Download Imaging cardiac activation sequence during ventricular tachycardia

Am J Physiol Heart Circ Physiol 308: H108–H114, 2015.
First published November 17, 2014; doi:10.1152/ajpheart.00196.2014.
Imaging cardiac activation sequence during ventricular tachycardia in a canine
model of nonischemic heart failure
Chengzong Han,1 Steven M. Pogwizd,2 Long Yu,1 Zhaoye Zhou,1 Cheryl R. Killingsworth,2 and Bin He1,3
1
Department of Biomedical Engineering, University of Minnesota, Minneapolis, Minnesota; 2Division of Cardiovascular
Disease, Department of Medicine, University of Alabama at Birmingham, Birmingham, Alabama; and 3Institute for
Engineering in Medicine, University of Minnesota, Minneapolis, Minnesota
Submitted 24 March 2014; accepted in final form 13 November 2014
imaging; electrocardiography; mapping; heart failure; ventricular
tachycardia; canine model
CONGESTIVE HEART FAILURE (HF) afflicts ⬃5 million individuals
in the United States alone with nearly half of deaths occurring
suddenly, primarily from ventricular tachycardia (VT) degenerating to ventricular fibrillation (27). Cardiac catheter ablation
has been widely used in the intervention of VTs in patients
with structural heart diseases (39). However, some VTs are
hemodynamically unstable, making standard electroanatomical
catheter-based mapping (5, 33) difficult. Moreover, because of
the invasive nature of standard mapping, the ablation procedure can be prolonged with considerable fluoroscopy exposure
to both the patient and the physician.
Address for reprint requests and other correspondence: B. He, Dept. of
Biomedical Engineering, Univ. of Minnesota, Minneapolis, MN 55455
(e-mail: [email protected]).
H108
Recently, cardiac resynchronization therapy (CRT) has
evolved as an important treatment option to improve the
cardiac mechanical performance by restoring the cardiac electrical synchrony in congestive HF patients (35). Invasive catheter-based mapping of endocardial activation also emerges as
an important tool to study the electrophysiological mechanism
of CRT in congestive HF patients.
Noninvasive cardiac electrical imaging techniques provide a
novel and much needed means (3, 9, 38), which offers the
potential to help with delineation of the underlying arrhythmia
mechanisms related to HF and to facilitate corresponding
selections of patient-specific treatment options for HF and its
related arrhythmias. Previously we have rigorously evaluated a
three-dimensional (3D) cardiac electrical imaging (3DCEI)
technique in imaging VTs in rabbit (9, 11) and in dog (10)
without structure diseases. This study aims to 1) investigate the
arrhythmia mechanisms of spontaneously occurring or norepinephrine (NE)-induced VTs in a canine model of nonischemic
HF by means of 3DCEI and 2) assess the performance of
3DCEI in imaging the activation patterns of VTs associated
with nonischemic HF with the aid of invasive 3D intra-cardiac
mapping (28 –30). We performed simultaneous body surface
potential mapping and 3D intra-cardiac mapping in a closedchest condition in HF dogs. The 3DCEI technique was applied
to characterize the dynamic activation patterns of both spontaneously occurring VTs and NE-induced VTs. The imaged
activation sequence was also compared with the simultaneously measured activation sequence from 3D intra-cardiac
mapping.
MATERIALS AND METHODS
HF canine model and experimental procedures. The experimental
protocol, approved by the Institutional Animal Care and Use Committees of the University of Minnesota and the University of Alabama
at Birmingham, was performed on three canines. HF in each dog was
produced by induction of aortic insufficiency, which was later followed by constriction of the abdominal aorta. Aortic insufficiency was
created via the carotid artery, in which a Fogarty balloon catheter was
used to perforate 1 to 2 valve leaflets under fluoroscopic guidance.
Aortic constriction was created through an abdominal incision in
which a suture was placed around the abdominal aorta so that the
systolic blood pressure gradient increased by 50 – 60 mmHg compared
with baseline measurement. The dogs survived for 22 ⫾ 5 mo, after
which a terminal mapping study was performed. Progression of HF
was assessed by left ventricle (LV) end-diastolic dimension, LV
end-systolic dimension, and LV fractional shortening using Doppler
echocardiography.
Simultaneous body surface potential mapping and 3D intracardiac mapping was performed on the dogs when HF developed.
The anesthesia regimen and the surgical procedure for the mapping
study were similar to the previous normal canine experiment (10).
0363-6135/15 Copyright © 2015 the American Physiological Society
http://www.ajpheart.org
Downloaded from http://ajpheart.physiology.org/ by 10.220.33.5 on August 3, 2017
Han C, Pogwizd SM, Yu L, Zhou Z, Killingsworth CR, He B.
