Exploratory and validation phases
The first 546 consecutive patients provided dataset of AP localization and QRS morphology types in 12 leads including 509 ECGs with full preexcitation and 547 ECGs with baseline preexcitation that were furnished to the WPW24.com database. Distribution of all APs in that cohort is presented in Figure 1 and Table 1 . Illustrative examples of how QRS polarity in a single ECG lead impacts distribution of APs along mitral and tricuspid annuli is presented inFigure 5 and 6 . Depending on the AP type/localization there were 2 to 5 optimized steps for ECG analysis. To reach the final diagnosis/distribution map a mean 3.0 +/-steps per ECG were needed.
For validation, the dataset of 260 ECGs were used, 49% of these ECGs were obtained prospectively, after the exploratory phase. The remaining 51% were obtained retrospectively albeit from a new centre that did not participate in the exploratory phase. Distribution of all APs in the exploratory and validation phases is presented in Figure 1 andTable 1 – no differences in the distribution of APs between the exploratory and validation datasets were found (p = 0.263).
To compare the exploratory cohort with the data from the literature a simplified categorization of AP regions had to be applied since the categorization and regions in every study are different, sometimes poorly defined and often with overlap of different regions among studies, making direct detailed comparison impossible (Supplementary Table 1 ). However, when all AP were grouped into four major categories (right free wall, left free wall, right posteroseptal and other septal) no difference in the distribution of AP among the exploratory cohort and the pooled cohort from the seven published studies (n = 1252), was present, p = 0.24.