Vol. 275, Issue 5, H1886-H1897, November 1998
SPECIAL COMMUNICATION
High-density epicardial mapping during current injection and
ventricular activation in rat hearts
Emilio
Macchi1,
Massimo
Cavalieri1,
Donatella
Stilli1,
Ezio
Musso1,
Silvana
Baruffi1,
Giorgio
Olivetti2,
Philip R.
Ershler3,
Robert L.
Lux3, and
Bruno
Taccardi3
1 Dipartimento di Biologia
Evolutiva e Funzionale, Sezione Fisiologia, and
2 Istituto di Anatomia ed
Istologia Patologica, Università degli Studi, 43100 Parma, Italy;
and 3 Cardiovascular Research
and Training Institute, University of Utah, Salt Lake City, Utah 84112
 |
ABSTRACT |
The purpose of this study is to report new
methods for manufacturing precision electrode arrays for recording
high-resolution potential distributions from epicardial surfaces of
small-animal hearts. Electrode arrays of 64 leads (8 × 8) and 121 leads (11 × 11) were constructed with a tulle substrate
to which insulated, fine silver wires (60-µm diameter) were attached
by knots at mesh node intervals of 540 × 720 µm. Insulation was
removed at the tips of the knots. Potential distributions and waveforms
were recorded from saline solutions and rat heart epicardium during ventricular paced beats and during passive current injection in the
diastolic interval. Electrical responses obtained from rat epicardium
compared favorably with those observed in studies of larger-animal
hearts, which used arrays having greater electrode spacing, and
revealed the effects of myocardial anisotropy. Epicardial potentials
measured early after stimulation in the region surrounding the pacing
site were interpreted in terms of potentials generated by an equivalent
quadrupolar source. We conclude that electrode arrays for epicardial
mapping of small hearts can be constructed with sufficient ease and
precision to allow detailed study of fiber structure and
electrophysiology in these hearts in normal and pathological conditions.
high-resolution cardiac mapping; epicardial electrode arrays; epicardial potential distributions in rat heart; cardiac anisotropy; ventricular paced beat
| |
INTRODUCTION |
KNOWLEDGE OF THE RELATIONSHIP between myocardial fiber
architecture, spread of excitation, and the associated extracellular potential is necessary for correctly interpreting ECG signals. This
relationship has generally been studied by recording epicardial potentials during pacing in normal dog hearts (28, 33), in hearts that
underwent remodeling, with nontransmural necrosis, or in the presence
of healed infarcts (16, 35). This same relationship has hardly been
studied in the heart of small animals despite their widespread use for
evaluating cardiac function in normal and pathological conditions and
for therapeutic purposes because of the need for high-resolution
electrode arrays to explore in detail a significant portion of the
limited epicardial surface.
The aim of the present study was to perform in vivo measurement of
epicardial potentials in the rat heart during current injection and
ventricular activity, with a view toward identifying fiber orientation
and architecture (19, 21) and defining patterns of ventricular
excitation in normal and pathological hearts. The rat was utilized in
our investigation in consideration of the finding that it has been used
more than any other animal for electrophysiological studies in vitro
and for ECG analysis in various heart conditions (3, 4, 12, 31, 32). A
preliminary study (2) utilizing 2-mm interelectrode distance epicardial
arrays over the entire ventricular surface revealed some limitation in
the resolution of potential distribution details that could be observed
on the ventricular surface of rat hearts during normal and paced
activity. To improve the resolution of detail on the limited surface of rat epicardium, epicardial electrode arrays with interelectrode distance as small as 540 µm were constructed. These electrode arrays
require a higher degree of accuracy in manufacturing than do
lower-resolution arrays because epicardial potential distributions are
affected to a greater extent by non-uniformities of array geometry,
electrode contact area, and electrode polarization when the
interelectrode distance is reduced.
We describe 1) the technique used to
manufacture high spatial resolution electrode arrays for recording good
quality extracellular potentials over limited epicardial areas,
2) in vitro tests of electrode
arrays by recording the potential distribution in response to a current
injection in saline solution, and 3)
the epicardial potential distribution during unipolar current injection
and ventricular activity in the rat heart in vivo.
The results indicate that high-resolution mapping arrays can be
successfully used on the hearts of small animals and that the features
observed in maps during current injection and ventricular activity are
similar to those observed in larger hearts. Thus the technique may be
exploited to estimate the passive and active electric properties of the
myocardium in normal hearts and in the presence of myocardial structure
altered as a consequence of various heart conditions such as
hypertrophy, ischemia, and infarction.
| |
METHODS |
Construction of Electrode Array
In a first attempt, electrode arrays for recording epicardial
potentials in rat hearts were fabricated by fastening the uninsulated ends of 60-µm-diameter silver wires to a patch of nylon sock at interelectrode distances of 500 µm as previously described for lower-resolution arrays in dog hearts (1). Silver wire was produced by
Metalli Preziosi (Paderno Dugnano, Italy), and an insulating layer of acetalic polyvinyl was applied by INVEX
(Quattordio, Italy). These high-density arrays exhibited two defects:
irregular geometry and large electrode contact surface area. Irregular
array geometry, as evidenced by uneven interelectrode distance,
resulted from nonuniform tension applied to elastic filaments of the
sock during manual fastening of electrodes. On the other hand, a large electrode contact surface area affected the distribution of current injected at the epicardium, which flowed preferentially through the
low-resistance epicardial liquid layer, and only a small fraction of
the current passed through the myocytes.
To overcome these difficulties, the previously used technique was
modified by adopting tulle as the stiff substrate for electrodes in
place of the nylon sock. Tulle is a sheer machine-made net with
hexagonal, rectangular, or rhombic mesh. It is made of cotton or nylon
and is used chiefly for veils and evening dresses and also for wrapping
nougats. The type of tulle we used has 20-denier nylon filaments
corresponding to 50-µm-diameter, 540 × 360-µm rectangular
mesh openings, with single filaments along one direction, and pairs of
closely apposed filaments in the orthogonal direction (Manifattura
Beccalli, Bosisio Parini, Italy) (Fig.
1A).
The relative stiffness of tulle net allows the construction of a
uniform electrode array with sufficient flexibility to conform to the
rounded epicardial surface. Uniform size knots with a loop diameter of
~250 µm (Fig. 1B) were obtained
at regular positions on the mesh (Fig. 1,
A and
C) according to the following
procedure. The end of the insulated silver wire was lightly lubricated
with petrolatum and fastened to the inelastic nylon substrate under a
dissecting microscope by pulling both ends of the knot with moderate
tension. The short end of the knot was subsequently twisted around the
long end and cut close to the loop. Thus the short end of the knot was
constrained away from the array to prevent damage of the myocardial
cells during epicardial recording. After all the knots were fastened, the electrode array was inverted, and the insulation was removed from a
limited area of each loop by means of a stripping paste (dichloromethane and methanol; Baldini Vernici, Porcari, Italy). Thus
each electrode consisted of an entirely insulated loop with an exposed
surface area of ~200 × 10
6
cm2 (Fig. 1,
B and
C; see APPENDIX
A) on the epicardial side of the array. At the end of
construction, silver wires from each row of the array were fastened
together into bundles. Tulle electrode arrays consisting of 8 × 8 and 11 × 11 rows and columns, with interelectrode distances of
1.08 × 1.08 mm and 540 × 720 µm, respectively, were
fabricated (Fig. 1, A and
C).
