A Model for the Real Time Detection of Ventricular Late Potentials Oriented to Embedded Systems Implementation

In this paper, it is proposed a new dynamic model for the real time detection of ventricular late potentials (VLP) in high-resolution electrocardiogram signal (HREGC). The VLPs are very small intensity signals addressing dangerous incoming cardiac pathologies, not detectable without a proper processing of the electrocardiographic signal. The model is based on the time-frequency (TF) representation of the HRECG, and on the wavelet denoising. It is suitable for the implementation in an embedded system useful for the real time acquisition and processing of the HRECG signal, based on FPGA devices. The model was successfully verified on a wide database of HRECGs. Then, it has been successfully compiled by using the DSPbuilder compiler by Altera, in a .sof file to program an Altera’s FPGA device of the Cyclone V family.