scispace - formally typeset

Journal ArticleDOI

A comparison between ultralow-frequency ballistocardiograms and those secured by an improved high-frequency technique, with studies to explain remaining differences.

01 Jul 1962-American Heart Journal (Elsevier)-Vol. 64, Iss: 1, pp 79-100

TL;DR: The advance in ballistocardiographic instrumentation which has been so rapid and so encouraging in recent years has been due primarily to the use of certain physical principles, but on the assumption that well-known physical formulae could be properly applied to the vibration problems of the human body, a new viewpoint emerged.

AbstractThe advance in ballistocardiographic instrumentation which has been so rapid and so encouraging in recent years has been due primarily to the use of certain physical principles. On the assumption that well-known physical formulae could be properly applied to the vibration problems of the human body, a new viewpoint emerged. This included a well-based criticism of the high-frequency (HF) ballistocardiograph, ¹⁻² i.e., that the movement which took place between body and table was introducing an error, and that the vibration properties of this movement between body and table led to an undue magnification of certain components of the recorded forces, those delivered in resonance with the body's own vibration properties, and undue attenuation of others, those above the body's resonance frequency. It was proposed to avoid or minimize such errors by using another type of instrument, the ultralow-frequency (ULF) ballistocardiograph.³⁻⁶

Summary (2 min read)

Instruments

  • Because their room had a low ceiling, their bed---like Rappaport's instrument, and like Henderson's table of 50 years agois displaced laterally by a pair of pins (AC, Pig. 3), sharp at each end and 13.7 cm. long.
  • This system renders their records almost altogether free of building vibration.
  • Of great value has been a secondary electrical circuit, with dry cells and a milliammeter, which indicates when one of the magnets touches the inside of its coil, an error of technique very likely to pass unnoticed without this warning device, and capable of causing marked distortion of the ballistocardiogram.
  • On the assumption that the body moves methods of tightening the subject the as a unit, the physical characteristics of restoring force and damping are about the coupling between body and table can doubled.

Results

  • Comparison of the ULF and HF force records ift healthy persons.
  • The measurements secured in the 30 men were used to construct a grand average normal male ballistocardiogram of the HF type, and another of the ULF type.
  • The tips of the H and I waves of the HF records follow those of the ULF records by an average of 0.012 and 0.022 second, respectively, and these are significant differences.
  • In the first group are the differences in size and shape of the waves.
  • If this were true, the authors should be able to start with the record secured by one instrument and, through the use of physical principles, compute the record of the other.

It became possible only recently

  • With the development of the digital computer.
  • The authors took as a starting point a typical complex secured by the low-frequency instrument, a complex midway between the largest and smallest ones of the respiratory cycle.
  • The results (Figs. 12,D and 13,D) show clearly that the curves thus constructed from measurements made on low-frequency ballistocardiograms bear a close resemblance in shape and amplitude to the HF ballistocardiograms (Figs. 10 and 11 ) secured experimentally on the same subjects.
  • Obviously, therefore, the authors have a clear understanding of the reason for the major points of difference between the two types of records,9 the average differences of wave height and timing shown in Fig. 7,A and B .
  • Theoretical studies on the effect of loose masses within the body.

Discussion

  • (a) The HF force tracing is distorted by movement of the body on the table, although it should be noted that because of recent improvements this movement is much smaller and the distortion caused is much less than that sometimes seen on records secured by older instruments with high natural frequency.
  • It is far from certain that all the slurs and notches so commonly seen in ULF ballistocardiograms have their origin in the circulation, and, if not, a better estimate of circulatory abnormalities might be made if they were not recorded.
  • The chief argument in favor of a circulatory origin for the notches so often seen in ULF record lies in their apparent movement with respiration.
  • Chiefly because of the difference in J-wave amplitude, the over-all amplitudethe vertical distance between the tips of the I and J waves-is larger in HF than in ULF records.

