Klasifikasi Jenis Eksim Menggunakan Metode You Only Look Once

Abstract

Dermatitis, also known as eczema, is an inflammatory skin condition that often appears on dry skin areas and is caused by various factors. Eczema causes itching, redness, and discomfort that can significantly affect the patient’s quality of life. Many people tend to underestimate this condition, mistaking it for ordinary itching, which often leads to delays in treatment. Without proper diagnosis and care, eczema can worsen and lead to complications such as secondary skin infections due to excessive scratching. One solution to support eczema diagnosis is the application of artificial intelligence-based technology. The You Only Look Once (YOLO) algorithm, based on Convolutional Neural Network (CNN) architecture, offers high speed and accuracy in detecting and classifying various forms of eczema from skin images. The dataset used in this study consists of 538 images in total, with the types of eczema divided into three main classes: contact eczema, atopic eczema, and nummular eczema. The use of the YOLOv8 algorithm for skin eczema detection produced excellent results, achieving an accuracy of 87.5% on training and validation data.