Figure 1.
Sequence diagrams of GRE, UGRE and QISS.
In GRE, no cardiac triggering is used. Here, the TIAMO trigger changes the excitation mode directly after every excitation pulse, which means that both modes needed for a complete slice are acquired in an interleaved fashion. Saturation pulses are applied every TR. In UGRE, the cardiac trigger event starts the acquisition of a complete slice in a single shot. Due to TIAMO, this acquisition has to be repeated to acquire the same slice with the second mode. Venous saturation RF pulses are applied only sparsely. QISS uses three different saturation pulses to prepare the image contrast: In the imaging slice to suppress background tissue, a travelling venous saturation pulse below the imaging slice and a fat saturation pulse in the imaging slice. After the venous saturation, a time interval QI allows unsaturated arterial blood spins to enter the imaging slice. No saturation pulses are applied during the single-shot slice acquisition. The duration of the trigger events is shorter than pictured in these diagrams.
Table 1.
Sequence parameters.
Figure 2.
Complete MIPs of GRE, UGRE and QISS of one volunteer.
MIPs of all acquired stations merged together manually in one volunteer for GRE, UGRE, and QISS. In this volunteer, the two stations covering the upper part of the thighs (marked by white double-headed arrows) used individual RF shims to prevent complete artery signal dropout. Vessel signal fluctuation in GRE could be mostly avoided in UGRE and QISS due to cardiac triggering (grey arrows). Tissue signal intensity fluctuation in QISS is most probably due to time delay induced by heartbeat triggering giving the surrounding tissue time to relax. White arrow in QISS shows vein visible in QISS sequence, as only one of the TIAMO RF shim modes could be used, preventing complete venous suppression.
Table 2.
Averaged VBR of all volunteers.
Figure 3.
Signal variation flow artifact.
Signal variation flow artifact (white arrow in GRE) shown for consecutive axial slices at position of the grey arrow in Fig. 1 compared to the corresponding images in UGRE and QISS. In UGRE large arteries seem to be slightly broadened (white arrow). In the QISS images, B1 inhomogeneities are considerably more pronounced as TIAMO could not be used (white arrows). For GRE and UGRE, especially the smaller vessels can be more easily delineated than for QISS (grey arrows). The upper row shows the slice position superior to the images of the lower row.
Figure 4.
Artifacts in GRE, UGRE and QISS.
The upper row visualizes the flow artifact in GRE images, aliasing in UGRE, and B1 inhomogeneity in QISS, in each case marked by a white arrow. Due to use of TIAMO, the B1 inhomogeneity could be ameliorated in GRE and UGRE. Zoomed images of the left thigh in the lower row show GRE, where the flow artifact in the phase-encode direction and a small signal dropout inside the vessel lumen are clearly visible (grey arrow), UGRE and QISS. For UGRE, slight broadening of the arteries can be observed (grey arrow). In the QISS image, blurring of the vessel walls is clearly observed (grey arrows). Artery circumference can be delineated best in UGRE.
Figure 5.
MIPs showing the first station of the thigh. The upper row shows images acquired using TIAMO with the CP+ and CP2+ modes. In the case of QISS, only CP+ was used. In the lower row, dedicated TIAMO shims were calculated after the acquisition of B1 maps. In the case of QISS, only one of the two modes could be used. One can clearly see that the left superficial femoral artery can be delineated throughout the entire station with the dedicated B1 shims (grey arrows), whereas the finger arteries lost signal (short white arrow in GRE and UGRE).