Spectrally Selective B1 Insensitive T2 Preparation Sequence for 3T Imaging

R. Nezafat, J. A. Derbyshire, R. Ouwerkerk, M. Stuber, E. R. McVeigh Lab of Cardiac Energetics, Division of Intramural Research, NHLBI, NIH, DHHS, Bethesda, MD, United States, Biomedical Engineering, Johns Hopkins University, Baltimore, MD, United States, Radiology, Johns Hopkins University, Baltimore, MD, United States Introduction: T2 magnetization preparation (T2 Prep) and spectrally selective fat suppression (FatSat) sequences are commonly used to enhance imaging contrast [1]. Increased B1 inhomogeneity at 3T necessitates a preparation sequence that is more robust with respect to B1 inhomogeneity. We propose a spectrally selective, B1 insensitive, T2 Prep sequence for imaging at 3T. Numerical simulations, phantom and in-vivo images acquired using this technique are presented. Theory: T2 Prep: A novel B1 and B0 insensitive T2 magnetization preparation sequence is presented which results in a uniform T2 Prep across the imaging field of view. This sequence is based on the simultaneous frequency and amplitude modulation of the applied RF pulses [2]. A segmented BIR4 with insertion delays of t = T between pulse segments 1 and 2, 3 and 4 respectively was calculated for a net flip angle of -360 or 0 as shown in Fig 1. This segmented pulse forms a B1 insensitive T2 prep weighting sequence and was described earlier for use as a zero or double quantum filter [3]. The delay T can be set to achieve a desired contrast between different tissues. Fat Sat: we modified this segmented BIR4 further by inserting an extra delay δτ between the BIR4 segments 1 and 2 to suppress the signal from fat. The equal delay Τ in a segmented BIR4 causes the acquired phase in the first delay to be compensated in the second delay. However, by introducing non-equal delays, the acquired phase in the first segment will not be compensated in the second delay, which results in an additional frequency offset dependent phase. The phase difference between two main components of fat and water can be set to 90 by adding a time delay difference of δτ = 1/4∆f, in which ∆f is the resonance frequency difference of water and fat. The water component will return to longitudinal magnetization by the adiabatic half passage component of the pulse in the last segment. However, the adiabatic half passage will not return the fat signal to the longitudinal axis, instead the fat components will remain in the transverse plane. Subsequently, this transverse fat signal is dephased by a spoiling gradient, resulting in suppression of the fat signal without affecting the water signal. Method: Simulations & Phantom Studies: To illustrate the response of the magnetization to the preparation sequence in the presence of B 1 and B0 inhomogeneity, computer simulations were performed to calculate the normalized longitudinal magnetization Mz/Mequ after experiencing the T2 Prep and T2 Prep/FatSat pulse. To investigate the T2 weighting, B1 and B0 variation, and T2 Prep/Fat Sat efficiency, phantom studies were performed using (i) a water phantom containing 5 tubes filled with different T2 species