Imaging cardiac activation sequence during ventricular tachycardia in
a canine model of nonischemic heart failure. Am J Physiol Heart Circ
Physiol 308: H108 –H114, 2015. First published November 17, 2014;
doi:10.1152/ajpheart.00196.2014.—Noninvasive cardiac activation
imaging of ventricular tachycardia (VT) is important in the clinical
diagnosis and treatment of arrhythmias in heart failure (HF) patients.
This study investigated the ability of the three-dimensional cardiac
electrical imaging (3DCEI) technique for characterizing the activation
patterns of spontaneously occurring and norepinephrine (NE)-induced
VTs in a newly developed arrhythmogenic canine model of nonischemic HF. HF was induced by aortic insufficiency followed by aortic
constriction in three canines. Up to 128 body-surface ECGs were
measured simultaneously with bipolar recordings from up to 232
intramural sites in a closed-chest condition. Data analysis was performed on the spontaneously occurring VTs (n ⫽ 4) and the NEinduced nonsustained VTs (n ⫽ 8) in HF canines. Both spontaneously
occurring and NE-induced nonsustained VTs initiated by a focal
mechanism primarily from the subendocardium, but occasionally
from the subepicardium of left ventricle. Most focal initiation sites
were located at apex, right ventricular outflow tract, and left lateral
wall. The NE-induced VTs were longer, more rapid, and had more
focal sites than the spontaneously occurring VTs. Good correlation
was obtained between imaged activation sequence and direct measurements (averaged correlation coefficient of ⬃0.70 over 135 VT
beats). The reconstructed initiation sites were ⬃10 mm from measured initiation sites, suggesting good localization in such a large
animal model with cardiac size similar to a human. Both spontaneously occurring and NE-induced nonsustained VTs had focal initiation in this canine model of nonischemic HF. 3DCEI is feasible to
image the activation sequence and help define arrhythmia mechanism
of nonischemic HF-associated VTs.
H109
IMAGING CARDIAC ACTIVATION SEQUENCE DURING VT IN HF DOGS
n
CC ⫽
冑
M
៮E兲共AT M ⫺ AT
៮
兲
i
兺 共ATiE ⫺ AT
i⫽1
冑
n
n
M 2
៮E兲2 ⫻
៮
兲
兺 共ATiE ⫺ AT
兺 共ATiM ⫺ AT
i⫽1
i⫽1
,
(1)
where n is the number of grid points of the heart model and ATiE and
ATiM are the noninvasively estimated activation time and measurement
៮E and
constructed activation time at the i-th myocardial grid point. AT
M
៮
are their respective mean values. Common biostatistics rules
AT
were used to interpret the value of CC (4). The CC values from 0 to
0.25 or from 0 to ⫺0.25 were considered the absence of correlation,
CC values from 0.25 to 0.50 or from ⫺0.25 to ⫺0.50 were considered
to be poor correlation, CC values from 0.50 to 0.75 or ⫺0.50 to ⫺0.75
were considered to be moderate to good correlation, and CC values
from 0.75 to 1 or from ⫺0.75 to ⫺1 were considered to be very good
to excellent correlation (4). The localization error (LE), which is
defined as the distance between the site of earliest activation from 3D
intra-cardiac measurements and the center of mass of the myocardial
region with the earliest imaged activation time, was computed to
evaluate the performance of 3DCEI in localizing the origin of activation. Statistical significance of differences was evaluated by Student’s t-test (paired or unpaired), and a P value ⬍0.05 was considered
statistically significant.
RESULTS
Experimentation and modeling. With HF, LV end-diastolic
dimension increased by 47% (from 3.70 ⫾ 0.08 cm to 5.43 ⫾
0.26 cm), LV end-systolic dimension increased by 63% (from
2.20 ⫾ 0.16 to 3.59 ⫾ 0.35 cm), and LV fraction shortening
decreased by 16% (from 41 ⫾ 3% to 34 ⫾ 3%) (all P ⬍ 0.05).
During the simultaneous body surface potential mapping and
3D intra-cardiac mapping, spontaneously occurring PVCs and
couplets were recorded in all three dogs, whereas spontaneous
VTs were recorded in two dogs. VTs were also induced using
NE in all three dogs. For all three animals, the realistic
geometry canine heart-torso model was constructed from
UFCT images obtained after the mapping study. The canine
ventricular myocardium was tessellated into 28,871 ⫾ 4,309
evenly spaced grid points. The spatial resolution of the ventricle models was 2 mm.
Spontaneously occurring VTs and NE-induced nonsustained
VTs. Data analysis was performed on four episodes of spontaneously occurring VTs (total of 24 VT beats) and eight episodes of NE-induced nonsustained VTs (total of 111 VT beats)
recorded during the mapping study. Table 1 shows the comparison between spontaneously occurring VTs and NE-induced
VTs in HF dogs. The NE-induced VTs were longer (14 ⫾ 2 vs.
