Chaos Theory: An Emerging Tool for Arrhythmia Detection

The heart is an important muscular organ of the human body which pumps blood throughout the body. It is essential for human life. Timely and accurate assessment of the functioning of the heart has great relevance for reducing the death rate due to cardiac diseases around the world. If the heart is not able to pump blood smoothly, then heart diseases are likely to appear. These heart diseases are known as arrhythmia. Electrocardiogram (ECG) is a diagnostic tool for assessing the functioning of heart non-invasively. It not only detects cardiovascular diseases, but also examines breathing pattern and mental stress. ECG appears in the form of an electrical signal that comprises of P-QRS-T waves and is captured by pasting electrodes on the surface of the skin in a conductive medium. Features of these wave components, such as clinical frequencies, heart rate (HR) measurement, RR interval measurement, spectral components, non-linearity, trajectory identification, and amplitudes, help doctors to diagnose cardiac arrhythmias accurately. This paper presents a computer aided diagnosis (CAD) system to extract non-linearity and trajectory patterns using the theory of chaos analysis to aid cardiologists diagnose arrhythmia accurately. ECG signal is non-stationary and non-linear in nature due to which it contains multiple time-varying frequencies. So a more reliable and accurate technique like time–frequency transform such as short-time Fourier transform (STFT) etc. is needed. In this paper, STFT is used, which is an efficient technique to observe frequency contents of small non-linear segments in time domain. It is used to determine the sinusoidal frequency and phase content of the local sections of a signal. R-peak is very crucial for classifying cardiac arrhythmia. Therefore, STFT is used for detecting R-peaks and their frequency contents. Due to limited time–frequency resolution, STFT usually misses on some information. Therefore, there is a need of supplementing the existing research on the ECG signal interpretation by using non-linear techniques. These gaps have motivated us to use chaos theory (analysis) for ECG signal analysis. The non-linear techniques are expected to yield supplementary clues about the non-linearities in the considered segment. In chaos analysis, the sketched trajectories represent the flow of the system where each trajectory involves a subregion of the phase space known as an attractor. In phase space, set of points depicts complete status of the cardiac cycles through which the system migrates over time. An attractor showcases the best preview according to the initial conditions and time delay dimension. The shape of an attractor may be oval, egg-shaped, circular and in some cases corn-type. It increases the decision capability of the proposed system by identifying correct arrhythmia type. For validating this research work, physioNet database [Massachusetts Institute of Technology-Beth Israel Hospital Arrhythmia database (M-BArr DB), Ventricular Tachyarrhythmia database (VT DB)] and real time database (R T DB) have been used. The proposed technique has been evaluated on the basis of sensitivity (Se) and positive predictive value (PPV). Se of 99.92% and PPV of 99.93% are obtained for the considered databases (M-BArr DB, VT DB and RT DB). Chaos theory together with STFT has proved itself as a good approach that reduces the occurrence of spurious outcomes and has demonstrated the properties that are typical outlook of deterministic chaotic systems.