Various recent denoising techniques for ECG signals were surveyed and their performance compared using benchmark datasets such as the MIT-BIH Cardiac Arrhythmias Database.
This survey reviews and compares recent denoising techniques for ECG signals to improve the accuracy of cardiac disorder diagnosis.
Most of the cardiac disorders are diagnosed by analysis of electrocardiogram (ECG) of the subject. Noise sources in ECG can either be cardiac or extra cardiac, resulting in the distribution of artifacts throughout the original signal. Non-ideal conditions such as electromagnetic interference caused by power cables of the monitoring equipment and muscle or electrode movements corrupt the ECG. This leads to inaccurate analysis of fatal cardiac diseases such as arrhythmias making noise removal from the ECG signals necessary before using them for diagnosis purposes. PLI, Baseline Wander, motion artifact (MA) and muscle artifacts (EMG) are the most common types of noise contaminating an ECG signal. This paper presents a survey of some of the major denoising techniques from recent years used in biomedical signal processing for de-noising ECGs. The techniques were simulated and tested on various renowned benchmark datasets mainly MIT-BIH Cardiac Arrhythmias Database. Performance comparisons have also been discussed in the paper.
Butt et al. (Wed,) conducted a review in Cardiac disorders / Arrhythmias. Denoising techniques was evaluated on Performance comparisons of denoising techniques. Various recent denoising techniques for ECG signals were surveyed and their performance compared using benchmark datasets such as the MIT-BIH Cardiac Arrhythmias Database.