6 ⫾ 3 beats long, P ⬍ 0.05) and were also more rapid than the
Table 1. Summary of ectopic activity during mapping study
in heart failure dogs
Mapping
Spontaneous VT
NE-induced VT
Dog No.
No. of Beats
Cycle Length, ms
No. of Beats
Cycle Length, ms
1
2
3
Mean
4⫾1
–
8⫾7
6⫾3
395 ⫾ 49
–
380 ⫾ 95
388 ⫾ 11
16 ⫾ 13
13 ⫾ 13
14
14 ⫾ 2
247 ⫾ 12
203 ⫾ 24
246
232 ⫾ 25
Values are means ⫾ SD. VT, ventricular tachycardia; NE, norepinephrine.
AJP-Heart Circ Physiol • doi:10.1152/ajpheart.00196.2014 • www.ajpheart.org
Downloaded from http://ajpheart.physiology.org/ by 10.220.33.5 on August 3, 2017
Up to 128 repositionable body surface potential map (BSPM)
electrodes were uniformly placed to cover both the anterior-lateral
chest up to the mid-axillary line and the posterior chest. The heart
was exposed via median sternotomy, and up to 47 transmural
plunge-needle electrodes were inserted in the LV and right ventricle (RV). Each LV plunge-needle electrode contains four bipolar
electrode-pairs (inter-electrode distance of 500 ␮m), each separated by 2.5 mm (28, 29), and each RV electrode contains eight
bipolar electrode-pairs with an inter-electrode distance of 500 ␮m
(30). There were no significant changes in heart rate or mean
arterial blood pressure after electrode insertion (heart rate, 99 ⫾ 8
to 101 ⫾ 2 beats/min; mean arterial blood pressure, 82 ⫾ 17 to 76 ⫾ 5
mmHg; both P ⫽ not significant). The chest and skin were then
carefully closed with silk suture, and the mapping electrode wires
were externalized above and below the sternotomy incision. Bipolar electrograms were continuously recorded from all electrodepairs together with body surface potentials from surface electrodes.
Spontaneously occurring ventricular arrhythmias including premature ventricular complex (PVC), couplet, and VT were recorded
during the mapping study. NE was later infused at 1.6 – 6.25
␮g·kg⫺1·min⫺1 to induce VTs in all three animals. At the completion of simultaneous mapping study, two sets of ultra-fast
computed tomography (UFCT) images were obtained on the living
animal to obtain the anatomical information. One without IV
contrast (with slice thickness of 3 mm) was used to construct the
torso model and extract the location of BSPM electrodes. Another
one with IV contrast (with slice thickness of 0.33 mm) was
obtained for construction of a detailed heart model and 3D localization of plunge-needle electrodes. The plunge-needle electrodes
were then carefully localized as previously described (9, 10) by
replacing each with a labeled pin. A postoperative UFCT scan of
the formalin-fixed canine heart was subsequently performed to
further facilitate precise 3D localization of the transmural electrodes.
3DCEI and data analysis. The physical-model-based 3DCEI
approach was used to noninvasively reconstruct the activation
sequence throughout the 3D ventricular myocardium. The forward
modeling and inverse computation of the 3DCEI approach were
previously described (8 –10, 21). Briefly, the realistic geometry
heart-torso model was constructed from UFCT images for each
animal. The 3D ventricular myocardium was discretized into
thousands of grid points. A distributed equivalent current density
model was used to represent the cardiac electrical sources within
the ventricular myocardium. Derived from the bidomain theory
(23, 35a), potentials measurable over the body surface are linearly
related to the myocardium equivalent current density distribution
given a tessellated geometrical heart-torso model that includes
myocardial tissue, blood masses, lungs, and torso. The QRS
segment for each beat was extracted to construct the BSPM matrix.
The start of Q wave and end of S wave were annotated from both
the root mean square ECG and butterfly plot of all ECGs. A
spatiotemporal regularization technique and lead-field normalized
weighted minimum norm estimation (37) were used to solve the
inverse problem to reconstruct the time course of local ECD at
each myocardial site. The activation time at each myocardial site
was determined as the instant when the time course of the estimated local ECD reached its maximum magnitude (21).
Numerical data are presented as means ⫾ SD. The performance of
3DCEI was evaluated by comparing the noninvasively imaged activation sequence with the measured activation sequence obtained from
3D intra-cardiac mapping. The measured activation sequence was
constructed in the same way as we did in normal dogs (10). The
Pearson’s product moment correlation coefficient (CC) was computed
to quantify the agreement of overall activation pattern between the
measured activation sequence and the noninvasively imaged activation sequence. The CC is defined as
H110
IMAGING CARDIAC ACTIVATION SEQUENCE DURING VT IN HF DOGS
the figure) had a slightly different activation pattern where the
initiation site has shifted to the low LV apex. Figure 2D shows
another activation pattern for the VT beat X26 (also indicated
in the red box of ECG lead II in the figure), where the initiation
site moved to posterior base of LV. Figure 2E shows VT beat
X30 had a focal initiation site at the subepicardium of inferior
lateral wall of LV. Such dynamic shift of initiation site has
been well captured in the noninvasively imaged activation
sequence.
