Automated classification of human atrial fibrillation from intraatrial electrograms

An Evaluation of Phase Analysis to Interpret Atrial Activation Patterns during Persistent Atrial Fibrillation for Targeted Ablation

BACKGROUND

Phase analysis has been used to identify and localize atrial fibrillation (AF) sources for targeted ablation. We previously demonstrated that repetitive wannabe reentry (incomplete reentry) often generated an apparent stable rotor using phase analysis. The misinterpretation caused by phase analysis using atrial electrograms (AEGs) may result from detecting inaccurate time points at phase inversion (π to -π) in the instantaneous phase waveform converted from AEG. The purpose of this study was to evaluate the accuracy of phase analysis to detect atrial activations recorded from the high-density mapping of AF in patients with persistent and long-standing persistent (LSP) AF.

METHODS AND RESULTS

During open heart surgery, we recorded activation from both atria simultaneously using 512 electrodes in 7 patients with persistent and LSP AF. The phase analysis was compared to manual measurements during 4 s of data. For the accuracy of activation sequence maps, a successful recording site was defined as having ≤4 mismatched activation times during the 4 s. In all AF episodes, the accuracy of the phase analysis was only 82% of the total number of activation times due to either activation time differences (14.7%), under-sensing (2.7%), or over-sensing (0.6%). Only 67.9% of the total recording sites met the requirement of a successful recording site by phase analysis. In unsuccessful recording sites, AEG characteristics were relatively irregular cycle length (CL), complex AEG, and double potential AEG.

CONCLUSION

The phase analysis was less accurate in recording sites with a relatively irregular CL, complex AEG, or double potential AEG. As a result, phase analysis may lead to the misinterpretation of atrial activation patterns during AF. A visual review of the original AEG is needed to confirm the detected AF sources of phase analysis before performing targeted ablation.

Semi-supervised clustering of fractionated electrograms for electroanatomical atrial mapping

BACKGROUND

Electrogram-guided ablation procedures have been proposed as an alternative strategy consisting of either mapping and ablating focal sources or targeting complex fractionated electrograms in atrial fibrillation (AF). However, the incomplete understanding of the mechanism of AF makes difficult the decision of detecting the target sites. To date, feature extraction from electrograms is carried out mostly based on the time-domain morphology analysis and non-linear features. However, their combination has been reported to achieve better performance. Besides, most of the inferring approaches applied for identifying the levels of fractionation are supervised, which lack of an objective description of fractionation. This aspect complicates their application on EGM-guided ablation procedures.

METHODS

This work proposes a semi-supervised clustering method of four levels of fractionation. In particular, we make use of the spectral clustering that groups a set of widely used features extracted from atrial electrograms. We also introduce a new atrial-deflection-based feature to quantify the fractionated activity. Further, based on the sequential forward selection, we find the optimal subset that provides the highest performance in terms of the cluster validation. The method is tested on external validation of a labeled database. The generalization ability of the proposed training approach is tested to aid semi-supervised learning on unlabeled dataset associated with anatomical information recorded from three patients.

RESULTS

A joint set of four extracted features, based on two time-domain morphology analysis and two non-linear dynamics, are selected. To discriminate between four considered levels of fractionation, validation on a labeled database performs a suitable accuracy (77.6 %). Results show a congruence value of internal validation index among tested patients that is enough to reconstruct the patterns over the atria to located critical sites with the benefit of avoiding previous manual classification of AF types.

CONCLUSIONS

To the best knowledge of the authors, this is the first work reporting semi-supervised clustering for distinguishing patterns in fractionated electrograms. The proposed methodology provides high performance for the detection of unknown patterns associated with critical EGM morphologies. Particularly, obtained results of semi-supervised training show the advantage of demanding fewer labeled data and less training time without significantly compromising accuracy. This paper introduces a new method, providing an objective scheme that enables electro-physiologist to recognize the diverse EGM morphologies reliably.

Dynamic time warping applied to estimate atrial fibrillation temporal organization from the surface electrocardiogram

Atrial fibrillation (AF) is the most commonly diagnosed arrhythmia in clinical practice. However, the mechanisms responsible for its induction and maintenance still are not fully understood. To this respect, analysis of the electrical activity organization within the atria could play an important role in their proper interpretation. Although many algorithms to quantify AF organization from invasive electrograms can be found in the literature, a reduced number of indirect estimators from the standard ECG have been proposed to date. Furthermore, these surface methods can only yield a global AF organization assessment, blurring the possible information that each individual fibrillatory (f) wave may provide. To this respect, the present manuscript proposes a novel method for direct and short-time AF organization estimation from single-lead surface ECG recordings. Through the computation of morphological variations among f waves, the temporal arrhythmia organization is estimated. The f waves are individually extracted and delineated from the atrial activity signal, making use of a dynamic time warping approach. The proposed algorithm was tested on real AF surface recordings in order to discriminate atrial signals with different organization degrees, obtaining a diagnostic accuracy higher than 88%. In addition, its performance was validated by comparison with two temporal organization measures from invasive unipolar electrograms of both atria, providing statistically significant linear correlations between invasive and non-invasive estimates. As a consequence, new standpoints are opened through this work in the non-invasive analysis of AF, where the individualized study of each f wave could assess short-time AF organization, would improve the understanding of AF mechanisms and become useful for its clinical treatment.

An algorithm to measure beat-to-beat cycle lengths for assessment of atrial electrogram rate and regularity during atrial fibrillation

INTRODUCTION

Experimental models have demonstrated that atrial fibrillation (AF) may be due to one or more rapid drivers (source) producing AF. These drivers may be characterized by rapid and regular cycle lengths (CLs), producing fibrillatory conduction to the rest of the atria. The ability to reliably identify such drivers would be invaluable. The purpose of this study was to develop and validate a CL variability detection (CLVD) analysis capable of accurately determining beat-to-beat CLs of atrial electrograms (AEGs) during AF, and then to compare this analysis with dominant frequency (DF) analysis.

METHODS AND RESULTS

We analyzed 6 episodes of AF in 6 dogs (sterile pericarditis model) due either to a single, stable left atrial reentrant circuit, or unstable reentrant circuits causing fibrillatory conduction to the rest of the atria. During AF, AEGs were recorded simultaneously from 400 to 420 electrodes on both atria. CLs from over 20,000 AEGs were manually measured, and compared to CLs detected using both the CLVD and DF analyses. There was significant correlation between (1) CLs measured manually and the CLVD analysis (mean CL: correlation coefficient [CC]= 0.96, standard deviation [SD]: CC = 0.89); and (2) mean CL measured manually and the DF analysis (CC = 0.84). However, there was poor correlation between SD of CLs measured manually and the organization index (OI) by DF analysis (CC =-0.59).

CONCLUSION

The CLVD analysis was validated as being accurate for detecting both rate and degree of regularity of AEGs during AF, and more accurate than DF analysis.

Conduction velocity restitution of the human atrium--an efficient measurement protocol for clinical electrophysiological studies

Conduction velocity (CV) and CV restitution are important substrate parameters for understanding atrial arrhythmias. The aim of this work is to (i) present a simple but feasible method to measure CV restitution in-vivo using standard circular catheters, and (ii) validate its feasibility with data measured during incremental pacing. From five patients undergoing catheter ablation, we analyzed eight datasets from sinus rhythm and incremental pacing sequences. Every wavefront was measured with a circular catheter and the electrograms were analyzed with a cosine-fit method that calculated the local CV. For each pacing cycle length, the mean local CV was determined. Furthermore, changes in global CV were estimated from the time delay between pacing stimulus and wavefront arrival. Comparing local and global CV between pacing at 500 and 300 ms, we found significant changes in seven of eight pacing sequences. On average, local CV decreased by 20 ± 15% and global CV by 17 ± 13%. The method allows for in-vivo measurements of absolute CV and CV restitution during standard clinical procedures. Such data may provide valuable insights into mechanisms of atrial arrhythmias. This is important both for improving cardiac models and also for clinical applications, such as characterizing arrhythmogenic substrates during sinus rhythm.

