VitalAI Healthcare Platform - Frontend

Welcome to the frontend repository of VitalAI, a cutting-edge healthcare platform that aims to provide fast and reliable medical care services to patients through the power of Artificial Intelligence.

Table of Contents

• Introduction
• Features
• Tech Stack
• Google Collaboratory Notebooks
• Screenshots Demo
• Contributing

Introduction

VitalAI is a revolutionary healthcare system that connects patients with licensed medical professionals through a user-friendly web and mobile application. Our platform incorporates AI health assistants and advanced analytics insights to deliver personalized and efficient medical solutions to patients in need.

Features

• Real-time GPS Locator: Instantly locate patients during emergencies or in need of quick medical help.
• AI Health Assistant: Advanced AI-powered tools provide accurate diagnosis and personalized healthcare recommendations.
• Fast Medical Care Access: Connect patients to licensed medical professionals for quick and reliable care.
• User-friendly Interface: Intuitive design for easy navigation and seamless user experience.
• Secure Authentication: Protects sensitive user information with encrypted authentication.

Tech Stack

Our frontend is built using modern technologies and libraries to ensure a smooth and responsive user interface:

• React: A popular JavaScript library for building user interfaces.
• TypeScript: Adds static typing to JavaScript for enhanced developer productivity.
• React Router: Handles navigation and routing within the application.
• Dayjs: A fast and lightweight library for parsing, validating, manipulating, and formatting dates.
• React Query: Enables efficient data fetching and synchronization for improved performance.
• Tailwind CSS: A utility-first CSS framework for building custom and responsive designs.
• Material-UI: Provides a rich set of UI components for a modern and polished look.

Google Collaboratory Notebooks

• Bot question & answering
  • Data preparation Notebook : Dataset used was the MedQuAD: Medical Question Answering Dataset which covers 37 question types (e.g. Treatment, Diagnosis, Side Effects) associated with diseases, drugs and other medical entities such as tests. Read article here
  • Indexing vector database for bot similarity search and data retrieval Notebook: Pinecone free vector database for ML was used.
  • Creating main pipeline Notebook : The flax-sentence-embeddings/all_datasets_v3_mpnet-base model was used as the embedding/retrieval model and vblagoje/bart_lfqa served as the Generative model.
• Disease Prediction Model
  • Model training pipeline Notebook
• Drug Recommendation Model
  • Model training pipeline Notebook

Screenshots Demo

This section displays some of the screens of the project:

• Home page - patients view

  

• View health practitioners details - patients view

  

• List of available health professionals by Organization - patients view

  

• User interaction with medical bot

  

• Disease prediction model

  

• Disease prediction model results(patients view)

  

• Home page - health practitioners view

  

• Drug recommendation (health practitioners view)

  

• Drug recommendation model results(health practitioners view)

  

• Realtime health practitioner and patients chat(health practitioners view)

  

Contributing

I welcome contributions to improve VitalAI! Feel free to open issues, submit pull requests, or provide feedback to help us enhance our platform.

For more information about the backend of VitalAI, please visit the Backend repository or the AI Flask Sever .If you have any questions or need support, contact me on [email protected]. Thank you for being a part of the VitalAI community!