Figure 3 shows a 14-beat polymorphic VT induced by NE
for dog 3. Five initiation sites were identified for this VT. As
shown in Fig. 3A, the first VT beat initiated at posterior basal
lateral wall of RV, and the initiation site has shifted to middle
lateral of LV for the following six VT beats (Fig. 3B for a
representative beat). The initiation site for eighth VT beat has
further moved to middle anterior LV wall, as shown in Fig. 3C.
The initiation sites were also identified at LV apex (Fig. 3D)
for beat 9 and beat 10 and at anterior LV apex (Fig. 3E) for
beats 11–14. The initiation sites were well identified in the
noninvasively imaged activation sequence.
Table 2 summarizes the spatial dispersion of focal initiation
sites and the quantitative comparison between imaged and
measured activation sequences for the mapped VT beats. The
NE-induced nonsustained VTs that were mapped had more
initiation sites, as compared with the spontaneously occurring
VTs (13 focal sites vs. 8 focal sites). Furthermore, out of the 16
focal sites, most initiation sites were located at apex (7 sites),
right ventricular outflow tract (2 sites), and left lateral wall (6
sites). The 3DCEI technique successfully detected and located
the dispersed focal initiation sites for these VT beats. Good
Fig. 1. A–C: comparison between measured activation sequence and imaged activation sequence for a 5-beat spontaneous ventricular tachycardia (VT) in dog
1. The activation sequence is color-coded from white to blue, corresponding to earliest and latest activation. The activation sequence is displayed on epicardial
and endocardial surfaces, respectively. The red box in ECG indicated the mapped beats. The initial site of activation is marked by a black asterisk and a purple
arrow. LV, left ventricle; RV, right ventricle; PVC, premature ventricular complex.
AJP-Heart Circ Physiol • doi:10.1152/ajpheart.00196.2014 • www.ajpheart.org
Downloaded from http://ajpheart.physiology.org/ by 10.220.33.5 on August 3, 2017
spontaneously occurring VTs (cycle length of 232 ⫾ 25 vs.
388 ⫾ 11 ms, P ⬍ 0.05). Both the 3D intra-cardiac mapping
and 3DCEI show these VT beats initiated by a focal activation,
mostly arising from different subendocardial sites of the ventricles.
Figure 1 shows an example of a spontaneously occurring
five-beat VT preceded by PVCs and sinus rhythm in a pattern
of bigeminy. As shown in Fig. 1A, a PVC beat initiated at
subendocardium of LV apex and the wavefront then terminated
at the basal posterior ventricles. Figure 1B shows the first VT
beat, which had the same activation pattern and same initiation
site as the preceding isolated PVC in Fig. 1A. Figure 1C shows
the second VT beat, where the initiation site has shifted to the
middle lateral wall of LV. This VT was maintained by shifting
the initiation site within the two sites (VT beat X1 and VT beat
X2 in Fig. 1, B and C) for the following three beats. The
noninvasively imaged activation sequence showed good agreement with the measured activation sequence, and the shift of
initiation site was well captured from the imaged maps.
Another example of a 31-beat VT induced by NE (in which
all VT beats were mapped and imaged) is shown in Fig. 2 for
dog 1. All the ectopic beats in this VT had focal activation
pattern. As shown in Fig. 2, A and B, the beginning of this VT
has been continuously firing in a monomorphic pattern at the
LV apex site slightly closer to anterior wall, which is similar to
the sites in the spontaneous VT in Fig. 1. However, starting
from middle stage of this VT, other initiation sites were also
observed among different ectopic beats, which were not seen in
the mapped spontaneously occurring VTs. Figure 2C shows
that VT beat X17 (indicated in the red box of ECG lead II in
IMAGING CARDIAC ACTIVATION SEQUENCE DURING VT IN HF DOGS
H111
correlation was obtained between the imaged activation sequence and the simultaneous measurement (averaged CC of
⬃0.7 over 135 ectopic VT beats from 16 focal sites). Furthermore, the initiation sites were reconstructed to be ⬃10 mm
from measured sites, suggesting good localization in a large
animal model with cardiac size similar to a human.
