# iCardo: A Machine Learning Based Smart Healthcare Framework for Cardiovascular Disease Prediction

**Authors:** Nidhi Sinha, Teena Jangid, Amit M. Joshi, Saraju P. Mohanty  
**arXiv:** [2212.08022](https://arxiv.org/abs/2212.08022)  
**Date:** 2022-12-07  
**Summary page:** https://www.summarizepaper.com/en/paper/2212.08022v1/

## Key points

- Point of care services and medication have become simpler with efficient consumer electronics devices in a smart healthcare system.
- Early identification of cardiovascular disease (CVD) is crucial for preventing premature mortality.
- Machine learning has been used to predict CVD in the literature.
- This article focuses on selecting the best classifier model and feature sets for predicting CVD.
- The study compares seven classifiers using sixteen different feature sets.
- Four distinct feature sets were selected using Lasso, Tree-based algorithms, Chi-Square, and RFE methods containing five, ten, fifteen, and twenty features respectively.
- A total of 112 models were trained and evaluated to determine the most effective combination of feature set and classifier model.
- The SVM classifier with fifteen chosen features showed the highest overall accuracy.
- The proposed framework called iCardo utilizes machine learning to predict CVD in a smart healthcare system.
- Healthcare data is stored in the cloud and can be accessed by patients, caretakers, and healthcare providers through integration with the Internet of Medical Things (IoMT).
- This provides a valuable tool for anticipating cardiovascular disease and improving patient outcomes in smart healthcare systems.
- Overall, this study demonstrates the effectiveness of machine learning in predicting CVD using selected feature sets.

## Summary

The point of care services and medication have become simpler with efficient consumer electronics devices in a smart healthcare system. Cardiovascular disease (CVD) is a critical illness that can lead to heart failure, and early identification is crucial for preventing premature mortality. Machine learning has been used in the literature to predict CVD. This article focuses on selecting the best classifier model and feature sets for predicting CVD. The study compares seven classifiers using sixteen different feature sets. The original dataset had 56 attributes and 303 occurrences, with 87 instances of good health and the rest having CVD. Four distinct feature sets were selected using Lasso, Tree-based algorithms, Chi-Square, and RFE methods containing five, ten, fifteen and twenty features respectively. A total of 112 models were trained and evaluated to determine the most effective combination of feature set and classifier model. The SVM classifier with fifteen chosen features showed the highest overall accuracy. The proposed framework called iCardo utilizes machine learning to predict CVD in a smart healthcare system. The healthcare data is stored in the cloud and can be accessed by patients, caretakers, and healthcare providers through integration with the Internet of Medical Things (IoMT). This provides a valuable tool for anticipating cardiovascular disease and improving patient outcomes in smart healthcare systems. Overall, this study demonstrates the effectiveness of machine learning in predicting CVD using selected feature sets.