CLINICAL RESEARCH
Europace (2011) 13, 809–814
doi:10.1093/europace/eur034
Ablation for Atrial Fibrillation
Spatial relationship between left atrial roof
or superior pulmonary veins and bronchi or
pulmonary arteries by dual-source computed
tomography: implication for preventing injury of
bronchi and pulmonary arteries during atrial
fibrillation ablation
Yi-Gang Li 1*†, Mei Yang 1†, Yuhua Li 2, Qunshan Wang 1, Linwei Yu 2, and Jian Sun 1
1
Department of Cardiology, Xinhua Hospital, Shanghai Jiao Tong University School of Medicine, 200092 Shanghai, China; and 2Department of Radiology, Xinhua Hospital,
Shanghai Jiao Tong University School of Medicine, 200092 Shanghai, China
Received 3 November 2010; accepted after revision 19 January 2011; online publish-ahead-of-print 23 February 2011
Aims
Bronchi or pulmonary arteries (PAs) could be injured during atrial fibrillation (AF) ablation. Therefore, the aim of the
present study was to evaluate the spatial relationship between left atrial roof or superior pulmonary veins (PVs) and
neighbouring structures of AF patients and provide anatomic guidance for AF ablation to avoid injuring bronchi or
PAs.
.....................................................................................................................................................................................
Methods
A dual-source computed tomography (DSCT) scan was used to depict the left atrium (LA), PVs, and nearby strucand results
tures including bronchi and PAs in 58 patients with drug-refractory AF (mean age, 64 + 9 years). The distance
between LA roof or superior PVs (SPVs) and bronchi or PAs was measured. The average minimal distances from
the left, middle, and right points of the LA roof to the principal bronchi were 17.0 + 6.4, 23.7 + 5.1, and
23.2 + 7.7 mm, respectively. The LA roof was closer to the right PA (RPA) than the left PA (LPA) for more than
90% of patients. The average minimal distances from the left, middle, and right points of the LA roof to the PAs
were 8.3 + 5.0, 5.9 + 3.1, and 6.0 + 2.8 mm, respectively. The average minimal distances between the left superior
pulmonary vein and bronchi or LPA were 0.32 + 0.79 or 0.4 + 1.0 mm, respectively. The average minimal distances
between the right superior pulmonary vein and bronchi or RPA were 0.27 + 0.94 and 0.0 + 0.1 mm, respectively.
Both of the root parts of SPVs of most patients were in direct contact with branches of trachea and PAs.
.....................................................................................................................................................................................
Conclusion
Dual-source computed tomography provides important imaging information for determining the relationship
between LA, PVs, and neighbouring structures. Use of pre-procedural cardiac CT scans may help avoid ablationinduced injury of bronchi and PAs.
----------------------------------------------------------------------------------------------------------------------------------------------------------Keywords
Atrial fibrillation † Dual-source computed tomography † Left atrial roof † Pulmonary vein † Ablation
Introduction
Percutaneous catheter ablation with circumferential pulmonary
vein (PV) isolation using radiofrequency energy is increasingly
used for treating atrial fibrillation (AF).1,2 Ablation strategies
include circumferential PV ablation, linear ablation, complex fractionated electrogram ablation, etc.3 – 6 However, segmental
ablation is still used in practice in some laboratories, which may
* Corresponding author. Tel: +86 21 55964561, fax: +86 21 55964561, Email: [email protected]
†
Contributed equally to this work.
Published on behalf of the European Society of Cardiology. All rights reserved. & The Author 2011. For permissions please email: [email protected].
810
result in PV stenosis or other complications. For many AF patients,
especially persistent AF patients, left atrium (LA) roofline ablation
is necessary for maintaining sinus rhythm after the procedure. As
the LA roof and superior PVs (SPVs) are in very close vicinity to
bronchi and pulmonary arteries (PAs) that branch out around
them, radiofrequency energy application during ablation of roofline
and superior PVs may injure the nearby bronchi and PAs, which is
very rare but fatal.7 – 9
With the recent introduction of dual-source computed tomography (DSCT), which significantly improved temporal resolution
of ECG-gated multidetector-row CT,10 – 12 cardiac visualization of
AF patients with irregular or high heart rate is much more accurate
and reliable. The aim of the present study was to evaluate the anatomic relationship between the LA roof or SPVs and bronchi or
PAs using DSCT.
Methods
Patient selection
The study population consisted of 58 AF patients who underwent electrophysiological examination and catheter ablation procedure in
Shanghai Xinhua Hospital from March to October in 2008. All patients
had frequent AF episodes or serious symptoms while antiarrhythmics
were not effective enough or the patients could not tolerate the side
effects. All patients underwent cardiac DSCT scan in order to determine the anatomic relationship between LA/SPV and bronchi/PAs.
Those who had a medical history of rheumatic heart disease, valvular
heart disease, congenital heart disease, dilated cardiomyopathy, hypertrophic cardiomyopathy, heart surgery, or organic diseases of the
bronchus, lung, or oesophagus were not included. Those whose
cardiac images were not clear enough, especially LA and PV, were
also excluded.
Image acquisition and reconstruction
The patients underwent cardiac imaging using a 64-slice DSCT system
(Somatom Definition, Siemens Medical Systems, Forchheim,
Germany). Eighty millilitres non-ionic contrast medium iohexol
Y.-G. Li et al.
(Omnipaque, Amersham Health, Amersham, UK) were injected
through the antecubital vein and chased with 40 mL of saline by
using a power injector at a rate of 5 mL/s, after which the scanning
was initiated at the time that the CT value of arotic root increased
to 100Hu. Image acquisition was performed from the base of the
lungs to the apices during a single breath-hold. Image acquisition was
ECG-gated. A collimation of 0.6 mm was used. Scan data were transferred to the Syngo workstation (Leonardo, Siemens, Forchheim,
Germany) for reconstruction, measurement, and analysis. Cardiac
image reconstruction was undertaken using the series of ‘0.75 mm
slice thickness and best diastole’.
Measurement
Syngo workstation provides sagittal, coronal, and transverse sections
for investigation and measurement. The distance from the LA roof
to the bronchus or PA was measured from the coronal section
while sometimes sectional adjustment was necessary. In order to
avoid measurement difference between different researchers, all data
were taken by one person.
Parameters measured included LA dimension (LA1, LA2, and LA3),
shortest distance between the LA roof and bronchi or PAs, and shortest distance between the SPVs and bronchi or PAs. LA1 and LA3 were
defined as the maximal transverse diameter and superior – inferior
diameter of LA from coronal section, and LA2 defined as the
maximal antero-posterior diameter of LA from the sagittal section
(Figure 1). Left atrium roof was defined as the area between two
SPVs. When measuring the distance between two structures, it
referred to the distance between their inner surfaces that were identified by a change in signal density. Distance between the LA roof and
bronchi or PAs was measured at three locations of the LA roof, that is
the left point [the intersection point of the left superior pulmonary
vein (LSPV) and LA], the right point [the intersection point of the
right superior pulmonary vein (RSPV) and LA], and the middle point
(the midpoint of the above two points) (Figure 4). The distance from
the PV ostia to the position where SPV was closest to bronchus or
PA was also measured, which is defined as dB or dP, respectively
(Figures 3 and 4). Pulmonary arteries ostium was composed by
points where the change in curvature of the LA surface was most dramatic. When an antrum existed, PV ostium indicated the interface of
LA and antrum.
Figure 1 Coronal section and sagittal section of left atrium obtained by Syngo workstation. (A) Sagittal section showing measurement of LA2.
(B) Coronal section showing measurement of LA1 and LA3.
811
Preventing injury of bronchi and PAs during AF ablation
Statistical analysis
Statistical analyses were performed using SPSS11.5 (SPSS 11.5 for
Windows, LEAD Technologies Inc., Chicago, IL, USA). Values are
expressed as mean + standard deviation for continuous variables,
constituent ratio for categorical variables. Comparisons between the
two groups were made by a two-tailed Student’s t-test for continuous
variables. The difference was considered statistically significant when
P-value was ,0.05.
