LabVIEW ECG and Noise Simulator for Advanced Synthesis of Machine Learning Databases

This study aims to develop an advanced simulator of electrocardiogram (ECG) signals and overlaid noises (electromyogram, power-line interference, baseline wander, white noise), which benefits the extended functionality of the LabVIEW Biomedical Toolkit. Interactive interface handles the visualization of time and spectral behavior of ECG, noise and mixed ECG+noise signals from Physionet databanks or artificial signal generators with advanced morphological and frequency controls. Noise-free and noisy ECG databases with fixed or randomized behavior can be thereby created for the purpose of data augmentation in machine learning arrhythmia detection strategies, or for standard performance testing of diagnostic electrocardiographs.