DISCUSSION
In this study, we report a novel investigation of 3DCEI for
characterizing the global activation pattern and localizing origin of activation during both the spontaneously occurring VTs
and NE-induced nonsustained VTs in a novel irreversible
model of nonischemic HF. We showed that both spontaneously
occurring VTs and NE-induced nonsustained VTs in this HF
model were initiated by a focal mechanism from multiple
endocardial, and at times, epicardial sites, which are similar to
VTs in human HF heart. Good agreement was obtained between the noninvasively imaged activation sequence and its
directly measured counterpart, as quantified by a CC of ⬃0.70
and an LE of ⬃10 mm averaged over 135 ectopic VT beats.
These findings imply that 3DCEI is feasible in noninvasively
characterizing the spatial patterns of ventricular activation
sequences, localizing the arrhythmogenic foci on a beat-to-beat
basis, and helping define the arrhythmia mechanism in the
setting of nonischemic HF.
Much effort has been devoted to the development of noninvasive cardiac electrical imaging techniques to estimate the
equivalent cardiac sources from BSPM, by solving the ECG
AJP-Heart Circ Physiol • doi:10.1152/ajpheart.00196.2014 • www.ajpheart.org
Downloaded from http://ajpheart.physiology.org/ by 10.220.33.5 on August 3, 2017
Fig. 2. A–E: comparison between measured activation sequence and imaged activation sequence for a 31-beat VT induced by norepinephrine (NE) in dog 1. The
red box in ECG indicated the mapped beats. The initial site of activation is marked by a black asterisk and a purple arrow.
H112
IMAGING CARDIAC ACTIVATION SEQUENCE DURING VT IN HF DOGS
Downloaded from http://ajpheart.physiology.org/ by 10.220.33.5 on August 3, 2017
Fig. 3. A–E: comparison between measured activation sequence and imaged activation sequence of a 14-beat nonsustained polymorphic VT induced by NE for
dog 3. The red box in ECG indicated the mapped beats. The initial site of activation is marked by a black asterisk and a purple arrow.
inverse problem (1- 3, 6 –15, 17–19, 21, 22, 25, 26, 32, 34, 36,
38, 40). To translate the cardiac electrical imaging techniques
into a clinically powerful tool, it is important to show their
clinical relevance and validate such techniques under the various conditions of clinically relevant cardiac diseases. Our
previous study (10) provided validation for the 3DCEI approach to assess NE-induced VT in control dogs, but it provided no information on the performance of this approach in a
failing heart. The present work extended the validation of
3DCEI into a novel irreversible large animal model of non-
ischemic HF induced by combined pressure and volume overload (unlike the most commonly used model of nonischemic
HF, the rapid pacing model, which is reversible following
cessation of pacing). Although the mapping approach is similar
to the previous study, the development of a new arrhythmogenic HF model that is irreversible and the simultaneous
closed-chest mapping studies with concurrent 3DCEI imaging
followed by postmapping UFCT scan are extremely challenging series of studies to perform in a failing heart. Meanwhile,
the VTs induced by NE in control hearts activated more rapidly
AJP-Heart Circ Physiol • doi:10.1152/ajpheart.00196.2014 • www.ajpheart.org
IMAGING CARDIAC ACTIVATION SEQUENCE DURING VT IN HF DOGS
Table 2. Summary of dispersed focal initiation sites and
quantitative comparison between measured and imaged
activation sequences for VTs in heart failure dogs
Origin Sites
VT Type
Localization
Error, mm
0.64
7.9
0.60
0.72
15.1
11.4
0.65
0.64
0.52
5.5
15.2
11.6
0.78
0.87
0.63
4.4
16.3
9.5
0.69
9.8
0.74
10.6
0.70
0.78
0.59
0.78
15.3
3.5
13.5
9.7
Dog No. 1
LVA
Anterior BLW
BLW
RVA
Spontaneous VT,
NE-induced VT
NE-induced VT
Spontaneous VT,
NE-induced VT
Spontaneous VT
NE-induced VT
Spontaneous VT
Anterior BRW
MRW
LVA
NE-induced VT
NE-induced VT
NE-induced VT
Posterior BRW
Spontaneous VT,
NE-induced VT
Spontaneous VT,
NE-induced VT
NE-induced VT
NE-induced VT
Spontaneous VT
Spontaneous VT,
NE-induced VT
NE-induced VT
Posterior LVA
MLW
Dog No. 2
Dog No. 3
LVA
Anterior LVA
RVA
Posterior BLW
MLW
Posterior MLW
Mean
0.63
0.69 ⫾ 0.09
12.8
10.7 ⫾ 4.0
Values are means ⫾ SD. BLW, basal left wall; BRW, basal right wall;
MLW, middle left wall; MRW, middle right wall; LVA, left ventricle apex;
RVA, right ventricle apex.
than those in failing hearts and VTs initiated from relatively
fewer sites in control hearts. Furthermore, the altered substrate
of the failing heart (e.g., cardiac enlargement, hypertrophy,
altered shape, remodeling) may also make noninvasive imaging of cardiac activation more challenging. Nevertheless, we
showed that the imaging performance was consistent with our
previous findings in the control dogs both in term of the
correlation of global activation pattern and localization of focal
initiation sites in such a large animal model with cardiac size
similar to human. Meanwhile, the more dispersed VT focal
sites in the failing heart were also successfully detected and
localized. In addition, our validation results indicate that the
novel imaging principle, which is based upon basic biophysics,
is applicable to imaging cardiac activation in failing hearts.
