Predict2Protect: Machine Learning Web Application in Early Detection of Heart Disease

Area of Science:

• Cardiovascular Health
• Machine Learning in Healthcare
• Public Health Technology

Background:

• Heart disease is a leading global cause of mortality.
• Untimely diagnosis due to healthcare access and cost barriers contributes to poor outcomes.
• There is a need for accessible tools for early heart disease risk identification.

Purpose of the Study:

• To develop an accessible application for accurate heart disease risk prediction.
• To provide an easy-to-use interface for patients to assess their risk.
• To facilitate early detection and intervention for heart disease.

Main Methods:

• A machine learning model, Predict2Protect, was developed using Python.
• An open-source dataset of 1025 patients was preprocessed and split for model training and testing.
• Four machine learning models were evaluated, with a Decision Tree model selected for its performance.

Main Results:

• The Decision Tree model achieved 100% accuracy on training data and 95% accuracy on test data.
• The application, built with Streamlit, provides a 95% accurate heart disease risk assessment.
• The tool estimates the percentage risk of developing heart disease within the next year.

Conclusions:

• Predict2Protect offers a highly accurate and accessible method for heart disease risk assessment.
• The application can empower individuals with knowledge of their cardiovascular health.
• This tool has the potential to improve global access to early heart disease detection and treatment.