Results
Patient population
The subjects of this study were 58 patients with average age of
64 + 9 years. Twenty-five of all subjects were women while 33
were men. They were 41 paroxysmal AF patients (22 women
and 19 men) and 17 persistent AF patients (3 women and 14 men).
Table 1 Comparison of left atrium dimension between
patients with different atrial fibrillation type, gender,
and age
n
LA1 (mm)
LA2 (mm)
LA3 (mm)
................................................................................
Paroxysmal AF
41
66.1 + 12.5
41.4 + 7.5
55.4 + 6.2
Persistent AF
17
73.6 + 5.5
47.5 + 8.1
61.0 + 8.7
P value
Male
33
0.021a
71.1 + 6.3
0.008a
42.5 + 7.7
0.008a
58.4 + 7.5
Female
25
66.7 + 8.9
44.1 + 8.7
55.3 + 7.1
P value
Age ,70 years
36
0.031a
68.5 + 8.1
0.443
42.5 + 8.9
0.443
56.5 + 7.3
Age ≥70 years
22
70.3 + 7.3
44.4 + 6.6
58.1 + 7.7
0.399
0.397
0.430
P value
a
Persistent atrial fibrillation compared with paroxysmal atrial fibrillation, P , 0.05;
female compared with male, P , 0.05.
Left atrium dimension
The mean LA1, LA2, and LA3 were 69.2 + 7.8, 43.2 + 8.1, and
57.1 + 7.4 mm, respectively.
Difference in the LA dimension was examined between patients
with different AF type (paroxysmal or persistent AF), gender, and
age (Table 1). Persistent AF patients had significant greater LA
dimensions in all directions than paroxysmal AF patients. Men
had a significantly greater transverse diameter than women. No
significant difference in the antero-posterior and superior–inferior
diameter was found between men and women. Patients older or
younger than 70 years did not have significant difference in LA
dimension (Table 1).
The relationship between the left atrium
roof and bronchi or pulmonary arteries
The anatomic relationship between the LA roof, the left and right
principal bronchi, LPA, and RPA was depicted for each patient
(Figure 2). The trachea was located in front of the oesophagus
and was divided into the left and right principal bronchi above
the LA roof. The LA roof was not in close proximity to both principal bronchi as the RPA ran transversely between the LA roof and
the principal bronchi. The average minimal distance from the left,
middle, and right point of the LA roof to two principal bronchi
was 17.0 + 6.4, 23.7 + 5.1, and 23.2 + 7.7 mm, respectively. The
left point was nearest to the principal bronchus, which was significantly closer than the other two points (P , 0.001).
When comparing the distance with LPA and RPA, except that
10% of left points of LA roof were closer to the LPA, all the
other points of the LA roof (including left, right, and middle
points) were in closer proximity to the RPA. This is a result of
anatomy that the RPA ran transversely above the LA roof for
almost all patients while LPAs were located supero-left to the
LA roof (Figures 2–4). The average minimal distance from left,
middle, and right points of LA roof to PAs were 8.3 + 5.0,
5.9 + 3.1, and 6.0 + 2.8 mm, respectively. The left point was farthest away from PAs, which was significant compared with the
middle and right points, respectively (P , 0.001 and P ¼ 0.001).
The minimum among the distances between PAs and the three
points was defined as the minimal distance between the LA roof
Figure 2 Anatomic relationship between the left atrium roof,
left and right principal bronchi, left pulmonary arteries and right
pulmonary arteries (computed tomography scan images from
coronal section). Trachea divides into two principal bronchi
above the left atrium roof with some distance away from left
atrium roof, separated by pulmonary arteries. Pulmonary arteries
locate above the left atrium roof and superior pulmonary veins
and were in close proximity with them.
and PAs, which was 4.7 + 2.6 mm. Three patients had their LA
roofs in direct contact with PAs.
