Generador de Señales Electrocardiográficas con Morfologías Múltiples

摘要

In this paper is presented, a computational tool, to generate (in a controlled way) multiple morphologies of quasi-artificial electrocardiogram (ECG) beats, starting from the configuration and concatenation of six (6) default segments of an ECG beat. The tool allows: to add noises and line base drifts, to convolve the signal with up to three mathematical functions, and to introduce artefacts in any time position of a beat. The tool could be used for the training and evaluation of ECG patterns classifiers such as neural networks (NN), support vector machines (SVM) and others based on Artificial Intelligence techniques.