P-wave beat-to-beat analysis to predict atrial fibrillation recurrence after catheter ablation

Introduction

Identification of patients prone to atrial fibrillation (AF) relapse after catheter ablation is essential for better patient selection and risk stratification.

Purpose

The current prospective cohort study aims to validate a novel P-wave index based on beat-to-beat (B2B) P-wave morphological and wavelet analysis designed to detect patients with low burden AF, as a predictor of AF recurrence within a year after successful catheter ablation.

Methods

12-lead ECG and 10-minute vectorcardiogram (VCG) recordings were obtained from 138 consecutive patients scheduled for AF ablation. Pre-ablation B2B P-wave index, along with standard P-wave indices, clinical scores and patients history and physical examination parameters were evaluated as AF recurrence predictors.

Results

Univariate analysis revealed that patients with higher B2B P-wave index had a two-fold risk for AF recurrence (HR: 2.35, 95% CI: 1.24–4.44, p: 0.010). Prolonged P-wave, interatrial block, early AF recurrence, female gender, heart failure history, previous stroke, and CHA2DS2-VASc score ≥2 were also found to be related to higher recurrence rate. Multivariate analysis of predictors that can be assessed before ablation revealed that B2B P-wave index, along with heart failure history and history of previous stroke or transient ischemic attack are independent predicting factors of AF relapse.

Conclusion

B2B P-wave morphology and wavelet analysis, is a promising, non-invasive technique, able to identify patients prone to AF recurrence after pulmonary veins ablation. Further studies are needed to assess the predictive value of B2B index with greater accuracy and evaluate a possible relationship with atrial substrate analysis.

Funding Acknowledgement

Type of funding sources: Public Institution(s). Main funding source(s): Hellenic Society of Cardiology

Beat-to-beat P-wave analysis outperforms conventional P-wave indices in identifying patients with a history of paroxysmal atrial fibrillation, during sinus rhythm

Introduction

Atrial fibrillation (AF) is the most common arrhythmia and is associated with high risk of morbidity and mortality. In many patients, AF is of episodic character (paroxysmal AF – PAF), which makes the identification of these patients during sinus rhythm (SR) challenging.

Purpose

The aim of the present study is to compare the performance of beat-to-beat P-wave analysis with P-wave indices used as predictors of PAF, such as P-wave duration, area, voltage, axis, terminal force in V1, inter-atrial block or orthogonal type, in identifying patients with history of PAF during sinus rhythm.

Methods

Standard 12-lead ECG and 10-minute orthogonal ECG recordings were obtained from 40 consecutive patients with short history of PAF under no antiarrhythmic medication and 60 age- and sex- matched healthy controls. The P-waves on the 10-minute recordings were analyzed on a beat-to-beat basis and classified as belonging to a primary or secondary morphology according to previous study. Wavelet transform used to further analyze P-wave orthogonal signals of main morphology on a beat-to-beat basis.

Results

38 out of 327 studied features were found to differ significantly among the two groups. These features were tested for their diagnostic ability and receiver operating characteristic curves were ploted. Only 3 of them performed adequetly, with an area under curve (AUC) above 0.65; Two of them came from morphology analysis (percentage of beats following main morphology in axis X and Y) and one from wavelet analysis (max energy in high frequency zone -Y axis). Among standard P-wave indices, P-wave area in lead II was the one with the highest AUC (0.64).

Conclusion

Novel indices derived from beat-to-beat analysis outperform stadard P-wave markers in identifying patients with PAF history during sinus rhythm.

Funding Acknowledgement

Type of funding sources: None. ROC curves of most significant featuresAUC characteristics of P-wave indices

P-wave beat-to-beat morphology analysis outperforms conventional P-wave indices in detecting patients with paroxysmal atrial fibrillation

Funding Acknowledgements

Type of funding sources: None.

Background

Atrial fibrillation (AF) - the most common sustained cardiac arrhythmia - while not a life-threatening condition itself, leads to an increased risk of stroke and high rates of mortality. Early detection and diagnosis of AF is a critical issue for all health stakeholders.

