Possibilities of Multilayer Perceptron in Complexing Risk Factors of Diabetic Foot Syndrome

The levels of markers of endothelial dysfunction, MMP9, and polymorphisms of genes of vascular tone regulation factors, platelet receptors, vascular wall remodeling and prothrombotic factors, and markers of endothelial dysfunction were studied in patients with complicated and uncomplicated diabetes mellitus. It was found that the neural network analysis of these data makes it possible to predict the development of diabetic foot syndrome with an accuracy of 92.9%. The high accuracy of neural network analysis is due to the multifactorial nature of the diabetic foot syndrome.