Dynamic Magnetic Resonance Nephrography

PURPOSE In this volunteer study, 2 navigator-gated strongly T1-weighted saturation-recovery (SR) sequences, a turbo fast low angle shot (TurboFLASH) and a new true fast imaging in steady precession (TrueFISP) readout technique, were compared for suitability in dynamic magnetic resonance nephrography. MATERIALS AND METHODS Ten healthy volunteers (mean age 26.1 +/- 3.6) were equally divided into 2 subgroups. After bolus-injection of 3.75 mL of gadobutrol (approximately 0.05 mmol/kg body weight), slightly obliqued coronal single-slice images of the kidneys were recorded every 4-5 seconds during free breathing using 1 of the 2 sequences. Time-intensity curves were determined from manually drawn regions-of-interest over the kidney parenchyma. Both sequences were subsequently evaluated with regard to linearity of signal, signal to noise ratio (SNR), and time-dependent behavior of signal intensity curves. RESULTS : The TurboFLASH readout showed better linearity of the signal behavior as compared with the TrueFISP technique (TurboFLASH: no deviation from linearity down to T1 = 400 milliseconds; TrueFISP at T1 = 700 milliseconds: 12% deviation, at T1 = 400 milliseconds: 19%). The time-intensity curves of the TrueFISP sequence exhibited distinctly lower variability than the TurboFLASH approach. The SNR increased with TrueFISP by 3.4 +/- 0.5-fold for native renal parenchyma and by 3.3 +/- 0.9 for contrast-enhanced renal parenchyma. For split renal function evaluation, the linear regression to the signal increase in the first minutes after the first pass could be performed with higher reliability using the TrueFISP technique (increase of correlation coefficient by 17.1%). CONCLUSION A SR navigator-gated TrueFISP sequence seems most favorable for dynamic magnetic resonance nephrography due to the high signal yield and low curve variability.