Purpose

The aim of this study is to compare the performance of standard P-wave indices with beat-to-beat P-wave morphological variability parameters in identifying patients with history of Paroxysmal Atrial Fibrillation (PAF).

Methods

Three-dimensional 1000Hz ECG digital recordings of 10 minutes duration were obtained from a total of 39 PAF patients and 60 healthy individuals. Following artifacts and ectopic beats removal, P‑wave morphology analysis was performed based on the dynamic application of the k‑means clustering process and main and secondary P-wave morphologies were identified. The percentage of P-waves following the main or the secondary morphology in each lead was calculated, as well as established indices such as Advanced Interatrial Block, P-wave duration, axis and area, P-wave Terminal Force in lead V1 and Orthogonal Leads Type 1, 2 or 3.

Results

9 out of 24 parameters studied, were found to be significantly different between the two groups. 7 of these indices were derived from morphology analysis and 2 from P-wave area. Logistic regression revealed that the percentage of P-waves allocated to main morphology in X axis performed better than all other indices in identifying patients with PAF history from healthy volunteers in terms of total accuracy and F1 measure.

Conclusion

P-wave beat-to-beat morphology analysis can identify PAF patients during normal sinus rhythm more efficiently than standard P-wave indices. Abstract Figure.

219A machine learning classification algorithm to detect patients with paroxysmal atrial fibrillation during sinus rhythm

Background

Atrial fibrillation (AF) - the most common sustained cardiac arrhythmia - while not a life-threatening condition itself, leads to an increased risk of stroke and high rates of mortality. Early detection and diagnosis of AF is a critical issue for all health stakeholders.

Purpose

The aim of this study is to identify P-wave morphology patterns encountered in patients with Paroxysmal AF (PAF) and to develop a classifier discriminating PAF patients from healthy volunteers.

Methods

Three-dimensional 1000Hz ECG signals of 5 minutes duration were obtained through the use of a Galix GBI-3S Holter monitor from a total of 68 PAF patients and 52 healthy individuals. Signal pre-processing, consisting of denoising, QRS auto-detection, and ectopic beats removal was performed and a signal window of 250ms prior to the Q-wave (Pseg) was considered for every single beat. P‑wave morphology analysis based on the dynamic application of the k‑means clustering process was performed. For those Pseg that were assigned in the largest cluster, the mean P-wave was computed. The correlation of every P-wave with the mean P-wave of the main cluster was calculated. In case that it exceeded a prespecified threshold, the P-wave was allocated to the main morphology. For the remaining P‑waves, the same approach was followed once again, and the secondary morphology was extracted (picture). The P-waves of the dominant morphology were further analyzed using wavelet transform, whereas time-domain characteristics were also extracted.

A Support Vector Machine (SVM) model was created using the Gaussian Radial Basis Function kernel and the forward feature selection wrapper approach was followed. ECGs were allocated to the training, internal validation, and testing datasets in a 3:1:1 ratio.

Results

The percentage of P-waves following the main morphology in all three leads was lower in PAF patients (91.2 ±7.3%) than in healthy subjects (96.1 ±3.5%, p = 0.02). Classification between the two groups highlighted 7 features, while the SVM classifier resulted in a balanced accuracy of 91.4 ± 0.2% (sensitivity 94.2 ± 0.3%, specificity 88.6 ± 0.1%)

Conclusion

An Artificial Intelligence based ECG Classifier can efficiently identify PAF patients during normal sinus rhythm.

Abstract Figure.

244An automated beat exclusion algorithm to improve beat-to-beat P-wave morphology analysis

Background

A manually beat-to-beat P-wave analysis has previously revealed the existence of multiple P-wave morphologies in patients with paroxysmal Atrial Fibrillation (AF) while on sinus rhythm, distinguishing them from healthy, AF free patients.

Purpose

The aim of this study was to investigate the effectiveness of an Automated Beat Exclusion algorithm (ABE) that excludes noisy or ectopic beats, replacing manual beat evaluation during beat-to-beat P-wave analysis, by assessing its effect on inter-rater variability and reproducibility.