This represents an important finding in our effort in establishing noninvasive 3D cardiac activation imaging capability for
imaging cardiac diseases.
This study also shows the potential role of 3DCEI in complementing invasive cardiac mapping techniques to define the
arrhythmia mechanism. Previous invasive 3D intra-cardiac
mapping showed that spontaneously occurring PVCs and VTs
in a nonischemic HF rabbit model were initiated in the subendocardium by a nonreentrant mechanism (30). Focal mechanism was also observed in patients with idiopathic dilated
cardiomyopathy (28). In the present study, the findings for the
mechanism of spontaneously occurring arrhythmias were consistent with the previous invasive mapping results (20). Both
the spontaneously occurring VTs and the NE-induced nonsustained VTs initiated by a focal mechanism primarily from the
subendocardium. Most focal initiation sites were located at
apex, right ventricular outflow tract, and left lateral wall. These
findings indicate arrhythmias usually originated within the
subendocardium and may arise continuously from a single site
or consecutively from multiple sites located at Purkinje cells or
subendocardial myocardium. The spontaneously occurring
VTs could be attributed to the elevated activation of sympathetic nervous system in HF dogs. When compared with the
spontaneously occurring VTs, the NE-induced VTs were
faster, longer, and had more focal initiation sites. We also
identified a few VT beats induced by NE (3 out of 135 VT
beats) that initiated from the subepicardium. Furthermore, this
novel canine HF model exhibits no significant differences in
LV fibrosis compared with controls (S. M. Pogwizd, personal
communication). These mechanistic findings and their noninvasive assessment and characterization by 3DCEI will provide
the foundation for a wide range of studies in which this
noninvasive approach can be applied to other scenarios (such
as later studies in patients) where concurrent 3D intra-cardiac
mapping is not possible. The good correlation between the
measured and imaged activation sequences and the good localization accuracy of initiation sites imply that 3DCEI may
become an important clinical tool to benefit the clinical decision of the choice of the optimal interventional procedures
(e.g., during catheter ablation to help shorten the ablation time
and help with a critical question of whether an endocardial or
epicardial approach is needed, or during implantation of CRT
device to investigate the electrical synchronization for the
optimal placement of pacemaker leads) when treating HF
patients.
It is also noted that CC was used to quantify the overall
agreement between measured activation sequence and imaged
activation sequence, and we used common biostatistics rules to
interpret the value of CC. We realized that the interpretation of
the CC value could vary depending on the problem, and
therefore we compared the interpretation of the CC value with
other similar studies by other groups on other approaches of
cardiac inverse solutions (3, 16, 38). We believe that our
interpretation of CC value as good correlation is reasonable.
Conclusion
Both the spontaneously occurring and the NE-induced nonsustained VTs associated with nonischemic HF could have
focal activation pattern. The 3DCEI approach is feasible to
image the activation pattern and localize the initiation site of
nonsustained VTs in this animal model of nonischemic HF.
These findings could both contribute to the basic science
research regarding VT mechanism in HF as well as clinically
help in understanding better the mechanisms of VTs in patients
with nonischemic HF and thus choosing the patient-specific
optimal interventional procedures.
GRANTS
This work was supported in part by National Heart, Lung, and Blood
Institute Grants HL-080093 (to B. He) and HL-073966 (to S. M. Pogwizd) and
National Science Foundation Grant CBET-0756331 (to B. He). C. Han was
supported in part by a Predoctoral Fellowship from the American Heart
Association, Midwest Affiliate.
AJP-Heart Circ Physiol • doi:10.1152/ajpheart.00196.2014 • www.ajpheart.org
Downloaded from http://ajpheart.physiology.org/ by 10.220.33.5 on August 3, 2017
Correlation
Coefficient
H113
H114
IMAGING CARDIAC ACTIVATION SEQUENCE DURING VT IN HF DOGS
DISCLOSURES
B. He is an inventor of a patent granted to the University of Minnesota.
AUTHOR CONTRIBUTIONS
Author contributions: C.H., S.M.P., and B.H. conception and design of
research; C.H., S.M.P., L.Y., Z.Z., and C.R.K. performed experiments; C.H.,
S.M.P., L.Y., Z.Z., and B.H. analyzed data; C.H., S.M.P., and B.H. interpreted
results of experiments; C.H. prepared figures; C.H. drafted manuscript; C.H.,
S.M.P., and B.H. edited and revised manuscript; C.H., S.M.P., L.Y., Z.Z.,
C.R.K., and B.H. approved final version of manuscript.
