Convolutional Neural Network Model for Diagnosis of Low-Density Type Cholesterol

Excess Cholesterol is one of the most prevalent diseases of the human body system. Determining the right type of Cholesterol in the body is a significant challenge. This is because the human body contains two types of cholesterol: high-density lipoprotein (HDL) cholesterol and low-density lipoprotein (LDL) cholesterol. HDL cholesterol is beneficial to the body, promoting good health, while LDL cholesterol is detrimental to human health, as it builds up plaques in the walls of the arteries, preventing the free flow of blood in the coronary arteries and consequently causing heart failure. In this paper, a computer-aided Convolutional Neural Network (CNN) model for analyzing, diagnosing, and classifying medical image data of patients as high-density and low-density lipoproteins is presented. A case study of the model using the Kaggle Eye State classification datasets on cholesterol was carried out. Results from the study established an average performance of 96.6% and superiority over the decision tree and binary logic pattern models by 3.78% and 2.85% respectively. The validity of this result was tested with live data obtained from the Medical Centre, Bells University Clinic, Ota, Nigeria. These performance indices proved that the CNN model could be very useful for the diagnosis of cholesterol in any healthcare centre.