Spatial Magnitude, Orientation, and Velocity of
the Normal and Abnormal QRS Complex
By KATSUHIKO YANO, M.D., AND HUBERT V. PIPBERGER, M.D.
Downloaded from http://circ.ahajournals.org/ by guest on June 17, 2017
Records of the spatial velocity of the electrocardiogram have been recorded in a similar
fashion.8 More recently such time-based spatial records have also been obtained by means
of digital computation.9
In spite of the attractiveness of this new
type of spatial data display tlhese recording
procedures have not been used widely. Clinical applications have been very limited.10-'2
Since digital computer facilities together with
a large electrocardiographic library of normal
and abnormal records were available at this
laboratory, an attempt to appraise the diagnostic usefulness of spatial data display on
a time basis was made in the present study.
For evaluation 252 normal and 328 abnormal
records were selected. Statistical separation
between normal and abnormal findings in the
QRS complex was considered an adequate test
procedure for this purpose.
SPATIAL characteristics of the electrocardiogram have been increasingly emphasized since the advent of vectorcardiography. Progress in this field was hampered
originally, however, through introduction of
numerous lead systems with discrepancies in
lead performance.' Only the more recently
developed, corrected orthogonal leads resulted in relatively close agreement among systems, thus enhancing the comparability of
results.2
Commonly, spatial electrocardiographic
data are displayed in the form of vector loops
projected on three mutually perpendicular
planes. The oscilloscope beam is interrupted
at regular intervals for timing. It was realized
soon, however, that part of the time information cannot be recovered from such records.
Beginning and end of QRS loops are frequently hidden by superimposed P or T loops. Parts
of loops may be perpendicular to a given
plane and not be represented at all. Furthermore, time intervals between P, QRS, and T
loops cannot be measured. This becomes feasible when loops are recorded on running film.
Such a display, however, leads to severe distortions in vector directions.
A new type of spatial electrocardiographic
display on a continuous time basis was developed more recently.3-7 Curves of spatial magnitude and orientation have been recorded on
a time basis by means of analog computers.
The spatial orientation is usually expressed in
terms of azimuth and elevation angles. This
type of data display provides complete spatial
information without loss of time information.
Materials and Methods
From a library of more than 6,000 orthogonal
electrocardiographic records a limited number of
normal and abnormal tracings characteristic of
various diagnostic electrocardiographic entities
were selected for study. The distribution of these
samples is shown in table 1. In all normal subjects history and complete physical examination
did not reveal any signs or symptoms of past or
Table 1
Number of Electrocardiographic Records and Diagnostic Entities Used for the Study
Diagnostic categories
Normal records
Myocardial infarction, old
252
129
Anterior, 36
Posterodiaphragmatic, 55
Apical, 38
From the V.A. Eastern Research Support Center,
Left ventricular hypertrophy
Right ventricular hypertrophy
Left ventricular conduction defect
Right ventricular conduction defect
Mouint Alto Hospital, and the Department of Medicine, Georgetown University School of Medicine,
Washington, D. C.
Supported in part by a U. S. Public Health Service
Research Grant CD-00064-04 from the Division of
Chronic Diseases.
