Deep Hybrid Models for Accurate Skin Disease Diagnosis Using Medical Imaging

Abstract

The successful management of dermatological conditions relies on prompt identification; y et, d ermatologists are not readily accessible. This paper proposes a conversational AI chatbot that categorizes images using convolutional neural networks (CNNs) and diagnoses symptoms using natural language processing (NLP). Users may input symptoms or upload images to the algorithm, which delivers initial diagnoses and recommendations. Photographs of skin diseases are precisely categorized with a fine-tuned <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$R$</tex> esNet-18 m odel, whereas verbal descriptions are processed by an NLP-driven decision tree model. The chatbot, utilizing Google Dialogflow e nable real-time engagement, is implemented through a Flask-based API. Experimental data validate the system's efficacy as an assistance in dermatological examinations, demonstrating a classification accuracy of 94.1

Ähnliche Arbeiten