Iterative optimal design of PET experiments for estimating beta-adrenergic receptor concentration.

To estimate in vivo myocardial beta-adrenergic receptor concentration with sufficient precision and to reduce the experimental complexities in positron emission tomography (PET), an iterative optimal design method is applied. An initial three-injection protocol, utilising [F-18]-labelled (R)- and (S)-fluorocarazolol and unlabelled (S)-fluorocarazolol, is optimised for ligand dosages and administration times to maximise the precision of all model parameters using the D-optimal criterion. Using this experimental protocol, PET data are collected in porcine studies, and model parameters are estimated. All model parameters are identified with satisfactory precision. The in vivo myocardial beta-receptor concentration is 7.5+/-0.6 pmol x ml(-1), which corresponds to the in vitro result of 10.1+/-1.3 pmol x ml(-1). With more accurate parameter values, a simplified two-injection protocol is optimally designed, utilising only radiolabelled and unlabelled (S)-fluorocarazolol, based on a new criterion to maximise the precision of the beta-receptor concentration. This revised optimum design predicts that the in vivo beta-receptor concentration can be estimated with good precision but reduced experiment complexity.