Skip to content

Latest commit

 

History

History
121 lines (90 loc) · 4.27 KB

README.md

File metadata and controls

121 lines (90 loc) · 4.27 KB

Preventive Healthcare System

This project is an AI-powered Predictive Healthcare System that helps identify individuals at risk of developing chronic diseases such as diabetes, heart disease, or obesity. The system provides personalized recommendations for preventive care based on lifestyle and medical data.

Features

  • AI-based health risk prediction for diabetes, heart disease, and obesity.
  • User-friendly form for inputting health and lifestyle data.
  • Displays risk probabilities and provides feedback on how to manage risks.
  • Uses pre-trained machine learning models to assess the risk levels.
  • Integrated with Streamlit for interactive web-based applications.

Prerequisites

Ensure you have the following installed:

  • Python 3.7+
  • streamlit
  • pandas
  • scikit-learn
  • requests
  • Pillow (for image processing)

Install all the necessary dependencies using the following command:

pip install -r requirements.txt

Installation & Setup

  1. Clone the repository:
git clone https://github.com/aaarif796/AI-Powered-Preventive-Healthcare-System.git
cd AI-Powered-Preventive-Healthcare-System
  1. Download or Place Model Files:

Make sure to have the pre-trained model files:

  • label_encoders.pkl
  • lr_dt.pkl (Logistic Regression model for Diabetes)
  • lr_ht.pkl (Logistic Regression model for Heart Disease)
  • lr_ob.pkl (Logistic Regression model for Obesity)

Place these files inside the model folder.

  1. Add Images:

Place relevant images in the images folder for visual representation.

  1. CSS Styling:

The application uses a custom CSS file for styling. Ensure you have the style.css file in the style folder.

Usage

  1. Run the Application:

Use Streamlit to launch the app with the following command:

streamlit run app.py
  1. Input Data:

Fill out the form with your general health and lifestyle details (e.g., age, exercise habits, smoking history, etc.).

  1. Receive Feedback:

The app will predict your risk level for diabetes, heart disease, and obesity based on the data you provide. It will also offer personalized advice based on the risk level.

Folder Structure

├── images
│   ├── healthcare.webp
├── model
│   ├── label_encoders.pkl
│   ├── lr_dt.pkl
│   ├── lr_ht.pkl
│   ├── lr_ob.pkl
├── style
│   ├── style.css
├── app.py
├── README.md
├── requirements.txt

Model Details

  • label_encoders.pkl: Used to encode categorical data.
  • lr_dt.pkl: Logistic Regression model for predicting the risk of diabetes.
  • lr_ht.pkl: Logistic Regression model for predicting heart disease risk.
  • lr_ob.pkl: Logistic Regression model for obesity risk.

Acknowledgments

This application was developed as part of the TechXcelerate 2024 challenge, focusing on developing a predictive healthcare system using machine learning and AI.

will be adding two models, XGBoost and Naive Bayes, to the analysis of the diabetes dataset( diabetes data.csv) or enhanced predictive modeling.

Sprint 2 focusing on developing a predictive healthcare system using machine learning and AI. Sprint 2 will be adding a model to analyse the heart disease dataset. More input to come, after I explore the dataset. sprint2 Run the heart_complete.csv dataset on Jupiter notebook changed the numerical values on the line plot and bar plot: layout (3,3) to (4,4), because I was getting errors Addded the Xgboost and Naive bayes model to boost the accuracy. The Xgboost gave me an error on the jupiter notebook, but when I run it on the colab google notebook it ran fine. create an app for predicting heart disease Uploaded the new heart_complete.ipynb file, with the Xgboost and Naive Bayes added. I will work on it even more to fix the error and upload the latest version of the file. Transfer updated Jupyter Notebook to GitHub Prepare for the Presentation Created an app for Predicting heart disease Uploaded the app on Github Built the app using Flask. But I couldn't get it to produce/output a result. Then ran it with streamlit code and it worked. Summary of the Project: findings Uploaded the Sprint3 Power Point Slide Screenshot of the app with Flask Screenshot of the app with Streamlit