Download The Electric Heart Center for the QRS Complex in Cardiac Patients

The Electric Heart Center for the QRS
Complex in Cardiac Patients
By GEORGE E. SEIDEN,
M.D.
A modification of Frank's cancellation data-matching method for location of the QRS electric
heart center in the human was applied to 40 patients, most of whom had advanced cardiac disease.
The equivalent dipole locations for the QRS complex were found to cluster in a 5 cm. by 5 cm.
square area forward and to the left in the transverse plane. Correlation of electric heart center
location with orthodiagraphic estimate of the center of ventricular mass was poor in many of the
cases.
Downloaded from http://circres.ahajournals.org/ by guest on July 31, 2017
T
tion of the equivalent dipole for the QRS complex in 40 patients, most of whom suffered from
advance cardiac disease, in order to assess the
design requirements for an orthogonal lead
system. The method of Frank 8 was used after
slight modification which enabled us to shorten
the experimental time. Dipole location was
compared in most subjects with the respective
anatomic location of the center of ventricular
mass, as determined by orthodiagraphy, to
discover to what degree the relatively short
fiuoroscopic procedure could approximate results of the more lengthy electric heart center
determination.
HE feasibility of measuring QRS dipole
components in patients by use of body
surface leads is indicated by experimental evidence in support of theoretical assumptions of homogeneity of body tissues1' 2
and fixed location equivalent dipole representation of ventricular depolarization.'"6 Construction of leads to measure dipole components has
been based on representation of the body as a
sphere with centric dipole, or on other simplified schemes. Application of these schema results in large errors as has been demonstrated
by Burger and van Milaan," Schmitt, 7 and
Frank. 1 They concur that body parameters of
torso shape and dipole location must be taken
into account in the construction of accurate
orthogonal leads, and indeed that dipole location is a decidedly more critical parameter than
body shape. Frank has described a method for
locating the dipole in humans, which was accurate to ± }/i cm. in a single normal subject.8
His method employs a precision, four electrode
cancellation technic that fits potentials around
the chest of a human to dipole potentials produced in a homogeneous torso model with
known dipole location.
The present study was made in an attempt
to discover experimentally the anatomic loca-
METHOD
In order to study cardiac patients, Frank's heart
center method was simplified. The 7 to 10 hours of
experimental time was shortened to 2 to 3 horn's by
modifying the method for acquiring heart level cancellation data. Standard chest electrodes were used
instead of needles and a protractor rather than
a tailor made vest was employed to locate chest electrode positions. Frank's method consisted of placing
electrodes 1 and 2 of the four electrode cancellation
system (fig. 1) at fixed positions on the chest at the
level of the center of the heart. They were connected
across a potentiometer with a fixed tap R, which
served as afixedreferenceterminal during the entire
procedure. Electrode 4 was placed at heart level, and
its pattern examined with respect to R. Search electrode 3 was used to explore the body surface at the
same heart level for a mirror pattern of that recorded
from 4. Cancellation was attempted by connecting 3
and 4 across a potentiometer whose tap S could be
adjusted for smallest voltage between S and reference
R. Further body exploration with electrode 3 and
variation of potentiometer top S were conducted
until there was minimal QRS potential difference
between R and S. After achieving a cancellation in
this manner, electrode 4 could be moved to a new
From the Edward B. Robinette Foundation and
Medical Clinic, Hospital of the University of Pennsylvania, Philadelphia, Pa.
This investigation was supported in part by Grant
H-339, U. S. Public Health Service.
An abstract of this paper entitled "Anatomic Location of the Electric Heart Center in Pationts," appeared in Circulation, 12: 773, 1955.
Received for publication December 27, 1955.