Circulation, Volume XXIX, January 1964
Number of cases
Total
107
105
32
31
31
580
YANO, PIPBERGER
108
Downloaded from http://circ.ahajournals.org/ by guest on June 17, 2017
present cardiovascular disease. Records were classified in pathologic categories only when reliable
evidence for specific diagnoses existed. In the
series with old myocardial infarction, for instance,
a well-documented history of infarction had to
be available, including diagnostic electrocardiograms at the time of the acute episode. Cases in
the category of ventricular hypertrophy were included only when an underlying disease known
to lead to ventricular overload was present. The
series with right ventricular hypertrophy differed
from that of a general hospital population. Of the
available Veterans' population 91 per cent had
chronic cor pulmonale due to obstructive emphysema. The remaining cases had either pulmonary
tuberculosis or bronchogenic carcinoma. Records
were classified as ventricular conduction defects
when the QRS duration exceeded 0.12 second,
which was found previously to be the normal
limit for simultaneously recorded orthogonal
leads.13
In each subject three corrected orthogonal leads
(Frank's lead system 14) were recorded simultaneously on magnetic analog tape with FM
channels. Subsequently the records were digitized
as described previously.'5 A digital computer
(IBM 7090) was used for processing and analysis
of the data. Details of the computational procedures were reported previously.9' 10
Two different procedures were used for ob-
taining curves of spatial magnitude, orientation,
and velocity. The first one consisted of a continuous plot of these data on a time scale as
described previously by others.3-8
It was soon realized, however, that such a display leads to difficulties for statistical correlations (fig. 1). Discrimination between normal
and abnormal requires comparison between individual records and normal standards. Initial parts
of QRS need to be compared with the initial part
of the normal QRS complex. The same holds true
for the terminal part of QRS. Owing to the normal variability in QRS duration, such comparisons proved impractical, however. When the beginning of a given QRS complex was lined up in
time with the onset of a noirmal standard, representing a mean of the normal control, the ends
of these two complexes differed in time in most
instances. This difficulty could be partly overcome when both beginning and end of QRS
were lined up in time with the normal standard.
This procedure leads, however, either to a gap
or to an overlap between the initial and terminal
QRS portion (fig. 1). In order to overcome this
difficulty it was decided to normalize QRS in
time. Regardless of its duration, the QRS complex was divided in time in 10 equal parts. Each
of the 11 points obtained represents an instantaneous vector. Connecting the points by a line then
leads to new curves of spatial magnitude, orien-
NORMALIZED TIME
REAL TIME
A
A
vr~~~~rr
0
SECONDS
0.03
0.06
E
0.09
r
u
X
0.12
t
2
.1
,5
f
,
0.06
0.03
0
SECONDS
o 0 ' ' o0.03 '
0.06
0
*
.
.
.
1.
.
5Y
.
'to
710
'O
0
Figure 1
On the left, curves of spatial magnitude of two QRS complexes with different durations are
shown in real time (0.08 and 0.12 second). Comparison between initial and terminal half of
QRS is shown in the middle diagram. Onsets and ends of the complexes are lined up in time
for this comparison. This type of representation leads either to a gap or to an overlap between
initial and terminal half. On the right, the QRS duration was normalized by dividing the complex in time in 10 equal parts. Corresponding parts of the two complexes are lined up for
correlation without distortions due to gaps or overlaps. For further detail, see text.
Circulation, Volume XXIX, January 1964
109
NORMAL AND ABNORMAL QRS COMPLEX
AZIMUTH
ELEVATION
BACK
HEAD
270`
-90.
00
~~~~1800
RIGHT
0'
LEFT
90°
group.
+90Q
FEET
FRONT
Figure 2
Scales used for
curves
study. The various ranges had to be computed on
the basis of 96 percentile limits.17 This was necessary because most of the findings were not
normally distributed. Raniges based on a meani
+2 standard deviationis are, therefore, not an
optimal representation of data. An estimate of
the diagnostic usefulness of the various curves
was obtained by determining percentages of cases
outside normal ranges for each pathological sub-
of spatial orientation.
Downloaded from http://circ.ahajournals.org/ by guest on June 17, 2017
tation, or velocity. Time normalization and curve
plotting are performed automatically. In a previous study it was found that normalization in
time does not decrease the power of diagnostic
discrimination.16
Scales for azimuth and elevation angles are
shown in figure 2.
In order to evaluate configurations of the various curves, mean values for each instantaneous
vector were computed. Thus, mean curves could
be obtained for each pathologic entity under
Results
In figures 3 to 5 mean normal curves of
spatial magnitude, velocity, and orientation
are shown together with the ranges for normal
findings. These normal ranges were based on
96 percentile limits and time-normalization.