Long-term biatrial recordings in post-operative atrial fibrillation

Although atrial fibrillation (AF) is a common complication of cardiac surgery, its pathophysiology remains unclear. The study of post-operative AF demands for the recording of cardiac electrical activity in correspondence of AF onset and progression. Long-term recordings in post-surgery patients could provide this information, but, to date, have been limited to surface signals, which precludes a characterization of the arrhythmic triggers and substrate. In this study we demonstrate the feasibility of a continuous long-term recording of atrial electrical activities from the right and left atria in post-surgery patients. Local atrial epicardial electrograms are acquired by positioning temporary pacing wires in the right and left atria at the end of the intervention, while three day recordings are obtained by a digital holter recorder, adapted to epicardial signal features. The capability of the system to map local atrial activity and the possibility to obtain quantitative information on atrial rate and synchronization from the processed epicardial signals are proven in representative examples. The quantitative description of local atrial properties opens new perspective in the investigation of post-surgery AF.

Evaluation of atrial fibrillation induced during anesthesia with fentanyl and pentobarbital in German Shepherd Dogs with inherited arrhythmias

OBJECTIVE

To determine the type of atrial fibrillation induced by use of 2 pacing protocols during fentanyl and pentobarbital anesthesia before and after administration of atropine and to determine the organization of electrical activity in the left and right atria during atrial fibrillation in German Shepherd Dogs.

ANIMALS

7 German Shepherd Dogs.

PROCEDURES

Extrastimulus and pacedown protocols were performed before and after atropine administration. Monophasic action potential spectral entropy and mean dominant frequency were calculated during atrial fibrillation.

RESULTS

Atrial fibrillation occurred spontaneously in 6 of 7 dogs. All 7 dogs had atrial fibrillation induced. Sustained atrial fibrillation occurred in 13 of 25 (52%) episodes induced by the extrastimulus protocol and in 2 of 12 episodes of atrial fibrillation induced by pacedown. After atropine administration, sustained atrial fibrillation did not occur, and the duration of the nonsustained atrial fibrillation (6 episodes in 2 dogs of 1 to 26 seconds) was significantly shorter than before atropine administration (25 episodes in 7 dogs of 1 to 474 seconds). The left atrium (3.67 +/- 0.08) had lower spectral entropy than the right atrium (3.81 +/- 0.03), indicating more electrical organization in the left atrium. The mean dominant frequency was higher in the left atrium in 3 dogs.

CONCLUSIONS AND CLINICAL RELEVANCE

Atrial fibrillation developed spontaneously and was induced in German Shepherd Dogs under fentanyl and pentobarbital anesthesia. Electrical activity was more organized in the left atrium than in the right atrium as judged by use of spectral entropy.

An automatic system for the analysis and classification of human atrial fibrillation patterns from intracardiac electrograms

This paper presents an automatic system for the analysis and classification of atrial fibrillation (AF) patterns from bipolar intracardiac signals. The system is made up of: 1) a feature-extraction module that defines and extracts a set of measures potentially useful for characterizing AF types on the basis of their degree of organization; 2) a feature-selection module (based on the Jeffries-Matusita distance and a branch and bound search algorithm) identifying the best subset of features for discriminating different AF types; and 3) a support vector machine technique-based classification module that automatically discriminates the AF types according to the Wells' criteria. The automatic system was applied on 100 intracardiac AF signal strips and on a selection of 11 representative features, demonstrating: a) the possibility to properly identify the most significant features for the discrimination of AF types; b) higher accuracy (97.7% using the seven most informative features) than the traditional maximum likelihood classifier; and c) effectiveness in AF classification also with few training samples (accuracy = 88.3% with only five training signals). Finally, the system identifies a combination of indices characterizing changes of morphology of atrial activation waves and perturbation of the isoelectric line as the most effective in separating the AF types.

Assessment of the dynamics of atrial signals and local atrial period series during atrial fibrillation: effects of isoproterenol administration

Background

The autonomic nervous system (ANS) plays an important role in the genesis and maintenance of atrial fibrillation (AF), but quantification of its electrophysiologic effects is extremely complex and difficult. Aim of the study was to evaluate the capability of linear and non-linear indexes to capture the fine changing dynamics of atrial signals and local atrial period (LAP) series during adrenergic activation induced by isoproterenol (a sympathomimetic drug) infusion.

Methods

Nine patients with paroxysmal or persistent AF (aged 60 ± 6) underwent electrophysiological study in which isoproterenol was administered to patients. Atrial electrograms were acquired during i) sinus rhythm (SR); ii) sinus rhythm during isoproterenol (SRISO) administration; iii) atrial fibrillation (AF) and iv) atrial fibrillation during isoproterenol (AFISO) administration. The level of organization between two electrograms was assessed by the synchronization index (S), whereas the degree of recurrence of a pattern in a signal was defined by the regularity index (R). In addition, the level of predictability (LP) and regularity of LAP series were computed.

Results

LAP series analysis shows a reduction of both LP and R index during isoproterenol infusion in SR and AF (R_SR = 0.75 ± 0.07 R_SRISO = 0.69 ± 0.10, p < 0.0001; R_AF = 0.31 ± 0.08 R_AFISO = 0.26 ± 0.09, p < 0.0001; LP_SR = 99.99 ± 0.001 LP_SRISO = 99.97 ± 0.03, p < 0.0001; LP_AF = 69.46 ± 21.55 LP_AFISO = 55 ± 24.75; p < 0.0001). Electrograms analysis shows R index reductions both in SR (R_SR = 0.49 ± 0.08 R_SRISO = 0.46 ± 0.09 p < 0.0001) and in AF (R_AF = 0.29 ± 0.09 R_AFISO = 0.28 ± 0.08 n.s.).

Conclusions

The proposed parameters succeeded in discriminating the subtle changes due to isoproterenol infusion during both the rhythms especially when considering LAP series analysis. The reduced value of analyzed parameters after isoproterenol administration could reflect an important pro-arrhythmic influence of adrenergic activation on favoring maintenance of AF.

A method for quantifying atrial fibrillation organization based on wave-morphology similarity

A new method for quantifying the organization of single bipolar electrograms recorded in the human atria during atrial fibrillation (AF) is presented. The algorithm relies on the comparison between pairs of local activation waves (LAWs) to estimate their morphological similarity, and returns a regularity index (rho) which measures the extent of repetitiveness over time of the detected activations. The database consisted of endocardial data from a multipolar basket catheter during AF and intraatrial recordings during atrial flutter. The index showed maximum regularity (rho = 1) for all atrial flutter episodes and decreased significantly when increasing AF complexity as defined by Wells (type I: rho = 0.75 +/- 0.23; type II: rho = 0.35 +/- 0.11; type III: rho = 0.15 +/- 0.08; P < 0.01). The ability to distinguish different AF episodes was assessed by designing a classification scheme based on a minimum distance analysis, obtaining an accuracy of 85.5%. The algorithm was able to discriminate among AF types even in presence of few depolarizations as no significant rho changes were observed by reducing the signal length down to include five LAWs. Finally, the capability to detect transient instances of AF complexity and to map the local regularity over the atrial surface was addressed by the dynamic and multisite evaluation of rho, suggesting that our algorithm could improve the understanding of AF mechanisms and become useful for its clinical treatment.

Suppression of atrial tachyarrhythmias by pacing

INTRODUCTION

Treatment of atrial tachyarrhythmias (ATs) remains difficult in many patients. Accordingly, new therapeutic approaches for AT suppression are evaluated. Atrial pacing may prevent ATs by modifying the electrophysiologic conditions required for sustained ATs.

METHODS AND RESULTS

New pacing algorithms for prevention of AT are aimed at permanent overdrive suppression of arrhythmic activity, reduction of dispersion of atrial refractoriness produced by short-long cycles, more aggressive overdrive pacing after spontaneous sinus conversion to prevent early reinitiation of ATs, and prevention of inadequate rate decay in patients with vagally induced ATs. AT prevention may be achieved by dedicated atrial pacing sites, e.g., pacing at the insertion of Bachmann's bundle or biatrial pacing, which compensates for interatrial conduction delay. Preexciting regions of critical conduction delay, pacing at the triangle of Koch or coronary sinus os, and dual-site right atrial pacing have shown antiarrhythmic effects. Atrial preventive pacing and pharmacologic treatment may work synergistically in the concept of hybrid therapy. To prevent atrial electrical remodeling, early termination of AT seems desirable. This may be achieved by implanted devices that automatically detect ATs and provide atrial antitachycardia pacing for organized ATs. Initial studies showed that regular AT can automatically be terminated in approximately 50% of treated episodes.

CONCLUSION

Pacing for prevention of AT and termination of organized AT episodes may become important steps within the concept of hybrid therapy of AT. However, their clinical efficacy and optimal patient selection remain to be evaluated in prospective, well-designed clinical trials.

