Spectral analysis for slow pathway characterization in atrioventricular nodal reentry tachycardia

Atrioventricular Nodal Reentrant Tachycardia (AVNRT) represents the most prevalent form of supraventricular tachycardia, arising from reentrant circuits near the atrioventricular node. This thesis exploits the spectral analysis of slow pathway signals to find some parameters that uniquely characterize these signals compared to other cardiac signals. Identifying slow pathway potentials is crucial as they are the target for ablation in treating AVNRT. However, current methods to determine the exact area for radiofrequency application only narrow down the potential zone, leaving the final selection strongly dependent on the electrophysiologist's expertise. Utilizing the CARTO 3 mapping system by Biosense Webster, we were able to extract 355 signals from 42 patients. The research introduces a multiparametric index composed of a better combination of parameters appropriately weighted by the Latin hypercube sampling method. The findings reveal that integrating spectral analysis into electrophysiological studies holds promise for advancing the treatment of AVNRT, proposing a shift towards a more precise approach. In particular, we could classify the signal into target (slow potentials) and non-target (other cardiac signal) with a sensibility of 80% and a specificity of 55%, significantly reducing the current false positive rate. The thesis underscores the potential of merging time and frequency domain analyses to refine signal interpretation, aiming at reducing procedural risks and improving patient outcomes. Furthermore, it opens avenues for future technological and methodological advancements in electrophysiology, including the development of diagnostic tools that incorporate AI for real-time signal classification, enhancing the efficacy and safety of ablation procedures.