Did you find this useful? Give us your feedback

...read more

Content maybe subject to copyright    Report

A
comparison between
ultralow-frequency ballistocardiograms and
those secured by an improved high-frequency technique,
with studies to explain rema,ining differences
Isaac Starr, M.D.
Philadelphia, Pa.
Abraham Noordergraaf, Ph.D.
Utrecht, Netherlands
T
he advance in ballistocardiographic
instrumentation which has been so
rapid and so encouraging in recent years
has been due primarily to the use of certain
physical principles. On the assumption
that well-known physical formulae could
be properly applied to the vibration prob-
lems of the human body, a new viewpoint
emerged. This included a well-based criti-
cism of the high-frequency (HF) ballisto-
cardiograph,1,2 i.e., that the movement
which took place between body and table
was introducing an error, and that the
vibration properties of this movement
between body and table led to an undue
magnification of certain components of
the recorded forces, those delivered in
resonance with the body’s own vibration
properties, and undue attenuation of
others, those above the body’s resonance
frequency. It was proposed to avoid or
minimize such errors by using another
type of instrument, the ultralow-frequency
(ULF) ballistocardiograph.3-6
The very considerable technical problems
were overcome in a number of laboratories,
and several types of excellent ultralow-
frequency ballistocardiographs have been
constructed; these differ only in technical
details and give essentially similar records.
With this experience before us the instru-
ment used in this study was constructed
by Mr. George Peirce.
The possession of such an instrument
permitted us to compare force ballisto-
cardiograms taken by the ultralow-fre-
quency technique, when acceleration of
the table is recorded, with force ballisto-
cardiograms taken with the high-frequency
technique, when displacement of the table
is recorded. We expected that a study of
the differences between the two force
records would provide important informa-
tion, because each instrument approached
the problem from a different direction, and
neither method seemed altogether free of
error. Thus, if the records secured by each
method closely agreed with one another,
we would have strong evidence that the
remaining errors were not important.
In addition, by testing large numbers of
healthy persons and patients by both
methods, we could not only compare and
define their comparative utility in the
From the Department of Therapeutic Research. University of Pennsylvania. Philadelphia, Pa., and the Department
of Medical Physics, Physical Laboratory, University of Utrecht, Utrecht. Netherlands.
This work was supported by Research Grant H-625 (C-8,9,10) from the National Heart Institute, National Institutes
of Health, United States Public Health Service.
Received for publication Nov. 17. 1961.
79

80
Starr a& Noordergraaf
Am. Hcort J.
July, 1962
force-Displacement of a ftxGd spring
(High freauency balEsto)
0
accelcrometcr
&
-,-I
Force=Mass X accczleration of CX~ object
free to move (Ultra Ion-freauency ballisto)
Fig. 1. Illustrations of the principles behind HF and
ULF ballistocardiographs. Our problem is to meas-
ure the force with which the little man pushes on
the ball suspended by a very long and light wire.
In the first method we place a strong spring behind
the ball. When the little man pushes the ball, this
spring is compressed a short distance which we
can measure (displacement measurement). Then,
after turning the spring upright, we add weights
until the spring is compressed the same distance.
So we compare the unknown force with a known
force. In the second method we leave the suspended
ball as free to move as is possible. Now, when the
little man pushes, the ball is accelerated. An instru-
ment mounted on the ball records the amount of
acceleration. The force applied can be calculated
from the product of the acceleration recorded and
the mass mox:ed, i.e., the total mass of ball and
accelerometer. Of course, the little man of this
figure applies a force external to the ball, whereas
the ballistocardiogram comes from forces having
their origin within the body.
detection of cardiovascular abnormality,
but also investigate the matter from a
practical point of view, by asking the
question, “Does ultralow-frequency battis-
tocardiography provide a better clinical
method than the high-frequency tech-
nique?”
Accordingly, during the past years
we have tested over 400 patients and
healthy persons by both methods; the
second test was made as soon as the first
was completed. An explanation of the
differences between the records secured
by the two instruments was soug-ht by
mathematical studies and by several series
of experiments.
Theory
When one is interested in the forces, the
theory behind the two types of ballisto-
cardiographs can be easily grasped by
reference to Fig. 1, which is designed to
demonstrate the two different ways of
measuring an unknown external force.
In ballistocardiography, one is interested
in the internal force, or what is propor-
tional to such force, the acceleration of
the body’s center of gravity, and the same
two methods can be used to detect it.
Instruments
The ultralolo-frequency ballistocardi-
ograph. The bed, made of a frame of hard
aluminum alloy (61 ST 1) tubing, support-
ing a hammock of Grade A airplane fabric,
was constructed for us by the late Dr.
M. B. Rappaport, and was essentially
similar to that which he has described.6
After we had added a footplate, magnets,
and fittings, the total weight was 3.3
kilograms. The general setup is shown in
Fig. 2; the mass and physical properties
are given in Table 1.
Unlike that illustrated in Rappaport’s
article,6 our bed was suspended from the
ceiling by four wires, each 182 cm. in
length. Because our room had a low ceiling,
our bed---like Rappaport’s instrument, and
like Henderson’s table of 50 years ago-
is displaced laterally by a pair of pins
(AC,
Pig.
3), sharp at each end and 13.7
cm. long. When these pins are in place, the
point of support of the suspension on this
same side is not directly over the end of
the pin (A), but over a point (B) 1.5 cm.
from this base. The natural frequency of