All patients were divided into two groups according to the
maximal superior –inferior diameter of LA (LA3): one group
with LA3 ≤ 57 mm and the other one with LA3 . 57 mm. The
distance between the LA roof and PA was measured for these
two groups. The left point of LA roof was closer to PA for the
group with LA3 . 57 mm (6.7 + 3.5 vs. 9.6 + 5.6 mm, P ¼
0.024). There was no significant difference in the distance from
812
Y.-G. Li et al.
Figure 3 Relationship between the left atrium roof, superior pulmonary veins, and bronchi (computed tomography scan images mildly
adjusted from the coronal section). (A) Right bronchus is in direct contact with the right superior pulmonary vein. (B) Left bronchus is in
direct contact with the left superior pulmonary vein. Arrows indicate where superior pulmonary veins and bronchi are closest. dBL and dBR
were measured from the arrows to superior pulmonary vein ostia.
the middle or right points of the LA roof to PAs, or minimal distance between the LA roof and PAs (P ¼ 0.103, 0.403, and
0.101, respectively).
The relationship between superior
pulmonary veins and bronchi
or pulmonary arteries
Branches of SPVs, principal bronchi, and PAs were in very close
proximity to each other in the neighbourhood of the LA roof
and SPVs (Figures 3 and 4).
LSPVs of 49 patients (85%) and RSPVs of 52 patients (90%) were
in direct contact with the bronchi. The average minimal distances
from the SPV root to the bronchi were 0.32 + 0.79 mm for
LSPV and 0.27 + 0.94 mm for RSPV, which were not significantly
different (P ¼ 0.708). dBL referred to dB for LSPV and dBR for
RSPV. The mean value of dBL was 22.7 + 7.5 mm with 95% confidence interval of 20.7–24.6 mm while the mean value of dBR was
16.9 + 8.2 mm, significantly shorter than dBL (P , 0.001), with 95%
confidence interval of 14.8 –19.0 mm (Figure 5).
LSPVs of 49 patients (85%) and RSPVs of 57 patients (98%) were
in direct contact with PAs. The average minimal distances from the
SPV root to the PAs were 0.4 + 1.0 mm for LSPV and 0.0 +
0.1 mm for RSPV, which was significantly different (P ¼ 0.007).
dPL referred to dP for LSPV and dPR for RSPV. The mean value
of dPL was 28.9 + 9.5 mm with 95% confidence interval of 26.4–
31.4 mm while the mean value of dPR was 19.9 + 9.4 mm, significantly shorter than dPL (P , 0.001), with 95% confidence interval
of 17.4–22.3 mm (Figure 5).
When the risk of injury to the bronchus or PA is considered
altogether, 95% confidence interval of dB and dP of the same
side should be combined. The shortest intervals were 20.7 –
31.4 mm from the LSPV ostium, and 14.8–22.3 mm from the
RSPV ostium to either bronchi or PAs.
Figure 4 Relationship between left atrium roof, superior pulmonary veins, and pulmonary arteries (computed tomography
scan images mildly adjusted from the coronal section). Right pulmonary arteries runs transversely above the left atrium roof while
the Left pulmonary arteries lies supero-left from the left atrium
roof. Both the Left pulmonary arteries and the Right pulmonary
arteries are in direct contact with root of superior pulmonary
veins. Arrows indicate where pulmonary arteries and superior
pulmonary veins are closest. dPL and dPR were measured from
the arrows to superior pulmonary vein ostia. Red points indicate
the left, middle, and right points of the left atrium roof, from
which the distance between the left atrium roof and bronchi or
pulmonary arteries was measured.
Discussion
Left atrium dimension
Our study showed a significant difference in LA dimension
between persistent and paroxysmal AF patients, indicating that
813
Preventing injury of bronchi and PAs during AF ablation
d 35
(mm)
95% confidence interval for LSPV
30
mean value for LSPV
95% confidence interval for RSPV
mean value for RSPV
25
20
15
10
Bronchus PA Bronchus PA
Figure 5 Mean distance and corresponding 95% confidence interval from pulmonary vein ostium to the position where superior pulmonary
veins were closest to bronchi or pulmonary arteries. Red signifies the left and green signifies the right. This chart shows clearly that no matter
bronchus or pulmonary arteries, the location where superior pulmonary vein is closest to bronchi or pulmonary arteries, it is nearer to pulmonary
vein ostium for the right side. It indicates that it is more risky to ablate in the interior wall of the right superior pulmonary vein near the ostium than
doing so for the left superior pulmonary vein.
LA gets dilated as AF develops. This result is consistent with the
known fact that atrial dilatation and AF may be mutually dependent
and constitute a vicious circle. Left atrium enlargement is an independent risk factor for the development of AF13 – 15 while AF itself
can also cause atrial dilatation.16,17
The relationship between the left
atrium roof and the bronchi
or pulmonary arteries
Trachea divides into left and right principal bronchi above PAs. For
the vast majority of patients, there was no neighbouring relationship between the principal bronchi and the LA roof. The distance
between LA roof and principal bronchi was long enough that LA
roof ablation could hardly injure principal bronchi. Therefore, it
can be concluded that most atrio-bronchial fistulae were possibly
not caused by roofline ablation.
For all people, the RPA ran to the right since ramifying from the
PA trunk, lying above the LA roof. This close anatomic relationship
between the LA roof and the PAs was demonstrated in our study,
especially the neighbouring relationship between the LA roof and
the RPA, inferring that roofline ablation may cause injury to PAs,
especially RPAs. Although there has not been any report about
PA injury during AF ablation, it still deserves serious attention for
the close anatomic relationship between the LA roof and the PA.
The dilation of LA dimension could affect the anatomic relationship between the LA and neighbouring structures and this impact is
showed in this study that the roof of LA with greater superior –
inferior diameter is closer to PA at the left point. On the other
hand, the distance between the right or middle point of the LA
roof and PA as well as the minimal distance between these two
structures was not affected by LA dilation. The result also showed
that the left point of the LA roof was the farthest one away from
PAs (8.3 + 5.0 mm) compared with the middle and right point.
Therefore, although our study indicates that LA dilation may affect
correlation between the LA roof, this impact is not enormous.
The relationship between superior
pulmonary vein and bronchi
or pulmonary arteries
To avoid PV stenosis, circumferential PV ablation is often performed outside the ostia, that is, ablation to the atrial wall. Since
three-dimensional electroanatomic system-guided circumferential
PV ablation could hardly be accurate enough to distinguish the
interface of the LA and PVs, ablation at the interior wall of PVs
still occurs occasionally. Moreover, when complete circumferential
PV isolation could not be achieved or it was difficult to identify
gaps, segmental ablation at the interior wall of PVs is still exercised
for complete electrical isolation in some labs. Therefore, knowledge about the relationship between the root part of SPVs and
nearby structures is helpful for avoiding injury. The close anatomic
relationship between the root of SPVs and PAs or bronchi was
clearly shown in this study, which further support many experts’
opinion against segmental ablation because this ablation method
is likely to cause PA or bronchus injury.
Although Wu et al. 18 have made an investigation about the close
relationship between the bronchi and PVs, they just gave an overall
description of the anatomic connection between those two structures. This study went deeper that besides the minimal distance
between the SPVs and bronchi, it focused more on the distance
away from the PV ostia when the SPVs were closest to the
bronchi or PAs, which is obviously more clinical significant, telling
electrophysiology (EP) doctors where it is highly risky to ablate.
When SPVs were closest to the bronchi or PAs, the closest position
was some distance away from PV ostia (dB or dP) which showed significant difference between left and right sides. Both dB and dP were
smaller for the right than those for the left. Moreover, the minimal
814
distance from RSPV to PA was significantly shorter than that from
LSPV to PA. These results indicate that it takes more risk when
performing ablation in the interior wall of root RSPV because it is
more possible to injury the neighbouring bronchi or PAs.