313
Circulation Retetmk, Volume TV, May 1966
314
ELECTRIC HEART CENTER FOR QRS COMPLEX
Downloaded from http://circres.ahajournals.org/ by guest on July 31, 2017
Fio. 1. Schematic representation of the four electrode cancellation system.
position at the transverse level, and the procedure
repeated so as to achieve further cancellations with
respect to fixed terminal R. Potentiometer settings
for the fixed tap of 1 and 2, and for the various electrode pairs which yielded cancellation against this
tap were plotted on transverse image loops8 obtained
from known di]X>le locations in a homogeneous
human shaped model. The transverse image loop on
which the plotted potentiometer settings formed the
tightest cluster around the setting of the fixed 1-2
tap established a dipole location in the model equivalent to that of the human subject. This method,
though time consuming, was remarkably accurate in
a single normal subject.8 Bod}' surface exploration
with electrode 3 was the most time consuming
feature.
Frank's procedure was modified by accomplishing
three or more cancellations with a minimum of body
surface exploration. The four electrodes were held by
a belt in fixed positions at the level of the center of
the ventricle, as estimated byfluoroscopy,so that
electrode 4 was anterior to the anatomic line joining
1 and 2, and electrode 3 was posterior to this line.
Cancellation was sought by systematic adjustment of
the potentiometers and, if required, minor adjustments in the level of the belt; in this manner the time
consuming body search for mirror patterns was circumvented. After one cancellation was achieved, one
or more of the electrode positions could be altered for
a new cancellation in which both potentiometers
were again used as search variables.
Potentiometer settings for each cancellation could
then be plotted on transverse image loops characteristic of different dipole locations. Since there was no
fixed terminal against which all cancellations were
made, there was no point on the image diagrams
around which the plotted potentiometer settings
were expected to cluster. Therefore, the criterion for
the selection of the image loop of bestfitwas the degree of proximity of the plotted pairs of potentiome-
ter settings for each independent cancellation. Expressed mathematically, the appropriate image loop
was that for which the sum of the squares of the
image vectors joining the plotted potentiometer
settings for the various cancellations was smallest.
This modification of Frank's method is subject to
difficulty in patients with narrow transverse vector
loops (when the long axis of the loop exceeds about
six times the short axis). In such patients, there will
be a direction in which wide separation of the plotted
pairs of potentiometer taps of some of the cancelling
electrode pairs will be encountered. The reason for
this is as follows: Cancellation depends upon minimum residual potential difference between cancelling
taps. Potential difference between cancelling taps is
expressed on image loops as the product of distance
between taps times the dipole component in the
direction of a line joining the taps. Therefore, apparent cancellation may occur with separate taps if they
are aligned in the direction of a very small dipole
component; that is, in a direction perpendicular to
the long axis of a very narrow vector loop. In 3 of the
40 patients who had narrow vector loops in the transverse plane, it was necessary to recognize the occurrence of separation of taps of cancelling pairs due to
small QRS dipole component in the direction of
separation in order to find an image diagram which
satisfied all the cancellation data. In these 3 patients,
Tran3versa
imago loop
FIG. 2. Transverse image loop (to scale) representative of model dipole location 23. Axes are marked in
image space units." For a patient with myocardial
infarction, the two potentiometer settings of each
cancellation have been plotted and joined by a solid
line to indicate their degree of proximity. The proximity of the plotted taps for each of six cancellations
indicates that this image loop is characteristic for the
patient. This method was used in 37 of the 40 patients
studied, and is independent of knowledge of transverse plane dipole components.
315
SEIDEN
Downloaded from http://circres.ahajournals.org/ by guest on July 31, 2017
Transverse
image loop
Fia. 3. Image loop identification in another patient
with myocardial infarction despite wide separation of
taps for 2 of 6 cancellations. This presumptive identification depends upon the demonstration of a small
dipole component in the direction of wide separation
of taps. Insert It shows relative peak amplitudes of
dipole components in indicated directions in this
case. This presumptive method was used in 3 of the
40 patients studied. Unliko the case in figure 2, this
method requires knowledge of transverse plane dipole
components.
the dipole was presumptively located by determining
an image loop for which all widely separated taps
were in a direction of small QRS dipole component.
