Ventricular abnormal potentials (VAPs) identification is a challenging issue, since they constitute the ablation targets in substrate-guided mapping and ablation procedures for ventricular tachycardia (VT) treatment. In this work, two approaches for the supervised classification of VAPs in bipolar intracardiac electrograms are evaluated and compared. To this aim, 954 bipolar electrograms were retrospectively annotated by an expert cardiologist. All signals were acquired from six patients affected by post-ischemic VT by the CARTO3 system at the San Francesco Hospital (Nuoro, Italy) during routine procedures. The first classification approach was based on a support vector machine trained and tested on four different features, extracted from both the time and time-scale domain, to identify physiological and abnormal potentials. Conversely, in order to assess the significance of the first approach and its features, in the second approach all the samples constituting a time-domain segment of each bipolar electrogram were given as input to a feed-forward artificial neural network. In both cases, the accuracy in VAPs and physiological potentials identification exceeded 79%, suggesting their efficacy and the possibility of VAPs automatic recognition without identifying peculiar features.

Supervised Classification of Ventricular Abnormal Potentials in Intracardiac Electrograms

Baldazzi G.;
2020

Abstract

Ventricular abnormal potentials (VAPs) identification is a challenging issue, since they constitute the ablation targets in substrate-guided mapping and ablation procedures for ventricular tachycardia (VT) treatment. In this work, two approaches for the supervised classification of VAPs in bipolar intracardiac electrograms are evaluated and compared. To this aim, 954 bipolar electrograms were retrospectively annotated by an expert cardiologist. All signals were acquired from six patients affected by post-ischemic VT by the CARTO3 system at the San Francesco Hospital (Nuoro, Italy) during routine procedures. The first classification approach was based on a support vector machine trained and tested on four different features, extracted from both the time and time-scale domain, to identify physiological and abnormal potentials. Conversely, in order to assess the significance of the first approach and its features, in the second approach all the samples constituting a time-domain segment of each bipolar electrogram were given as input to a feed-forward artificial neural network. In both cases, the accuracy in VAPs and physiological potentials identification exceeded 79%, suggesting their efficacy and the possibility of VAPs automatic recognition without identifying peculiar features.
File in questo prodotto:
Non ci sono file associati a questo prodotto.

I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.

Utilizza questo identificativo per citare o creare un link a questo documento: http://hdl.handle.net/11567/1061176
 Attenzione

Attenzione! I dati visualizzati non sono stati sottoposti a validazione da parte dell'ateneo

Citazioni
  • ???jsp.display-item.citation.pmc??? ND
  • Scopus 1
  • ???jsp.display-item.citation.isi??? 1
social impact