Among the different treatments for ventricular tachycardia (VT), catheter ablation guided by the electroanatomic mapping is increasingly adopted. Ablation targets are currently identified by visual inspection from the cardiologist during the mapping procedures. To improve the outcome and consequently reduce the VT recurrences, in this study, an automatic approach exploiting the deep learning on intracardiac bipolar electrograms (EGMs) is proposed. A balanced dataset composed of 752 physiological and 752 abnormal ventricular potentials (AVPs) was collected from nine post-ischemic VT patients. By adopting the synchrosqueezed wavelet transform, for each EGM a time-frequency representation was obtained to be used for feeding a siamese neural network, known to be usable on small dataset. Our findings revealed accurate classification capabilities, thus encouraging the introduction of deep learning tools in the recognition of AVPs and paving the way for their use in supporting clinicians for targeting arrhythmogenic sites.

Abnormal Ventricular Potentials Identification Using a Siamese Neural Network

Pitzus A.;Baldazzi G.;Raffo L.;Pani D.
2023-01-01

Abstract

Among the different treatments for ventricular tachycardia (VT), catheter ablation guided by the electroanatomic mapping is increasingly adopted. Ablation targets are currently identified by visual inspection from the cardiologist during the mapping procedures. To improve the outcome and consequently reduce the VT recurrences, in this study, an automatic approach exploiting the deep learning on intracardiac bipolar electrograms (EGMs) is proposed. A balanced dataset composed of 752 physiological and 752 abnormal ventricular potentials (AVPs) was collected from nine post-ischemic VT patients. By adopting the synchrosqueezed wavelet transform, for each EGM a time-frequency representation was obtained to be used for feeding a siamese neural network, known to be usable on small dataset. Our findings revealed accurate classification capabilities, thus encouraging the introduction of deep learning tools in the recognition of AVPs and paving the way for their use in supporting clinicians for targeting arrhythmogenic sites.
2023
abnormal ventricular potentials
arrhythmogenic sites identification
Cardiac electrophysiology
deep learning
ventricular tachycardia
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: https://hdl.handle.net/11584/399385
 Attenzione

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

Citazioni
  • ???jsp.display-item.citation.pmc??? ND
  • Scopus 0
  • ???jsp.display-item.citation.isi??? ND
social impact