Points at 9/10 of the QRS duration were
omitted because of a wide spread of normal
findings at this time. Meaningful comparisons
of normal and abnormal findings were, therefore, not possible in this late part of QRS. The
extent of the normal ranges based on the 96
percentile distributions are shown in table 2.
'able 2
Ranges of Spatial Magnitude, Velocity, and Orientation for Eight Instantaneous QRS
Vectors
Normalized
QRS
1/10
2/10
3/10
4/10
5/10
6/10
7/10
8/10
Spatial
magnitude
(mV.)
Spatial
(mV. / msec.)
0.13 ±0.05
0.04- 0.26
0.34 +0.15
0.11 - 0.71
0.68 ±0.29
0.25- 1.43
1.38 ±0.42
0.72- 2.49
1.60 ±0.51
0.68- 2.67
0.02 ±0.01
0.01 - 0.05
0.04 +0.02
0.01 - 0.08
0.09 +0.03
0.05- 0.17
0.11 ±0.04
0.04- 0.19
0.09 +0.04
0.03- 0.18
0.10 ±0.04
0.03- 0.18
0.06 +0.03
0.02- 0.13
0.04 ±0.02
0.01 - 0.06
1.15 ±0.39
0.48- 1.91
0.65 +0.26
0.23- 1.15
0.29 +0.14
0.07- 0.63
velocity
Azimuth
(degrees)
68.3 ± 36.7
19 - 111
94.9 + 26.4
53 - 148
147.5 + 29.6
93 - 220
191.1 ± 26.7
145 - 239
222.2 ± 30.2
177 - 280
257.9 ± 32.7
193 - 304
282.5 ± 25.1
221 - 322
288.7 ± 50.4
113 - 354
Elevation
(degrees)
Real time range
after QRS onset
(sec.)
-10.8 ±28.7 0.008 - 0.011
-53 - +45
- 1.9 ±20.5
0.014 -0.022
-36 - +38
21.6 ±15.2 0.23 -0.34
- 9 -+50
35.1 ±13.3 0.030 -0.045
+14 - +51
34.2 ±15.1 0.038- 0.056
+ 2 -+59
21.2 ±18.9 0.046- 0.067
-21 - +55
6.0 ±22.9 0.053 -0.078
-47 - +49
- 0.6 ±36.0
0.061 -0.090
-72 - +71
The mean and standard deviation is indicated for each vector on the upper line. The
upper and lower limits of ranges based on 96 per cent of each distribution are shown on the
second line. For comparison with real time ranges in seconds corresponding to each instantaneous vector are indicated in the last column.
Circulation, Volume XXIX, January 1964
110
YANO, PIPBERGER
mV/m sec.
201
SPATIAL MAGNITUDE
mv
SPATIAL VELOCITY
2.5
I5
1.5
1.0-
1.0
0 5Downloaded from http://circ.ahajournals.org/ by guest on June 17, 2017
0.5
0
i
o*0
a
a
ko
'10
10
O f
8/
4
-90*-
3600*
AZIMUTH
a
A
a
a
4
'10
80
10
ELEVATION
-60-
270° -30-
1800 -
90@ NORMAL
ANTERIOR INFARCT
-POSTERO - DIAPHRAGMATIC INFARCT
---APICAL INFARCT
-
0- a
-r6
- -
aT
l/lo
/10
U-
*
4/lo
5'10
Aa_
+
010
'10
q-d 1-
I*
_
Figure 3
/10l
IA
*
I
4/l10t
10
Average QRS curves of spatial magnitude, velocity, and orientation for three types of myocardial infarct. The shaded area indicates the range of normal. Mean curves of the different groups
are identified in the right lower corner. These means were computed from the records listed
in table 1. For further detail, see text.
Circulation, Volume XXIX, January 1964
NORMAL AND ABNORMAL QRS COMPLEX
mV
2.5
111
mV/m se CS
SPATIAL
0.20 1
SPATIAL MAGNITUDE
VELOCITY
-
0.15
20
1.5
0.10
1.0
-
-
0O05
Downloaded from http://circ.ahajournals.org/ by guest on June 17, 2017
05
0 4-
.