View larger version (116K):
[in this window]
[in a new window]
|
Fig. 1.
A: tulle net has 360 × 540-µm
rectangular mesh opening with single nylon filaments along rows and
adjoining pairs of nylon filaments along columns. Each electrode is
represented by a thick oblique segment at intersections of columns and
alternate rows. B: schematic drawing
of a silver wire knot. se, Short end of knot; le, long end of knot;
Dl, loop diameter
(250 µm); Dw,
silver wire diameter (60 µm). Black surface, estimated electrode
contact surface area equal to 200 × 106
cm2 (see APPENDIX
A). C:
photomicrograph of tulle array composed of 2 electrodes. Diameter of
tulle nylon filament is 50 µm, and interelectrode distance is 540 µm. Insulation is stripped at lower surface of electrodes.
Magnification, ×120.
|
|
Electrode Chloriding
Silver electrodes were chlorided to obtain uniform impedance values for
all array electrodes. The small surface area of the electrodes makes
empirical methods of chloriding difficult to use. Optimal chloriding
current density and time were determined by measuring electrode
impedance-frequency characteristics for different amounts of chloriding
currents expressed in milliamperes times seconds per square centimeter
(10, 20). Thus electrode chloriding consisted of the following steps.
First, each electrode was briefly chlorided, and its resistance was
monitored on the oscilloscope to verify that its value was in the
expected range. Subsequently, all electrodes were simultaneously
chlorided with optimal current density and time, and electrode
impedance was reduced and made stable in the 30-k
range (for details
see APPENDIX A). After preparing
>500 electrodes, as described in Construction of
Electrode Array, we were able to obtain electrodes of
reproducible size, shape, and impedance.
Mapping System
Electrodes were connected to AC-coupled, variable-gain differential
amplifiers of a 256-channel mapping system (8). The reference electrode
was positioned at the insulating boundary of a cylindrical saline bath
for in vitro measurements or on the root of the aorta in animal
experiments. Data were recorded at a bandwidth of 0.03-500 Hz,
input impedance of 1012 , and
sampling rate of 1 kHz/channel. No additional filtering was used to
minimize waveform distortion. The continuous flow of data from the
experiment, at the overall rate of 256 kHz with 12-bit resolution, was
handled by a double buffering technique, and data were stored
sequentially on a Quadra 800 Macintosh hard disk in real time.
Recording length could be selected between 1 s and the time interval
corresponding to maximum available contiguous disk space. The entire
set of waveforms was continuously monitored on the oscilloscope screen
during recording sessions. Calibration factors were computed by feeding
a reference triangular signal to all amplifier inputs and correcting
for equal mean square value on all output signals. The triangular
signal had a peak value spanning the amplifier output range for the
selected gain, 70-ms period, and 10-s duration to obtain significant
values of the calibration factors.
Data Analysis
Data were displayed off-line as waveforms and equipotential contour
lines on the graphic terminal and on a laser printer for quality
control. Contour lines were computed with a custom-written program. The
electrode array was triangulated by dividing the rectangular mesh with
lower left-upper right diagonals. Contour lines were drawn in each
triangle with linear interpolation. The choice of the diagonal may
affect the results of the interpolation locally in the presence of
steep spatial potential gradients. However, this simple representation
allows rapid detection of electrodes having poor contact. Contour lines
were not plotted in the six triangles having the current injection
electrode as the vertex, which was disconnected from the corresponding
amplifier during stimulation.
In Vitro Tests
The electrode array was immersed in a 0.9% NaCl solution. Unipolar
biphasic current pulses, 5-ms duration and 50- to 100-µA intensity,
were injected through one electrode, and potentials were recorded from
all other electrodes. Current injection and potential recording were
sequentially repeated for each array electrode. Data recorded in the
saline solution were displayed as waveforms and potential distributions
to evaluate electrode array performance. Waveform distortion on all
electrodes revealed polarization of the current injection electrode,
whereas waveform distortion at a single electrode indicated a high
impedance value of that electrode. On the other hand, potential
distributions revealed geometry irregularities within the array as
deviations from circular equipotential lines, with the center at the
current injection electrode. The finite volume conductor and position of the common return current electrode affected the shape of circular concentric equipotential lines.
In Vivo Recordings
Studies were done on eight healthy 1-yr-old rats of either sex weighing
300-400 g, anesthetized intraperitoneally with 5 µg/kg body
weight of fentanyl citrate and 250 µg/kg body weight of droperidol (Leptofen, Farmitalia-Carlo Erba, Milan, Italy). Additional amounts of
anesthetic were administered during the experiment as needed. Under
artificial respiration (rodent ventilator 7025, Ugo Basile, Comerio,
Italy), the heart was exposed through a longitudinal sternotomy. Body
temperature was maintained constant at 37°C with infrared lamp
radiation. The sternum was covered with a plastic sheet to maintain the
heart in a moist and constant- temperature environment. All procedures
performed on the animals conformed to the guiding principles of the
Veterinarian Animal Care and Use Committee of the University of Parma
(Parma, Italy). In each experiment, unipolar epicardial electrograms
were recorded with sock or tulle electrode arrays containing 64 or 121 electrodes. A 10-mm silver spiral electrode was sutured to the aortic
root as a reference for all unipolar recordings. The arrays were
usually positioned on the anterior surface of the heart and were kept moist by periodic addition of small amounts of warm saline (37°C). Unipolar stimuli were delivered between a single electrode on the array
and a common return current electrode, a heavily chlorided silver
spiral, inserted into the chest wall. Stimulus duration was 5 ms during
subthreshold current injection and 1 ms or less during pacing, with the
stimulus strength just above threshold. During the subthreshold current
injection, one of the unipolar electrograms was used to trigger a
stimulator, with a programmable delay of output current pulse of
variable phase, duration (100 µs to 10 ms) and intensity (10 µA to
10 mA) to be delivered just after the end of the T wave (4-channel
biomedical stimulator model 425 and biphasic stimulator model 220, Crescent Electronics, Sandy, UT). To ensure stable positioning of the
recording array and to minimize pressure over the epicardial surface,
the wire bundles were suspended by means of a horizontal rod that
reduced mechanical stress over the epicardium. At the end of the
experiment, the location of the array was marked by inserting pins into
the tissue at each corner of the array.
