City Research Online

Abstract

This thesis has described, and illustrated the use of, causal probabilistic network (CPN) techniques in the modelling of lipid and lipoprotein metabolism. The models constructed enable prediction of the health outcome associated with disorders of lipid and lipoprotein metabolism, i.e. an individual's "potential for atheroma", and as such could be used as part of a strategy for decision support in the management of hyperlipidaemia.

CPN techniques are shown to be appropriate in modelling lipid and lipoprotein metabolism where relatively few data exist to support hypotheses concerning metabolic pathways. These techniques enable uncertainty and conditional independence assumptions to be included in model formulation, limiting the amount of data necessary to construct a model; a high level of uncertainty being represented where few data exist.

The modelling process has highlighted numerous areas of controversy concerning the physiology of lipid and lipoprotein metabolism. In particular questions have been raised concerning the mechanisms involved in reverse cholesterol transport and in the metabolism of very low density lipoproteins (VLDL) and low density lipoproteins (LDL). In addressing these issues, via the modelling process, it has been possible to illustrate consistency between data from numerous studies, both experimental and epidemiological. In effect, the models provide a summary of current knowledge which can be used to direct further research toward areas of controversy or areas where few data exist describing lipid and lipoprotein metabolism, and in the interpretation of the results of new research describing metabolic pathways.

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

Export

Downloads