Europace Advance Access published online on November 12, 2008
Europace, doi:10.1093/europace/eun307
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
CLINICAL RESEARCH
Evaluation of spatiotemporal organization of persistent atrial fibrillation with time- and frequency-domain measures in humans
1 Department of Cardiology, Rikshospitalet University Hospital, Sognsvannsveien 20, 0027 Oslo, Norway; 2 Division of Cardiology, Department of Medicine, Johns Hopkins University School of Medicine, Baltimore, MD, USA
Manuscript submitted 6 August 2008. Accepted after revision 21 October 2008.
* Corresponding author. Tel: +47 2307 2016, Fax: +47 2307 3917, Email: rfb123{at}hotmail.com
 |
Abstract |
|---|
 Top Abstract IntroductionMethodsResultsDiscussionConclusionFundingAcknowledgementsReferences |
|---|
Aims: Areas with complex fractionated electrograms are commonly targeted during ablation of persistent atrial fibrillation (AF). These signals are, however, found in most sampled areas of the left atrium (LA), implying the need for further differentiation.
Methods and results: Electrograms were recorded over 60 s at eight different LA endocardial sites in 10 patients with persistent AF, using a fully automated algorithm. These were analysed in sequential 2 s segments for activity, mean amplitude, continuous activity percentage, and dominant frequency (DF). All three time-domain measures differed significantly between the LA sites (P < 0.001), whereas DF did not. Activity, continuous activity percentage, and mean activity–amplitude were highest in the mid-coronary sinus and lowest on the posterior wall. In a pairwise analysis, there were significant differences in activity between all locations (P < 0.001–0.044). To visualize the spatiotemporal activity patterns, activity was plotted against amplitude. This revealed distinct activity patterns with large intra- and inter-individual differences.
Conclusion: There are significant activity gradients and distinct activity patterns within the LA in humans with persistent AF. Further work is required, however, to determine whether these findings signify areas with different roles and importance in AF maintenance.
Key Words: Atrial fibrillation, Mapping, Electrograms, Spectral analysis, Catheter ablation
|
Introduction |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionConclusionFundingAcknowledgementsReferences |
|---|
Several mechanisms are considered to be involved in the perpetuation of atrial fibrillation (AF), such as incessantly firing foci and/or multiple re-entrant wavelets in the left atrium (LA).1
,2 Persistence of AF also results from electrical and structural remodelling of the LA due to AF itself.3 Konings et al.4 demonstrated using high-density epicardial mapping that fractionated atrial potentials were associated with slow conduction, functional conduction block, and pivot points. Nademanee et al.5,6 reported an approach targeting such sites with complex fractionated atrial electrograms (CFAEs), which resulted in AF termination and long-term maintenance of sinus rhythm. Others, however, have not obtained comparable results due in large part to differences in CFAE definition and interpretation.7,8 Even so, many of the approaches for persistent AF involves some sort of CFAE ablation or electrical defragmentation.9–11
To facilitate and enhance the precision of AF substrate mapping, automated software recognition of CFAEs has been developed (e.g. CFAE Software Module, Carto XP System, Biosense Webster, Diamond Bar, CA, USA and Fractionation Mapping Module, Ensite System, St Jude Medical Inc., St Paul, MN, USA). However, in a recent study, exploring the first of these algorithms, as much as 80% of the LA endocardial surface was defined as CFAE sites.12 Similar results were found in a study evaluating the second algorithm.13 This is also in accordance with the findings of Jaïs et al.,14 who in an earlier study mapped the regional distribution of complex electrical activity and found it in most LA regions. These findings differ, at least to some extent, from the more clustered distribution of CFAE complexes reported by Nademanee et al.5 In this study, many of the cases required only a limited number of RF applications for AF termination. One reason for this may be that the operator's visual interpretation included additional and important facets of signal dynamics not grasped by the automated algorithm.
