Hybrid radial basis function with firefly algorithm and simulated annealing for detection of high cholesterol through iris images

Abstract

Cholesterol is a lipid (fat) produced by the liver and is required to build and maintain cell membranes. Cholesterol is also important for the metabolism of fat soluble vitamins. This important lipid is found in human blood. Excess cholesterol (high cholesterol) can cause health problems such as being a factor of coronary heart disease that responsible for the heart attacks, liver or kidney disease. Observation of iris pattern can detect several types of diseases, one of which is high cholesterol. The purpose of this research is to detect whether someone is exposed to high cholesterol or not, through iris images based on firefly algorithm, simulated annealing, and radial basis function. Firefly algorithm and simulated annealing are used in the unsupervised learning process in radial basis function neural networks. The stages of high cholesterol detection process are images processing namely grayscale process, thresholding, histogram equalization, segmentation, and detection process is using radial basis function neural network. The percentage success rate of the recognition pattern of iris images for detecting high cholesterol is 89%.