Mediscan-AI

Project Overview

Mediscan-AI is an award-winning, clinical-grade AI health platform designed to provide accessible medical triage and real-time health monitoring. Developed during a 24-hour hackathon, the system bridges the gap between patient symptoms and preliminary medical insights.

Advanced AI Triage & Sensing

• Intelligent Symptom Analysis: Integrates OpenAI/Claude APIs to process natural language symptoms and provide detailed clinical explanations and risk assessments.
• Non-Contact Heart Rate Detection: Utilizes rPPG (Remote Photoplethysmography) via OpenCV and Python to detect heart rate fluctuations through a standard camera feed.
• Bilingual Voice Interface: Supports full functionality in both English and Tamil, featuring voice input to ensure accessibility for diverse user groups.
• Automated Medical Reporting: Instantly generates downloadable PDF reports containing prediction history and vital signs for physician review.

Technical Architecture

Frontend: Modern, responsive Dark Mode UI built with JavaScript (ES6+), CSS3, and HTML5.
Backend: Scalable Node.js and Express API architecture for high-speed data processing.
AI & Signal Processing: Python-based OpenCV scripts for real-time video signal processing and OpenAI for LLM-driven diagnostics.
Database: Hybrid storage using MongoDB for persistent history and LocalStorage for session states.