Detection of Cardiovascular Disease Using Ensemble Feature Engineering With Decision Tree

Introduction

Cardiovascular disease (CVDs) is one of the deadliest diseases in the world. In 2019, WHO and GBD reported that 30% deaths occurred due to CVDs and it is also expected that 23.6 million people may get affected by the end of 2030 (Joshi & Rienks 2021). In the last few decades, the number of death due to CVDs has increased noticeably from 15.2% to 28.1% in India (Tichenor and Sridhar 2019). Indian public medical system is still incapable of preventing non-communicable diseases including CVDs. Due to lack of the five A’s like awareness, access, absence, affordability, and accountability in India, the efficient healthcare facility is not adequate (Kasthuri et al 2018). Medical investigation and therapeutic assistance are still unavailable in rural areas which hamper human health. Despite the government initiatives of medicare for needy people, the majority of Indian populace does not have preventive health care benefits. All these factors are associated with slow therapy and increased morbidity and mortality (Peiris & Ghosh 2021). Similarly, few developed countries including China, America has a higher percentage of CVDs death. As a result, CVDs are a prime reason for health loss for all regions of the world (GRF Collaborators 2018, Roth & Mensah 2020).

Cardiovascular diseases are a cluster of disorders of heart and blood vessels, which are typically associated with human behavioural factors (Jennum et al 2021). The most significant behavioural risk factors include poor diet, physical latency, tobacco and liquor consumption. The effect of hazardous factors may increase blood pressure, glucose level, blood lipids, overweight and obesity. These factors indicate an increase in cardiac disorder and other complexity (Sabzmakan & Morowatisharifabad 2014). The most accurate method for CVDs diagnosis is coronary computed tomography angiogram, echocardiogram, and magnetic resonance imaging. These tests are relatively expensive, time-consuming and not readily available in low resource regions (Abbara & Blanke 2016). To address these issues, an intelligence model is highly demanded which is used for accurate and early detection of CVDs. Therefore, the availability of such services is essential to save physicians time and reduce the financial burden for unprivileged patients. Additionally, this model acts as a pre-alarming system for normal people who have a probability of growing cardiac disorder (Haq et al 2021).