MediAlert

Overview

This project integrates two innovative healthcare solutions into one platform:

1. Adverse Medical Event Prediction from Phone Calls: Leverages AI and NLP to analyze patient-doctor conversations, detect early signs of potential adverse events, and generate actionable risk reports for healthcare providers.
2. Doctor Appointment Booking System: Provides a seamless, user-friendly mobile app for patients to book appointments, view doctor availability, and receive confirmation and reminder notifications.

By combining these functionalities, our platform improves patient safety through early detection and streamlines access to quality healthcare.

Features

• AI-Powered Analysis:
  • Converts audio conversations to text using Deepgram.
  • Analyzes text with AWS Comprehend Medical to detect potential adverse events.
  • Generates detailed risk reports for healthcare providers.
• Doctor Dashboard:
  • Web-based interface built with React.js and TailwindCSS.
  • Secure login and data management.
  • Visualization of risk analysis and downloadable reports.
• Appointment Booking System:
  • Flutter-based mobile app for easy booking.
  • Real-time scheduling with conflict resolution.
  • Geo-location services to locate nearby doctors.
  • Automated email confirmations and reminder notifications.

System Architecture

Doctor Dashboard

• Frontend: Built with React.js and TailwindCSS, deployed on Vercel for a responsive web interface.
• Backend: A Node.js and Express server handles API requests, user authentication and integration with AI services.
• AI Processing:
  • Deepgram API converts patient call audio to text.
  • AWS Comprehend Medical analyzes the text to extract medical insights and detect adverse events.
  • An AI/ML layer processes and cross-references data (e.g., FDA guidelines) to generate detailed risk reports.
• Database: MongoDB stores user profiles, doctor details, and risk reports.

Appointment Booking Module

• Mobile App: Developed in Flutter for cross-platform compatibility, allowing patients to search for doctors, book appointments, and view schedules.
• Backend Integration: The same Node.js server handles appointment scheduling, preventing double bookings and managing real-time data.
• Notifications: Real-time notifications are sent via Firebase Cloud Messaging, and email confirmations/reminders are managed by Nodemailer.
• Geo-Location: Utilizes geo-location APIs to assist patients in finding nearby doctors.

Tech Stack

• Frontend (Web): React.js, TailwindCSS
• Mobile App: Flutter
• Backend: Node.js, Express.js
• Database: MongoDB
• AI/NLP: AWS Comprehend Medical, Deepgram
• Authentication: Firebase Authentication
• Notifications: Firebase Cloud Messaging, Nodemailer
• Deployment: Vercel (Web), Render (Backend)

Hosted Links

• Doctor Dashboard: Vercel Deployment

• Patient Mobile App Figma Design: Figma Design

• Video Presentation: Youtube Video

• PPT Presentation: PPT

Test Cases

• Adverse Event Detection Test Cases: Adverse Event Detection Test Document.md
• Doctor Appointment Test Cases: Doctor-Appointment Test Documentation.md

Deployment

• Frontend: Deployed on Vercel.
• Backend: Deployed on Render.
• Mobile App: Built APKs/iOS builds using Flutter’s build tools.

License

This project is licensed under the MIT License. See the LICENSE file for details.