Morphological Study
After epicardial potential recording with the tulle electrode array was
completed, in five rats, the hearts were directly fixed in 10%
buffered Formalin for histological examination of the explored
epicardial region, and fiber direction was determined. In the remaining
three rats, in addition to fiber direction, the thickness of the
pericardial layer was measured. To this end, the abdomen was opened,
the abdominal aorta was cannulated with a PE-200 catheter filled with
phosphate buffer (0.2 M, pH 7.4) and heparin (100 UI/ml), and the
coronary vasculature was perfused for 10 min with phosphate buffer and
fixed with a solution containing 2% paraformaldehyde and 2.5%
glutaraldehyde for an additional 10 min. Small fragments of the
epicardial surface of the left and right ventricular myocardium close
to the area covered by the electrode array were postfixed in 1% osmium
tetroxide, dehydrated in acetone, and embedded in araldite. Thin
sections were cut and stained with uranyl acetate and lead citrate, and
pictures were taken at ×3,500 in a Zeiss 9M electron microscope
and printed at a final calibrated magnification of ×10,000.
Measurements of the visceral pericardial thickness were obtained from
eight pictures of each heart at 10-µm regular intervals.
| |
RESULTS |
In Vitro Tests
To simulate the in vivo conditions, the tulle electrode array was
positioned over the free surface of a fine circular sponge immersed in
saline solution. The lower side of the array was in contact with the
saline solution, and all electrodes exhibited uniform contact area.
Biphasic unipolar current pulses through one electrode generated
good-quality potential waveforms (Fig. 2A)
at all other electrodes. Potential distributions (Fig.
2B) generated by current injections
at selected electrodes at the center of the 8 × 8 array
demonstrated that the tulle electrode array has uniform geometry.
Equipotential lines were circular, with the center at the current
injection electrode, and exhibited only minor irregularities.
Specifically, the potential distributions displayed symmetry about the
midlines and main diagonals of the 4 × 4 map array (Fig.
2B). Moreover, the distributions
from the 16 maps were similar when spatially aligned to the stimulus
site. Potential distribution due to a current injection through the center electrode of the higher resolution 11 × 11 tulle array is
displayed in Fig.
3A. The
slightly defective symmetry of the circular equipotential lines
recorded in vitro by the different electrode arrays was
due to the influence of the saline-bounded volume conductor and finite
distance of the common return current electrode from the electrode
array.
View larger version (26K):
[in this window]
[in a new window]
|
Fig. 2.
A: potential waveforms recorded in
response to biphasic unipolar current injection in 0.9% NaCl solution
by 8 × 8 tulle electrode array with 1.08-mm interelectrode
distance, covering 0.53-cm2
surface area. Top waveform, 1-s
recording from 1 of the array electrodes during current injection
through electrode 29; rectangle, 45-ms
time interval for display of the array waveforms below. Positive and
negative current pulses had 5-ms duration, 100-mA intensity, and 20-Hz
frequency. Missing waveform at electrode
29 identifies injection electrode. Highest recorded
potential value was 13.7 mV. B:
potential distributions recorded in the same solution as in
A during positive phase (potential
waveform at bottom) of unipolar
current injection through a series of 16 electrodes in central region
of 8 × 8 tulle array. Potential distribution due to current
injection through electrode 29 (3rd
row, 2nd column) refers to waveforms in
A. In each map, spacing between
equipotentials is 1.6 mV, value of highest equipotential is 6.4 mV, and
potential value at current injection point is missing.
Bottom right square below maps shows 8 electrodes surrounding current injection electrode ( ) and oblique
triangulation of rectangular mesh. Equipotentials were not drawn within
the 6 triangles having current injection electrode as a vertex.
|
|
View larger version (48K):
[in this window]
[in a new window]
|
Fig. 3.
In vitro (A) and in vivo
(B) potential distributions recorded
with higher-resolution 11 × 11 tulle electrode array during
anodal current injection through center electrode
(electrode 61;  ). +, Positive
potential values at array sites. Equipotentials were not drawn within
the 6 triangles having current injection electrode as vertex. In each
map, nos. at bottom
(left to
right), absolute potential maximum,
increment between positive equipotentials, and absolute potential
minimum (all in µV). Reference electrogram refers to
electrode 60, just above stimulated
point. Vertical bar, time instant during anodal pulse.
|
|
In Vivo Recordings
A total of eight rats were used in this study. Four rats were used to
test and develop the sock electrode array, which proved unsatisfactory,
as explained in METHODS. These results
were discarded. The other four rats were used to test the tulle array.
Two rats enabled us to demonstrate the suitability of the 8 × 8 tulle electrode array both in vitro and in vivo. Similarly, two rats
were used to test the 11 × 11 tulle electrode array. This number
of experiments is sufficient to demonstrate the feasibility of the
method but is too small to justify a statistical analysis.
As soon as the tulle electrode array was positioned on the anterior
ventricular epicardium (Fig.
4A),
unipolar electrograms exhibited injury potentials (ST-T elevation; Fig.
4B), with a pattern similar to
monophasic action potentials. After a few minutes, ST-T elevation
diminished (Fig. 4C) to give rise to
stable, good quality epicardial electrograms. The ST-T elevation was
most likely due to injury caused by pressure exerted by the electrode
array on the delicate and thin visceral pericardium. In the rat heart, the visceral pericardium is composed of thin layers of collagen and a
mesothelial layer (Fig. 5). These layers
may contain vessels, nerves, and lymphatics. In our animals, the
thickness measured from the pericardial myocytes to the epicardial
surface varied from 1.0 to 10.0 µm, with an average value 3.5 ± 1.2 µm.
View larger version (17K):
[in this window]
[in a new window]
|
Fig. 4.
A: schematic representation of
anterior aspect of rat heart with 11 × 11 tulle electrode array
positioned over anterior ventricular surface. RV and LV, right and left
ventricular surfaces, respectively; RA and LA, right and left atrial
appendixes, respectively; PA, pulmonary artery; A, aorta; IVS,
interventricular septum. Inset within
11 × 11 electrode array refers to 3 × 5 electrode subset for
display of electrograms in B and
C. Nos. at corners, electrode no.
B: unipolar electrograms
exhibited ST-T elevation as soon as electrode array was positioned on
epicardium. Box: perspective view of perimeter of electrode
array composed of 3 × 5 electrode subset identified by top
left and bottom right corner electrode numbers. Numbers on
electrograms relate to corresponding electrodes at corners of 3 × 5 electrode subset. C: after 30 min,
stable, good quality epicardial electrograms were recorded.
|
|
View larger version (191K):
[in this window]
[in a new window]
|
Fig. 5.
