Turk J Elec Eng & Comp Sci (2015) 23: 2089 – 2106 c ⃝ T ¨ UB ˙ ITAK doi:10.3906/elk-1305-41 Turkish Journal of Electrical Engineering & Computer Sciences http://journals.tubitak.gov.tr/elektrik/ Research Article A comparative study of two different FPGA-based arrhythmia classifier architectures Ahmet Turan ¨ OZDEM ˙ IR * , Kenan DANIS ¸MAN Department of Electrical and Electronics Engineering, Erciyes University, Melikgazi, Kayseri, Turkey Received: 05.05.2013 • Accepted/Published Online: 17.07.2013 • Printed: 31.12.2015 Abstract: Early diagnosis of dangerous heart conditions is very important for the treatment of heart diseases and for the prevention of sudden cardiac death. Automatic electrocardiogram (ECG) arrhythmia classifiers are essential to timely diagnosis. However, most of the medical diagnosis systems proposed in the literature are software-based. This work focused on the hardware implementation of a mobile artificial neural network (ANN)-based arrhythmia classifier that is implemented on a field programmable gate array (FPGA) as a single chip solution, as an alternative to various software models of ANNs. Due to the parallel nature of ANNs, hardware implementation of ANNs needs a large amount of chip resources. In order to create an ANN structure in an FPGA, the dimensions of the ANN structure must be reduced; therefore, a data reduction algorithm was employed in the training phase and ECG features and consequently the ANN structure size was reduced with principal component analysis. An eight-input ANN-based arrhythmia classifier that has one hidden layer with two neurons and one output layer with one neuron was implemented on a single-chip FPGA. In this work, two different classifiers were consequently implemented in both 32-bit floating and 16-bit fixed point numerical representations on the same FPGA. Key words: Arrhythmia, field programmable gate arrays, artificial neural networks, principal component analysis, classification 1. Introduction Cardiovascular diseases (CVDs) are one of the major causes of disability in adults as well as the main causes of death in both developed and developing countries. They claim 17.1 million lives per year according to the World Health Organization [1]. CVDs usually end with cardiac arrest that is primarily caused by electrical abnormalities of the heart called arrhythmias. Early diagnosis of dangerous heart conditions is very important for the treatment of heart diseases and for the prevention of sudden cardiac death. Therefore, automatic electrocardiogram (ECG) arrhythmia classifiers are essential to timely diagnosis. Arrhythmias can be divided into two main groups, the first including life-threatening arrhythmias such as ventricular tachycardia (VT) and ventricular fibrillation (VF). These can cause sudden death because the heart cannot pump blood properly and cannot send enough blood to the brain and the rest of the body. They need to be immediately terminated by a defibrillator [2]. Various studies on automatic VT and VF detectors have been carried out [3–6]. The second group of arrhythmias including premature ventricular contraction (PVC) does not need immediate treatment by a defibrillator but still needs therapy to prevent further complications. Recent studies showed that PVC is an indicator for increased risk of sudden cardiac death [7]. Since PVC arrhythmias * Correspondence: aturan@erciyes.edu.tr 2089