Comparison of ULF and HF ballistocardiograms 8 1
Table I. Physical properties of our new HF
and ULF tables, without a subject
I I
HF
ULF
table table
Mass (Kg.)
Resonance frequency
when
36 3.3
loaded with 74-Kg. iron
bars (cycles/set.) 12.7 0.12
Damping ( y0 of critical value) 5.0 10
Table II. Masses which move with the HF
table, not including the subject
Table top and footplate
Aluminum bracing
Picker Flexicast for shoulder yoke
Picker Flexicast for small of back
Nonslip pad
23 Kg.
1
10
0.6
1.4
36 Kg.
the system is about 0.12 cycles per second,
a value which varies a little with differences
in the stretch in the support wires caused
by differences in the weights of the pa-
tients. No mechanism to provide ad-
ditional damping has been used.
The electrica apparatus used is identical
to that described by Rappaport, except
that we use two bar magnets and two
coils. The corresponding poles of the
magnets are placed in opposite directions,
so that the signal adds but the interference
substracts. The resulting voltage is fed
into circuits,
described by Rappaport,6
capable of integrating and differentiating
it, but we have used only the differentiating
circuit in
this investigation. A record of
acceleration was secured by means of the
amplifier and recorder of a Sanborn Twin-
Viso instrument. Up to this point, our
equipment is only slightly modified from
that described by Rappaport.
Because the coils are supported from the
floor, we found it very difficult to get the
record sufficiently free of building vibra-
tions. After much experimentation we
now use the following arrangement. Three
small wooden pieces in contact with the
floor support a loo-Kg. concrete block.
Above this block is a wooden board the
position
of which can be adjusted by three
screws. On this board lie a 15cm. thickness
of folded nylon blankets and another board
on which the coils rest. This system renders
our records almost altogether free of build-
ing vibration.
Of great value has been a secondary
electrical circuit, with dry cells and a
milliammeter, which indicates when one
of the magnets touches the inside of its
coil, an error of technique very likely to
pass unnoticed without this warning de-
vice, and capable of causing marked dis-
tortion of the ballistocardiogram.
We use a calibrator designed by Dr.
Walter Gamble.16 Two forces, each of
140 Gm.,
one directed headward, the
other footward, are allowed to act al-
ternately on the table by a pendulum that
interrupts first one and then the other.
I
\ I
Fig. 2. Our ULF ballistocardiograph.
A,
The table.
B, “Mine jacks,” columns of heavy pipe extending
between floor and ceiling to support the frame from
which the table is suspended. Only the base is
shown in the figure. C, Support wires.
D,
Lateral
support. F, Coils shown withdrawn from magnets
E. G, Footboard.

82
Starr and Noordergraaf
Am. Heart J.
July, 1962
RT RT
-P- -P-
I I
I
I
VERf ICAL
Fig. 3. Diagram which shows details of the suspen-
sion of our ULF instrument. To conserve space, the
vertical dimension has been reduced in respect to
the horizontal. AB = 1.5 cm.; angle (Y = 0.5 degrees.
This produces a series of square waves of
about 0.4 second duration in the base line
of the record. By this means the height
of any wave of the ballistocardiogram can
be related to the forces applied.
The newest h,i,gh-frequency instrument.
The design of our latest instrument stem-
med from the desire to take ballistocardio-
grams with the subject tilted as well as
horizontal. This instrument, made for us
by the Technitrol Engineering Company,
is shown and briefly described in Fig. 4.
The masses and vibration properties are
given in Tables I and II.
From the first, our standard technique
of getting the subject tight on the table
has been to have him lie on it with his feet
in contact with the footplate and his knees
bent. Then, by straightening his knees the
body was forced headward, putting tension
on the clothing and skin of his back as
well as increasing the pressure of his heels
on the footplate.
This technique has been improved as
follows, A pad of thin rubber-like material,
ordinarily used to prevent small rugs from
slipping on a polished floor, has been
placed on the table top; this has proved to
be an important addition when the table
Table III. Effect of tightening the subject on the HF table. The frequency and damping of move-
ment between subject and table under various conditions
Subject
Conditions
Damping,
Frequency as
defined
/ (c.fu.) by X,/XI
J.U. Lying on nonslip pad, feel free of footplate 5.8
0.60
(mass 8.5 Kg.)
On nonslip pad, feet against foot plate 8.1
0.52
On nonslip pad, tight between Flexicast shoulder yoke and
footplate 7.6
0.28
F.X.E.
(mass 66 Kg.) Lying on bare table, feet free of footplate 5.4
0.41
Lying on nonslip pad, feet free of footplate 5.0 0.51
Lying on nonslip pad, feet tight against footplate 7.1
0.35
On pad, tight between Flexicast shoulder yoke and footplate 7.2
0.33
Same, but tighter
8.2 0.68
B.P.
(mass 72 Kg.) On pad, tight between Flexicast shoulder yoke and footplate,
tilted 1.5 degrees 7.2 0.34
Same,
level 7.0 0.72
Same, Flexicast soft 6.8 0.52
Lying loose on table, feet free of footplate 5.0 0.48

