A seven-compartment model for ferrokinetics is presented here and an optimization algorithm of the gradient type for the evaluation of its parameters is described. The reliability of the parameter estimates is tested calculating marginal standard deviations by computing the information matrix throug
Modeling, identification and parameter estimation of bilirubin kinetics in normal, hemolytic and Gilbert's states
โ Scribed by Claudio Cobelli; Mario Frezza; Claudio Tiribelli
- Publisher
- Elsevier Science
- Year
- 1975
- Tongue
- English
- Weight
- 890 KB
- Volume
- 8
- Category
- Article
- ISSN
- 0010-4809
No coin nor oath required. For personal study only.
โฆ Synopsis
Human metabolism of bilirubin is investigated by plasma disappearance curves of 3H-bilirubin in normal, hemolytic, and Gilbert's states.
A two-compartment model of bilirubin kinetics is presented. Structural identifiability of the model is investigated. Model's parameter estimation from experimental data is performed in the least-squares sense with a procedure particularly suitable for this criterion function. Some problems related to the estimation of parameters of compartmental models are discussed. Several parameters of clinical interest are derived from the model.
The system has been simulated on a digital computer using the CSMP simulation language, provided with the estimation algorithm. Results obtained on a sample of two clinical situations (Gilbert's syndrome, hemolytic syndromes) and on a control group are presented. Clinical inferences drawn from the parameters obtained by the model are discussed: (i) in Gilbert's syndrome the metabolic defect possibly concerns the intrahepatic binding of bilirubin; (ii) if alterations in radiobilirubin kinetics occur under hyperhemolitic conditions, they possibly indicate that the excretory capacity of the liver has been exceeded.
๐ SIMILAR VOLUMES
A new approach to the dynamic modeling of chemical kinetics is presented. The technique is based on systematic planning of computer experiments, which allows an empirical model for the computed responses to he developed in the space of parameters. The empirical equations which are obtained provide c