Ultra-high frequency for the assessment of cardiac resynchronization therapy

Cardiac resynchronization therapy (CRT) involves the use of a pacing system for the purpose of resynchronizing cardiac contraction. Ideal candidates to receive CRT are patients with optimal medical therapy who demonstrate a reduced ventricular ejection fraction of about 35 % and a QRS complex duration greater than or equal to 120 ms. Clinical data collected from patients presenting with heart failure with symptoms associated with class II, III or IV according to the classification table proposed by the New York Heart Association (NYHA) show that CRT can significantly reduce the rate of hospitalization and mortality. CRT can be delivered by devices equipped with an automatic cardioversion and defibrillation system, taking the name CRT-D, or by resynchronization systems equipped with only the pacemaker function, named CRT-P. Thus, CRT is a treatment option, delivered mostly in a biventricular manner, to correct cardiac abnormalities in patients with impaired ventricular contraction. An innovative tool to provide evaluation and optimization of CRT is Ultra-High Frequency Electrocardiography (UHF-ECG), introduced in 2017 by Jurak et al. UHF-ECG allows, unlike standard ECG, acquisition of the ECG signal at very high frequencies, up to 2 kHz. Through the use of this new technology, it is possible to obtain a spatiotemporal representation of the ECG signal; in fact, the UHF-ECG signal represents the distribution of ultra-high frequency oscillations both spatially and temporally. The most advantageous tool for improving CRT settings is the analysis of UHFQRS maps, which graphically represent, in the form of a color map, the ventricular activation pattern. The purpose of this thesis is to evaluate, after processing UHF-ECG signals and obtaining UHFQRS maps, the use of a new spatial feature, called Area in Pixel, in the context of CRT. Specifically, UHF-ECG signals recorded from 19 patients with QRS complex durations greater than or equal to 120 ms belonging to classes II and III according to the NYHA classification and having CRT systems implanted for at least six months at the time of measurement, were used in this study. Data were recorded during sinus rhythm (pacing off) and during six different pacing configurations for each patient, including atrial pacing, sequential left ventricular (LV) or right ventricular (RV), and biventricular (BiV) pacing. These data were recorded, in collaboration with the Czech Technical University in Prague, through the Procardio-8 acquisition system with a sampling rate of 1 kHz. Processing of the UHF-ECG signals was performed in MATLAB environment, while statistical analysis was conducted on Microsoft Excel. The values of Area in Pixel were obtained considering all 19 patients and evaluating different electrode arrangements, in detail the six precordial leads, the 48 anterior leads, the 48 posterior leads and the 96 leads. In addition, an attempt to calculate the new spatial parameter by varying the energy threshold on the UHFQRS maps was considered, as it was considered appropriate to observe whether there was useful information on ventricular electrical conduction even at slightly lower frequencies. After calculating the mean and standard deviation values of Area in Pixel, it was obtained that these values are lower and with less variability in biventricular pacing configurations, supporting the idea that biventricular pacing is the most beneficial in CRT. Next, Area Under the Curve (AUC) values were calculated as percentages, showing that the Area in Pixel parameter has high discriminatory power mostly for electrode arrangements related to the six precordial leads and the 48 anterior leads. Thus, it is evident from this study that Area in Pixel is promising in the context of CRT evaluation, as this parameter provides information on the spatial distribution of the ventricular activation pattern.