Parameter Estimation in 2D Magnetic Resonance Biexponential Relaxometry Through Use of Inversion-Recovery Null Points

Two-dimensional magnetic resonance transverse relaxometry (2D-MRR) provides a means of characterizing water pools of different mobility. It has been applied to tissue analysis in biomedical applications as well as to materials science. When implemented using an inversion-recovery sequence, the 2D-MRR experiment permits, in principle, nulling of a selected term in a 2D biexponential model. If this were achieved through appropriate selection of the experimentally defined inversion time TI, the resulting signal would be monoexponential and decay parameters could be readily defined. While exact nulling cannot be achieved in practice, TI values may be identified that result in decay curves that are statistically classified as monoexponential according to the Bayesian information criterion (BIC). We have developed an algorithm, Custom Objective Function for Fitting Exponential Equations (COFFEE), based on this concept that permits higher accuracy in the estimation of 2D-MRR parameters. A further advance, Enhanced Spacing of Points for Reduced Errors with Sensibly Selected Objectives (ESPRESSO), uses prior information to tailor TI value selection to enrich the set of transient signals with monoexponential signals to further improve parameter estimation.