W
.
o0
X
a
8,
4
0
't
to
0
'0
360°
-
g
X
1
aX
4,
'0
10
-90
.
al
8/0
-
ELEVATION
AZIMUTH
-60
2700°
-300-
1800°
00-
+30
9oo0
+ 6 '-
NORMAL
LVH
-----' RVH
n'
v
a
a
0
a
1X
X
4,
10
a
+94
2
r --u
-----r
4/
84
10
10
Figure 4
Average QRS curves of spatial magnitude, velocity, and orientation for cases with left and
and right ventricular hypertrophy. The shaded area indicates the range of normal. Mean curves
of the different groups are identified in the right lower diagram. For further detail, see text.
Circulation, Volume XXIX, January 1964
-,
8Z10
112
mV
w-w
YANO, PIPBERGER
mV/m sec.
SPATIAL MAGNITUDE
2.5
-
2.0
-
I.5
-
1. 0
-
Downloaded from http://circ.ahajournals.org/ by guest on June 17, 2017
0.5
-
0
0
360°
I
I
I
*I
0.15
VELOCITY
-
0.10
0.05-
-
u
v
a
a
v
8,
410
/10
5 SPATIAL
0.20
---
--0
-
AZI MUTH
10
-906
4/,10
8
'10
4,
8,I0o
ELEVATION
-60
2700j
-30
1800
-
0
1
1
a11
90@
1
11
11
NORMAL
+60
a1
1
-
1L----
-
_o
0
A
'lo
4
10
1
U-U---
4- Qn
Ir
= V IA--
8
0
10
0
LVCD
RVCD
a
10oli
Figure 5
Average QRS curves of spatial magnitude, velocity, and orientation for cases with left and
right ventricular conduction defects. The shaded area indicates the range of normal. Mean
curves of the different groups are identified in the right lower diagram. For further detail, see
text.
Circulation, Volume XXIX, January 1964
NORMAL AND ABNORMAL QRS COMPLEX
Downloaded from http://circ.ahajournals.org/ by guest on June 17, 2017
The standard deviations of the mean values
are included.
The group with myocardial infarcts in figure 3 was divided into three subgroups according to locations of the infarcts (anterior,
postero-diaphragmatic, and apical). This
classification was based on conventional interpretation of the orthogonal electrocardiogram and vectorcardiogram.
The spatial magnitude curves of all types
of infarcts were found lower than that of the
normal control group. This was particularly
obvious in the middle portion of QRS. Similar
findings were obtained for the spatial velocity,
which was also decreased mainly in the middle part of QRS.
As expected, anterior and apical infarcts
deviated from normal spatial orientation predominantly in azimuth curves, whereas postero-diaphragmatic infarcts were characterized by abnormalities in the head-foot direction depicted in elevation curves.
The curves of spatial magnitude in ventricular hypertrophy deviated from normal in
opposite directions (fig. 4). Whereas the mean
of the cases with left ventricular hypertrophy
exceeded the normal average, the mean of
the right ventricular hypertrophy records was
found below the normal mean. The spatial
velocity of the latter group was also moderately lower than normal. The series with left
113
ventricular hypertrophy was not significantly
different from normal.
A deviation from normal in spatial orientation was found for the group with right
ventricular hypertrophy in the second half of
QRS. It has to be kept in mind, however, that
this series consisted almost exclusively of cases
with chronic cor pulmonale due to obstructive
emphysema. A series with right ventricular
hypertrophy cases of an average hospital population might have also shown deviations in
anterior direction.
The groups with left and right ventricular
conduction defects (fig. 5) both exceeded the
normal average in spatial magnitude in the
first part of QRS. A marked drop in magnitude
was observed for right ventricular conduction
defect in the middle portion. As expected, the
spatial velocity curves were mainly below the
norm. It was interesting to note that the
azimuth curves for left ventricular conduction
defect and right ventricular conduction defect
both deviated from the normal average in
the early part of QRS. For right ventricular
conduction defect the major deviation was
seen in the second half. Elevation curves deviated only slightly in the mid-portion.