Figure 2 illustrates dipole location by the modification of Frank's method based on the proximity of
cancelling tap pairs as plotted on the image diagram
characteristic of the subject's QRS electric heart
center. Figure 3 illustrates a case in which dipole
location was pi-esumptively accomplished despite
separation of plotted taps due to small dipole
component.
PROCEDURE
A standard V lead electrocardiogram was recorded
for each patient, and orthodiagraphic trace of heart
borders in anteroposterior and lateral views (when
possible) was made. By means of a rubber belt, electrodes were positioned around the patient's body at
thefluoriscopicallydetermined level of the center of
the ventricular mass. The patient's body position
was kept constant by a specially constructed chair
with an adjustable back brace. DC resistance
between each electrode site and ground was made
less than 5000 ohms by brisk skin rubbing. Cancellation was then attempted as depicted in fiigure 1, by
systematic adjustment of potentiometer taps S and
RR to give minimum potential difference. If calculated cancellation coefficients1 were less than 10 per
cent, potentiometer settings, and electrode positions
with respect to an angular coordinate system marked
on the patient's chest8 wererecorded.A new cancellation was then attempted for different electrode sites
at belt level. Three or more cancellations were made
for each heart center determination. If coefficients
greater than 10 per cent were obtained, improvement
was attempted by upward or downward adjustment
of the belt by one or two inches.
The image counterparts of potentiometer taps of
cancelling electrode pairs were plotted on transverse
image loops9 for various dipole locations. The proper
dipole location was indicated by the geometric construction for which the sum of the squares of the distances between pairs of cancelling taps was smallest.
In 3 coses with discordant cancellations due to narrow transverse plane loops, distances between the
separated taps of discordant cancellations were
normalized prior to calculation of the sum of the
squares. This was accomplished by multiplying these
distances by the ratio of the dipole component in
their direction over the largest measured QRS component in the transverse plane. Heart centers thus
determined were represented by a point on the
anatomic transverse section of the torso model of
Frank.
Fluoroscopically determined heart centers were
located on the model's transverse section as follows:
In the frontal and saggital planes, orthodingraphically determined distances between midline and
mid-point between borders of the cardiac silhouette
were made applicable to the model by multiplying
them by a normalization factor consisting of the ratio
of model diameter to patient diameter for the
particular plane in question.
RESULTS
The number of cancellations used for each
heart center determination are listed in table 1.
In all, 188 different cancellations were accomplished at heart level in 43 patients. Cancellation coefficients, expressed as 4 times the
residual, divided by the sum of the potential
differences between the member electrodes of
the cancelling pairs, are also listed in the table.
Coefficients averaged 18 per cent or less in all
patients, 15 per cent or less in 97 per cent of
patients and 10 per cent or less in 63 per cent of
patients.
The validity of the heart center determination procedure was indicated by the degree of
fit in each case between subject data and data
obtained from a model with a known dipole
location. The degree offit,in turn, is indicated
by the average of the distances between cancelling taps as plotted on the image diagram
ELECTRIC HEART CENTER FOR QRS COMPLEX
316
TABLE 1.—Continued
TABLE 1.—Heart Center Location Data for Cardiac
<
Z
Normal.
Downloaded from http://circres.ahajournals.org/ by guest on July 31, 2017
Congenital heart
disease. Right
ventricular
hypertrophy.
Hypertension.
Left ventricular hypertrophy-
>
W
I-o 3
<
5
S
10
14
S
21
04
04
43
4
14
10
15
13
19
30
42
27
03
23
40
32
34
44
40
32
32
6
4
3
4
4
5
5
8
S
5
7
32
2ll
^repeat
3
16
3
16
26
19
24
27
14
45
21
24
14
13l
8
3
3
3
S
4
10
17
17
Coronary artery
disease. Right
bundlo branch
block.
