FURI | Spring 2025

Leveraging Data Science Techniques to Classify Eczema in Medical Imaging

Eczema is a skin condition affecting millions of people, and its diagnosis heavily depends on dermatologists’ clinical expertise. This study explores the effectiveness of deep learning methods, specifically Convolutional Neural Networks (CNNs), in diagnosing eczema from medical images, compared to traditional machine learning methods like Support Vector Machines (SVMs). This study optimizes a CNN-based model to enhance classification accuracy. Performance is assessed using accuracy, precision, recall, and F1 score. Preliminary findings show that deep learning outperforms traditional approaches in accuracy and feature extraction. This research advances AI-driven medical diagnostics, with potential applications in telemedicine and improved healthcare accessibility.