Computer-Aided Diagnosis of Cardiac Arrhythmias

Computer-Aided Diagnosis of Cardiac Arrhythmias

Markos G. Tsipouras (University of Ioannina, Greece), Dimitrios I. Fotiadis (University of Ioannina, Greece, Biomedical Research Institute-FORTH, Greece, & Michaelideion Cardiology Center, Greece) and Lambros K. Michalis (University of Ioannina, Greece)
DOI: 10.4018/978-1-60566-026-4.ch107
OnDemand PDF Download:
No Current Special Offers


In this chapter, the field of computer-aided diagnosis of cardiacdiac arrhythmias is reviewed, methodologies are presented, and current trends are discussed. Cardiac arrhythmia is one of the leading causes of death in many countries worldwide. According to the World Health Organization, cardiovascular diseases are the cause of death of millions of people around the globe each year. The large variety and multifaceted nature of cardiac arrhythmias, combined with a wide range of treatments and outcomes, and complex relationships with other diseases, have made diagnosis and optimal treatment of cardiovascular diseases difficult for all but experienced cardiologists. Computer-aided diagnosis of medical deceases is one of the most important research fields in biomedical engineering. Several computer-aided approaches have been presented for automated detection and/or classification of cardiac arrhythmias. In what follows, we present methods reported in the literature in the last two decades that address: (i) the type of the diagnosis, that is, the expected result, (ii) the medical point of view, that is, the medical information and knowledge that is employed in order to reach the diagnosis, and (iii) the computer science point of view, that is, the data analysis techniques that are employed in order to reach the diagnosis.
Chapter Preview

Review Of The Proposed Methods

There are several aspects that can be addressed in order to review the proposed methods for computer-aided diagnosis of cardiac arrhythmias. The type of the diagnosis is the most important since cardiac arrhythmia is a very complex problem, having several different characteristics that need to be considered before reaching a safe diagnosis. Also, the electrocardiogram (ECG) analysis that is employed for this purpose is another important aspect. Finally, the data analysis and classification algorithms that are used define the accuracy and robustness of each approach.

Type of Diagnosis

Concerning the type of the diagnosis, two main approaches have been followed in the literature: (i) arrhythmic episode classification, where the techniques focus on the total episode and not on a single beat, and (ii) beat-by-beat classification, in which each beat is classified into one of several different classes related to arrhythmic behavior. Arrhythmic episode classification was performed in most of the methods proposed early in the literature, addressing mainly the discrimination of one or more of ventricular tachycardia (VT), ventricular fibrillation (VF), and atria fibrillation (AF) from normal sinus rhythm (NSR). More recent approaches mainly focus on beat-by-beat classification. In each case, a much larger number of different types of cardiac arrhythmic beats are considered. A combination of these two different approaches has been proposed by Tsipouras (Tsipouras, Fotiadis, & Sideris, 2005), where beat-by-beat classification was initially performed and the generated annotation sequence was used in order to detect and classify several types of arrhythmic episodes.

Key Terms in this Chapter

Heart Rate Variability (HRV): The alterations of the heart rate between consecutive heartbeats.

Fuzzy Logic: Derived from fuzzy set theory dealing with reasoning that is approximate rather than precisely deduced from classical predicate logic.

Electrocardiogram (ECG): Recording of the electrical activity of the heart.

Atrial Premature Contractions (APCs): Single or paired extrasystoles that originate in the atrials.

Premature Ventricular Contractions (PVCs): Single or paired extrasystoles that originate in the ventricles.

Artificial Neural Network (ANN): An interconnected group of artificial neurons that uses a mathematical or computational model for information processing based on a connectionist approach to computation. It has the ability to learn from knowledge, which is expressed through interunit connection strengths, and can make this knowledge available for use.

RR-Interval Signal: The signal representing the durations between consecutive R waves of the ECG.

QRS Detection: Procedure for detecting the QRS complexes in the ECG signal.

Atrial Fibrillation (AF): Disorganized, high-rate atrial electrical activity.

Ventricular Tachyarrhythmias: Repetitive forms of three or more consecutive ventricular ectopic beats.

Complete Chapter List

Search this Book: