Key result
Neural networks outperformed other machine learning models such as random forest and naive bayes in predicting 10-year cardiovascular risk, achieving an accuracy of 97%.
Why the study?
Do machine learning models accurately predict 10-year cardiovascular risk using vital parameters and ECG signals?
Do machine learning models accurately predict 10-year cardiovascular risk using vital parameters and ECG signals?
Machine learning models, particularly neural networks, demonstrate high accuracy (97%) in predicting 10-year cardiovascular risk using clinical and ECG parameters.
Answers are grounded in this paper and cite their sources.
Should not yet change cardiovascular risk assessment; leaves open prospective validation of neural network models.
Cardiovascular diseases (CVD) are termed to be the group of diseases related to the human heart, which are very dangerous and risky. It affects human life on various levels depending upon its complexity rate. Various Machine Learning (ML) techniques have been on the top of the list to evaluate the cardiovascular risk. Different factors such as age, height, blood pressure, smoking habits, cholesterol levels etc have been considered in risk analysis. Many clinical parameters have been looked out thoroughly along with the variations in the ECG signals and other heart related conditions in humans to evaluate and determine the CVD risk factors. The risk that a particular person will be prone to, in the next 10 years will be determined by considering the abovementioned factors. Vital parameters dataset and few parameters of ECG signals are used carefully in the prediction process to build a well-defined machine learning based CVD risk score model. Established machine learning algorithms, feature selection algorithms and classifiers such as random forest and naive bayes are used. The performance metrics of the models are compared. The risk that the person is prone to, is divided into low, moderate, and high. The features that are extracted from the ECG are used to feed the predictors. Neural networks outperform other models with 97% accuracy. The proposed machine learning based models will play a maj or role in assisting the doctors in diagnosing the CVD.
No takes yet. Share an insight, caveat, or question.
Jayasudha et al. (2021) studied Cardiovascular disease risk. Neural networks vs. Other machine learning models (random forest, naive bayes) was evaluated on Model accuracy for predicting 10-year CVD risk. Neural networks outperformed other machine learning models such as random forest and naive bayes in predicting 10-year cardiovascular risk, achieving an accuracy of 97%.
Synapse has enriched 10 related papers. Consider them for comparative context: