Some extensions to the exhaustive-modeling approach to structural identifiability (Q1063530)

After reviewing graphical criteria for testing structural controllability and structural observability for linear compartmental models, the paper presents some new extensions to the exhaustive-modeling approach for analysing identifiability. It is shown how this analysis may be aided by extending the matrices to include the environment as an extra compartment, and by the grouping together of several model constraints in a sensibly chosen fashion. These ideas are illustrated by an identifiability analysis of a five- compartment model for the distribution of bone-seeking radionuclides in the body.