Optimal experiment design for PET quantification of receptor concentration.

Abstract

The mathematical models used to analyze positron emission tomography (PET) data obtained for receptor quantitation have many unknown parameters which must be estimated from the data. Obtaining unique and precise estimates of the model parameters from PET data is difficult as a result of the complex interdependence of the parameters. Here the authors address the task of estimating the concentration of myocardial beta-adrenergic receptors using unlabeled and (18)F-labeled S(-)-fluorocarazolol as the receptor ligand. For a three-injection study the authors have optimized the ligand injection times and dosages using the D-optimal criterion for estimating receptor concentration. They found that in optimizing a three-injection experimental design, the dose of ligand in the third injection approaches zero so that the optimal three-injection design is actually a two-injection experiment. Using this optimal experiment, the authors demonstrate estimates of receptor concentration that are almost five times as precise as compared to an empirically designed three-injection experiment.