Use of dual-source computed
tomography for atrial fibrillation ablation
Dual-source computed tomography has a much wider scale of
examination and better quality of visualization than traditional
multislice CT. Based on the 64-slice CT, DSCT adds a second
X-ray tube and a corresponding detector so that it realizes an
ideal temporal resolution for cardiac visualization of 83 ms.19
With the adaptive ECG-gated technology, it is no longer a
problem to give a cardiac scan for those who have very high
heart rate or irregular heart rhythm, especially for AF patients.
In most centres, pre-procedure cardiac CT scans are used just
for cardiac visualization and integration with the reconstructed
cardiac image by three-dimensional electroanatomic mapping.20 –
22
However, in order to reduce the risk of injury to surrounding
structures, the images should also be used for evaluation of the
spatial relationship between LA, PVs, and surrounding structures
like bronchi, PAs, and oesophagus.18,23 Moreover, information
such as thickness of the posterior wall of the LA, thickness of
the myocardium at the mitral isthmus region, and the spatial
relationship between the LA and coronary sinus or left circumflex
is also very helpful.24,25 This anatomic information could be very
important for guidance of the ablation procedure. With knowledge
of the relevant anatomic information of the specific patient, EP
doctors could be more aware of the risky locations during
energy application to avoid complications.
Conclusion
Roofline ablation may carry the risk of PA (especially RPA) injury
while ablation in the interior wall of SPVs could be more possible
to cause injury to bronchial or PA branches. Cardiac DSCT scan
provides important imaging information for determining the
relationship between LA, SPVs, and neighbouring structures. Preprocedural analysis of cardiac CT scan images may help reduce
risk of complications.
Conflict of interest: none declared.
Funding
This study was supported by grants from the National Natural Science
Foundation of China (No. 30670831 and No. 30871082), and the
Shanghai Committee of Science and Technology, China (No.
10XD1402800).
References
1. Chugh A, Morady F. Atrial fibrillation: catheter ablation. J Interv Card Electrophysiol
2006;16:15 –26.
2. Ernst S, Ouyang F, Löber F, Antz M, Kuck KH. Catheter-induced linear lesions in
the left atrium in patients with atrial fibrillation: an electroanatomic study. J Am
Coll Cardiol 2003;42:1283 –5.
3. Fassini G, Riva S, Chiodelli R, Trevisi N, Berti M, Carbucicchio C et al. Left mitral
isthmus ablation associated with PV isolation: long-term results of a prospective
randomized study. J Cardiovasc Electrophysiol 2005;16:1150 –6.
Y.-G. Li et al.
4. Haı̈ssaguerre M, Hocini M, Sanders P, Sacher F, Rotter M, Takahashi Y et al. Catheter ablation of long-lasting persistent atrial fibrillation: clinical outcome and
mechanisms of subsequent arrhythmias. J Cardiovasc Electrophysiol 2005;16:
1138 –47.
5. Sanders P, Jaı̈s P, Hocini M, Hsu LF, Scavée C, Sacher F et al. Electrophysiologic
and clinical consequences of linear catheter ablation to transect the
anterior left atrium in patients with atrial fibrillation. Heart Rhythm 2004;1:
176 –84.
6. Jalife J. Rotors and spiral waves in atrial fibrillation. J Cardiovasc Electrophysiol 2003;
14:776 – 80.
7. Schmidt M, Nölker G, Marschang H, Gutleben KJ, Schibgilla V, Rittger H et al. Incidence of oesophageal wall injury post-pulmonary vein antrum isolation for treatment of patients with atrial fibrillation. Europace 2008;10:205 –9.
8. D’Avila A, Ptaszek LM, Yu PB, Walker JD, Wright C, Noseworthy PA et al. Left
atrial–esophageal fistula after pulmonary vein isolation. Circulation 2007;115:
432 –3.
9. Doshi RN, Kaushal R, Cesario DA, Shivkumar K. Atriobronchial fistula formation
as a devastating complication of left atrial catheter ablation for atrial fibrillation.
Heart Rhythm 2006;3:59.
10. Scheffel H, Alkadhi H, Plass A, Vachenauer R, Desbiolles L, Gaemperli O et al.
First experience in a high pre-test probability population without heart rate
control. Eur Radiol 2006;16:2739 –47.
