Detecting Urinary Tract Infection (UTI) thru Analytes level using Convolutional Neural Network and Support Vector Machine

Abstract

Abstract. Early detection of Urinary Tract Infection (UTI) is vital since if the infection reaches the kidneys, a more severe illness may occur. Quite often, people with urinary tract infection aren't aware of their problem because there is a type of this infection that only manifests its symptoms when its already severe. This served as inspiration to the researchers to make this study. The study aims to give diagnosis to anyone who is interested to know their condition. It is a desktop application that can provide an analysis to the user's urine test strip using mobile phone's camera. The researchers used experimental method in this study. The system was developed using Python Programming Language utilizing algorithms such as Convolutional Neural Network (CNN) image processing and Support Vector Machine (SVM) for classification to attain the system's accuracy and reliability in detecting UTI. Researchers conducted an experiment to 65 participants. The result of the study has come up with the overall accuracy rate of 96.03% and overall reliability rate of ≥ 0.9 which interprets to excellent.