From the various curves in figures 3 to 5
it becomes obvious that a good part of the
mean curves of pathologic groups does not
exceed the limits of normal. Only part of these
Table 3
Diagnostic Recognition Rates of Various Curves of Spatial Data Display
Diagnostic categories
Spatial
magnitude
Recognition rate (per cent frequency)
Azimuth
Spatial
Elevation
velocity
angles
angles
Myocardial infarction
Anterior
Posterodiaphragmatic
Apical
Left ventricular hypertrophy
Right ventricular hypertrophy
Left ventricular conduction defect
Right ventricular conduction defect
Total
47
39
55
62
69
71
75
56
89
55
53
74
65
84
87
100
100
100
61
67
76
All eight instantaneous vectors of each tracing
89
56
92
81
91
89
47
63
74
84
70
were tested against normal ranges. If one
of these vectors were found outside normal limits, the record was considered abnormal
and included in the over-all recognition rate for the given type of tracing. No attempt to
separate pathologic ranges was made and the diagnostic recognition rate was limited to discrimination between normal and abnormal.
or more
Circulation, Volume XXIX, January 1964
YANO, PIPBERGER
114
Table 4
Number of Records That Would Not Have Been Diagnosed if Any of the Curves Had
Been Omitted
Diagnostic categories
Number of cases recognized in one curve only
Elevation
Azimuth
Spatial
Spatial
angles
angles
magnitude
velocity
4
10
0
0
Myocardial infarction
2
3
3
3
Left ventricular hypertrophy
0
1
6
1
Right ventricular hypertrophy
1
1
1
1
Left ventricular conduction defect
0
0
0
0
Right ventricular conduction defect
14
14
4
5
Total
This listing indicates the relative information content of the various tracings. It was highest
for the curves of spatial orientation.
Downloaded from http://circ.ahajournals.org/ by guest on June 17, 2017
curves exhibited characteristic configurations
that are useful for diagnostic purposes. Evaluations of the contour of curves, however, must
necessarily remain descriptive with the wellknown pitfalls of this type of analysis. A more
quantitative approach was chosen, therefore,
for further evaluation of the data. Percentages
of cases exceeding normal limits were determined for each pathologic group. These
limits of normal were based on 96 percentile distributions. Percentages of cases
exceeding these normal ranges are shown in
table 3. Eight instantaneous vectors of each
QRS complex were used for computation of
diagnostic recognition rates. If at least one
datum point exceeded normal limits, the
record was included in this recognition rate.
For the group with myocardial infarcts the
curves of spatial orientation led to the highest
recognition rates. The spatial magnitude was
least informative for this entity.
Both in left ventricular hypertrophy and
right ventricular hypertrophy the spatial orientation also contributed most to diagnostic
recognition. It could be expected that the left
ventricular hypertrophy series exceeded normal magnitude limits in a substantial number
of cases (62 per cent). It was very surprising
to find an even higher recognition rate for the
right ventricular hypertrophy group; 69 per
cent of the latter series showed at least one
instantaneous vector with a spatial magnitude below normal limits.
Diagnostic recognition rates were generally
very high in records with ventricular conduc-
tion defects. The spatial orientation was again
most informative but the spatial magnitude
and velocity data led also to relatively high
recognition rates. It must be kept in mind
here that the QRS duration was normalized
and that the prolongation of this complex
was not taken into account.
A further evaluation of the diagnostic usefulness of the data is shown in table 4. The
high recognition rates in table 3 make it obvious that most records were overdiagnosed
by use of all curves. Table 4 shows, therefore,
the number of records that would not have
been diagnosed if any one of the curves had
been omitted. This would have been the case
in 28 tracings (8.5 per cent) without curves
of spatial orientation. The curves of spatial
magnitude and velocity showed a considerably lower information content. Only four
and five cases, respectively, would not have
been recognized as abnormal without these
curves (1.2 and 1.5 per cent).
