A Fast Method for Obtaining a “Continuous” Sample of the Inversion-Recovery Curve Using Spatially Swept Adiabatic Inversion

The measure of the longitudinal, or T, , relaxation time in nuclear magnetic resonance experiments is a common diagnostic tool. There have been several methods proposed for measuring this parameter. Most of these methods measure the recovery of longitudinal magnetization after inversion or saturation. Conventional ( 180”-Ti90”) experiments measure the signal strength for a single temporal point Ti on the recovery curve. These protocols are typically long, due to the necessarily long times between NMR experiments (TR), which must be on the order of five times the T, constant or more. Many new techniques require only one experiment to obtain several points on the relaxation curve (Z-3). These are based on a technique proposed by Look and Locker (4) and later by Kaptein et al. (5) in which each data-collection experiment involves one inversion followed by N small-tip-angle RF pulses which “probe” the signal strength over time. The change in signal strength for each of the N acquisitions is correlated with the difference in Ti , allowing the relaxation rate to be sampled at N points in one data-collection period. Another group of T, methods is based on the equilibration of magnetization under rapid excitation (6, 7). These methods are also much faster than conventional inversion-recovery methods. A common quality of all of the above techniques is that they give a limited number of discrete measures of longitudinal relaxation. The method described in this Note is able to give a much greater sampling of the relaxation curve in a single experiment. It correlates Ti with a spatial dimension by adiabatically inverting spins in the presence of a magnetic-field gradient. An example of the pulse sequences used is shown in Fig. 1. The change in frequency of the inverting RF is constant (and adiabatic) over time, thus causing an inversion of spins which is temporally linear with distance in the spatial dimension. If the region of the sample which is inverted by this pulse is homogeneous in T,, then the inversion-recovery curve over time will be represented as a continuous inversion-recovery curve over space. The amount of the relaxation curve visible in the experiment is equal to the sweep time. A simple spin-echo (or gradient-echo) experiment with a readout gradient in the same direction as the inversion will allow one to discriminate points along the inversion-recovery curve at high resolution. The choice of adiabatic inversion parameters is governed by three major equations. These equations are given with respect to the effective field shown in Fig. 2, where H,