Abstract

Iridology is an analytical technique based on the body’s disease and weakness in the shape and structure of the iris of the eye (located around the pupil). Iridology analysis is usually done manually by iridology practitioner or by someone who is experienced in iridology because iridology can be learned. The purpose of this study is to detect high cholesterol or normal cholesterol using artificial neural network training and data input using Gray Level Co-Occurrence Matrix (GLCM) texture comparison method. Image input with 300×300 pixel incorporated into the program to do preprocessing stages such as grayscale, noise remover, image contrast, image exposure polar, and cropping the image. From the results of preprocessing, the average value of statistical data is calculated using GLCM methods with the distance is two pixels. Based on the results of testing the training data, the percentage of program accuracy is 97.5%. Based on the results of testing other image training, the accuracy percentage is 95%. The accuracy of testing image based on a medical examination is 81.81%.

Read Full Abstract: document01