Thin section of ventricular pericardium from a normal rat heart
demonstrating thickness of different layers of tissue separating
myocyte (right half) from epicardial
surface (left margin). Horizontal bar,
1 µm. Magnification, ×10,000.
|
|
Epicardial current injection.
Potential distributions recorded over the anterior ventricular surface
in response to unipolar anodal current injections with the 8 × 8 (Fig. 6) and 11 × 11 (Fig. 3B) electrode arrays displayed
elliptic equipotential lines, with the center at the stimulation point
and the common major axis parallel to fiber direction at the injection
site (19, 21). In vivo maps in Fig. 6 correspond to in vitro maps in
Fig. 2B because of current injection through the same 16 electrodes in the central part of the 8 × 8 array. Anisotropic
epicardial potential distributions were obtained even by injecting
low-density currents.
View larger version (62K):
[in this window]
[in a new window]
|
Fig. 6.
Epicardial potential distributions generated by anodal current
injections through the same 16 electrodes of 8 × 8 tulle array
positioned over anterior ventricular surface as in Fig.
2B. Spacing between equipotentials in
all maps is 700 µV, and value of highest equipotential is 4.9 mV.
Positive pulse in electrogram at
bottom refers to anodal current
injection of 5-ms duration and 100-µA intensity delivered during
sinus rhythm just after end of T wave. Major axis of elliptical
equipotential lines identifies average fiber direction in region
surrounding injection point. Bottom
right square below maps is defined in Fig.
2B.
|
|
Sinus rhythm. Epicardial potentials
recorded during sinus rhythm (as well as during ectopic activity; see
Ectopic activity) demonstrated the
ability of high-density electrode arrays to capture significant
features of the epicardial activation observed in hearts of larger
dimensions. The general time course and spatial distribution of
electrical events on the epicardium for the different activations were
similar in all experiments, although the details of the activation
pattern varied. Sinus rhythm electrograms in Fig. 4, sinus rhythm
isopotential maps in Fig. 7, and ectopic beat isopotential maps in Fig. 8 were
recorded by the 11 × 11 electrode array over the anterior
ventricular surface of the same rat.
View larger version (79K):
[in this window]
[in a new window]
|
Fig. 7.
Epicardial potential distributions recorded during sinus rhythm
activation by 11 × 11 tulle electrode array positioned over
anterior ventricular surface of rat heart (Fig.
4A).
A: potential map referring to 4 ms
from QRS onset. Dotted line, IVS (IVS line in Fig.
4A). In each map, reference
electrogram at bottom represents 50 ms. Vertical bar over reference electrogram is time instant (in ms)
from QRS onset to which map refers. Nos. at bottom
left, potential maximum (in µV) and electrode number
of potential maximum (in parentheses); nos. at bottom
right, potential minimum (in µV) and electrode number
of potential minimum (in parentheses). Step between equipotentials, 1 mV; continuous equipotentials, positive values including zero; dotted
equipotentials, negative values; + and  , position of absolute
potential maximum and minimum, respectively. Arrows in
C, 2 breakthrough points.
|
|
View larger version (54K):
[in this window]
[in a new window]
|
Fig. 8.
Potential distributions recorded during 12-ms interval after pacing
(A-H)
at center (electrode 61; ) of 11 × 11 tulle array positioned on anterior ventricular surface of
heart (Fig. 4A).  in
A: a subset of electrodes of tulle
array with 2-mm interelectrode distance. In each map, nos. at
bottom
(left to
right), absolute potential maximum,
increment between positive equipotentials, absolute potential minimum,
and increment between negative equipotentials (all in µV). Reference
electrogram, corresponding to 230 ms, refers to
electrode 60, just above stimulated
point. Vertical bar, elapsed time (in ms) after pacing.
|
|
The potential distributions of the sinus rhythm activation sequence of
normal epicardial tissue are illustrated in Fig. 7. Potential
distributions were entirely positive during the early stages of
activation on the surface explored as displayed in Fig. 7A (4 ms after QRS onset). At 5 ms
(Fig. 7B), one or more potential depressions appeared in the lower portion of the right ventricle. The
potential values in these depressions reached 10 mV at 6 ms
(Fig. 7C), and the region of densely
packed equipotential lines was considered to be the electrical
manifestation of an underlying activation wave front moving toward the
left ventricle and basal region of the right ventricle. At the same
time, one or more potential depressions appeared in the ventral aspect
of the right ventricle (Fig. 7C,
arrows). These events are the expression of activation wave fronts
emerging at two sites of the right ventricular surface (breakthrough
points). Meanwhile, another activation wave front moved from the free
wall of the left ventricle toward the right ventricle (Fig. 7,
D and
E, top
right corners). The activation of the ventral aspect of
the right ventricular surface was completed through merging, over the
interventricular septum, of two wave fronts, one coming from the right
ventricle and the other coming from the left ventricle (Fig.
7F). The maximum potential jump of
epicardial potentials across the wave front during sinus rhythm activation was 25 mV (Fig. 7C). The
sequence of events described was stable for several hours.
Ectopic activity. Ectopic beats were
elicited by delivering unipolar cathodal pulses at various sites of the
anterior ventricular surface. Pacing rate was slightly higher than
sinus rhythm, current density was just above threshold, and pulse
duration was 1 ms to avoid overlapping of the stimulus and early
activation potentials. Maps
A-H in Fig. 8
were recorded during paced activation from the center of the electrode
array. Two milliseconds after pacing (Fig.
8A), potential patterns appeared,
with negative potentials surrounding the pacing site and two positive
maxima located on opposite sides of the central region. A straight line
joining the two potential maxima was parallel to the fiber direction
near the stimulated point as verified by histological examination. Equipotential lines in the negative region were elongated, with the
major axis perpendicular to fiber direction near the pacing site at
this early stage of propagation, and the ratio between the absolute
values of the potential minimum and maximum was 1.5. Open circles in
Fig. 8A identify a subset of
electrodes of the 11 × 11 array with a 2-mm interelectrode
distance sampling the same area explored by the high-resolution array.
A lower-resolution electrode array fails to record clear-cut potential
patterns during the early stages of activation after epicardial pacing
as in dog hearts where well-defined potential patterns appeared only at 5-8 ms after the stimulus when the ratio between the potential minimum and maximum was ~6 (33). During the subsequent 4 ms (Fig. 8,
B-E),
the eccentricity of early negative equipotentials gradually shifted in
a direction parallel to the fibers. Eight milliseconds after pacing
(Fig. 8F), the negative
equipotentials became clearly elliptical, with the major axis parallel
to the fiber direction and the ratio between the potential minimum and maximum increased to ~6. The subsequent pattern of ectopic activation (Fig. 8, G and
H) was one of expanding negative
potential ellipses, with the major axis approximately parallel to the
fiber orientation. At this time, the two positive potential regions
that initially appeared on opposite sides of the central negative
region underwent changes, pointing to a progressive expansion and
rotation in a counterclockwise (CCW) direction (Fig. 8,
G and
H), whereas the two maxima
maintained their initial position, moving along a straight line. At 12 ms after pacing, the lower positive region moved completely outside the
array boundary (Fig. 8H), followed 6 ms later by the upper positive region (data not shown). The expansion
and rotation of the positive potential regions after epicardial pacing
are in agreement with previous findings in dog hearts (33).
