Skip to content

Repository files navigation

MediAI – AI-Powered Healthcare Risk Prediction Assistant

Overview

MediAI is a healthcare analytics and machine learning project designed to predict heart disease risk using patient health data. The project leverages a Random Forest Classification model to analyze medical attributes and classify patients into risk categories while providing confidence scores for each prediction.

The solution integrates data analytics, predictive modeling, and business intelligence visualization through an interactive Power BI dashboard, enabling healthcare professionals and analysts to better understand patient risk patterns and support data-driven decision-making.


Project Objectives

  • Predict heart disease risk using machine learning techniques.
  • Identify high-risk and low-risk patient groups.
  • Analyze healthcare trends across demographics and medical indicators.
  • Evaluate model confidence and prediction reliability.
  • Develop an interactive dashboard for healthcare analytics and reporting.

Technologies Used

Programming and Analytics

  • Python
  • Pandas
  • NumPy

Machine Learning

  • Scikit-Learn
  • Random Forest Classifier

Data Visualization

  • Power BI
  • DAX

Data Storage

  • CSV
  • Pickle (.pkl)

Dataset Description

The dataset contains patient medical information commonly used for heart disease assessment, including:

  • Age
  • Gender
  • Chest Pain Type
  • Resting Blood Pressure
  • Cholesterol Level
  • Fasting Blood Sugar
  • Resting ECG Results
  • Maximum Heart Rate
  • Exercise-Induced Angina
  • ST Depression
  • Slope
  • Number of Major Vessels
  • Thalassemia
  • Heart Disease Target Variable

Machine Learning Workflow

Data Preparation

  • Data Cleaning
  • Missing Value Verification
  • Feature Selection
  • Data Transformation

Model Development

  • Train-Test Split
  • Random Forest Classification
  • Model Evaluation

Model Evaluation

  • Accuracy Score
  • Confusion Matrix
  • Classification Report
  • Confidence Score Analysis

Dashboard Features

Key Performance Indicators

  • Total Patients
  • High-Risk Patients
  • Low-Risk Patients
  • Average Confidence Score

Analytical Visualizations

  • Risk Distribution Analysis
  • Risk by Gender
  • Risk by Age Group
  • Confidence by Chest Pain Type
  • Average Confidence by Age Group
  • Patient Risk Analysis Table

Interactive Filters

  • Gender
  • Age Group
  • Risk Level
  • Chest Pain Type

Dashboard Preview

MediAI – AI Powered Healthcare Risk Prediction Assistant

Key Insights

  • Identified high-risk and low-risk patient populations using predictive analytics.
  • Analyzed disease risk patterns across age groups and gender categories.
  • Evaluated prediction confidence levels for healthcare decision support.
  • Provided interactive healthcare intelligence through visual analytics.

Business Value

  • Supports early identification of potential heart disease cases.
  • Enables data-driven healthcare decision-making.
  • Assists healthcare professionals in patient risk assessment.
  • Enhances healthcare analytics and reporting capabilities.

Project Structure

MediAI-AI-Powered-Healthcare-Risk-Prediction-Assistant/
│
├── MediAI_Analytics.ipynb
├── medi_ai_final.csv
├── medi_ai_model.pkl
├── medi_ai_dashboard.pbix
├── dashboard.png
├── README.md
└── requirements.txt

Future Enhancements

  • Real-time patient risk prediction.
  • Deployment using FastAPI or Streamlit.
  • Explainable AI (XAI) integration.
  • AI-powered healthcare recommendation engine.
  • Healthcare chatbot integration.

Author

Lubna Shireen R

Data Analytics | Machine Learning | Business Intelligence | Power BI


License

This project is licensed under the MIT License.

About

MediAI is an AI-powered healthcare risk prediction assistant that uses Machine Learning and Power BI to predict heart disease risk, analyze patient health patterns, provide confidence score insights, and visualize healthcare analytics through an interactive dashboard.

Resources

Stars

0 stars

Watchers

0 watching

Forks

Releases

Packages

Contributors

Languages