique. The latter makes it especially interesting for ultra-high field applications where the TR in ASL is limited by the high SAR evoked by the inversion. So far, simulations of the efficiency were performed for field strengths up to 3 T [2–3]. However, the shorter T2 of arterial blood at 9.4 T requires reassessing of the labeling process at this field strength. Subjects and Methods: Bloch simulations were performed in Matlab for different blood velocities (vblood = 0.01 - 1 m/s), transmit fields (B1 +=2.5, 3.5 and 4.5 lT) and gradients (Gz = 0-10 mT/m) assuming T1 = 2400 ms [5], T2 = 30 ms [6] at 9.4 T. For comparison, calculations were also performed with 3 T relaxation times (T1 = 1330 s [7], T2 = 250 ms [8]). The efficiencies of cases with an adiabaticity \1 were set to zero. In order to take the temporally varying blood volume and the parabolic velocity profile into account, the total efficiency was calculated as described in Ref. [2]. Two flow weighting models were investigated: I. A two period model [2]; II. A detailed representation of vblood over time adapted from [4]. A constant transmit field was assumed since the influences of the B1 + shape of a surface coil on the inversion are negligible [3]. Results: Figure 1a shows the overall efficiencies for three transmit field strengths at 9.4 T and Fig. 1b the corresponding values at 3 T. For a given B1 +, the more detailed model II (b) requires slightly stronger gradients than model I (a) to achieve the optimal performance. The flow weighted inversion efficiency during the cardiac circle is depicted in Fig. 2. Table 1 comprises the efficiencies for some B1 +/Gz combinations. Discussion/Conclusion: In comparison to 3 T, stronger gradients are required for labeling at ultra-high field. However the lower efficiency will be compensated by the longer longitudinal relaxation time at 9.4 T. One possibility to reduce the T2 effect may be the utilization of local non-linear gradient coils [3]. The strong dependency on vblood complicates the quantification of DC-CASL especially in patients.