Mathematical Modeling
To attempt interpreting epicardial potential patterns 2 ms after pacing
(Fig. 8A), we computed the
potentials generated by a linear quadrupole (Fig.
9), represented by two opposite dipoles separated by a small distance (18), immersed in an infinite homogeneous
anisotropic monodomain (see APPENDIX
B). The linear quadrupole is assumed to be an
equivalent generator of the activation wave front a few milliseconds
after pacing. The quadrupolar potential distribution in Fig.
9A was generated by two collinear,
opposing dipoles separated by 1 mm on a uniform grid, with points
spaced at a distance between grid points
(d) of 0.5 mm at a distance of 0.125 mm from the plane of the quadrupole. Simulated potentials were
displayed in a plane at a short distance from the sources because the
potential distribution in the source plane is characterized by the
presence of equipotential lines that cluster all around the sources due
to the steep potential gradient surrounding these points. On the
contrary, in a plane at a short distance from the sources, the
potential gradient decreases and is similar to the gradient of measured
potentials. Another reason for displaying potentials at a finite
distance from the sources is that the linear quadrupole is an
equivalent source that reconstructs the potential distribution at a
distance from the wave front. Potential distributions were also
computed at the same short distance from the quadrupole when
d was decreased by a factor of two (d = 0.25 mm; Fig. 9B) and four (d = 0.125 mm; Fig.
9C), and the explored area was reduced to the
inner squares B and
C, respectively, in Fig.
9A. Simulation results indicate that
equipotential lines in the interdipolar region are always elliptical,
with the major axis parallel to the quadrupole axis (Fig. 9,
B and
C) and that only inadequate spatial
sampling (Fig. 9A) fails to reveal
this pattern. Lower-value negative equipotential lines outside the
interdipolar region were always elongated, with the major axis
perpendicular to the dipole axis (Fig. 9,
A-C). Potential patterns similar to the ones displayed in Fig.
9C were generated by two opposing
dipoles oriented along a diameter and evenly spaced from the center of
a circular conducting medium (7).
View larger version (57K):
[in this window]
[in a new window]
|
Fig. 9.
A: potential distribution generated by
a linear current quadrupole, superposition of 2 dipoles, separated by 1 mm, in an infinite homogeneous anisotropic monodomain. Each dipole is
obtained as superposition of a point source-sink pair separated by
0.125 mm and centered at a grid point. Quadrupole is parallel to fibers
that are oriented horizontally
(x-axis), with axial symmetry
( y = z) and anisotropy ratio
(x/y = x/z = 2). Dotted inner squares B and
C represent areas displayed in
B and
C, respectively, with higher
resolution. Potentials were computed on a plane at a distance
(d) = 0.125 mm from plane of
quadrupole. B and
C: potential distributions generated
by the same linear quadrupole in A in
a plane at d = 0.125 mm from
quadrupole when distance between grid points was reduced 2 times in
B (d = 0.25 mm) and 4 times in C
(d = 0.125 mm).
D: 21 × 21 grid points uniformly
spaced (d mm), where potentials were
computed. ,  , and  pairs, dipole positions separated by 1 mm
when grid points are spaced at d = 0.5, 0.25, and 0.125 mm, respectively
(A-C,
respectively). In each map, nos. at
bottom
(left to
right), potential maximum, step
between positive equipotentials, potential minimum, and step between
negative equipotentials (all in arbitrary units). Note that ratio
between absolute values of potential minimum and maximum decreases to 1 by increasing grid resolution due to prevailing influence of single
dipole potentials.
|
|
The similarity between recorded (Fig.
8A) and simulated (Fig.
9A) potential patterns indicates
that 0.5-mm spatial sampling fails to explore, with fine details, the
area surrounding the stimulated epicardial point at an early stage of
activation and only shows negative elliptic equipotentials, with the
major axis perpendicular to local fiber direction. Because simulated
potentials were computed in a plane at a short distance from the plane
of the quadrupolar source, it follows from the similarity between recorded and computed potentials that the epicardial array of electrodes is also at a short distance from myocytes. We cannot quantify how much the visceral pericardium and the epicardial liquid
layer separately affect extracellular potential, which is a fraction of
transmembrane potential as a function of extracellular and
intracellular resistance. Because the electrodes are separated by only
a few micrometers from the myocytes through the visceral pericardium of
rat heart, it is most likely that the smoothing action of the
epicardial liquid layer plays a significant role in decreasing the
amplitude of the extracellular potentials and spatial potential gradients.
| |
DISCUSSION |
The results of the present investigation demonstrate for the first time
that, with a tulle net of 0.39 cm2
with 121 electrodes, the potential field created by current injection and propagating wave fronts can be measured on an area covering ~20,000 epicardial myocytes, the average dimensions of which in the
adult rat heart are 120 µm in length and 16 µm in diameter (17).
Thus the average number of epicardial myocytes, depending on fiber
orientation and neglecting the interstitial space, is 10 beneath the
electrode contact area, approximately equal to 20,000 µm2 (see
APPENDIX A), and is 200 in each 540 × 720-µm quadrant of the array. Such a high resolution makes it
possible to describe the effects of myocardial structure on cardiac
electrical events close to the microscopic scale in normal or
pathological conditions. Recorded epicardial potential patterns in rat
hearts were similar to those obtained in dog hearts during spontaneous
and paced ventricular activation (1, 33). As previously found in dogs
(18, 33), potential patterns observed early in activation after paced
beat can be interpreted in terms of an equivalent quadrupole model.
Construction of the High Spatial Resolution Electrode Array
The high-density electrode array made from a patch of nylon sock that
we used in preliminary experiments was unsatisfactory because of
irregular geometry arising from overcrowding of the silver wire knots
in a small area and an excessive electrode contact surface area. The
tulle electrode array, although less elastic, was sufficiently flexible
to follow the curvature of the epicardial surface of the heart and
established close contact with the epicardium without constraining the
ventricular mechanics. In addition, the stiffness of the tulle allowed
for the construction of electrodes with a limited surface area.
Although stripping only a limited surface area of the insulated silver
wire required the visual inspection of each electrode under the
stereomicroscope and was time consuming, the procedure was essential to
avoid short-circuiting effects in the presence of the epicardial liquid
layer. The impedance of the small contact area was minimized by
chloriding the electrodes. The good performance of the tulle array was
particularly apparent when measuring epicardial potentials during
subthreshold current injection through one of the electrodes. In these
measurements, polarization of the current injection electrode was
minimized by using biphasic current pulses.
