Sources of artifact and systematic error in quantitative snapshot of FLASH imaging and methods for their elimination

Snapshot FLASH preceded by a preparation sequence is used to measure NMR parameters. Some preparation information is lost during snapshot FLASH acquisition as the magnetization evolves towards the steady state. This problem is often addressed by assuming that image intensity is proportional to the central k-space line, which may be sampled immediately after preparation using a centric-reordered scheme. Although some limitations of this assumption have been discussed, they are often ignored in parameter measurement. In this study, magnetization evolution effects in snapshot FLASH were investigated by computer simulation and phantom experiments in the context of T2 measurement. It was found that the standard analysis may lead to significant error in parameter measurement. Some improvement may be obtained using a more sophisticated analysis to take account of magnetization evolution. However, at the expense of a 50% reduction in signal intensity, accurate values may be obtained by a novel method that uses crusher gradients to eliminate evolved magnetization.