An advanced machine learning system designed to predict early signs of diseases which is based on patient data and symptoms. This tool aims to assist healthcare professionals in early diagnosis and intervention.
https://drive.google.com/file/d/1SfR7Ju-S3XmiD2aQVLdwAcb1JbX6wGEC/view?usp=sharing
Features
Predictive analysis for multiple diseases User-friendly web interface for inputting patient data High accuracy prediction model Detailed results with confidence scores Responsive design for desktop and mobile devices
Installation
Prerequisites
Python 3.7 or higher pip (Python package installer) Git
Setup Instructions
Clone the repository: Copygit clone https://github.com/Devdath-code/Early-Disease-Prediction-Model.git cd Early-Disease-Prediction-Model
Install required dependencies: Copypip install -r requirements
Download the model file:
The model file (model.pkl) is not included in the repository due to its large size. Download it from https://drive.google.com/file/d/1tDdLTys9V6rWPF94wz5OPDssY7VkIpYQ/view?usp=sharing Place the downloaded file in the root directory of the project.
Usage
Start the application: Copypython app.py
Open your web browser and navigate to: Copyhttp://localhost:5000
Enter patient data in the provided form and submit to get prediction results.
Model Information
This model uses a Random Forest classifier trained on a comprehensive healthcare dataset containing patient records with various symptoms and diagnostic outcomes. The model processes key features including:
- Patient demographic information (age, gender)
- Vital signs (blood pressure, heart rate, temperature)
- Laboratory test results (blood count, cholesterol levels, glucose levels)
- Reported symptoms and their duration
- Family history indicators
- Lifestyle factors (smoking, alcohol consumption, physical activity)
- Accuracy: 87.5%
- Precision: 85.3%
- Recall: 83.9%
- F1 Score: 84.6%
The model was validated using 5-fold cross-validation on a holdout test set comprising 20% of the original dataset.
Limitations
This model is designed as a supportive tool for healthcare professionals and should not replace professional medical diagnosis. Results should be interpreted by qualified medical practitioners.
Project Structure
disease_prediction_model/
├──app.py # Flask web application
├──model.pkl # Trained model
├──requirements # Required Python packages
├──templates/ # HTML templates for web interface
└──[other directories] # Additional project files
Future Enhancements
Integration with electronic health records Mobile application development Additional disease prediction capabilities Explainable AI features for better interpretation