EpiDetect

EpiDetect is a web application developed using the MERN stack, designed to predict skin diseases from images captured using your web camera. It uses a fine-tuned ResNet50 model for accurate skin disease detection.

Features

• User Profile Management: Users can view and update their profile information.
• Skin Disease Prediction: Upload or capture images using the web camera to predict skin diseases.
• Dashboard: Access different functionalities from a central dashboard.
• Blog Creation: Users can create and manage their blog posts.
• Contact Form: Users can send messages through the contact form.
• Prediction Records: View and download prediction records in PDF format.

Screenshots

User Profile

Predict Skin Disease

Contact Us

Dashboard

Prediction Records

Create Blog Post

Login

Technology Stack

• Frontend: React.js, HTML, CSS
• Backend: Node.js, Express.js
• Database: MongoDB
• Machine Learning: Python, TensorFlow, Keras (ResNet50 model)
• Other Tools: JWT for authentication, multer for file uploads

Setup Instructions

Prerequisites

• Node.js
• MongoDB
• Python
• TensorFlow and Keras

Installation

1. Clone the repository:

   git clone https://github.com/fatimaazfar/EpiDetect.git

2. Navigate to the project directory:

   cd EpiDetect

3. Install backend dependencies:

   cd server
   npm install

4. Install frontend dependencies:

   cd ../client
   npm install

5. Set up environment variables for server:

   • Create a .env file in the server directory.
   • Add the following environment variables:

     MONGO_URI=your_mongodb_connection_string
     JWT_SECRET=your_jwt_secret

6. Run the backend server:

   cd ../server
   node server.js

7. Run the frontend server:

   cd ../client
   npm start

8. Access the application at http://localhost:3000.

Usage

1. User Registration and Login: Register a new account or log in with existing credentials.
2. Profile Management: View and update your profile information.
3. Skin Disease Prediction:
   • Navigate to the Predict page.
   • Upload an image or capture one using your web camera.
   • Click on Predict to get the prediction results.
4. Blog Creation: Navigate to the Create Blog page to create and manage your blog posts.
5. Contact Us: Use the contact form to send messages.
6. Prediction Records: View and download your prediction records in PDF format.

Future Work

I will soon be incorporating a chatbot to recommend natural solutions and treatments for light-scale skin diseases, causes of the disease, and anything relevant to the disease that the patient could find useful.

Contributing

Contributions are welcome! Please fork the repository and submit a pull request for any improvements or bug fixes.

License

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

Contact

For any inquiries, please contact Fatima Azfar at [email protected].