Array Dimensions
Many groups (e.g., Refs. 6, 33) have studied epicardial potentials with
arrays characterized by different interelectrode distances and number
of electrodes. These techniques have been useful in understanding
large-scale spatial distribution of wave fronts and potentials as a
function of time. The description of the finer details of conduction,
however, requires higher-resolution recording. Spach and colleagues
(26, 27) have suggested that a complete understanding of microscopic
propagation requires mapping at a resolution approaching the dimension
of the individual myocytes. Electrode arrays used in dog hearts usually
sample epicardial potentials with a 2-mm or larger interelectrode
distance. One previous study (6) described the activation sequence over
dog epicardium with a high-density electrode array, with the electrodes evenly spaced 350 µm on a rigid mesh. High-resolution mapping of dog
epicardium was also reported for the interpretation of activation times
(22) and for in vivo estimation of cardiac transmembrane current (36).
The array used in our study has the highest reported spatial resolution
for a flexible epicardial electrode array. The high level of resolution
enabled us to display early activation patterns 2 ms after the onset of
the epicardial pacing stimulus (Fig.
8A). Despite this level of
resolution, the early potential distributions failed to detect the
central equipotential lines that are elongated along the fiber
direction as revealed by the theoretical model that simulates an
equivalent potential distribution generated by a linear quadrupole
(Fig. 9).
The use of a miniature flexible electrode system for epicardial mapping
may be extended to mouse hearts, in consideration of the growing
interest in genetically modified murine models (14). The anterior
aspect of a mouse heart measures ~3.5 × 4 mm. The shortest
interelectrode distance that can be obtained with the tulle used is 360 µm, and a 7 × 10 electrode array with a 540 × 360-µm
interelectrode distance can cover an epicardial surface area of 3.24 × 3.24 mm2. However, the
flexibility of such an array has to be tested in new experiments.
Epicardial Potential Response to Current Injection
Current injection has been used to assess passive electrical properties
of the myocardium such as interstitial and "gross tissue"
anisotropic resistivity (13, 23). Early work by Woodbury (38)
demonstrated anisotropic influence of myocytes on passive current flow
from one injection electrode. Clerc (5) measured longitudinal and
transverse intracellular and interstitial resistivities of an in vitro
calf trabecula preparation. Roberts and colleagues (23, 24) obtained
values for the tissue resistivities in vivo with a method similar to
the "four-electrode technique" (29). However, resistivity values
measured by various investigators are inconsistent (25). Kleber and
Riegger (15), who measured the electrical properties of
arterially perfused rabbit papillary muscle, suggested that the
relatively large differences in measured parameters may reflect the
shunting effect of the current through the thin superficial liquid
layer. Our results in rat hearts in vivo confirmed that the surface of
the current injection electrode in contact with the epicardial liquid
layer greatly affects measured potential distributions in response to
current injection, and the shunting effect of the fluid is minimized by
reducing the thickness of the liquid layer and the electrode area in
contact with the underlying myocytes. On the other hand, by recording epicardial potential distributions in response to a current injection over small epicardial areas, the passive electrical properties of
cardiac muscle may be estimated with greater detail. Specifically, the
eccentricity of elliptical equipotential lines obtained by high-density
epicardial mapping may provide information regarding the anisotropic
electrical properties of the tissue on the basis of the bidomain model
(13, 21). In addition to the measurements of tissue resistivity, the
technique identifies myofiber orientation at points of unipolar (21) or
bipolar (19) current injection.
High-density epicardial mapping may also provide useful information in
pathological conditions. In a previous study (34), myocardial
resistivity was shown to change dramatically when local ischemia is induced. Using a method based on the four-electrode technique, Steendijk et al. (30) demonstrated that, within
2 min of coronary occlusion, myocardial anisotropic electrical
resistivity increased and returned to the control value after
reperfusion. Fallert et al. (9) showed, with the same method, that
impedance mapping revealed significantly different values for normal,
ischemic, and infarcted tissues. Thus it is tempting to anticipate that high-quality, high-resolution epicardial potential recordings will make
it possible to define the passive electrical properties of cardiac
muscle in normal and pathological conditions. Epicardial potential
distributions, which we have shown to be measurable with our electrode
array, are known to be altered in a number of cardiac diseases, such as
myocardial infarction, ischemia, and conduction disturbances.
In addition, the distribution of injected currents and related
potentials, which we have shown to be measurable with our array, is
known to be altered in hearts with myocardial ischemia and
infarctions (9, 30, 34). Moreover, disparity of repolarization, an
arrhythmogenic condition, can be inferred from the distribution of the
QRST area on the epicardium. This variable, too, can be measured with
our electrode array. For instance, the recently described relationship
between beat-to-beat variability of ventricular repolarization and the
amount of ventricular fibrosis in rat hearts (31) may be examined more
closely by means of high-resolution epicardial mapping. Finally, very
little is known about the changes in myocyte volume and shape after
different loads are imposed on the myocardium and the potential
distributions are recorded on the epicardial surface. Similarly,
variations in the amount and composition of the interstitium may
greatly affect the distribution of electrotonic currents during action potential propagation along different directions, altering excitation potential patterns measured at the epicardium. In particular, during
ventricular reentrant tachycardia, high-resolution epicardial maps may
help in studying, with more details, the configuration of reentry
pathways as recently revealed in canine hearts (37).
Epicardial Ventricular Activation
Epicardial potentials recorded during sinus rhythm and stimulated
activity in the rat heart were consistent with the electrical activity
measured in the dog heart under similar conditions. Particularly, during sinus rhythm, multiple breakthrough points were present on the
anterior ventricular surface and initiated wave fronts, which moved
along preferential directions probably related to the myocardial fiber
direction (1). These wave fronts collided, thus terminating the
activation, in the basal region of the ventricles. Early after
epicardial pacing, the position of the epicardial minimum and two
maxima revealed the orientation of myocardial fibers near the pacing
site, whereas at later stages of activation, the CCW rotation and
expansion of the positive areas correlated with the helical spread of
excitation through CCW-rotating intramural fibers as previously
demonstrated in dog hearts (33, 35).
