readme.md

metadata

title: MediScan API
emoji: 🏥
colorFrom: blue
colorTo: green
sdk: docker
pinned: false

MediScan AI Backend API

AI-powered skin disease classification API using EfficientNetV2-S with AES-256 encryption.

Features

• 🧠 AI-powered skin disease classification
• 🔍 Grad-CAM visualization for interpretability
• 🔒 AES-256-CBC encryption for medical images
• 👤 User authentication and history tracking
• 📊 Real-time statistics and analytics

Model

Uses NishantFOT/MediScanB from Hugging Face Hub.

Frontend

Deployed on Netlify: https://mediscan-codex.netlify.app

API Endpoints

• POST /api/register - User registration
• POST /api/login - User authentication
• POST /api/analyze - Image analysis with AI
• GET /api/history - Get user's analysis history
• GET /api/stats - Get user statistics
• GET /api/health - Health check

Technology Stack

• Flask (Backend)
• PyTorch + EfficientNetV2-S (AI Model)
• SQLAlchemy (Database)
• Cryptography (AES-256 Encryption)
• Hugging Face Hub (Model Hosting)