On the convergence of the generalized linear least squares algorithm (Q2568626): Difference between revisions

The authors present theoretical considerations concerning the generalized linear least squares (GLLS) algorithm, first introduced by \textit{D. Feng} and \textit{D. Ho} [Parametric imaging algorithms for multicompartmental models dynamic studies with positron emission tomography, in K. Uemura, N. A. Lassen, T. Jones, and I. Kanno (eds.), Quantification of Brain Function:Tracer Kinetics and Image Analysis in Brain PET, Elsevier Science Publishers, Amsterdam,The Netherlands 127--137 (1993)]. They consider the inverse problem of parameter estimate of a linear continuous dynamic system by introducing in GLLS both Gauss-Newton and alternating Gauss-Newton algorithms for solving it. Numerical experiments in the field of clinical positron emission tomography are also presented.