<p>Heart attacks remain a prominent contributor to mortality rates and impose a considerable economic burden, thereby exerting a profound influence on the overall state of global health. The utilisation of machine learning (ML) and artificial intelligence (AI) technologies has greatly enhanced the advancements in the field of early detection and prevention of heart disease. The potential mitigation of myocardial infarction risk can be significantly enhanced through the implementation of personalised medication management and tailored lifestyle advice. Intelligent healthcare systems, characterised by their sophisticated functionalities, provide individuals with the invaluable advantage of real-time monitoring and alerts, thus enabling the efficient management of their cardiovascular health. The dataset encompasses a comprehensive range of variables, encompassing vital statistics, genetic information, behavioural propensities, and medical records. In this work, a smart healthcare system is proposed which will be helpful to analysis the medical history of patients to predict the chances of heart attacks. A dataset of patients has been collected from online sources and classifiers such as naive bayes (NB), decision table (DT), J48 and random forest (RF) used to analysis this dataset based on ECG and ST_slop parameters of selected dataset. For performance and predictive capabilities of classifiers, margin curves of 4 classifiers has been analysed based on opted attributes on same dataset. It has been found that the accuracy results of random forest are 99.89% which are better than the others. The study showed that how smart health care system is helpful for early prediction of heart attacks and can reduce the mortality rate.</p>