Principal component analysis and cluster analysis for measuring the local organisation of human atrial fibrillation

The distribution of atrial electrogram types has been proposed to characterise human atrial fibrillation. The aim of this study was to provide computer procedures for evaluating the local organisation of intracardiac recordings during AF as an alternative to off-line manual classification. Principal component analysis (PCA) reduced the data set to a few representative activations, and cluster analysis (CA) measured the average dissimilarity between consecutive activations of an intracardiac signal. The data set consisted of 106 bipolar signals recorded on 11 patients during electrophysiological studies for catheter ablation. Performances of PCA and CA in distinguishing between organised (type I) and disorganised (type II/III, Wells criteria) were assessed, in comparison with manual reading, by evaluating the predictive parameters of the classification analysis. Both methods gave high accuracy (92% for PCA and 89% for CA), confirming the feasibility of on-line characterisation of AF. Sensitivity was lower than specificity (81% against 98% for PCA, and 77% against 97% for CA), with seven out of eight misclassifications of PCA in common with CA. Differences between manual and computer analysis may be related to the higher resolution of PCA and CA in the measurement of the organisation of atrial activations. These procedures are suitable for providing automatic (by CA) or semi-automatic (by PCA) measures of the extent of local organisation of AF in the pre-ablation treatment phase.

Prevalence, characteristics and clinical implications of regular atrial tachyarrhythmias in patients with atrial fibrillation: insights from a study using a new implantable device

OBJECTIVES

This study prospectively analyzed atrial tachyarrhythmia (AT) organization and antitachycardia pacing (ATP) success in patients with an implanted device for AT therapy.

BACKGROUND

In patients with atrial fibrillation (AF), the incidence of regular, slow ATs amendable by ATP is unknown.

METHODS

Forty patients with previously documented AT (70% with AF) received a new pacemaker with atrial electrogram (AEG) storage and atrial ATP capabilities for standard pacing indications. The AEGs acquired during the first month (study phase 1) were classified into high (type I), intermediate (type II) and low (type III) degrees of organization. Atrial ATP was then activated, and treated AT episodes were retrieved three and six months after implantation (study phase 2).

RESULTS

Of 824 AEGs retrieved before ATP activation (study phase 1), 351 (43%) were classified as type 1, 47% as type II and 10% as type III. Episodes of AT starting as type I (35%) and type II or III (65%) maintained their type over 1 min in 73%. All patients with an exclusive history of AF also showed type I AEGs. In 361 subsequently treated AT episodes (study phase 2), ATP was successful in 62% of type I and 34% of type II episodes, but not in type III (p < 0.0001).

CONCLUSIONS

The majority of patients with a history of AF show not only disorganized but also highly organized AT episodes, which can be successfully terminated by ATP.

Linear and non-linear analysis of atrial signals and local activation period series during atrial-fibrillation episodes

Linear and non-linear indexes for the characterisation of the dynamics in atrial signals (AS) and local atrial period (LAP) series are assessed in different atrial fibrillation (AF) episodes as defined by Wells. Parameters include the linear index obtained from the cross-correlation function (CCF) between ASs and the non-linear synchronisation (S) index based on the mutual corrected conditional entropy (MCCE). Regularity (R) was computed on single-lead AS. In addition, the level of predictability (LP) and the regularity of LAP series were computed. It was found that the level of synchronisation between ASs decreased passing from type-I to type-II AF when using linear (CCF: 0.90 +/- 0.10 against 0.44 +/- 0.18; p<0.001) and non-linear (S: 0.22 +/- 0.10 against 0.05 +/- 0.03; p<0.001) indexes. The regularity index (in normal sinus rhythm (NSR): 0.30 +/- 0.08; in AF-I: 0.19 +/- 0.10; in AF-II: 0.09 +/- 0.02; NSR against AF-I p<0.001; AF-I against AF-II p<0.001) and level of predictability (in NSR: 65 +/- 18; in AF-I: 27 +/- 13; in AF-II 7 +/- 6; NSR against AF-I p<0.001; AF-I against AF-II p<0.001) significantly decreased in the LAP series passing from NSR to AF-II. The proposed parameters succeeded in discriminating the different dynamics which characterised AS and LAP series during different kinds of AF episodes.