Concerning the Derivation and Representation of Parameter Images in NMR Imaging

Abstract

In this work several aspects of the derivation and representation of parameter images in NMR imaging are taken into consideration. First we discuss the use of statistical parameter estimation techniques to determine the minimum variance bound (MVB) on the precision of any parameter determined from an NMR experiment and give some practical examples. Then we explore the effect of experiment design, i.e. the choice of sequence type and timings, on the precision of the estimated parameters achievable per unit experiment time. In the third section we compare some practical methods to extract p, T₁ and T₂ images from a set of partial saturation, saturation recovery or inversion recovery experiments which to a greater or less extent take into account the fact that a multiple echo/multiple repetition time experiment provides a set of largely independent T₁ and T₂ measurements. In the final section a provisional attempt is made to develop a non-linear gray scale for the display of ρ, T₁ and T₂ images which takes into account the estimated precision of the displayed parameter values.