ABSTRACT

Background: Cardiovascular diseases are the leading causes of death in the world, caused by disorders of the heart and blood vessels. Among all the heart related diseases, IHD is more prevalent in Indian population. Naïve Bayes algorithms refers to a classification technique that depends on the application of Bayes theorem and which is the best model in machine learning. It is relatively simple to build a model to obtain the estimated probability for a prediction and capable of handling extremely large datasets. Methodology: A retrospective study was designed to access the 7304 echocardiography records of patients who underwent transthoracic echocardiography at Department Cardiology, JSS Hospital in the year 2016. A model was developed with ECHO database using Naive Bayes classifier. The dataset consists of 6191 patients without IHD and 1113 patients with IHD along with their ECHO parameters. Results: The model can be trained using naivebayes, e1071 and caret packages, which allows us to perform Naïve Bayes in a powerful and scalable architecture. The final output displays that a Naive Bayes classifier was built which has the capability of predicting whether a person suffers from IHD or not, with an accuracy of approximately 95%. The value of Kappa Statistc (0.810), Precision (0.838), F – Score (0.839) and ROC (0.969). Conclusion: The Naïve Bayes model is implemented in R-studio as an application, which takes ECHO parameter as an input and it was tested for its accuracy (95 %) in predict disease risk. Keywords: ECHO database, Ischemic Heart Disease, Naïve Bayes Classifier, R studio
Publication date: 01/01/2023

https://ijbpas.com/pdf/2023/January/MS_IJBPAS_2023_6797.pdf

Download PDF
https://doi.org/10.31032/IJBPAS/2023/12.1.6797