System identification of tumor growth described by a mixed effects model
Résumé
System identification of treated tumor growth is addressed in this paper. Three main difficulties are examined: (i) the determination of a suited dynamical model structure (modeling problem), (ii) the inter-individual variability of the therapeutic responses (population identification problem or longitudinal data analysis) and (iii) the effects of some categorical factors on the model parameters. To solve these problems, a mixed effect model of tumor growth, a two step identification approach and an estimation algorithm based on expectation maximization, are proposed and applied to in vivo data. A double effect of treatments on the tumor volume responses is pointed out.
| Origine | Fichiers produits par l'(les) auteur(s) |
|---|
Loading...