Pade–Laplace analysis in the fitting of multi-exponential nuclear magnetic resonance relaxation decay curves

Abstract

Pade–Laplace analysis has been proposed as a method for identifying the number of exponential components in a multi-exponential decay. The usefulness of this method has been examined by computer simulations, for data sets typical of T_1ρ measurements in NMR. Emphasis was placed on resolving exponential functions with similar time constants in the presence of experimental noise. For three exponential decays Pade–Laplace analysis was found to be only on a par with non-linear least-squares (NLLS) fitting at identifying the three components. Similarly no improvement was seen in the accuracy of the fitted parameters over those found in NLLS fitting.