4
5
6
15
12
Rheumatic heart
disease. Mitral
stenosis.
4
8
11
51
23
21
322
151
03
12
23
56
3-
31
24
27
27
24
27
35
16
29
20
4
4
6
3
6
7
12
10
11
8
10
12
9
12
Takaynsu's disease.
5
8
Luetic heart
disease. Left
ventricular hypertrophy.
3
4
5
S
Coronary artery
disease. Anterior wall infarction.
3
12
12
13
3
3
5
81
3
W
Coronary artery
disease. Posterior wall infarction.
10
6
7
32
18
21
32
23
22
3-
Coronary artery
disease. T wave
abnormality.
14
10
9
7
3
9
43
14*
27
10
32
26
81
03
34
23
12
23
211
31
135
36
2-
* Distances between taps were normalized in
starred cases, as explained in text.
f Code number refers to location in the transverse
plane of Frank's human shaped model. The first
digit indicates location along X axis; the second
along Z axis (fig. 4).
>repeat
21* 03
51 02
14 04
9
Diagnosis
natomi
code n
|g
8
5
3
4
4
Coronary artery
disease. Left
bundle branch
block.
•g
lectric heart center c
numbe rt
Co
umber
plane i
Diagnosis
verage canceillation
coemci ent ini per cen
inaverse
lations
§
rt centei
rt
Patients
70\
nj./repeat 20
00
42
53
11
72\
_0J repeat
203
42
30
30
13
11
2-
34
35
20
32
123
30 43
11 31
18* 11
13
10
27
representative of the subject's dipole location.
The smaller the average distance, the better the
fit; average of zero indicates a perfectfit.The
average of the distance between taps for each
of forty cases is shown in table 1. The mean
average distance is 24 units of image space.9
The resolution of the location procedure is
indicated by the loss offitincurred when patient data in each case is matched to model
data for dipole locations 1 and 2 cm. distant
from the best location. For locations 1 cm.
distant from the determined location, the mean
average distance between taps is 29 units; for
locations 2 cm. distant, the mean is 51 units.
In each case, the dipole locations 1 and 2 cm.
distant were selected conservatively in the direction giving data—fitting closest to the best
fit. This analysis indicates that resolution of the
location procedure is ± 1 cm. in the transverse
plane.
The method was successful in 40 patients.
The determined locations are shown on a transverse anatomic section of the human torso
model in figure 4. Ninety-three per cent of results fall within a 5 cm. by 5 cm. area. The
center of the cluster is at anatomic location 22.
317
SEIDEN
,1-1,cm.
-T
1
1
X location number
1 2 3 4 5 6 7 6
| |1 | 1 | 1 | 1-|
1 1 1
l-7cm.
-1-04- • 3 • 2 •
i-L * * » •
•
14 • • •
4 4 *J
IS
tS)
Downloaded from http://circres.ahajournals.org/ by guest on July 31, 2017
E
D
Front
FIG. 4. Electric heart center anatomic locutions for
40 patients are plotted to scale on a transverse section
of the left anterior quadrant of the Frank model of a
human torso. More than one result at a given point is
indicated by number alongside the point. Location
code numbers are spaced 1 cm. apart. Thorax width
is 33 cm.; thorux depth is 25 cm.
Heart center locations for the 40 patients are
given in table 1.
In 4 of the 40 patients, an independent repeat
determination was done 2 to 3 months after the
initial study, which yielded locations in 3 cases
that were within 1 cm. of the original location,
and in the remaining case was within 2 cm.
In all, 43 different individuals were studied.
In 3 patients it was not possible to match cancellation data from the transverse level with
available model data. Average cancellation coefficients in these cases were 15 per cent, 17 per
cent, and 19 per cent. It has not been possible
to restudy those cases.
Orthodiagraphic estimates of the centers of
ventricular mass are listed in table 1, and compared with the electric heart center location.