Discussion
The introduction of time-based curves of
spatial parameters of the electrocardiogram
has been a significant step in the development
of optimal displays of electrocardiographic
data. This development represented a significant improvement over conventional vectorcardiographic loop representations because
of the addition of reliable time information
to the spatial data contained in vector loops.
The search for optimal display methods for
spatial electrocardiogram information has beCirculation, Volume XXIX, January 1964
NORMAL AND ABNORMAL QRS COMPLEX
Downloaded from http://circ.ahajournals.org/ by guest on June 17, 2017
come particularly important, since it could be
shown that the major part of diagnostic electrocardiogram information is contained in
spatial data.'6-18 The question remains, however, whether time-based curves of spatial
magnitude, velocity, and orientation represent
an optimal method of spatial data representation. There can be no question that this
display contains all the spatial data possibly
available from the electrocardiogram. At any
given point spatial magnitude, orientation,
and velocity can be accurately read from the
graphs. The question to be answered remains,
however, whether this information can be
readily extracted from the curves.
A first approach to the evaluation of the
described recording procedures consists of
analysis of t-heir configuration. The average
curves of figures 3 to 5 exhibit the typical
deviations from the normal mean. A rather
large variability in curve configuration was
found, however, for each entity. This variability did not allow a clear and meaningful
separation of the groups for diagnostic classification. Evaluation of configurations, furthermore, leads necessarily to descriptive analysis,
which will always be subject to subjective
judgments. Following the common trend of
modern electrocardiography toward quantitative analysis, a different approach appears in-
dicated.
Quantitative analysis is feasible when consecutive data points are compared with normal ranges. As outlined above and illustrated
in figure 1, this is not possible without normalization in time. Such a procedure cannot be
performed by the analog computers described
for obtaining such curves.3-8 Although part of
the time information is lost through normalization, this procedure was not found to decrease diagnostic discrimination power.'6 For
practical use it follows that each of the consecutive data points has to be compared with
the normal control. This point-by-point comparison needs to be performed not only with
the normal range but also with each other
pathologic group in order to arrive at a differential diagnosis.
The use of 96 percentile ranges leads by
definition to 4-per cent false positive findings
Circulation, Volume XXIX, January 1964
\_*i.s,
115
for each datum point. When 8 points of each
curve were used, the percentage of false positives increased to 25 per cent in a single
curve. With 4 curves used together the false
positive rate becomes prohibitive.
The representation of spatial orientation by
curves of azimuth and elevation angles is a
necessity because spatial direction in a Cartesian coordinate system needs to be defined
by two angles. Ranges of spatial orientation
are commonly projected on the surface of a
globe as shown in figure 6. Typically such
ranges were found to be elliptical. Projections
of ranges on azimuth and elevation scales lead
to a distortion of the original range, which
now becomes a rectangle. Findings in the
corners of the rectangle will be included in
the range and represent false negative findings. The statistical pitfalls of time-based
spatial data display have not been considered in reported clinical applications of
this method.'0-'2
Of the various graphic methods for recording and displaying spatial electrocardiogram
\
+
Figure 6
- -1
Typical range of normal instantaneous vectors projected on the surface of a globe. Most of these ranges
comprising 96 per cent of findings are elliptical in
form. The range of normal in terms of azimuth and
elevaton angles is indicated by the shaded area. Any
unknown abnormal vector falling in the shaded area
will be interpreted as normal although its direction
deviates from the elliptical normal range. Such vectors
represent false negative findings.
YANO, PIPBERGER
116
by digital computation. Their numerical expression can be used for complex statistical
analyses in multidimensional vector space.9' 16
Such computations are strictly numerical and
not easily expressed in graphic form. Advantages of visual displays as described in the
present report are given up in strictly numerical data-handling procedures. These advantages have to be weighed against the
greater efficiency of complex numerical classification methods.