REFERENCES
AJP-Heart Circ Physiol • doi:10.1152/ajpheart.00196.2014 • www.ajpheart.org
Downloaded from http://ajpheart.physiology.org/ by 10.220.33.5 on August 3, 2017
1. Armoundas AA, Feldman AB, Mukkamala R, Cohen RJ. A single
equivalent moving dipole model: an efficient approach for localizing sites
of origin of ventricular electrical activation. Ann Biomed Eng 31: 564 –
576, 2003.
2. Barr RC, Spach M. Inverse calculation of QRS-T epicardial potentials
from body surface potential distributions for normal and ectopic beats in
the intact dog. Circ Res 42: 661–675, 1978.
3. Berger T, Pfeifer B, Hanser FF, Hintringer F, Fischer G, Netzer M,
Trieb T, Stuehlinger M, Dichtl W, Baumgartner C, Pachinger O,
Seger M. Single-beat noninvasive imaging of ventricular endocardial and
epicardial activation in patients undergoing CRT. PLoS One 6: e16255,
2011.
4. Dawson B, Trapp RG. Basic and Clinical Biostatistics (4th ed). New
York: Lange Medical Books/McGraw-Hill, 2004.
5. Gepstein L, Hayam G, Ben-Haim SA. A novel method for nonfluoroscopic catheter-based electroanatomical mapping of the heart: in vitro and
in vivo accuracy results. Circulation 95: 1611–1622, 1997.
6. Greensite F, Huiskamp G. An improved method for estimating epicardial
potentials from the body surface. IEEE Trans Biomed Eng 45: 98 –104,
1998.
7. Gulrajani RM, Roberge FA, Savard P. Moving dipole inverse ECG and
EEG solutions. IEEE Trans Biomed Eng 31: 903–910, 1984.
8. Han C, Liu Z, Zhang X, Pogwizd SM, He B. Noninvasive threedimensional cardiac activation imaging from body surface potential maps:
a computational and experimental study on a rabbit model. IEEE Trans
Med Imaging 27: 1622–1630, 2008.
9. Han C, Pogwizd SM, Killingsworth CR, He B. Noninvasive imaging of
three-dimensional cardiac activation sequence during pacing and ventricular tachycardia. Heart Rhythm 8: 1266 –1272, 2011.
10. Han C, Pogwizd SM, Killingsworth CR, He B. Noninvasive reconstruction of the three-dimensional ventricular activation sequence during pacing and ventricular tachycardia in the canine heart. Am J Physiol Heart
Circ Physiol 302: H244 –H252, 2012.
11. Han C, Pogwizd SM, Killingsworth CR, Zhou Z, He B. Noninvasive
cardiac activation imaging of ventricular arrhythmias during drug-induced
QT prolongation in the rabbit heart. Heart Rhythm 10: 1509 –1515, 2013.
12. He B, Li G, Zhang X. Noninvasive imaging of cardiac transmembrane
potentials within three-dimensional myocardium by means of a realistic
geometry anisotropic heart model. IEEE Trans Biomed Eng 50: 1190 –
1202, 2003.
13. He B, Li G, Zhang X. Noninvasive three-dimensional activation time
imaging of ventricular excitation by means of a heart-excitation-model.
Phys Med Biol 47: 4063–4078, 2002.
14. He B, Wu D. Imaging and visualization of 3-D cardiac electric activity.
IEEE Trans Inf Technol Biomed 5: 181–186, 2001.
15. Huiskamp G, Greensite F. A new method for myocardial activation
imaging. IEEE Trans Biomed Eng 44: 433–446, 1997.
16. Khoury DS, Berrier KL, Badruddin SM, Zoghbi WA. Three-dimensional electrophysiological imaging of the intact canine left ventricle using
a noncontact multielectrode cavitary probe: study of sinus, paced, and
spontaneous premature beats. Circulation 97: 399 –409, 1998.
17. Lai D, Liu C, Eggen MD, Iaizzo PA, He B. Equivalent moving dipole
localization of cardiac ectopic activity in a swine model during pacing.
IEEE Trans Inf Technol Biomed 14: 1318 –1326, 2010.
18. Li G, He B. Localization of the site of origin of cardiac activation by
means of a heart-model-based electrocardiographic imaging approach.
IEEE Trans Biomed Eng 48: 660 –669, 2001.
19. Liu C, Eggen MD, Swingen CM, Iaizzo PA, He B. Noninvasive
mapping of transmural potentials during activation in swine hearts from
body surface electrocardiograms. IEEE Trans Med Imaging 31: 1777–
1785, 2012.
