Cardiovascular Disease Prediction Using Data Mining Techniques: A Proposed Framework Using Big Data Approach

Introduction

Heart is an important organ in human body which draws blood through blood veins of circulatory system. Factors responsible for the risk of CVDs are smoking, alcohol, obesity, high BP (blood Pressure).

Figure 1.

Causes of Heart Disease

As per WHO (World Health Organization, around 20 million people die around the world due to cardiovascular disease. If proper medical care is given, many lives can be saved. Heart disease prediction is one of the difficult tasks in field of health sciences.

In 2015, an expected 17.7 million humans died from cardiovascular disease which is 31% of total worldwide deaths. An estimated 7.4 million people’s deaths out of total deaths were due to coronary heart sickness and estimated 6.9 million deaths have been due to heart stroke. We know that heart disease is topmost reason for deaths of humans all across the globe. Every one of the five deaths is due to heart disease. Almost 1/2 folks residents of US have minimum one risk component for heart sickness along with high blood strain, obesity, sedentary lifestyle, high blood sugar etc.

Coronary illness is a common sickness these days. Presently due to expanding costs of coronary illness, there was a need to build up another framework which can foresee heart sicknesses in a simple and less expensive way. Information mining systems and learning disclosure assumes a significant part in such discoveries. We know that experts who treat heart disease keep data records of patient. To extract the useful information these datasets are analyzed. Therefore data mining turns out to be a productive and fruitful tool.

Data mining is clearly defined as “The repetitive and interactive manner of coming across logical, novel, beneficial, and understandable knowledge (patterns, models, guidelines etc.) in huge databases”

(Fayyad et al, 1996; Kurgan et al, 2006; Gobel et al, 1999). The discovery of knowledge in databases is simply well-described step comprising numerous distinct movements. Information mining is the important stage that comes within the uncovering of hidden however beneficial information from significant databases.

Health Science information mining is a field which connects with parcel of imprecision and vulnerability. Providing value facilities at reasonable price is the main concern of the healthcare framework. Inappropriate medical outcome may prompt terrible results. Medical science is very large and medical conclusions are made depend on medical doctor’s knowledge not on the data covered up in the database which in some cases will result in improper results and also changes the nature of service to the patients (Wasan et al, 2006). Health checkup past data consists of a number of investigations needed to make a diagnosis of a specific illness .It is conceivable to accomplish the benefit of information mining in medicinal services via utilizing it as a shrewd investigative instrument. The specialists in the healthcare system recognize and anticipate the disease by making use of Information mining systems (Setiawan, 2009).

Researchers have suggested that the use of statistical mining in figuring out powerful remedies for patients can enhance practitioner overall performance. Experts investigated and came to result that by applying specific information mining strategies for the examination of cardiovascular sickness to find out which information mining method can offer prominent exactness. Many distinct data mining strategies were already used to help medical care departments to detect coronary heart ailment (Dewan et al, 2015; Chandna, 2014).