41
W
41
Downloaded from http://circ.ahajournals.org/ by guest on June 17, 2017
Figure 7
An unknown vector AB is compared with the mean
vector of a normal range, AC. The vector difference
AB-AC identifies the deviation from the normal mean
by one term. Such a vector difference expresses deviations both in spatial magnitude and orientation. In
time-based graphic displays, correlations of three
curves are necessary to express the same vector difference. Such correlations lead to a substantial number of false positive and negative findings.
data time-based curves of spatial magnitude.
orientation, and velocity appear as the most
inclusive ones. Their superiority to recordings
of scalar leads and vector loops should be obvious because the information of both is contained in one type of data display. Difficulties
arise, however, in the process of extracting the
information, as explained above. In digital
computations it was found that determination
of vector differences obviate most of the pitfalls of graphic displays (fig. 7). The vector
difference between an individual vector and
mean vectors of various diagnostic categories automatically leads to a differential
diagnosis based on the magnitude of these
vector differences.16 This one term comprises
both the information of spatial magnitude and
orientation. Since spatial velocity did not contribute more than 1.5 per cent in diagnostic
recognition rates, it is questionable whether
this type of data display justifies the relatively high cost of analog equipment which is required for its recording.
Vector differences are conveniently obtained
Summary
Curves of spatial magnitude, orientation,
and velocity of the QRS complex were obtained by digital computation from 252 normal and 328 abnormal orthogonal electrocardiograms (Frank system). An attempt was
made to evaluate the diagnostic usefulness
of this type of spatial data display. Because
of interindividual variability in QRS duration
it was not possible to compare accurately individual records with normal standards. Normalization in time was, therefore, necessary
by dividing each QRS complex in time in 10
equal parts. Mean curves were computed for
the normal control group and eight pathologic
entities. Although differences in configuration
between the variouis groups became evident,
a large overlap between the various ranges
prevented an efficient classification for diagnostic purposes.
A point-by-point separation between normal ranges and individual curves showed high
recognition rates for abnormalities without
specific diagnostic classification. The highest
information content appeared to be present
in curves of spatial orientation.
Since time-normalization and numerical statistical analysis of multiple points are not
feasible with presently available analog computers, the question is raised whether digital
computation is not more efficient in diagnostic
classification of electrocardiograms.
References
1. PIPBERGER, H. V., AND LILIENFIELD, L. S.: Ap-
plication of corrected electrocardiographic lead
systems in man. Am. J. Med. 25: 539, 1958.
2. PIPBERGER, H. V.: Current status and persistent
problems of electrode placement and lead systems for vectorcardiography and electrocarCirculation, Volumtne XXIX, January 1964
117
NORMAL AND ABNORMAL QRS COMPLEX
Downloaded from http://circ.ahajournals.org/ by guest on June 17, 2017
diography. Progr. Cardiovas. Dis. 2: 248, 1959.
3. MCFEE, R.: Trigonometric computer with electrocardiographic application. Rev. Sci. Instr.
21: 420, 1950.
4. SAYERS, B. McA., SILBERBERG, F. G., AND DURIE,
D. F.: Electrocardiographic spatial magnitude
curve in man. Am. Heart J. 49: 323, 1955.
5. ABILDSKOV, J. A., INGERSON, W. E., AND HISEY,
B. L.: Linear time scale for spatial vectorcardiographic data. Circulation 14: 556, 1956.
6. VON DER GROEBEN, J.: Spatial frequency distribution of the QRS loop as studied on 154 normal
individuals. Am. Heart J. 59: 875, 1960.
7. MooRE, A. D., HARDING, P., AND DOWER, G. E.:
Polarcardiograph. An analogue computer that
provides spherical polar coordinates of the
heart vector. Am. Heart J. 64: 382, 1962.
8. HELLERSTEIN, H. K., AND HAMLIN, R.: QRS component of the spatial vectorcardiogram and of
the spatial magnitude and velocity electrocardiograms of the normal dog. Am. J. Cardiol.
