Measurement of short and ultrashort T2 components using progressive binomial RF saturation

A new magnetic resonance (MR) method for measuring T2 relaxation times in tissues is proposed. The method is based on a T2 selective saturation period followed by sampling of the remaining longitudinal magnetization. Saturation of the longitudinal magnetization is accomplished by a single binomial RF pulse of zeroth order with a constant flip angle. The T2 selectivity is controlled by the RF pulse duration. A full T2 spectrum can be obtained by performing a series of measurements with varying RF pulse duration. On a conventional 1.5 T system this approach allows detection of T2 components as short as several hundred microseconds. A major limitation is the method's susceptibility to resonance offsets. At typical offsets of 0.1-0.2 ppm the sensitivity of the method is limited to a T2 range below 20 ms, which corresponds to an RF pulse duration shorter than 50 ms. The new method was used to acquire T2 spectra from the liver of pigs in vitro on a conventional 1.5 T system. We observed a short T2 component around 17 ms and an ultrashort T2 component in the range of 0.9-1.1 ms. Numerical simulations and in vitro measurements suggest that resonance offsets have effects that require further investigation.