In summary, the major advantage of high-density epicardial mapping with
tulle electrode arrays is that stable recordings can be obtained for
several hours from the same epicardial area, and the electrode array
can be easily made and is durable. Moreover, because the epicardial
potential estimate reflects the spatial average over an electrode area
of ~200 × 106
cm2, our technique is particularly
useful to also explore small areas of myocardium in the hearts of
larger animals. The high sensitivity of this methodology should provide
information on the electrical correlates of the anatomic changes
occurring in several heart conditions such as hypertrophy, myocardial
ischemia, and infarction.
| |
APPENDIX A |
Optimal Electrode Chloriding
Geddes et al. (10) have studied the properties of silver-silver
chloride electrodes as a function of the thickness of the chloride that
was plated on the silver. For a piece of silver metal with a given
surface, thin layers of chloride reduced the impedance of the
electrode. When the layer of chloride was too great, the electrode
impedance started to increase again. The lowest electrode impedance was
produced when the plating current density was limited to ~5
mA/cm2 and the thickness of
chloride layer was limited in the range of 500-2,000
(mA · s)/cm2.
Contact area of our array electrodes was approximately equal to
one-half of the lateral surface of one-fourth of the loop of the knot
(Fig. 1B). Thus the electrode
contact area was estimated 1/2
Dw · 1/4Dl 
20,000 µm2 = 200 × 106 cm2, where
Dw is the 60-µm silver wire diameter and
Dl is the 250-µm loop diameter,
so that the 200-nA current intensity through the electrode surface area
corresponds to 1 mA/cm2 chloriding current density. Optimal
chloriding current density and time for tulle array electrodes was
estimated by evaluating electrode impedance in the 10-Hz to 10-kHz
range for different chloriding current densities and time intervals. It
was found that a 100-nA current intensity decreased electrode impedance to an average minimum value of 30 k in the 10-Hz to 10-kHz range for
a chloriding time interval of >1,200 s (Fig.
10). However, a current intensity of 200 nA or greater attained slightly lower impedance values that started to
increase after 600 s. Thus a current intensity of 100-200 nA
flowing through the electrode area during a 20-min interval was assumed
to be the optimal amount of chloriding current, corresponding to
a chloride layer thickness (i.e., charge density) of 600-1,200
(mA · s)/cm2.
View larger version (19K):
[in this window]
[in a new window]
|
Fig. 10.
Electrode impedance vs. frequency for bare electrode and after 1, 3, 6, 10, 15, and 20 min of chloriding, with a 125-nA current intensity
through electrode surface area of 200 × 106
cm2, corresponding to a current
density of 0.625 mA/cm2. Stable
impedance values of 30-50 k were obtained in 10- to 100-Hz
frequency range after 20 min of chloriding.
|
|
It has been reported (11) that by overchloriding for 2 min and then
dechloriding for 30 s a silver electrode, so as to deposit a layer of
chloride corresponding to two or three times the optimal amount of
chloriding current for that electrode, the resistance is lowered and
made more stable than that previously described. We also verified, for
our array electrodes, the validity of the proposed technique that was
particularly useful to stabilize electrode impedance during
long-lasting measurements of epicardial potentials.
| |
APPENDIX B |
Quadrupole Potentials in an Infinite Anisotropic Monodomain
The potential distribution generated by a linear current quadrupole was
computed as superposition of potentials generated by two equal strength
current dipoles, a short distance apart, aligned in the same direction
and with opposite orientation. The potential distribution generated by
a current dipole was computed as the superposition of potentials
generated by two equal-strength, opposite polarity, point current
sources (source and sink) separated by a small distance.
The field of a point current source is described by the equation
divJ = 4I
(r),
with current source I at the origin of
coordinates where divJ is the divergence of current density J in a volume conductor, (r) is the Dirac delta
function and r is the vector distance between source point
and field point. In an infinite homogeneous anisotropic medium,
Ji = ikEk = ik(
/xk),
where i, k = 1, 2, 3 and
x1 = x, x2 = y, x3 = z,
J1 and Ek are
the rectangular components of current density J and electric
field E, respectively,
ik is the conductivity tensor, is the scalar potential, and
/xk denotes partial
differentiation of with respect to
xk, and if we choose
x-,
y-, and
z-axes parallel to the principal axes
of conductivity tensor ik, we
obtain Poisson's equation for potential
|
(B1)
|
If we introduce new variables x = x'
,
y = y'
, and z = z'
, the equation changes into the following
form
|
(B2)
|
which is similar to Poisson's equation in an infinite
homogeneous isotropic medium if we substitute I' = I/
. Thus its solution is
|
(B3)
|
The infinite homogeneous anisotropic volume conductor is
characterized by fibers parallel to the
x-axis, axial symmetry about the
x-axis, and the anisotropic
conductivity ratio
(x)/(y) = (x)/(z) 
1.
| |
ACKNOWLEDGEMENTS |
We thank Carolina Panizzi, Matteo Gazza, and Michele Miragoli
(Dipartimento di Biologia Evolutiva e Funzionale, Università degli Studi, Parma, Italy) for experimental data analysis and electrode
chloriding measurements.
| |
FOOTNOTES |
This work was supported by grants from the Italian Ministry of
University and Scientific and Technological Research and the Italian
National Research Council; National Heart, Lung, and Blood Institute
Grant R01-HL-43276-09; and awards from the Nora Eccles Treadwell
Foundation and the Richard A. and Nora Eccles Harrison Fund for
Cardiovascular Research.
The costs of publication of this
article were defrayed in part by the
payment of page charges. The article
must therefore be hereby marked
"advertisement"
in accordance with 18 U.S.C. §1734 solely to indicate this fact.
Address for reprint requests: E. Macchi, Dipartimento di Biologia
Evolutiva e Funzionale, Sezione Fisiologia, Università degli
Studi, 43100 Parma, Italy.
Received 26 March 1998; accepted in final form 27 July 1998.
| |
REFERENCES |
1.
Arisi, G.,
E. Macchi,
S. Baruffi,
S. Spaggiari,
and
B. Taccardi.
Potential field on the ventricular surface of the exposed dog heart during normal excitation.
Circ. Res.
52:
706-715,
1983[Abstract/Free Full Text].
2.
Arisi, G., K. Signoretto, M. Manghi, and E. Macchi.
Potential and isochrone maps on the rat's ventricular surface
during normal and ectopic beat (Abstract). Int. Congr.
Electrocardiol. 19th Lisbon Portugal 1992, p. 103. (Int. Symp. Vectorcardiogr. 33rd)
3.
Assayag, P.,
F. Carrè,
B. Chevalier,
C. Delcayre,
P. Mansier,
and
B. Swynghedauw.
Compensated cardiac hypertrophy: arrhythmogenicity and the new myocardial phenotype. I. Fibrosis.
Cardiovasc. Res.
34:
439-444,
1997[Abstract/Free Full Text].
4.
Carrè, F.,
Y. Lessard,
P. Coumel,
L. Ollivier,
S. Besse,
Y. Lecarpentier,
and
B. Swynghedauw.
Spontaneous arrhythmias in various models of cardiac hypertrophy and senescence of rats. A Holter monitoring study.
Cardiovasc. Res.
26:
698-705,
1992[Medline].
5.
Clerc, L.
Directional differences of impulse spread in trabecular muscle from mammalian heart.
