An iterative technique for refinement of RF pulses for magnetic resonance imaging

Recent advances in non-Fourier-based magnetic resonance imaging (MRI) methods have treated a need for spatially-selective RF pulses. In applications such as wavelet encoded imaging, narrow profile excitations are necessary for effective image reconstructions. In this paper we discuss problems with Fourier-based imaging, and then present an iterative technique which is implemented on a host imaging computer. This method of RF pulse optimization produces considerably accurate pulses for certain target profiles.