In the X axis direction (from right toward left),
the average fluoroscopic heart center location
was 2.5 cm. to the left of position 00 in the
model, with standard deviation of 1.2 cm., as
compared with average electric heart center
location of 2.3 cm. to the left of model position
00, with standard deviation of 2.0 cm. In the Z
axis direction (from front to back), the average
fluoroscopic location was 2.5 cm. forward of
position 00 in the model, with a standard deviation of 2.1 cm. as compared with average
electric heart center of 2.1 cm. forward of 00,
with standard deviation of 1.2 cm.
Agreement of locations as detei mined by
both methods was good in some individual
cases and poor in others (table 1). The average
distance between electric and fluoroscopic centers in individual cases was for the X direction
2.0 cm. with standard deviation of 1.6 cm., and
for the Z direction, 1.6 cm., with standard deviation of 1.3 cm.
DISCUSSION
This method of electric heart center determination (for QRS complex) depends upon the
validity of single fixed location dipole representation of ventricular depolarization. The validity of this assumption is indicated by examining
the cancellation coefficients in this study. The
average coefficient was 12 per cent as opposed
to 0 per cent for perfect dipole representation.
Some other sources of error in electric heart
center determination include (1) inhomogencities within the patient's body, (2) error in
location of the dipole level in patients, (3) error
in location of angular electrode sites on the
body, (4) errors due to variation in patient's
posture and respiration during the procedure
and (5) difference in configuration between
torso model and subject's body build. Nevertheless, despite these errors, this method was
capable of resolution of dipole location to within
± 1 cm. in the transveise plane.
Fluoroscopic estimate of the center of ventricular mass suffers from numerous well known
errors, chief of which is the difficulty of determining locations of the ventricles within the
cardiac silhouette. In this study, no attempt
was made to do this; rather the midpoint
between the cardiac borders was assumed to be
the center of the ventricles, and errors inherent
in this procedure were accepted. In lateral
view, orthodiagraphic technique is more difficult because the heart borders are difficult to
discern. For many of the patients, it was not
possible to make an accurate estimate of the
heart center in the sagittal view.
In cases no. 13, 25, and 31, marked discrepancy (5, 5, 4 cm., respectively) between
fluoroscopic and electric heart center locations
might be attributed to the presence of aneurys-
318
ELECTRIC HEART CENTER FOR QRS COMPLEX
Downloaded from http://circres.ahajournals.org/ by guest on July 31, 2017
mal left auricles of mitral stenosis and insufficiency. The fluoroscopic estimate of the ventricular center, would be expected to be
markedly in error to the right, since it depends
on the determination of the midpoint between
the borders of the cardiac silhouette, which in
these cases, are made up largely of the aneurysmal auricles. In the remainder of the cases,
discrepancies were usually less, and there was
no apparent explanation for them, other than
errors inherent in the methods.
The practical value of electric heart center
data secured in this study is that such data
may be useful in developing an orthogonal lead
system for vectorcardiography. Construction of
accurate orthogonal leads can be accomplished
for a particular individual after his electric
center has been determined, but this procedure
is impractical for general clinical use. As a more
practical approach, lead systems relatively insensitive to individual differences in dipole
location have been developed here and elsewhere. This study of the distribution of heart
centers helps to establish the design requirements for such systems, and enables an estimate
to be made of their effectiveness in mitigating
pronounced effects of variation of dipole location.
SUMMARY
Anatomic locations of electric heart centers
for the QRS complexes of 40 cardiac patients
were determined, by a modification of Frank's
method, with a precision of d= 1 cm. Ninetythree per cent of results clustered in a 5 cm. by
5 cm. square area forward and to the left in the
transverse plane. The average location was at
anatomic model position 22.
Average location of fluoroscopically determined heart centers correlated well with average location for QRS electric heart centers.
Correlation in many individual cases was poor.
This information may serve as a basis for
design and appraisal of orthogonal vectorcardiographic lead systems.