In order to explore other modalities to visualize CFAE sites and their dynamics and to investigate whether it was possible to further differentiate these areas, we designed an automated software algorithm for AF electrogram (EGM) evaluation to be used both on- and off-line.
|
Methods |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionConclusionFundingAcknowledgementsReferences |
|---|
Study patients, instrumentation, and mapping
The study sample consisted of 10 patients referred to our institution for catheter ablation of persistent, symptomatic drug-refractory AF. None of the patients had undergone a prior AF ablation procedure. All patients were on either amiodarone or sotalol, and they were maintained on their respective anti-arrhythmic medications for at least 3 months post-ablation. Table 1 summarizes their clinical characteristics. The study protocol was approved by our institutional review board. Before the AF ablation procedure, each patient gave written informed consent. A history, physical examination, electrocardiogram, and echocardiogram were obtained. Patients were given oral anticoagulation at least 1 month before the procedure and underwent transoesophageal echocardiography on the day of ablation to exclude LA thrombi.
|
Patients were studied under moderate conscious sedation, heparinized, and instrumented, as described previously.12
,15,16 A three-dimensional LA surface reconstruction was created using the Carto XP electroanatomical mapping system (Biosense Webster) and an 8 mm mapping and ablation catheter (Navistar, Biosense Webster).
Electrogram sampling and analysis
Bipolar endocardial EGMs were obtained from the antrum of the right and left superior and inferior pulmonary veins (RSPVa, RIPVa, LSPVa, LIPVa, respectively), the central portions of the roof (RO), posterior wall (PW), atrial septum (AS), and the floor—as reflected by the mid-segment of the coronary sinus (CSm; 5 to 6 o'clock position in the LAO view). The catheter used in this location was a deflectable 6F catheter with an electrode size/spacing of 1/2-5-2 mm (C.R. Bard, Lowell, MA, USA). The signals were bandpass-filtered with cutoff values at 30 and 400 Hz. The mapping catheter was kept at each location for at least 60 s. Stable catheter position and endocardial contact were ensured by fluoroscopic visualization of the catheter and by inspection of the position of the catheter tip icon displayed in the three-dimensional map. The CSmid recordings were visually inspected with respect to ventricular far-field contamination, which was not observed. The reason for the consistency of this finding may have been the use of a long sheath via the right femoral vein, pushing this part of the CS catheter towards the ‘roof’ of the CS. Electrograms recorded at each site were stored on a standard electrophysiology recording system (EP Med, C.R. Bard) and later transferred to an in-house programmed software (LabVIEW, National Instruments, Austin, TX, USA) for an entirely automated time- and frequency-domain analysis of the spatiotemporal signal characteristics. The reason for this was three-fold: because of the large amount of data; to avoid any selection bias and operator-dependent interpretation; and to enable on-line use.
Previous work has shown that EGM segments <6 s were required to obtain the resolution necessary to observe the varying states of organization of the AF signal.17 Although preparatory analyses showed nearly identical results when 2 and 4 s segments were used, 2 s segments were chosen to facilitate the detection of areas with a high degree of temporal variability and to allow a more rapid exploration of the LA when used on-line.
Figure 1 illustrates how each of the 30 consecutive 2 s segments of the bipolar endocardial LA EGMs was processed. First, the signal was rectified. All peaks of the bipolar EGM deflections that exceeded 50 µV (0.05 mV) were identified and tagged. The general noise level in all recordings was below 5 µV. Secondly, if the interval between two or more successive (tagged) deflections was <50 ms, this segment was defined as continuous activity, and the duration was marked with a yellow underscore. Based on this, three time-domain measures were calculated. These were (i) activity, defined as the number of deflections >50 µV/s; (ii) continuous activity percentage, defined as the continuous activity duration in percentage of each 2 s segment; and (iii) mean activity amplitude, defined as the mean of the deflection amplitudes >50 µV in each 2 s segment. Thirdly, the power spectrum of each successive 2 s segment was calculated by the fast Fourier transformation (FFT), and the dominant frequency (DF) was automatically detected and displayed. The selection of these measures was based on previous work by several authors.5,14,18,19
|
Before applying the 2048-point FFT, each signal segment was low-pass filtered with a 20 Hz cutoff, multiplied with a Hanning window, and zero-padded to obtain a frequency resolution of 0.1 Hz.20
,21 The frequency with the highest power had to be >10% greater than the second highest to be designated dominant.
Statistical analysis
Data are presented as counts, percentage, mean ± standard deviation (SD), median, interquartile range (IQR), and coefficient of variation (CV; defined as IQR/median) where appropriate. Because the EGM data were not normally distributed, all analyses were performed with non-parametric methods. Friedman's test was applied to estimate overall significance, and Wilcoxon's rank sum test was used for the post hoc pairwise comparisons. P-values were adjusted for multiple comparisons with Holm's sequential Bonferroni method. The differences in temporal variation were evaluated with the two-sample Kolmogorov–Smirnov test.
Estimates (e.g. IQR or DF) based on less than five observations were not reported. Statistical analyses were performed using the Stata 10 statistical software (Stata Corporation, College Station, TX, USA). All tests were two-tailed, and adjusted P-values <0.05 were considered significant.
|
Results |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionConclusionFundingAcknowledgementsReferences |
|---|
In the 10 study subjects with persistent AF, a total of 78 adequate 60 s recordings were acquired from the eight targeted endocardial LA locations, resulting in 2340 2 s recording segments for analysis. Because the time- and frequency-domain measures obtained in the antral regions of the superior and inferior pulmonary veins on either side did not differ significantly, the data obtained from RSPVa and RIPVa and from LSPVa and LIPVa were combined for the groupwise analyses of spatial gradients and temporal variation and denoted RPVa and LPVa. The visualization and analyses of the spatiotemporal activity patterns, in contrast, were performed for each patient at each location. Two patients presented to the laboratory in sinus rhythm (after previous cardioversions). In these, AF was induced by atrial burst pacing.
Spatial gradients
Overall, there was a significant difference in activity (i.e. deflections >50 µV/s) recorded from RPVa, RO, PW, CSm, AS, and LPVa (P < 0.001). The activity was highest in CSm, followed by AS, RO, LPVa, RPVa, and PW [25 (12–32), 20 (4–34), 18 (8–24), 12 (6–23), 9 (4–14), and 6.5 (1–14) deflections/s, respectively] (values are given as median and IQR). In a pairwise analysis, there were significant differences in activity between all locations compared (P < 0.001–0.044) (Figure 2A).
|
Likewise, there was overall a significant difference among the continuous activity percentages recorded from RPVa, RO, PW, CSm, AS, and LPVa (P < 0.001). The continuous activity percentage was greatest in CSm, followed by RO, AS, LPVa, RPVa, and PW [36.5 (17–51.5), 30 (8–46), 28 (3–59), 18 (4–40.5), 10.5 (2–22), and 7 (0–20.5)%, respectively]. In a pairwise analysis, there were significant differences in activity between most locations (P < 0.001–0.012), except between RO and AS and between CSm and AS (P = 0.683 and 0.107, respectively) (Figure 2B).
Furthermore, there was a significant difference between the activity amplitudes recorded from RPVa, RO, PW, CSm, AS, and LPVa (P < 0.001). The amplitude was greatest in CSm, followed by AS, RO, LPVa, RPVa, and PW [133.5 (95.5–172.5), 103 (68–132), 95 (74–128), 92 (72–119.5), 82 (71–97), and 76 (62–93) µV, respectively]. In a pairwise analysis, the amplitudes were significantly different between most locations (P < 0.001), except between RO and AS, RO and LPVa, and between AS and LPVa (P = 0.781, 0.282, and 0.309, respectively) (Figure 2C).
For DF, there was overall no significant difference between those recorded from RPVa, RO, PW, CSm, AS, and LPVa [5.2 (4.5–5.9), 5.3 (4.7–5.8), 5.0 (4.6–5.6), 5.1 (3.9–5.8), 5.3 (4.1–6.1), and 5.2 (4.2–6.1)Hz, respectively; P = 0.906] (Figure 2D).
Temporal variation
In order to assess temporal variability in activity, continuous activity percentage, mean activity amplitude, and DF among different sites, we computed the CV (defined as IQR/median) (data not shown).
For activity, the CV ranged from 80% in CSm to 200% on PW. The corresponding figures for continuous activity percentage were 95% (CSm) to 293% (PW) and for activity amplitude 32% (RPVa) to 62% (AS). In a pairwise analysis, the temporal variation in activity was significantly different between all locations compared (P = <0.001–0.008), whereas for continuous activity percentage, it was significant between all locations (P = 0.002–0.008), except between RO and CSm. In addition, the variation in amplitude was significant between all locations compared (P = 0.002–0.040).
For DF, the CV ranged from 20 to 39%. In a pairwise analysis, the temporal variation in DF was not significantly different between any of the sites compared, except between RO and PW (P = 0.008).
Besides the large differences in temporal variability between the sites when compared groupwise, there were also large disparities within the study subjects with regard to which sites having the smallest and largest temporal variability.
Spatiotemporal activity patterns
To further visualize these differences in spatiotemporal activity patterns within the LA, activity was plotted against mean activity amplitude for each of the 30 consecutive 2 s recording segments for each study subject by location (AA plot).
Figure 3 shows a typical example of recordings from one of the study subjects obtained at RSPVa and LSPVa. Figure 3A shows the first 10 s of these two recordings. In Figure 3B, activity is plotted against mean activity amplitude for each of the 30 consecutive 2 s recording segments. Note the distinctly different spatiotemporal activity patterns recorded at these two LA sites. The recording from LSPVa (black circles) shows a clustered pattern, whereas that from RSPVa (red circles) is much more scattered. Figure 3C shows the differences in temporal variation of the continuous activity percentage throughout the 60 s recordings; whereas Figure 3D shows the corresponding data for DF.
|
Figure 4 shows the AA plots for all the 10 study subjects by LA location, demonstrating highly different activity patterns. At some locations, there were minimal variations in both activity and amplitude (e.g. in LSPVa in study subject no. 1, on AS in no. 2, and in CSm in no. 8). At some sites, the activity varied more than the amplitude (e.g. in RIPVa in no. 9, in LSPVa in no. 3, and in RO in no. 10). At other locations, there were considerable variations in both measures (e.g. in CSm in no. 1 and in RIPVa in no. 7).
|
To obtain a single measure for the degree of variation in the AA plot, the square root of the product of the CV values for activity and amplitude was calculated at each site for each patient and designated AA variation index. The AA variation index is shown in the upper left corner of each graph in Figure 4. This index differed by a factor of 2–8 between the LA sites when studied groupwise, whereas it differed by a factor of 1–7 within the study subjects. The variance in activity and amplitude within each subject was tested between the sites with highest and lowest AA variation indices. For activity, the variances were significantly different in eight subjects (80%) (P = 0.002–0.036), whereas for amplitude, they were significantly different in nine subjects (90%) (P = 0.003–0.016).
|
Discussion |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionConclusionFundingAcknowledgementsReferences |
|---|
There are three main findings in this study. First, there are significant spatial activity gradients within the LA in patients with persistent AF. Secondly, there is a large degree of temporal variation of EGM characteristics at some locations, although it is much less in others. Thirdly, and possibly most important, is the potential of AA plot as an additional visualization mode to facilitate the differentiation of sites with fibrillatory activity.
Spatial gradients
There was an explicit activity gradient within the LA between the floor (as reflected by CSm) and the roof, septum, the pulmonary vein antral regions, and the PW. A left-to-right atrial gradient has previously been shown for DF in both experimental models and humans.19,22,23 In humans, however, no differences were found within the LA. We observed similar behaviour of DF, which may suggest that DF mapping is less sensitive to reflect spatiotemporal differences in activity during persistent AF, particularly with regard to fractionation, as recently pointed out by Haïssaguerre et al.24
The highest DF recorded in our study was 5.1 Hz, which seems to be low in a group of patients with persistent AF. This could, however, be a result of our DF algorithm, which might have been too conservative, especially in patients with highly fractionated activity. This probably also explains the observation of significant spatial activity gradients without parallel variation in DF gradients.
In contrast to Lazar et al.,19 Sanders et al.,25 and Atienza et al.,26 who found highest activity in the PV regions, we detected highest activity in the mid-portion of the CS. This incongruity may result from differences in signal acquisition and patient selection. In our study, the signals were sampled solely on an anatomical basis with no regard to signal characteristics. To avoid selection bias, no signals were rejected as spurious. Furthermore, Sanders et al.25 and Atienza et al.26 acquired the CS signals more distal in the CS, and this may have contributed to this difference. Jaïs et al.14 found a clearly lower activity in the distal CS and a dissociation between this area and the corresponding LA endocardial activity in one-third of their cases, while the more proximal part, the region recorded in our study, exhibited activity comparable with that in the septal region. Also, all the three above-mentioned studies had a considerable fraction of paroxysmal AF patients (58–64%). It has consistently been reported that in persistent AF, the DF is found more often in non-PV regions.14,25 This may also in part explain why they found the highest degree of activity in the PV regions.
Finally, in all the three studies, relevant drugs were withheld, except in Sanders et al.'s25 study in which amiodarone was not discontinued. This may also have contributed to the observed difference. Not all studies, however, have reported uniformly higher DF levels. Sahadevan et al.27 found that DFs in the LA range from 3.05 to 7.20 Hz in patients with permanent AF.
Left atrium is from several previous studies believed to be important in the maintenance of AF, in particular, the PW.28 If high activity per se reflects the importance of an area to perpetuate AF, the CSm, RO, AS, and LPVa/RPVa areas seem to be of greater importance than the PW in this study with persistent AF. This contrasts some reports, but is in accordance with the work of Nademanee et al.,5,6 who found the septal area to be crucial for perpetuating AF in many of the cases.
The essential role of the PW and PV ostia in initiating AF has been widely accepted after the landmark work of Haïssaguerre et al.,28 but the importance of the other areas of the LA in the maintenance of AF is less clear. High-frequency activity in other LA locations contributing to the maintenance of AF has been recently corroborated by Sanders et al.,25 who observed prolongation in the AF cycle length and termination of paroxysmal AF in humans during ablation of localized sites of high frequency activity. In the present study, we found the fastest and most regular activity (i.e. with lowest AA-variation index) not only in the posterior LA, but also present at most of the other sites evaluated.
Temporal variability
Great dispersion in temporal variability was observed for all measures with large inter- and intra-individual differences. This is in contrast to previous studies in humans,5,19,25 likely due to shorter recording periods and a lack of evaluating temporal stability explicitly. We evaluated temporal variability in this study for three reasons. First, there is a paucity of this type of data in humans with AF. Secondly, previous studies have largely assumed temporal stability without validating this postulate. Thirdly, and most importantly, temporal stability may contain essential additional information that can be used to distinguish the areas of primary importance for AF perpetuation from those that represent only secondary fibrillatory conduction away from such sites.
As observed for spatial gradients, there was also a clear difference between the time- and frequency-domain measures with respect to temporal variability. The main reasons for the lower variability in DF were most probably the same as for spatial gradients, i.e. the combination of our relatively conservative DF algorithm requiring 10% dominance, and the fact that DF analysis, in general, is less sensitive when the signals are more fractionated.
Spatiotemporal activity patterns
The atrial EGMs during AF are particularly difficult to interpret during real-time visual inspection alone because of the incessant changes in activation patterns. The distinct inter- and intra- individual differences in spatiotemporal activity patterns found in this study were not fully revealed until visualized by means of the AA plot.
The large intra- and inter-individual differences observed were also reflected by an AA-variation index differing by factors of 1.2–7.0 and 2.3–7.9, respectively. In spite of this large variability, there were some locations with little variation in both activity and amplitude in most study subjects [e.g. in LSPVa in study subject no. 7 (Figure 4)]; whereas in other locations, activity and/or amplitude varied substantially (e.g. in CSm in study subject no. 7).
In an experimental sheep model of AF, Skanes et al.29 found that the fastest periodic frequencies occurred in the posterior LA and hypothesized that functional or anatomically based re-entrant wave fronts, or rotors, were the source of this periodic activity and were the perpetuators of AF. Parallel findings have been made by others.22 In a canine model of AF, Morillo et al.30 targeted sites with short cycle length with cryoablation and observed termination of AF.
