Towards a full 3D MR parameter quantification by means of (demodulated) off-resonance insensitive balanced SSFP

Purpose/Introduction: Quantitative MR parameter mapping has been demonstrated to allow for improved diagnosis and staging of diseases[1]. In this work, a novel procedure for the simultaneous quantification of the relaxation times T1, T2, the relative spin-density PD and the off-resonance frequency Δω is proposed. To this end a series of different phase-cycled 3D-bSSFP measurements are employed. The contrast behavior of the bSSFP sequence depends on both T1 and T2 and varies strongly with the applied phase-cycle and the local off-resonance frequency. We show, that the data can be numerically fitted to the bSSFP signal equation allowing for the simultaneous extraction of T1, T2, PD and Δω. An advantage of this approach is that it provides an intrinsic compensation for bSSFP off-resonance effects. The methodology is validated in a phantom study and the applicability is demonstrated in-vivo. Subjects and Methods: A series of 3D-bSSFP-experiments with different phase-cycles are acquired. The corresponding signal equation is well known[2] and has been shown to resemble an ellipse-equation[3]. By fitting the data numerically onto the ellipse signal-equation T1, T2, PD and the local offresonance frequency Δω for each voxel in the 3D image can be obtained. Phantom experiments were performed on containers doped with different Gd-concentrations to provide a wide range of T1- and T2-values. In addition, in-vivo brain data has been acquired from a volunteers with an isotropic resolution of 1.3x1.3x1.3mm³ within 21min. Results: The obtained phantom parameter maps show good agreement with reference values. A comparison is depicted in the Fig.1. The parameter maps do not suffer from banding-artifacts due to “off-resonant”-spins despite the relatively long TR. In contrast to other methods T2-values small as 10 to 20ms can be measured in high accuracy. The in-vivo data in Fig.2 shows accurate T1, T2 as well as PD and Δω (not shown) maps. Off-resonance effects, usually present in the frontal cortex, are compensated. Discussion/Conclusion: A novel method for parameter quantification has been presented. The method is based on a single bSSFP-sequence with fixed TR and a single flip-angle. Only the phase-cycle is changed during the measurement. We have shown that full 3D-MR-parameter mapping in high isotropic resolution is feasible. We hypothesize that significant scan-time reductions can be achieved by employing acceleration techniques such as parallel imaging or Compressed Sensing, making the method an promising candidate for clinical applications