11. Knackstedt C, Muhlenbruch G, Mischke K, Bruners P, Schimpf T, Frechen D et al.
Imaging of the coronary venous system: validation of three-dimensional rotational
venous angiography against dual-source computed tomography. Cardiovasc Interv
Radiol 2008;31:1150 –8.
12. Bastarrika G, Arraiza M, Arias J, Broncano J, Zudaire B, Pueyo JC et al. Dualsource CT coronary angiography: image quality and optimal reconstruction interval. Radiologia 2009;J51:376 –84.
13. Vaziri SM, Larson MG, Benjamin EJ, Levy D. Echocardiographic predictors of nonrheumatic atrial fibrillation: the framingham heart study. Circulation 1994;89:
724 –30.
14. Psaty BM, Manolio TA, Kuller LH, Kronmal RA, Cushman M, Fried LP et al. Incidence of and risk factors for atrial fibrillation in older adults. Circulation 1997;96:
2455 –61.
15. Tsang TS, Barnes ME, Bailey KR, Leibson CL, Montgomery SC, Takemoto Y et al.
Left atrial volume: important risk marker of incident atrial fibrillation in 1655
older men and women. Mayo Clin Proc 2001;76:467–75.
16. Sanfilippo AJ, Abascal VM, Sheehan M, Oertel LB, Harrigan P, Hughes RA et al.
Atrial enlargement as a consequence of atrial fibrillation: a prospective echocardiographic study. Circulation 1990;82:792 –7.
17. Dittrich HC, Pearce LA, Asinger RW, McBride R, Webel R, Zabalgoitia M et al.
Left atrial diameter in nonvalvular atrial fibrillation: an echocardiographic study.
Stroke prevention in atrial fibrillation investigators. Am Heart J 1999;137:494–9.
18. Wu MH, Wongcharoen W, Tsao HM, Tai CT, Chang SL, Lin YJ et al. Close
relationship between the bronchi and pulmonary veins: implications for the prevention of atriobronchial fistula after atrial fibrillation ablation. J Cardiovasc Electrophysiol 2007;18:1056 –9.
19. Moser M, Bosserhoff AK, Hunziker EB, Tai CT, Chang SL, Lin YJ et al. Ultrastructural cartilage abnormalities in MIA/CD2RAP2deficient mice. Mol Cell Biol 2002;
22:1438 –45.
20. Tops LF, Bax JJ, Zeppenfeld K, Jongbloed MR, Lamb HJ, van der Wall EE et al.
Fusion of multislice computed tomography imaging with three-dimensional electroanatomic mapping to guide radiofrequency catheter ablation procedures.
Heart Rhythm 2005;2:1076 – 81.
21. Dong J, Calkins H, Solomon SB, Lai S, Dalal D, Lardo AC et al. Integrated electroanatomic mapping with three-dimensional computed tomographic images for
real-time guided ablations. Circulation 2006;113:186 –94.
22. Kistler PM, Earley MJ, Harris S, Abrams D, Ellis S, Sporton SC et al. Validation of
three-dimensional cardiac image integration: use of integrated CT image into electroanatomic mapping system to perform catheter ablation of atrial fibrillation.
J Interv Card Electrophysiol 2006;17:341 –8.
23. Kobza R, Auf der Maur C, Kurtz C, Hoffmann A, Allgayer B, Erne P. Esophagus
imaging for radiofrequency ablation of atrial fibrillation using a dual-source computed tomography system: preliminary observations. J Interv Card Electrophysiol
2007;19:167 –70.
24. Achenbach S, Ropers D, Kuettner A, Flohr T, Ohnesorge B, Bruder H et al.
Contrast-enhanced coronary artery visualization by dual-source computed tomography – initial experience. Eur J Radiol 2006;57:331 –5.
25. Chiribiri A, Kelle S, Götze S, Kriatselis C, Thouet T, Tangcharoen T et al. Visualization of the cardiac venous system using cardiac magnetic resonance. Am J
Cardiol 2008;101:407–12.