Key result
An artificial neural network predicts cardiovascular disease with ~89% accuracy using 13 clinical features.
Why the study?
Cardiovascular disease is a primary cause of increased mortality globally, making automated methods for early prediction of CVD critical to prevent the disease.
Artificial Neural Networks can achieve high accuracy (88.5%) in predicting cardiovascular disease using standard physiological attributes from public datasets.
Answers are grounded in this paper and cite their sources.
May support ML models for CVD prediction on public data; leaves open prospective validation before clinical use.
INTRODUCTION: Cardiovascular disease (CVD) is one of the primary causes of the increased mortality rate universally. Therefore, automated methods for early prediction of CVD are of utmost importance to prevent the disease. OBJECTIVES: In this study, we have pointed out the major advantages, drawbacks, and the scope of enhancing the prediction accuracy of the existing automated cardiovascular disease prediction methods. In addition to that, we have analyzed various combinations of attributes that can help in prediction at the earliest. METHODS: We have exploited various machine learning models to analyse their performances in predicting the CVD at the earliest. RESULTS: For a publicly available database, the Artificial Neural Network attained the highest accuracy of 88.5% and recall of 90%. CONCLUSION: We justified the notion that it will be beneficial to identify potential physiological and behavioural attributes to predict CVD accurately as early as possible.
No takes yet. Share an insight, caveat, or question.
Choudhury et al. (2023) studied Cardiovascular disease (n=303). Artificial Neural Network (ANN) vs. Logistic Regression, K-Nearest Neighbor, Random Forest was evaluated on Prediction accuracy using 13 features. The Artificial Neural Network attained the highest accuracy of 88.52% and recall of 90% in predicting cardiovascular disease using 13 clinical features from the UCI database.
Synapse has enriched 10 related papers. Consider them for comparative context: