2026 International Conference on Networks Computers and Communications (ICNCCom 2026)

With the limited availability of dermatologists in India and the increasing dependence on online self-diagnosis, there is a growing need for accessible and trustworthy AI-assisted healthcare solutions. Skin diseases are among the most common health conditions affecting people worldwide, yet early identification and timely consultation remain challenging, particularly in regions with limited medical accessibility. Many individuals rely on unverified internet sources for self-diagnosis, which can lead to misinformation, unnecessary anxiety, and inappropriate self-medication. To address these issues, this project presents an AI-powered mobile application for preliminary skin disease prediction using a custom-built Convolutional Neural Network (CNN) integrated with Explainable Artificial Intelligence (XAI) and an intelligent chatbot system. The proposed system is designed to analyze skin lesion images uploaded by users and classify them into multiple dermatological categories using a CNN model trained on the HAM10000 dataset. To improve transparency and user trust, the system incorporates Local Interpretable Model-Agnostic Explanations (LIME), which visually highlights the regions of the image that influenced the prediction made by the model. Unlike conventional black-box AI systems, the explainability component helps users better understand how the prediction was generated. In addition to disease prediction, the application integrates a GROQ API-based intelligent chatbot that provides general awareness, preventive precautions, and basic guidance related to the predicted skin condition. The chatbot is intentionally restricted from providing medications, prescriptions, or treatment recommendations, thereby encouraging users to seek professional medical consultation instead of relying on automated healthcare advice. The application is implemented as a mobile-based platform with a React Native frontend and a Flask backend for AI model inference and chatbot communication. Experimental results demonstrate that the proposed system can provide fast and accessible preliminary assessments while improving transparency and user interaction through explainable AI and conversational assistance. The project aims to serve as a supportive healthcare tool that promotes awareness, accessibility, and responsible use of AI in dermatological screening.

Skin Disease Detection, Deep Learning, Explainable AI, LIME, HAM10000, Mobile Health Systems, CNN, Healthcare Chatbot, Flask, React Native