Copyright © 2018Authors. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. International Journal of Engineering & Technology, 7 (2.17) (2018) 27-33 International Journal of Engineering & Technology Website: www.sciencepubco.com/index.php/IJET Research Paper Analysis of ECG Arrhythmia for Heart Disease Detection using SVM and Cuckoo Search Optimized Neural Network Dinesh D. Patil 1* , R. P. Singh 2 , Vilas M. Thakare 3 , Avinash K. Gulve 4 1* Research Scholar, Department of Computer Science and Engineering, 2 Department of Computer Science and Engineering 1,2 Sri Satya Sai University of Technology and Medical Sciences, Sehore-466001, India, 3 Department of Computer Science, Faculty in Engineering & Technology, Sant Gadge Baba Amravati University, Amravati-444601, India. 4 Department of Master of Computer Applications, Government College of Engineering, Aurangabad-431005, India. *Corresponding author E-mail:dineshonly@gmail.com Abstract This paper tried to address several topics concerning the analysis, synthesis and compression of the electrocardiogram signal (ECG) using the MIT database. We detect the R-wave by identifying the location of each interval delineating a QRS complex using unbiased and biased estimators. In the second part of the work, we segmented the signal into RR periods constituting the vectors of a data matrix, where we extracted its main components in order to reduce the size of the cardiac information, and then further reduced in addition the size by the use of a threshold on the signal. Then the classification is done for automatic detection of heart disease using Support Vector Machine (SVM) and Cuckoo Search Optimized Neural Network. ECG beats with 4 types of abnormalities (RBBB, APC, PVC and LBBB) from ECG records is retrieved from the MIT-BIH arrhythmia database. Analysis of the different groups shows the overall recognition perfor- mance was 99.50%. The worst is 99.63% for the RBBB class. Keywords: Cuckoo Search; Cardiovascular disease; Electrocardiograms; Neural Network; QRS; Support Vector Machine. 1. Introduction Cardiovascular disease (CVD) is a major public health problem. They come top of medical causes of death in India. According to WHO (World Health Organization), heart attacks and strokes are responsible for more than 80% of cardiovascular deaths. There are many risk factors: tobacco, sedentary lifestyle, obesity, high blood pressure, diabetes and genetic factors. The heart, the central organ of the cardiovascular system, can be affected by many diseases that can be either benign, like some tach- ycardia’s for example, or very serious, such as myocardial infarc- tion that causes 10% of deaths worldwide. Due to the scale of the problem, the follow-up of patients at risk becomes essential. Minor arrhythmias inform the physician about the patient's cardiac status. They must be detected in particular to prevent possible degenera- tion in severe arrhythmias [1]. The ElectroCardioGramme (ECG) is the most commonly per- formed examination because it is quick to put in place, inexpensive and above all non-invasive and therefore very inconvenient for the patient [2]. The electrocardiogram is a graphical representation of the electrical potential that controls the muscular activity of the heart. This po- tential is collected by electrodes placed on the surface of the body. It is presented as a series of repetitive (wave) deflections, each rep- resenting a phase of functioning of the heart. Each visible distortion on these waves can be attributed to cardiac dysfunction or arrhyth- mia. Cardiac arrhythmias mainly have their origins in the birth of the cardiac stimulus or in the conduction of the depolarization wave through the myocardium (path followed by the depolarization wave from its point of electrical activation). One way to detect cardiac disorders is to collect the cardiac electri- cal signal (ECG) by sensors and then analyze it. This analysis pre- sents both practical and theoretical issues for current research in pattern recognition and medicine. The objective pursued through this paper is to propose new methods of recognition of cardiac ar- rhythmias to help the doctor to read long-term recordings. The ECG is very often supplemented by a similar 24-hour exami- nation called "Holter", in which the patient can go about his or her usual activities. The main advantage of Holter recording over the short-term ECG is that it allows the detection of sporadic events that do not necessarily occur during the few seconds of ECG re- cording performed in a hospital setting, when the patient is at rest [3-6] In the case of the recognition of cardiac arrhythmias, the first step is to treat the raw signal coming from an often noisy Holter record- ing: to filter it to extract the useful signal, to transform it if neces- sary in the frequency domain or in the domain time scale to locate and segment areas of interest [1].These allow us to quantify and describe each beat. Note that these descriptors are those usually used by the cardiologist to make his diagnosis. At this point, each