AI-Powered Dermatological Diagnosis and Public Health Management System Using EfficientNetB3 and Conversational Agents

This study presents an AI-powered dermatological diagnosis and public health management system that combines the EfficientNetB3 deep learning model with conversational agents. The system analyzes skin lesion images to assist in early detection of skin diseases, particularly melanoma. The proposed model achieves strong performance with an overall accuracy of 0.865, precision of 0.858, recall of 0.862, and an F1-score of 0.860, outperforming baseline approaches. Additionally, the conversational agent provides user guidance, symptom-based interaction, and health awareness support. The framework demonstrates the potential of artificial intelligence to improve dermatological diagnosis and enhance healthcare accessibility.