Funding Acknowledgements Type of funding sources: Foundation. Main funding source(s): FWO-Flanders, KU Leuven internal starting grant Introduction and Purpose Cardiac electrograms (EGMs) are one of the most important recordings obtained during electroanatomical voltage mapping and lie at the basis for planning most clinical electrophysiological interventions. Despite its widespread use, the relation of EGM shape and amplitude to the underlying excitation patterns and properties of cardiac tissue is not completely understood. Recent clinical studies [1] have provided important new guidelines on the relation between EGM amplitudes and the thickness of myocardial walls. The aim of this study is to quantify the effect of the wall thickness on EGM amplitudes and duration using analytical and in-silico approaches. Methods We study bipolar EGMs both in-silico and analytically in a homogeneous slab of cardiac tissue (70 x 70 x L mm), where L = 2, 5, or 10 mm, with parallel fiber direction. Simulations were performed using the cardiac electrophysiology simulator openCARP [2]. Cardiac cells were described by the ten Tusscher-Panfilov 2006 model (TP06) [4] with epicardial tissue parameters. A plane wave propagating along the fiber direction was initiated. The extracellular voltage at 147 points arranged in a hemisphere around a point was measured to study the effect of bipolar electrode orientation (see Fig. 1A [3]). In addition, we developed an analytical approach to obtain an EGM, using an equivalent dipole representation of the depolarization wavefront and analytical evaluation of the corresponding integrals. Results Fig. 1B and 1C show the dependency of the EGM properties on the electrode orientation, as represented by the angles α (incidence angle) and β (angle between electrode and propagation direction) [3]. Solid lines represent data from a state-of-the-art numerical methodology, the dashed lines show our analytical estimations. Both the peak-to-peak amplitude and EGM width are well approximated by our theory for all orientations of the electrodes. Fig. 2 shows how the EGM is influenced by the myocardial wall thickness L. Both the amplitude and the duration are in good agreement with our theory. We observe that the amplitude as well as the width increase with the slab thickness, confirming the result in [1] but also delivering an accurate analytical expression for this change. It may thus allow to discriminate effects of thickness and other factors affecting the EGMs, such as substrate abnormalities, for example. Conclusion We developed an analytical approach which can correctly describe the amplitude, duration, and shape of the depolarization part of the EGM. Our theory agrees with the previous in-silico and clinical studies on the influence of catheter orientation [3,5], and wall thickness [1,3]. Subsequent work in this direction is expected to provide better guidelines for clinical interpretation of EGMs, accounting for the effects of the thickness of myocardial wall in the characterization of the substrate of cardiac arrhythmias.

Funding Acknowledgements Type of funding sources: None. Background Torsade de Pointes (TdP) is a ventricular tachycardia that is still incompletely understood. While the origin of the arrhythmia is generally accepted to be caused by triggered activity, the perpetuation of such an episode is still under debate. Purpose We compare the difference in reentrant activity between self- and non-terminating(ST and NT) TdP episodes. In addition, we analyzed the distribution of the triggered activity for preferred locations. Methods A total of 5 dogs with Chronic Atrioventricular Block developed 54 episodes of TdP. These episodes were examined for reentry loops and focal sources. First, to detect loops, we used Directed Graph Mapping, a novel mapping technique based on principles from network theory. Subsequently, the number of simultaneous loops and the presence of (bi-)ventricular reentry was compared between the ST and NT groups. Second, focal sources were accumulated into heatmaps. The distribution of sources was compared to a randomly distributed model for the detection of preferred locations. Results Concerning reentry, during the time at which loops are present, a typical ST episode shows an average of 1.33(±0.66) reentry loops present at one time compared to an average of 2.72(±1.48) for a typical NT episode. Looking at the (bi-)ventricular reentry, only 1 of the 4 ST episodes (25%) shows the presence of this type of reentry and only 7.12 % of the total reentry duration, while all 9 NT episodes (100%) develop (bi-)ventricular reentry to a greater degree than in the one ST episode (Between 10.10% and 69.62% of their total reentry duration). Regarding the triggered activity, we see locations in the heatmaps containing up to 23% of all detected focal sources in the entire system. Compared to a random distribution of focal sources, the chances to get as many sources in one location as shown in the data are close to 0%. Conclusion Reentry in NT TdP differentiates itself from the reentry in ST cases on the basis of the number of simultaneous loops in the system and the presence of (bi-)ventricular reentry. Triggered activity in CAVB dogs occurs at preferred locations in the ventricles. Therefore, ablating these preferred regions could be a possible strategy to prevent TdP.