Four injection profiles were evaluated in 40 subjects (10 each). The profiles were designed to provide a signal intensity plateau for a 20 second contrast-enhanced MR angiography scan, using gadobenate dimeglumine. The injection profiles were: 1) standard, non-diluted, 1.6 mL/sec; 2) diluted, 1.6 mL/sec; 3) diluted, bi-phasic, and 4) diluted, patient-tailored. All subjects received a total dose of 0.1 mmol/kg, divided between a test bolus and a full bolus. The “diluted” protocols included dilution of the total dose to 40 mL with normal saline. Signal intensity profiles were measured and analyzed for width of plateau (FW80M) and peak signal intensity.
In the TIP algorithm, the test bolus SI is converted to contrast agent concentration [CA] to give the impulse response function (IRF). This IRF is then convolved with arbitrary “full” injection profiles (injection rate vs. time) to predict the full bolus concentration profile, which is then converted to predicted SI for the specific CE-MRA scan parameters.1-3 The user specifies a desired plateau duration (eg, equal to the CE-MRA scan duration) and the automated tool calculates the desired, multi-phasic injection protocol.
To-date, 34 (of 40) subjects have been enrolled in this IRB-approved study. Each subject received a test bolus and one of four randomized full-bolus injections, with total dose of 0.1 mmol/kg gadobenate dimeglumine (Bracco Diagnostics, Milan). Injections were controlled with a multi-phasic power injector (Spectris Solaris, Bayer). Injection parameters (Table 1) were selected to provide 20 sec of signal enhancement (except the non-diluted, single-phase injection that represents a “typical” clinical injection). Signal intensity profiles were measured at 3T with a thick-slice, <1.5 sec temporal resolution, 3D SPGR acquisition with CE-MRA mimicking parameters: TR = 3.5 ms, TE = 1.5 ms, flip angle = 30°, FOV = 400 x 248 x 64 mm3, voxel volume = 2 x 2 x 8 mm3, covering the entire aorta (oblique sagittal orientation) to reduce contaminating inflow signal. The test bolus acquisition was similar with: TR = 2 ms, TE = 0.8 ms, flip angle = 7°. Each acquisition collected dynamic volumes for 100 sec. For both acquisitions, subjects were instructed to hold their breath as long as reasonable when contrast enhancement was first observed in the pulmonary arteries. Signal intensities were measured from a single ROI placed in the supra-celiac abdominal aorta.
Example full-bolus signal intensity profiles are shown in Figure 1. The clinical standard [non-diluted, 1.6 mL/sec] injection shows a peaked signal intensity profile with no “plateau” phase. The three diluted injections show elongated signal enhancement phases compared to the clinical standard injection. Quantitative results of bolus profile width (FW80M) and peak signal intensity (normalized to baseline) are shown in Figure 2. The preliminary (n=34) results show no significant differences in peak signal intensity, and a significant increase in FW80M for all three diluted protocols compared to the non-diluted protocol.
Bi-phasic and TIP injections, because of their high initial injection rate, produce a more rapid SI upslope (Figure 3). The more rapid upslope allows earlier triggering of the CE-MRA scan, which will reduce venous enhancement. This is especially important for longer scans with their overall slower injection rates and slower upslope, that are more prone to venous enhancement. Figure 4 displays TIP injection profiles and the corresponding predicted SI profiles for three different requested plateau durations (20, 40, and 60 sec). More injection phases are required to maintain longer signal intensity plateaus.
1. Wilson GJ, Maki JH. J Magn Reson Imaging, May 5 (2016) Epub doi: 10.1002/jmri.25298.
2. Wilson GJ, Woods M, Springer CS Jr, Bastawrous S, Bhargava P, Maki JH. Magn Reson Med (2014)72(6):1746-54.
3. Wilson GJ, Springer CS, Bastawrous S, Maki JH. Magn Reson Med, Jun 14 (2016) Epub doi: 10.1002/mrm.26284.