Comparison of Activation Times Estimation for Potential-Based ECG Imaging

Abstract

Activation times (AT) represent the sequence of cardiac depolarization and are one of the most important parameters of cardiac electrical activity. However, estimation of ATs is challenging due to multiple sources of noise. If ATs are estimated from signals reconstructed using electrocardiographic imaging (ECGI), additional problems can arise due to over-smoothing or ambiguities in the inverse problem. Resulting AT maps can show falsely homogeneous regions or artificial lines of block. As ATs are not only important clinically, but are also used for evaluation of ECGI, it is important to understand where these errors come from.We present results from a community effort to compare AT estimation methods on a common dataset of simulated ventricular pacings. ECGI reconstructions were performed in terms of transmembrane voltages as well as epi-endo and pericardial potentials, all using 2nd-order Tikhonov and 6 regularization parameters. ATs were then estimated by the participants and compared to the truth.While the pacing site had the largest effect on AT correlation coefficients (CC), there were also differences between methods and source models that were poorly reflected in CCs. Results indicate that artificial lines of block are most severe for purely temporal methods. Compared to the other source models, ATs estimated from transmembrane voltages are more precise and less prone to artifacts.

Inference of ventricular activation properties from non-invasive electrocardiography

“…We adopted the mean ± standard deviation of Pearson's correlation coefficient ( Bear et al., 2018 ; Schaufelberger et al., 2019 ; Serinagaoglu Dogrusoz et al., 2019 ) as a measure of disagreement between our inference predictions and the ‘target data’ ( Table 2 ).…”

A Patchwork Method to Improve the Performance of Current Methods for Solving the Inverse Problem of Electrocardiography

“…That is, the detection of focal activity, and epicardial breakthrough sites can be missing or misplaced. The problem is seen not only with clinical data [5] but also in torso tank experiments [1], [4] and in simulated data [7], [8] where forward models are well defined, suggesting that the issue lies with the inverse methods themselves.New inverse methods are therefore needed to improve the performance and the accuracy of ECGI in the reconstruction of focal activity, and epicardial breakthrough sites.…”

A Patchwork Inverse Method in Combination with the Activation Time Gradient to Detect Regions of Slow Conduction in Sinus Rhythm

“…ECGI has successfully been used to provide electroanatomic mapping of arrhythmogenic substrates [2] and detect responders to cardiac resynchronization therapy [3]. However, recent validation studies have demonstrated that current implementations may detect artificial lines of block that are not seen in contact electrograms [4][5][6]. This raises the question whether ECGI can detect and locate slow conduction zones when they are present.…”