ACKNOWLEDGMENT
The author expresses appreciation for the excellent
technical assistance of Miss Cynthia Hamilton, and
is grateful to Doctors Ernest Frank and Calvin
F. Kay for their encouragement and helpful suggestions.
SUMMARIO Ii\T IiYTERLINGUA
Le location anatomic del centres electric del
corde pro le complexo QRS esseva determinate
in 40 patientes cardiac per medio de un modification del methodo de Frank, resultante in un
precision de ± 1 cm. Novanta-tres pro cento
del resultatos se congregava in un area quadrate
de 5 x 5 cm. al fronte e al sinistra in le piano
transverse. Le location median esseva al
puncto 22 del modello anatomic.
Le location median del fluoroscopicamente
determinate centros cardiac esseva ben correlationate con le location median del centro
electric del corde pro le complexo QRS. In
multe casos individual iste correlation esseva
magre.
Le information hie presentate pote esser
usate como base del construction e evaluation
de orthogone systemas de derivationes vectocardiographic.
REFERENCES
1
FRANK, E., KAY, C. F., SEIDEN, G. E., AND KEIS-
MAN, R. A.: A new quantitative basis for electrocardiographic theory: The normal QRS complex.
Circulation 12: 406, 1955.
1
SCHWAN, H. P., KAY, C. F., BOTHWELL, P. T., AND
FOLZ, E. L.: Electrical resistivity of living body
tissues at low frequencies, Fed. Proc. 13: 131,
1954.
3
FRANK, E.: Measurement and significance of cancellation potentials on the human subject.
Circulation 11: 937, 1955.
* SCHMITT, 0. H., LEVINE, R. B., SLMONSON, E., AND
DAHL, J.: Electrocardiograph^ mirror pattern
studies: 1. Experimental validity tests of the
dipole hypothesis and of the central terminal
theory. Am. Heart J. 46: 416, 1953.
5
SIMONSON, E., SCHMITT, 0. H., LEVINE, R. B., AND
DAHL, J.: Electrocardiographic mirror pattern
studies: 3. Mirror pattern cancellation in normal
and abnormal subjects. Am. Heart J. 45: 655,
1953.
•BURGER, H. C , AND VAN MILAAN, J. B.: Heart
vector and leads, III: Geometrical Representation. Brit. Heart J. 8: 157, 1946.
7
SCHMITT, O. H., AND SIMONSON, E.: Symposium on
electrocardiography and vectorcardiography.
The present status of vectorcardiography. Arch.
Int. Med. 96: 574,1955.
8
FRANK, E.: Determination of the electrical center
of ventricular depolarization in the human heart.
Am. Heart J. 49: 670, 1955.
8
—: The image surface of a homogeneous torso. Am.
Heart J. 47: 747,1954.
The Electric Heart Center for the QRS Complex in Cardiac Patients
GEORGE E. SEIDEN
Downloaded from http://circres.ahajournals.org/ by guest on July 31, 2017
Circ Res. 1956;4:313-318
doi: 10.1161/01.RES.4.3.313
Circulation Research is published by the American Heart Association, 7272 Greenville Avenue, Dallas, TX 75231
Copyright © 1956 American Heart Association, Inc. All rights reserved.
Print ISSN: 0009-7330. Online ISSN: 1524-4571
The online version of this article, along with updated information and services, is located on the
World Wide Web at:
http://circres.ahajournals.org/content/4/3/313
Permissions: Requests for permissions to reproduce figures, tables, or portions of articles originally published in
Circulation Research can be obtained via RightsLink, a service of the Copyright Clearance Center, not the
Editorial Office. Once the online version of the published article for which permission is being requested is
located, click Request Permissions in the middle column of the Web page under Services. Further information
about this process is available in the Permissions and Rights Question and Answer document.
Reprints: Information about reprints can be found online at:
http://www.lww.com/reprints
Subscriptions: Information about subscribing to Circulation Research is online at:
http://circres.ahajournals.org//subscriptions/