City Research Online

Abstract

A comprehensive Teeneeaeca model of the human respiratory system, incorporating events within the respiratory cycle, has been developed. This has been built up in a systematic manner, starting from a simple description and adding in a step-by-step fashion additional features. At each stage of complexity, the model behaviour has been examined and the validity of the assumptions tested.

The initial models involved many gross physiological assumptions with many features lumped into single compartments. Nevertheless, some of the predictions were physiologically plausible. ‘These models highlighted, however, the inability of either central tissue control or arterial chemoreceptor control individually to provide an adequate description of all dynamic phenomena.

Subsequent model development included additional tissue compartmentation and the effects of arterial blood CO2, and 02 concentrations on cerebral blood flow and cardiac output.

Further extension involved the separate development of a ling gas exchange model which included the variations of both lung volume and gas concentrations due to inspiration and expiration. This was then combined with the circulatory model to provide a comprehensive description of the respiratory system within which instantaneous nee of variables such as arterial blood CO, and oxygen concentrations during the respiratory cycle could be studied. Both for breathing and for the excessive production of muscle tissue co, as occurs in exercise, model results were within the range of physiological acceptability.

After obtaining qualitatively acceptable results, two approaches to system identification and parameter estimation were adopted. The first involved functional minimisation using the system trajectory in the hyperspace of the model states. The second used pattern recognition techniques whereby parameters were adjusted in order to mitch selected features of the simulated response with the corresponding features of physiological test data. This novel approach may provide a powerful technique for the identification of complex systems and in the respiratory context enable alternative hypotheses regarding controller strategy to be evaluated.

Having obtained a satisfactory mathematical model of the human respiratory system, its simulation using a digital computer can provide predictions about the system's behaviour without the use of difficult and time consuming experimentation. Whilst same physiological test data will still be required to validate the model predictions, the understanding of the system dynamics provided by the model enables experiments to be designed much more effectively in order to maximise the information obtained from the minimum number of tests.

Publication Type:	Thesis (Doctoral)
Subjects:	Q Science > QA Mathematics > QA75 Electronic computers. Computer science
Departments:	School of Science & Technology > Department of Computer Science School of Science & Technology > School of Science & Technology Doctoral Theses Doctoral Theses

Export

Downloads