Schätzung der Input- und Response-Funktion bei numerischer Konvolution und Dekonvolution im Rahmen pharmakokinetischer Untersuchungen

Abstract

A numerical procedure for convolution and deconvolution of pharmacokinetic data is looked at as a statistical estimation problem. Estimators for the input-and response function based on a point-area convolution method are derived and their asymptotical properties are discussed. Several approaches to estimate the asymptotic variance of the estimators including jackknife and bootstrap techniques are discussed. A simulation study to investigate the finite sample performance of the estimators is presented.