Contents lists available at ScienceDirect Medical Hypotheses journal homepage: www.elsevier.com/locate/mehy A novel ECG signal classification method using DEA-ELM Aykut Diker a, ⁎ , Engin Avci b , Erkan Tanyildizi b , Mehmet Gedikpinar c a Bitlis Eren University, Department of Informatics, TR-13100 Bitlis, Turkey b Fırat University, Department of Software Engineering, TR-23100 Elazig, Turkey c Fırat University, Department of Electric-Electronic Engineering, TR-23100 Elazig, Turkey ARTICLE INFO Keywords: Electrocardiogram Differential Evolution Algorithm Extreme Learning Machine Pan-Tompkins technique ABSTRACT Electrocardiogram (ECG) signals represent the electrical mobility of the human heart. In recent years, computer- aided systems have helped to cardiologists in the detection, classification and diagnosis of ECG. The aim of this paper is to optimize the number hidden neurons of the traditional Extreme Learning Machine (ELM) using Differential Evolution Algorithm (DEA) and contribute to the classification of ECG signals with a higher accuracy rate. In this paper, publicly ECG records in Physionet was utilized. Pan-Tompkins technique (PTT) and Discrete Wavelet Transform (DWT) approaches were implemented to obtain characteristic properties which are PR period, QT period, ST period and QRS wave of ECG signals. Then, ELM was executed to the ECG samples. Lastly, DEA on software ELM was developed for the assign of the number of hidden neurons, which were used in the ELM algorithm. The performance criterions were used in order to compare the performance of the classification exerted. Concordantly, it was realized that the highest classification achievement values were reached to Accuracy 97.5% and values 93 of number of hidden neurons, with the practice improved with the DEA compared to conventional ELM. Introduction The Electrocardiogram (ECG) is an alteration of the human heart. These signals have happened in different critical areas. These sections are such as P, QRS complex and T are shown in Fig. 1 [1–4]. Throughout atrial depolarization, the P wave occurs. QRS wave that can be separated partitions happens [1,2]. Detection and classification of the ECG has become one of the most notable powerful and unique instruments in the ECG applications [5,6]. Furthermore, the timely diagnosis of heart disease is a very critic for humans suffers from heart trouble. Analysis of ECG is of a very critical factor in the determination of heart diseases [7]. Particularly, ECG classification has been made effective algorithms by using computer-aid systems [8]. Researchers have developed many works for the identification and separate as normal and anormal of the heart record for many years. For examples of classification techniques are applied by AdaBoost method [9], Radial basis function [10], Adaptive neuro-fuzzy [11], Convolutional neural network (CNN) [12–14], Extreme Learning Machine (ELM) approach [15,16]. Computer-aided based on ELM systems has been exerted in ECG and EEG classification, etc. because of the ability such as rapid learning and generalization [14–18]. Though, it has some difficulties such as the availability of local minima's, indefinite learning percentage, the election of the numeral of hidden neurons and overfitting [21]. In order to solve the disadvantage of ELM, there are various nature-inspired population-based techniques with global search abilities such as Differential Evolutional (DE) [19,20,22,23], Particle Swarm Optimization (PSO) [24], Genetic Al- gorithm (GA) [25], Ant Colony Optimization (ACO) [26], Artifical Bee Colony algorithm (ABC) [27]. The ECG diagnostic is employed gen- erally features obtained from the P wave, QRS complex and T wave which are crucial sections of heart records. Meanwhile correct defining of the heart signals is a significant and critical process for the experts which is to make a correct detection decision. By using machine learning systems is detected of the heart signal has a major effect on the diagnostic of heart disease. In this work, the effect of the traditional Extreme Learning Machine based on Differential Evolutional Algorithm (DEA) method is studied in the correct detected of heart activity. The innovation of our technique is the utilization of which is efficient op- timization method DEA by the most appropriate values to overcome the disadvantage (e.g selection of the number of hidden neurons) of tra- ditional ELM for ECG classification. Additionally, since correct feature subtraction, is necessary for proper ECG classification, a system that utilizes DWT and PTT of ECG signals for feature subtraction was op- erated in this paper. Besides, the computer-aided techniques effective- ness was examined with regard to Accuracy, Sensitivity, Specificity, and F-measure. The main goal of this paper is to provide increase https://doi.org/10.1016/j.mehy.2019.109515 Received 3 November 2019; Received in revised form 25 November 2019; Accepted 30 November 2019 ⁎ Corresponding author. E-mail addresses: eavci@firat.edu.tr (E. Avci), etanyildizi@firat.edu.tr (E. Tanyildizi), mgedikpinar@firat.edu.tr (M. Gedikpinar). Medical Hypotheses 136 (2020) 109515 0306-9877/ © 2019 Elsevier Ltd. All rights reserved. T