Similar paroxysmal short CL activities have been observed in humans in the pulmonary veins during AF ablation, which resulted in progressive slowing of the AF before termination.31 In a recent report, the same group demonstrated that ablation at sites displaying a greater percentage of continuous activity or a temporal activation gradient was associated with slowing or termination of chronic AF.32 Of note is the fact that these types of EGMs and responses were seen in all areas of the LA and that the predictive value for identifying the essential ablation sites was quite low.
The present study, however, suggests that the areas with complex and fragmented activity may be further differentiated. The visualization of sites and regions with distinctly different activity patterns that may represent areas of different importance in AF maintenance warrants further studies including systematic ablation and evaluation of outcome.
Study limitations
Because the signal recordings were performed as part of a time-consuming clinical procedure, the number of recording sites and acquisition time were limited. Therefore, we cannot comment on the spatiotemporal patterns of activation over longer recording periods. However, Skanes et al.29 found in the sheep model that transient periodic activation lasted for 3–4 s, which suggests that a 60 s recording at each site should suffice.
The mapping was performed with an 8 mm tip catheter. This may have affected the registered atrial rate by simultaneously recording two adjacent wavefronts, thereby attenuating the spatial differences.33 If so, this would not represent a systematic bias as all recordings were performed with the same type of catheter, except those recorded in CSm. It is also possible that the recordings with the lowest variation were obtained with the roving catheter tip perpendicular to the endocardial surface and those with greater variability with a more tangential orientation. However, as the same type of variability was observed in CSm, it makes this possibility less likely.
The finding of the highest activity in the CS-mid region, the only location the signals were recorded with a smaller electrode, calls in question if this can explain the high activity recorded. Measured centre-to-centre, the inter-electrode distance was 2 mm in the CS catheter and 5.5 mm in the 8 mm tip catheter, whereas the inter-electrode gap is 1 mm in both. According to the work of Baerman et al.,33 a closer inter-electrode spacing resulted in lower calculated atrial rates. Therefore, the effect of the smaller electrode size in the CS in this study would more probably be an underestimation, rather than an overestimation, of the level of activity.
|
Conclusion |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionConclusionFundingAcknowledgementsReferences |
|---|
There are significant spatial activity gradients within the LA in patients with persistent AF. The activity was highest in CSm and lowest on the PW. The most important finding in this study may, however, be the potential of this simple, robust, and fully automated algorithm to identify LA endocardial sites displaying distinctly different spatiotemporal patterns of complex and fractionated activity, which are not easily appreciated by real-time visual inspection alone. However, randomized ablation studies systematically targeting different types of activities will be required to delineate whether these various AF activity patterns actually reflect areas with different pathophysiological roles in AF maintenance.
|
Funding |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionConclusionFundingAcknowledgementsReferences |
|---|
This work was supported by a grant from Inger and John Fredriksens Heart Foundation, Oslo, Norway.
|
Acknowledgements |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionConclusionFundingAcknowledgementsReferences |
|---|
The authors thank Howard Lowery for technical support with signal acquisition.
Conflict of interest: H.C. has been a consultant for Ablation Frontiers, Biosense Webster, Boston Scientific, and St Jude Medical. R.F.B. has been a consultant for Biosense Webster. A.C. and R.D.B.: none declared.
|
References |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionConclusionFundingAcknowledgementsReferences |
|---|
[1] Arora R, Verheule S, Scott L, Navarrete A, Katari V, Wilson E, et al. Arrhythmogenic substrate of the pulmonary veins assessed by high-resolution optical mapping. Circulation (2003) 107:1816–21.
[2] Allessie MA, Lammers WJEP, Bonke FIM, Hollen J. Experimental evaluation of Moe's multiple wavelet hypothesis of atrial fibrillation. In: Cardiac Arrhythmias.—Zipes DP, Jalife J, eds. (1985) New York: Grune & Stratton. 265–75.
[3] Wijffels MC, Kirchhof CJ, Dorland R, Allessie MA. Atrial fibrillation begets atrial fibrillation. A study in awake chronically instrumented goats. Circulation (1995) 92:1954–68.
[4] Konings KTS, Smeets JLRM, Penn OC, Wellens HJJ, Allessie MA. Configuration of unipolar atrial electrograms during electrically induced atrial fibrillation in humans. Circulation (1997) 95:1231–41.
[5] Nademanee K, McKenzie J, Kosar E, Schwab M, Sunsaneewitayakul B, Vasavakul T, et al. A new approach for catheter ablation of atrial fibrillation: mapping of the electrophysiologic substrate. J Am Coll Cardiol (2004) 43:2044–53.
[6] Nademanee K, Schwab MC, Kosar EM, Karwecki M, Moran MD, Visessook N, et al. Clinical outcomes of catheter substrate ablation for high-risk patients with atrial fibrillation. J Am Coll Cardiol (2008) 51:843–9.
[7] Oral H, Chugh A, Good E, Wimmer A, Dey S, Gadeela N, et al. Radiofrequency ablation of chronic atrial fibrillation guided by complex electrograms. Circulation (2007) 115:2606–12.
[8] Oral H, Chugh A, Good E, Crawford T, Sarrazin JF, Kuhne M, et al. Randomized evaluation of right atrial ablation after left atrial ablation of complex fractionated atrial electrograms for long-lasting persistent atrial fibrillation. Circ Arrhythmia Electrophysiol (2008) 1:6–13.
[9] Haïssaguerre M, Sanders P, Hocini M, Takahashi Y, Rotter M, Sacher F, et al. Catheter ablation of long-lasting persistent atrial fibrillation: critical structures for termination. J Cardiovasc Electrophysiol (2005) 16:1125–37.[CrossRef][Web of Science][Medline]
[10] Oral H, Chugh A, Good E, Igic P, Elmouchi D, Tschopp DR, et al. Randomized comparison of encircling and nonencircling left atrial ablation for chronic atrial fibrillation. Heart Rhythm (2005) 2:1165–72.[CrossRef][Web of Science][Medline]
[11] Verma A, Novak P, Macle L, Whaley B, Beardsall M, Wulffhart Z, et al. A prospective, multicenter evaluation of ablating complex fractionated electrograms (CFE's) during atrial fibrillation (AF) identified by an automated mapping algorithm: acute effects on AF and efficacy as an adjuvant strategy. Heart Rhythm (2008) 5:198–205.[CrossRef][Web of Science][Medline]
[12] Scherr D, Dalal D, Cheema A, Cheng A, Henrikson CA, Spragg D, et al. Automated detection and characterization of complex fractionated atrial electrograms in human left atrium during atrial fibrillation. Heart Rhythm (2007) 4:1013–20.[CrossRef][Web of Science][Medline]
[13] Lin Y-J, Tai C-T, Kao T, Chang S-L, Wongcharoen W, Lo L-W, et al. Consistency of complex fractionated atrial electrograms during atrial fibrillation. Heart Rhythm (2008) 5:406–12.[CrossRef][Web of Science][Medline]
[14] Jaïs P, Haïssaguerre M, Shah D, Chouari S, Clementy J. Regional disparities of endocardial atrial activation in paroxysmal atrial fibrillation. Pacing Clin Electrophysiol (1996) 19:1998–2003.[CrossRef][Medline]
[15] Cheng A, Calkins H. A conservative approach to performing transseptal punctures without the use of intracardiac echocardiography: a stepwise approach with real-time video clips. J Cardiovasc Electrophysiol (2006) 17:459–66.[CrossRef][Web of Science][Medline]
[16] Vasamreddy CR, Jayam V, Lickfett L, Nasir K, Bradley DJ, Eldadah Z, et al. Technique and results of pulmonary vein angiography in patients undergoing catheter ablation of atrial fibrillation. J Cardiovasc Electrophysiol (2004) 15:21–6.[CrossRef][Web of Science][Medline]
[17] Everett TH, Kok LC, Vaughn RH, Moorman JR, Haines DE. Frequency domain algorithm for quantifying atrial fibrillation organization to increase defibrillation efficacy. IEEE Trans Biomed Eng (2001) 48:969–78.[CrossRef][Web of Science][Medline]
[18] Kuck KH, Ernst S, Cappato R, Braun E, Lang M, Ben-Haim SA, et al. Nonfluoroscopic endocardial catheter mapping of atrial fibrillation. J Cardiovasc Electrophysiol (1998) 9:S57–S62.[Web of Science][Medline]
[19] Lazar S, Sanjay D, Marchlinski FE, Callans DJ, Gerstenfeld EP. Presence of left-to-right atrial frequency gradient in paroxysmal but not persistent atrial fibrillation in humans. Circulation (2004) 110:3181–6.
[20] Ng J, Goldberger JJ. Understanding and interpreting dominant frequency analysis of AF electrograms. J Cardiovasc Electrophysiol (2007) 18:680–5.[CrossRef][Web of Science][Medline]
[21] Ng J, Kadish AH, Goldberger JJ. Technical considerations for dominant frequency analysis. J Cardiovasc Electrophysiol (2007) 18:757–64.
[22] Mansour M, Mandapati R, Berenfeld O, Chen J, Samie FH, Jalife J. Left-to-right gradient of atrial frequencies during acute atrial fibrillation in the isolated sheep heart. Circulation (2001) 103:2631–6.
[23] Harada A, Sasaki K, Fukushima T, Ikeshita M, Asano T, Yamauchi S, et al. Atrial activation during chronic atrial fibrillation in patients with isolated mitral valve disease. Ann Thorac Surg (1996) 61:104–11.
[24] Haïssaguerre M, Wright M, Hocini M, Jaïs P. The substrate maintaining persistent atrial fibrillation. Circ Arrhythmia Electrophysiol (2008) 1:2–5.
[25] Sanders P, Berenfeld O, Hocini M, Jaïs P, Vaidyanathan R, Hsu L-F, et al. Spectral analysis identifies sites of high-frequency activity maintaining atrial fibrillation in humans. Circulation (2005) 112:789–97.
[26] Atienza F, Almendral J, Moreno J, Vaidyanathan R, Talkachou A, Kalifa J, et al. Activation of inward rectifier potassium channels accelerates atrial fibrillation in humans. Evidence for a reentrant mechanism. Circulation (2006) 114:2434–42.
[27] Sahadevan J, Ryu K, Peltz L, Khrestian CM, Stewart RW, Markowitz AH, et al. Epicardial mapping of chronic atrial fibrillation in patients. Preliminary observations. Circulation (2004) 110:3293–9.
[28] Haïssaguerre M, Jaïs P, Shah DC, Takahashi A, Hocini M, Quinou G, et al. Spontaneous initiation of atrial fibrillation by ectopic beats originating in the pulmonary veins. N Engl J Med (1998) 339:659–66.
[29] Skanes AC, Mandapati R, Berenfeld O, Davidenko JM, Jalife J. Spatiotemporal periodicity during atrial fibrillation in the isolated sheep heart. Circulation (1998) 98:1236–48.
[30] Morillo CA, Klein GJ, Jones DL, Guiraudon CM. Chronic rapid atrial pacing: structural, functional electrophysiological characteristics of a new model of sustained atrial fibrillation. Circulation (1995) 91:1588–95.
[31] Haïssaguerre M, Sanders P, Hocini M, Hsu L-F, Shah DC, Scavee C, et al. Changes in atrial fibrillation cycle length and inducibility during catheter ablation and their relation to outcome. Circulation (2004) 109:3007–13.
[32] Takahashi Y, O'Neill MD, Hocini M, Dubois R, Matsuo S, Knecht S, et al. Characterization of electrograms associated with termination of chronic atrial fibrillation by catheter ablation. J Am Coll Cardiol (2008) 51:1003–10.
[33] Baerman JM, Ropella KM, Sahakian AV, Kirch JA, Swiryn S. Effect of bipole configuration on atrial electrograms during atrial fibrillation. Pacing Clin Electrophysiol (1990) 13:78–87.[CrossRef][Medline]
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  M. Duytschaever and R. Tavernier What is needed for a good 'Decaf'? Europace, March 1, 2009; 11(3): 278 - 279. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||