PrecisionCardio: A Comprehensive Machine Learning Approach for Accurate Prediction of Heart Failure Trajectory

Accurate prediction of cardiac disease can prevent potentially fatal situations. For providing cardiologists with assistance to predict the trajectory of heart failure (HF), we introduce a system based on machine learning that employs medical patient history to estimate the severity of disease advan...

Full description

Saved in:

Bibliographic Details
Published in:	2024 IEEE 30th International Conference on Telecommunications (ICT) pp. 1 - 4
Main Authors:	Hossain, Muhammad Anwar, Bin Shawkat, Siam, Sharif, Kazi Shaharair, Hossain, Md Ikbal, Asmani, Herick, Rahman, Md Minhazur
Format:	Conference Proceeding
Language:	English
Published:	IEEE 24-06-2024
Subjects:	Accuracy Cardiac disease Cardio Vascular Disease Heart Heart Disease Logistic Regression Predictive models Random Forest Classifier Risk Analysis Support Vector Classifier Support vector machines Trajectory Vectors
Online Access:	Get full text
Tags:	Add Tag No Tags, Be the first to tag this record!

Description
Summary:	Accurate prediction of cardiac disease can prevent potentially fatal situations. For providing cardiologists with assistance to predict the trajectory of heart failure (HF), we introduce a system based on machine learning that employs medical patient history to estimate the severity of disease advancement. Using a labeled dataset comprising 918 records, we evaluated six models based on its performance in predicting heart disease in this study. Comparable results to our own were obtained from studies that utilized longitudinal multi-class prediction and utilized much larger datasets. A number of machine learning strategies for predicting cardiovascular diseases using patient data on important health factors are described in this work. We presented six unique classification techniques that were applied to develop the prediction models: Support Vector Machine(SVM), NaAve Bayes (NB), Random forest Classifier(RFC), K-Nearest Neighbours(KNN), Logistic Regression and Gradient Boosting Classifier. Before these models were constructed, the data under-went preprocessing and feature selection. The evaluation of the models incorporated the following metrics: F1-score, accuracy, precision, and recall. The SVM model achieved the maximum degree of accuracy, 92.39%.
ISSN:	2993-1916
DOI:	10.1109/ICT62760.2024.10606019