Skip to content

README.md

Latest commit

 

History

History
103 lines (84 loc) · 3.26 KB

README.md

File metadata and controls

103 lines (84 loc) · 3.26 KB
Raw

MediGuide

A Flask-based web application that helps users diagnose diseases and receive recommendations based on their symptoms. MediGuide leverages trained machine learning models and pre-loaded datasets to predict possible diseases and provide relevant guidance.

Features

  • Symptom-based disease prediction using an SVC machine learning model.
  • Detailed descriptions of predicted diseases.
  • Precautions, medications, diets, and workouts tailored to specific diseases.
  • User-friendly web interface with routes for prediction, contact, developer information, blog, and more.

Setup and Installation

Prerequisites

  • Python 3.x
  • Flask
  • NumPy
  • Pandas
  • Pickle
  • Pre-trained SVC model (svc.pkl) and datasets placed in the local models and datasets folders respectively.

Installation Steps

  1. Clone the repository:

    git clone https://github.com/Shreyash021104/MediGuide.git
cd MediGuide

  2. Install the required dependencies:

    pip install -r requirements.txt

    Note: Create a requirements.txt file if it does not exist:

    Flask
numpy
pandas

  3. Ensure that the following files exist in their respective directories:

    • models/svc.pkl - Pretrained Support Vector Classifier model.
    • datasets/symtoms_df.csv - Symptom dataset.
    • datasets/precautions_df.csv - Precautions dataset.
    • datasets/workout_df.csv - Workout recommendations dataset.
    • datasets/description.csv - Disease descriptions dataset.
    • datasets/medications.csv - Medications dataset.
    • datasets/diets.csv - Dietary recommendations dataset.

Running the Application

  1. Start the Flask development server:

    python main.py

  2. Open your web browser and navigate to:

    http://127.0.0.1:5000

How to Use the Application

  1. Enter the symptoms in the input field separated by commas (e.g., headache, fever, cough).
  2. Click "Predict" to receive the predicted disease and recommendations.
  3. Navigate through other pages, such as "About", "Contact", "Developer", or "Blog", using the navigation bar.

Project Structure

MediGuide/
├── datasets/            # CSV files containing symptoms, precautions, workouts, medications, etc.
├── models/              # Pre-trained model (svc.pkl).
├── templates/           # HTML templates for the web UI.
├── main.py              # Application entry point.
└── static/              # Static assets.

Core Functions and Logic

helper(dis)

  • Retrieves detailed information about a specific disease, including:
    • Description
    • Precautions
    • Medications
    • Dietary recommendations
    • Suggested workouts

get_predicted_value(patient_symptoms)

  • Processes user-entered symptoms into a vector format.
  • Predicts the disease using the pre-trained SVC model.

Routes

  • /: Main page for symptom input and disease prediction.
  • /about: Information about the application.
  • /contact: Contact page.
  • /developer: Information about the developer.
  • /blog: Blog section.

Author

  • Shreyash021104

Contributing

Feel free to fork the repository and submit pull requests to suggest improvements or add new features.

License

This project is licensed under the MIT License.