Kinetic parameter estimation from compartment models using a genetic algorithm

Kinetic parameters were estimated from a three-compartment fluorodeoxygluco se model with three rate constants using a genetic algorithm. The performan ce of the genetic algorithm was investigated by simulation studies, in whic h brain time-activity data (TAD) were generated using cited mean values of rate constants and the plasma TAD obtained from positron emission tomograph ic studies. The accuracy of kinetic parameter estimation using the genetic algorithm was compared with that using the non-linear least-squares (NLSQ) method. The margin of error in the parameters estimated using the genetic a lgorithm tended to be smaller than that obtained by the NLSQ method. Althou gh not statistically significant at a noise level of 5% in the brain TAD, t he difference between the two methods became significant for all parameters at a noise level of 15% or higher. Our results suggest that the genetic al gorithm is a promising means of estimating kinetic parameters from compartm ent models, because it is more robust against statistical noise than the NL SQ method and it can be rendered highly parallel far processing. ((C) 1999 Lippincott Williams & Wilkins).