20. Liu C, Pogwizd SM. Focal mechanisms underlying spontaneous and
norepinephrine-induced ventricular tachycardia in a new arrhythmogenic
canine model of nonischemic heart failure. Heart Rhythm 9: S407, 2012.
21. Liu Z, Liu C, He B. Noninvasive reconstruction of three-dimensional
ventricular activation sequence from the inverse solution of distributed
equivalent current density. IEEE Trans Med Imaging 25: 1307–1318,
2006.
22. Martyn Nash P, Chris Bradley P, Paterson DJ. Imaging electrocardiographic dispersion of depolarization and repolarization during ischemia:
simultaneous body surface and epicardial mapping. Circulation 107:
2257–2263, 2003.
23. Miller WT, Geselowitz DB. Simulation studies of the electrocardiogram.
I. The normal heart. Circ Res 43: 301–315, 1978.
25. Mirvis DM, Keller FW, Ideker RE, Cox JW Jr, Dowdie RF, Zettergren DG. Detection and localization of multiple epicardial electrical
generators by a two-dipole ranging technique. Circ Res 41: 551–557,
1977.
26. Ohyu S, Okamoto Y, Kuriki S. Use of the ventricular propagated
excitation model in the magnetocardiographic inverse problem for reconstruction of electrophysiological properties. IEEE Trans Biomed Eng 49:
509 –519, 2002.
27. Packer M. Sudden unexpected death in patients with congestive heart
failure: a second frontier. Circulation 72: 681–685, 1985.
28. Pogwizd SM, McKenzie JP, Cain ME. Mechanisms underlying spontaneous and induced ventricular arrhythmias in patients with idiopathic
dilated cardiomyopathy. Circulation 98: 2404 –2414, 1998.
29. Pogwizd SM. Focal mechanisms underlying ventricular tachycardia during prolonged ischemic cardiomyopathy. Circulation 90: 1441–1458,
1994.
30. Pogwizd SM. Nonreentrant mechanisms underlying spontaneous ventricular arrhythmias in a model of nonischemic heart failure in rabbits.
Circulation 92: 1034 –1048, 1995.
32. Pullan AJ, Cheng LK, Nash MP, Bradley CP, Paterson DJ. Noninvasive electrical imaging of the heart: theory and model development. Ann
Biomed Eng 29: 817–836, 2001.
33. Schilling RJ, Peters NS, Davies DW. Simultaneous endocardial mapping
in the human left ventricle using a noncontact catheter: comparison of
contact and reconstructed electrograms during sinus rhythm. Circulation
98: 887–898, 1998.
34. Skipa O, Nalbach M, Sachse F, Werner C, Dössel O. Transmembrane
potential reconstruction in anisotropic heart model. Int J Bioelectromagn
4: 17–18, 2002.
35. Spragg DD, Dong J, Fetics BJ, Helm R, Marine JE, Cheng A,
Henrikson CA, Kass DA, Berger RD. Optimal left ventricular endocardial pacing sites for cardiac resynchronization therapy in patients with
ischemic cardiomyopathy. J Am Coll Cardiol 56: 774 –781, 2010.
35a.Tung L. A Bidomain Model for Describing Ischemic Myocardial DC
Potentials (Dissertation). Cambridge, MA: Massachusetts Institute of
Technology, 1978.
36. Wang L, Zhang H, Wong K, Liu H, Shi P. Physiological-modelconstrained noninvasive reconstruction of volumetric myocardial transmembrane potentials. IEEE Trans Biomed Eng 57: 296 –315, 2010.
37. Wang JZ, Williamson SJ, Kaufman L. Magnetic source images determined by a lead-field analysis: the unique minimum-norm least-squares
estimation. IEEE Trans Biomed Eng 39: 665–675, 1992.
38. Wang Y, Cuculich PS, Zhang J, Desouza KA, Vijayakumar R, Chen
J, Faddis MN, Lindsay BD, Smith TW, Rudy Y. Noninvasive electroanatomic mapping of human ventricular arrhythmias with electrocardiographic imaging. Sci Transl Med 3: 98ra84, 2011.
39. Zeppenfeld K, Schalij MJ, Bartelings MM, Tedrow UB, Koplan BA,
Soejima K, Stevenson WG. Catheter ablation of ventricular tachycardia
after repair of congenital heart disease: electroanatomic identification of
the critical right ventricular isthmus. Circulation 116: 2241–2252, 2007.
40. Zhang X, Ramachandra I, Liu Z, Muneer B, Pogwizd SM, He B.
Noninvasive three-dimensional electrocardiographic imaging of ventricular activation sequence. Am J Physiol Heart Circ Physiol 289: H2724 –
H2732, 2005.