Methods

Beat-to-beat P-wave morphology analysis was performed on 34 ten-minute ECG recordings of patients with a history of AF. Each recording was analyzed independently by two clinical experts for a total of four analysis runs; once with ABE and once again with the manual exclusion of ineligible beats. The inter-rater variability and reproducibility of the analysis with and without ABE were assessed by comparing the agreement of analysis runs with respect to secondary morphology detection, primary morphology ECG template and the percentage of both, as these aspects have been previously used to discriminate PAF patients from controls.

Results

Comparing ABE to manual exclusion in detecting secondary P-wave morphologies displayed substantial (Cohen"s k = 0.69) to almost perfect (k = 0.82) agreement. Area difference among auto and manually calculated main morphology templates was in every case <5% (p < 0.01) and the correlation coefficient was >0.99 (p < 0.01). Finally, the percentages of beats classified to the primary or secondary morphology per recording by each analysis were strongly correlated, for both main and secondary P-wave morphologies, ranging from ρ=0.756 to ρ=0.940 (picture)

Conclusion

The use of the ABE algorithm does not diminish inter-rater variability and reproducibility of the analysis. The primary and secondary P-wave morphologies produced by all analyses were similar, both in terms of their template and their frequency. Based on the results of this study, the ABE algorithm incorporated in the beat-to-beat P-wave morphology analysis drastically reduces operator workload without influencing the quality of the analysis.

Abstract Figure.

P322Understanding the multiple P-wave morphologies in paroxysmal atrial fibrillation, during sinus rhythm, using computer simulation

Background

Atrial Fibrillation (AF) is the most common atrial arrhythmia. The initiation and perpetuation of AF are related to atrial remodeling affecting the electrical and structural atrial characteristics. The beat-to-beat analysis of the P-wave morphology (PWM), during sinus rhythm (SR), revealed the existence of a secondary PWM, while the proportion of the P-waves which follow the secondary morphology is higher in patients with a history of paroxysmal AF (pAF). This observation has led to the hypothesis that the multiple PWM may be the result of a transient shift in the stimulus origin, possibly within the broader anatomical region of the sinoatrial (SA) node, and it is the atrial electrical remodeling that contributes to more frequent P-waves following a secondary morphology in patients with pAF.

Purpose

To better understand the pathophysiology of AF there is a need to link different levels of analysis, in order to interpret macroscopic observations, through a surface electrocardiogram, with changes occurring at cell and tissue level. Towards this direction, computational modeling can be used as it is a non-invasive and reproducible method of analyzing the electrical activity of the heart.

Methods

The CRN atrial model was used, and a two-dimensional geometry of the atrial architecture was considered, including the major anatomical structures, like Crista Terminalis, Pectinate Muscles and Pulmonary Veins. Using existing knowledge, the CRN model was adapted to describe the ionic properties of the atrial structures as well as the electrical remodeling occurring under pAF conditions. Several scenarios were considered related to the extent of the electrical remodeled tissue and Heart Rate (HR) values. The stimulation protocol was designed as 5 stimuli originated at a specific point within the SA node area whereas the sixth stimulus originated either at the same location or 1 mm far from the previous one. The temporal variations of the atrial activation as a result of the transient shift of the sixth stimulus origin were computed.

Results

In electrically remodeled tissue, the displacement of the excitation site within the SA node resulted in a significant increase of the differences in atrial activation compared to healthy tissue, and the greater the spatial extent of the remodeling the greater the differences in the completion of the electrophysiological processes. In addition, increased HR or HR variability led to the increase of the differences especially when electrical remodeling coexists.

Conclusions

The observed differences in atrial substrate activation can explain the increased number of P-waves that match a secondary PWM in pAF patients during SR, while a future perspective is to use PWM as a marker to estimate the electrical remodeling extent in the atrial tissue. These results underline the need to link the macroscopic findings to the suspected microscopic electrical activity in order to better understand the pathophysiology of AF.