Development of FIR Models for the Central Nervous Control of the Cardiovascular System

Introduction

The cardiovascular system consists of two primary components. On the one hand, the hemodynamics are a part of it. The hemodynamics are comprised of the heart and the blood vessels. The model of the hemodynamics describes, how blood is being pumped through the heart and the blood vessels.

The hemodynamics can be viewed as a hydraulic problem, which can be modeled quite well by means of ordinary differential equations (ODEs). The ODE model of the hemodynamics used in this project had first been developed in ACSL by Montserrat Vallverdú in her Ph.D. dissertation. The model was adapted to her needs by Àngela Nebot in her Ph.D. dissertation [1,2]. Finally, the model was reinterpreted by me as a bond graph model. This version of the model was coded in Dymola/Modelica [3].

On the other hand, also the central nervous control (CNC) of the hemodynamic system forms part of the cardiovascular system. The model of the CNC describes the control mechanisms that regulate the flow of blood through the hemodynamic system.

The physiological mechanisms implementing the CNC in the body are not completely understood. The qualitative model of the CNC used in this project consists of five separate control loops that regulate, based on the blood pressure in the arteries of the brain (carotid sinus pressure), the heart rate, the myocardiac contractility, the peripheric resistance, the venous tone, and the coronary resistance.

The CNC was modeled originally by Montserrat Vallverdú in her Ph.D. dissertation using NARMAX model. Àngela Nebot replaced the NARMAX models in her Ph.D. dissertation by corresponding FIR models.

Both model types were identified on the basis of measurement data of Valsalva maneuvers obtained from patients with a heart catheter. These data were recorded in a hospital in Barcelona. Whereas the NARMAX models were only able to reproduce the low-frequency component of the measurement data that is caused by the respiration of the patient, the FIR models were perfectly capable of also identifying the high-frequency component of the measurement data that is caused by the heart rate. Consequently, the FIR models turned out to be considerably more accurate.


Most Important Publications

  1. Nebot, A., F.E. Cellier, and M. Vallverdú (1998), Mixed Quantitative/Qualitative Modeling and Simulation of the Cardiovascular System, Computer Methods and Programs in Biomedicine, 55(2), pp.127-155.

  2. Nebot, A., F. Mugica, F.E. Cellier, and M. Vallverdú (2003), Modeling and Simulation of the Central Nervous System Control with Generic Fuzzy Models, Simulation, 79(5), pp.648-669.

  3. Cellier, F.E. and A. Nebot (2005), Object-oriented Modeling in the Service of Medicine, Proc. 6th Asia Simulation Conference, Beijing, China, Vol.1, pp. 33-40.

  4. Cellier, F.E. (2003), Inductive Modeling, Unterlagen zur Vorlesung über die mathematische Modellierung physikalischer Systeme, Vorlesung 33.

Sponsors


Deutsche Version
Homepage


Last modified: January 22, 2006 -- © François Cellier