An unhealthy lifestyle can impact body health, causing, for example, metabolic syndrome and hypercholesterolemia. Metabolic syndrome, including high blood sugar, causes diabetes mellitus (DM). Abnormal cholesterol levels cause hypercholesterolemia (HC), which contributes to various forms of cardiovascular disease (CVD). According to the WHO, DM and CVD are the main causes of death in the top ranking. Early detection of DM and HC is necessary to reduce mortality from these two diseases. This paper aims to provide an alternative for early detection of DM and HC, with a non-invasive method based on the iris image. The main contribution of this paper is to use the 3D-GLCM algorithm to detect two diseases simultaneously. The iris image processing is carried out in several stages, namely: 1) Pre-processing is done by converting the RGB image to grayscale and image improvement using the AHE method; 2) Feature extraction is carried out using the 3D-GLCM method with six statistical characteristics; 3) Classification is carried out by training and testing on Dataset I and Dataset II. The method effectiveness is compared to the number of gray levels, namely 16, 32, 64, 128, and 256. Based on the five levels of gray, the best value is shown from the number of gray levels of 256, with the value obtained. These values are sensitivity value (0.9375), specificity value (0.0208), and accuracy (0.9844). The results showed that the higher the gray level of the image database used, the higher the sensitivity and accuracy values, while the lower the gray level indicates the specificity value.

Authors:

Rinci Kembang Hapsari, Miswanto Miswanto, Riries Rulaningtyas, Herry Suprajitno

Faculty of Sciences and Technology, Airlangga University, Surabaya, Indonesia