Determination of dyslipidemia by discrete and pseudo-discrete features

The necessity of controlling the lipid state of the organism originates from the association of dyslipidemia with atherosclerosis, heart diseases, diabetes mellitus, etc. In the clinical practice, the determination of the exact type of dyslipidemia is not straightforward due to certain problems: daily biological fluctuations, dietary effects, strong overlapping in characteristic plasma lipid levels in some classes of type llb, III and IV, frequent inavailability of a complete set of diagnostic measurements, as well as certain overlapping in the borderline values. These provide the objective to create a computer program for diagnostics* of dyslipidemia in 16 classes. The apparatus of Bayes classification has been applied using 13 discrete and 7 pseudo-discrete features. The proposed system is suitable for monitoring the treatment of dyslipidemia as well as for the purposes of students' education and post doctoral training.