J. Physiol. (Lond.)
255:
335-346,
1976[Abstract/Free Full Text].
6.
Cohen, M. L.,
R. H. Hoyt,
J. E. Saffitz,
and
P. B. Corr.
A high density in vitro extracellular electrode array: description and implementation.
Am. J. Physiol.
257 (Heart Circ. Physiol. 26):
H681-H689,
1989[Abstract/Free Full Text].
7.
De Ambroggi, L.,
and
B. Taccardi.
Current and potential fields generated by two dipoles.
Circ. Res.
27:
901-911,
1970[Abstract/Free Full Text].
8.
Ershler, P. R.,
B. W. Steadman,
R. F. Wyatt,
R. L. Lux,
and
J. A. Abildskov.
A 64 lead online system for clinical and experimental mapping of ventricular activation sequences.
In: Computers in Cardiology., edited by Kenneth L. Ripley. Los Alamitos, CA: IEEE Computer Society Press, 1984, p. 417-419.
9.
Fallert, M. A.,
M. S. Mirotznik,
S. W. Downing,
E. B. Savage,
K. R. Foster,
M. E. Josephson,
and
D. K. Bogen.
Myocardial electrical impedance mapping of ischemic sheep hearts and healing aneurysms.
Circulation
87:
199-207,
1993[Abstract/Free Full Text].
10.
Geddes, L. A.,
L. E. Baker,
and
A. G. Moore.
Optimum electrolytic chloriding of silver electrodes.
Med. Biol. Eng.
7:
49-56,
1969[Medline].
11.
Grubbs, D. S.,
and
D. S. Worley.
New technique for reducing the impedance of silver-silver chloride electrodes.
Med. Biol. Eng. Comput.
21:
232-234,
1983[Medline].
12.
Hart, G.
Cellular electrophysiology in cardiac hyperthrophy and failure.
Cardiovasc. Res.
28:
933-946,
1994[Free Full Text].
13.
Henriquez, C. S.
Simulating the electrical behavior of cardiac tissue using the bidomain model.
Crit. Rev. Biomed. Eng.
21:
1-77,
1993[Medline].
14.
James, J. F.,
T. E. Hewett,
and
J. Robbins.
Cardiac physiology in transgenic mice.
Circ. Res.
82:
407-415,
1998[Abstract/Free Full Text].
15.
Kleber, A. G.,
and
C. B. Riegger.
Electrical constants of arterially perfused rabbit papillary muscle.
J. Physiol. (Lond.)
385:
307-324,
1987[Abstract/Free Full Text].
16.
Kramer, J. B.,
J. E. Saffitz,
F. X. Witkowski,
and
P. B. Corr.
Intramural reentry as a mechanism of ventricular tachycardia during evolving canine myocardial infarction.
Circ. Res.
56:
736-754,
1985[Abstract/Free Full Text].
17.
Liu, S.,
D. R. Hilbelink,
and
A. M. Gerdes.
Regional changes in hemodynamics and cardiac myocyte size in rats with aortocaval fistulas. 2. Long-term effects.
Circ. Res.
69:
59-65,
1991[Abstract/Free Full Text].
18.
Macchi, E.,
G. Arisi,
and
B. Taccardi.
Identification of ectopic ventricular foci by means of intracavitary potential mapping: a proposed method.
Acta Cardiol.
47:
421-433,
1992[Medline].
19.
Macchi, E., M. Manghi, G. Di Cola, M. Groppi, M. Pennacchio, P. Colli-Franzone, L. Guerri, B. Taccardi, and R. L. Lux.
Identification of ventricular fiber architecture from electrotonic
epicardial potentials elicited by dipolar stimulations: numerical and
experimental models. Med. Biol. Eng.
Comp. 34, Suppl. 1:
83-84, 1996.
20.
Moore, K. B.,
T. Kimball,
and
B. Steadman.
Silver-silver chloride plunge electrode needles and chloriding monitor.
IEEE Trans. Biomed. Eng.
37:
532-535,
1990[Medline].
21.
Muzikant, A. L.,
and
C. S. Henriquez.
Paced activation mapping reveals organization of myocardial fibers: a simulation study.
J. Cardiovasc. Electrophysiol.
8:
281-294,
1997[Medline].
22.
Ndrepepa, G.,
E. B. Caref,
H. Yin,
N. El-Sherif,
and
M. Restivo.
Activation time determination by high-resolution unipolar and bipolar extracellular electrograms in the canine heart.
J. Cardiovasc. Electrophysiol.
6:
174-188,
1995[Medline].
23.
Roberts, D. E.,
L. T. Hersh,
and
A. M. Scher.
Influence of cardiac fiber orientation on wavefront voltage, conduction velocity, and tissue resistivity in the dog.
Circ. Res.
44:
701-712,
1979[Abstract/Free Full Text].
24.
Roberts, D. E.,
and
A. M. Scher.
Effect of tissue anisotropy on extracellular potential fields in canine myocardium in situ.
Circ. Res.
50:
342-351,
1982[Abstract/Free Full Text].
25.
Roth, B. J.
Electrical conductivity values used with the bidomain model of cardiac tissue.
IEEE Trans. Biomed. Eng.
44:
326-328,
1997[Medline].
26.
Spach, M. S.,
P. C. Dolber,
and
J. F. Heidlage.
Influence of the passive anisotropic properties on directional differences in propagation following modification of the sodium conductance in human atrial muscle: a model of reentry based on anisotropic discontinuous propagation.
Circ. Res.
62:
811-832,
1988[Abstract/Free Full Text].
27.
Spach, M. S.,
W. T. Miller,
D. B. Geselowitz,
R. C. Barr,
J. M. Kootsey,
and
E. A. Johnson.
The discontinuous nature of propagation in normal canine cardiac muscle: evidence for recurrent discontinuities of intracellular resistance that affect the membrane currents.
Circ. Res.
48:
39-54,
1981[Free Full Text].
28.
Spach, M. S.,
W. T. Miller III,
E. Miller-Jones,
R. R. Warren,
and
R. C. Barr.
Extracellular potentials related to intracellular action potentials during impulse conduction in anisotropic canine cardiac muscle.
Circ. Res.
45:
188-204,
1979[Abstract/Free Full Text].
29.
Steendijk, P.,
G. Mur,
E. T. van der Velde,
and
J. Baan.
The four-electrode resistivity technique in anisotropic media: theoretical analysis and application on myocardial tissue in vivo.
IEEE Trans. Biomed. Eng.
40:
1138-1147,
1993[Medline].
30.
Steendijk, P.,
A. D. van Dijk,
G. Mur,
E. T. van der Velde,
and
J. Baan.
Effect of coronary occlusion and reperfusion on local electrical resistivity of myocardium in dogs.
Basic Res. Ca
Top
Abstract
Introduction
