HackOrbit-Final-CODING-ERA-

VitalCheck: AI-Powered Healthcare Platform

1. Project Overview

VitalCheck is an advanced healthcare platform that empowers doctors to manage patients, register new cases, and leverage cutting-edge AI for disease prediction and report analysis.

Key Features

• Doctor dashboard with secure patient management
• AI-powered predictions for brain tumors, lung cancer, skin diseases, chest tuberculosis, and heart disease from blood reports
• Appointment scheduling and calendar
• Patient image upload and management
• JWT-based authentication and role-based access

---

2. AI Models & Usage

Supported AI Predictions

• Brain Tumor Detection (MRI)
• Lung Cancer Detection (CT)
• Chest Tuberculosis Detection (X-ray)
• Skin Disease Classification (Image)
• Heart Disease Prediction (Blood Report PDF)

Model Training & Data

• Brain Tumor: Trained on MRI images, 4 classes (glioma, meningioma, notumor, pituitary). See AI/train_brain_tumor_MRI.py.
• Lung Cancer: Trained on CT scans, 3 classes (benign, malignant, normal). See AI/train_lung_cancer_CT.py.
• Chest Tuberculosis: Trained on X-ray images, 2 classes (normal, tuberculosis). Dataset details in AI/dataset/TB_Chest_Radiography_Database/README.md.txt.
• Skin Disease: Trained on a large, multi-class dataset using transfer learning (MobileNetV2). See AI/train_skin_disease.py.
• Heart Disease: Trained on structured blood report data, predicts risk from PDF extraction. See AI/BloodReport_HeartDiseasedetectionModel.py.

How AI is Used

• The AI server (AI/Server.py) exposes REST endpoints:
  • /predict for image-based predictions (brain, lung, chest, skin)
  • /predict-blood for PDF blood report analysis
• Models are loaded at startup; predictions are made on uploaded files and return class and confidence.
• Example usage:
  • Upload an MRI image to /predict with scanType=MRI and bodyPart=brain for brain tumor prediction.
  • Upload a blood report PDF to /predict-blood for heart disease risk.

---

3. Backend Architecture

• Tech Stack: Node.js, Express, MongoDB, Mongoose, JWT, Multer, Nodemailer
• Key Endpoints:
  • /login, /register, /forgot-password, /reset-password for authentication
  • /register-patient (with image upload), /patients (CRUD) for patient management
  • Appointment endpoints for scheduling and viewing
• Security:
  • JWT authentication for all sensitive routes
  • Doctor-patient association: Each patient is linked to the registering doctor; only their patients are visible
  • Auto-generated patient credentials (email/password) shown to the doctor after registration
• Image Handling:
  • Patient images are uploaded and stored; default image fallback is provided

---

4. Frontend Architecture

• Tech Stack: React, Vite, MUI, Axios, React Router, React Calendar
• Features:
  • Doctor dashboard: View, register, update, and delete patients
  • Patient registration form (no email required; credentials auto-generated)
  • Modal to copy patient credentials after registration
  • Appointment calendar for managing patient appointments
  • AI prediction UI: Upload images or PDFs, view results with confidence scores
  • Patient image preview and fallback to default
• Authentication:
  • JWT token stored and sent with all API requests
  • Role-based UI: Only show patients and appointments for the logged-in doctor

---

5. Setup & Installation

Prerequisites

• Node.js
• Python 3.8+
• MongoDB

AI Server

cd AI
pip install -r requirements.txt
python Server.py

Backend

cd Backend
npm install
node server.js

Frontend

cd frontend
npm install
npm run dev

---

6. Usage Guide

• Register/Login as Doctor: Access the dashboard
• Register Patient: Fill the form, upload image (optional), copy credentials
• AI Predictions: Use the prediction UI to upload scans or reports
• Appointments: Schedule and view appointments for your patients
• Patient Management: View, update, or delete your patients only

---

7. Data & Model Sources

• Chest Tuberculosis Dataset: Details & Citations
• Other Models: Trained on public datasets (see training scripts for details)

---

8. Security & Best Practices

• All sensitive actions require JWT authentication
• Patient data is only accessible to the registering doctor
• Passwords are securely hashed and never exposed except at registration
• Image uploads are validated and size-limited

---

9. Project Structure

AI/           # AI models, training scripts, and Flask server
Backend/      # Node.js/Express backend, MongoDB models, API endpoints
frontend/     # React frontend, UI components, pages, assets

---

10. Contributing & License

• PRs and issues welcome!
• Cite datasets and models as required by their licenses

---

Ready for production?

• Secure environment variables for secrets and DB credentials
• Use HTTPS in deployment
• Set up proper CORS and rate limiting