MATLAB based ECG Signal Analysis

摘要

An Electrocardiogram Signal is a bioelectrical Signal which records the heart’s electrical activity versus time. It’s a crucial diagnostic tool for assessing heart functions. The first detection of arrhythmia is essential for cardiac patients. ECG arrhythmia is often defined as any gaggle of conditions during which the guts’ electrical activity is irregular and may cause the heartbeat to be slow or fast. It can happen during a healthy heart and be of minimal consequence, but it’ll also indicate a Significant problem that results in stroke or sudden cardiac death. As the ECG Signal is a non-stationary Signal, the arrhythmia may occur randomly within the time-scale, which suggests, the arrhythmia symptoms might not show up all the time. Still, they would manifest at certain irregular intervals during the day. Thus, automatic classification of arrhythmia is critical in clinical cardiology, especially for treating patients within the medical care unit. This project implements a MATLAB platform simulation tool to detect abnormalities within the ECG Signal and calculate heartbeat. The ECG Signal is downloaded from the MIT-BIH Arrhythmia Database. Since this Signal contains some noise and artifacts hence preprocessing of the ECG Signal is performed first. The preprocessing of ECG Signal is performed with Wavelet toolbox’s help wherein baseline wandering, denoising, and removal of high frequency and low frequency is performed to enhance SNR ratio of ECG Signal. The Wavelet toolbox and FarukUYSAL are additionally used for feature extraction of ECG Signals. Classification of arrhythmia is predicated on basic classification rules. The entire project is implemented on the MATLAB platform. The performance of the algorithm is evaluated on MIT-BIH Database.