6: 1049, 1960.
9. PIPBERGER, H. V.: Use of computers in interpretation of electrocardiograms. Circulation Research 11: 555, 1962.
10. ABILDSKOV, J. A., HISEY, B. L., AND INGERSON,
W. E.: Magnitude and orientation of ventricular excitation vectors in the normal heart
and following myocardial infarction. Am. Heart
J. 55: 104, 1958.
11. TOOLE, J. G., VON DER GROEBEN, J., AND SPIVACK,
A. P.: Calculated tempero-spatial heart vector
in proved isolated left ventricular overwork.
Am. Heart J. 63: 537, 1962.
12. TOOLE, J. G., VON DER GROEBEN, J., AND SPIVACK,
A. P.: Periodic abnormalities of the temperospatial QRS vector in isolated right ventricular
overwork. Am. Heart J. 65: 77, 1963.
13. PIPBERGER, H. V.: Normal orthogonal electrocardiogram and vectorcardiogram; with a
critique of some commonly used analytic criteria. Circulation 17: 1102, 1958.
14. FRANK, E.: Accurate, clinically practical system
for spatial vectorcardiography. Circulation 13:
737, 1956.
15. PIPBERGER, H. V., FREis, E. D., TABACK, L., AND
MASON, H. L.: Preparation of electrocardiographic data for analysis by digital electronic
computer. Circulation 21: 413, 1960.
16. PIPBERGER, H. V., STALLMANN, F. W., YANO, K.,
AND DRAPER, H. W.: Digital computer analysis of the normal and abnormal electrocardiogram. Progr. Cardiovasc. Dis. 5: 378, 1963.
17. PIPBERGER, H. V., STALLMANN, F. W., AND BERSON, A. S.: Automatic analysis of the P-QRS-T
complex of the electrocardiogram by digital
computer. Ann. Int. Med. 57: 776, 1962.
18. PIPBERGER, H. V.: Analysis of electrocardiograms
by digital computer. Methods Inform. Med. 1:
69, 1962.
Reaction to Harvey's Discovery of the Circulation
Harvey's book was not an overwhelming success. Some members of the
College of Physicians spoke of it with complimentary recognition. But at least
one of his colleagues went on teaching the movement of the blood in the hall
of barbers and surgeons for years as if he had never even heard of him, although
he knew Harvey and his views very well.
He had to put up with many mocking remarks from other sources, he was
called "circulator," as the attendants on quack doctors at fairs were called,
for there was too frequent mention of the circular movement of the blood in
his book. His practice fell off appreciably. Was it not a matter to be weighed
very carefully whether to turn to a physician who was held in contempt by a
number of professors and barbers of great distinction?
The apothecaries did not value his prescriptions, which was very unusual in
the case of a court physician. In one case he-was even sued for overlooking a
fracture in one of his patients.-TIBOR DOBY, M.D. Discoverers of Blood Circulation. From Aristotle to the Times of Da Vinci and Harvey. New York, AbelardSchuman, 1963, p. 208.
Circulation, Volume XXIX, January 1964
Spatial Magnitude, Orientation, and Velocity of the Normal and Abnormal QRS
Complex
KATSUHIKO YANO and HUBERT V. PIPBERGER
Downloaded from http://circ.ahajournals.org/ by guest on June 17, 2017
Circulation. 1964;29:107-117
doi: 10.1161/01.CIR.29.1.107
Circulation is published by the American Heart Association, 7272 Greenville Avenue, Dallas, TX 75231
Copyright © 1964 American Heart Association, Inc. All rights reserved.
Print ISSN: 0009-7322. Online ISSN: 1524-4539
The online version of this article, along with updated information and services, is
located on the World Wide Web at:
http://circ.ahajournals.org/content/29/1/107
Permissions: Requests for permissions to reproduce figures, tables, or portions of articles
originally published in Circulation 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 is online at:
http://circ.ahajournals.org//subscriptions/