Volume 64
Number 1
Comparison of ULF and HF ballistocardiograms X3
top is of polished metal as in our newer
instruments. Also, we have used a patented
device sold by the Picker X-ray Company
under the name of
Flexicast,
and designed
originally for the purpose of fixing various
parts of the body for x-ray therapy. It is
a kidney-shaped bag of rubber filled with
a granular substance which is easily
moulded at atmospheric pressure, but
which, when air is pumped out of the
bag, becomes stony hard. Two perforated
aluminum tubes attached to the table
footplate, and a movable crosspiece held
in place by pins through each tube, serve
to fix the Flexicast in position. For a while,
we also used a small bag of Flexicast under
the small of the back, but recently we
have discarded it.
Therefore, in our latest technique the
subject lies on the table on top of the
nonslip pad with his feet on the footplate
and his knees flexed. The Flexicast, while
soft, is moulded around his neck and
shoulders and held in place by the cross-
piece adjusted to the length of the subject.
The Flexicast is then hardened by con-
nection to the vacuum line of the building.
Then the subject, by straightening his
knees, compresses himself between foot-
plate and the shoulder yoke of the Flexi-
cast.
By this technique the force of compres-
sion can be made as great as any subject
can stand ; we have used forces as high as
70 kilograms. But too much compression
is painful; it also tends to arch the back
and so may defeat its purpose of tighten-
ing the subject on the table by diminishing
the area of contact. Also, a.nything which
causes discomfort may defeat our purpose
by altering the circulatory forces that we
are measuring. Hence, we aim at the
tightest attachment consistent with com-
plete comfort and adjust the position of
the crosspiece until this is obtained.
Theoretical comparison of the high-fre-
quency
and ultralow-frequency insfruments.
The physical properties of the two instru-
ments are given in Table I.
One also needs to know the physical
properties of the attachment of the sub-
ject’s body to the tables. The latter data
were found by experiments conducted on
3 healthy subjects as follows.
The table of the high-frequency instru-
ment was clamped to the frame, and the
subjects lay on it with a light bar strapped
to their shins. A screen attached to this
shin bar partly interrupted a light beam
playing on a phototube; both light source
and phototube were attached to the table.’
Fig.
4. Our newest HF ballistocardiograph, shown
tilted, and in the horizontal position in which it is
usually employed.
A,
The table. Neither the S-cm.
suspension nor the strong restraining spring is
shown. The table is made of 24 st aluminum, and
with the footplate and bracing weighs 24 kilograms.
B, The footplate. C. Movable crossoiece and DiDes
L I
which attack it to ‘footplate, used ‘to support the
shoulder yoke.
D,
Main table frame of cold rolled
steel, weighing 110 kilograms.
E, Base frame of
cold rolled steel, also
weighing 110 kilograms.
F,
The lifting mechanism, which weighs about 90
kilograms. Between base frame and floor are 4 pads
of corrugated rubber and 2 of cork, each about 7
by 7 cm. and 7 mm. thick.
Fig. 5. Diagram of the movement of the body on a
HF table, or on any other immobile surface, when a
force applied headward or footward is suddenlv
released.

Figures (9)
Citations
More filters

Journal ArticleDOI
TL;DR: Preliminary results suggest that the relationship between local and central disturbances is highly dependent on both the individual and the location where the accelerometer is placed on the body and that these differences can be resolved via calibration to accurately measure changes in cardiac output and contractility from a wearable sensor.
Abstract: Recent advances have led to renewed interest in ballistocardiography (BCG), a noninvasive measure of the small movements of the body due to cardiovascular events. A broad range of platforms have been developed and verified for BCG measurement including beds, chairs, and weighing scales: while the body is coupled to such a platform, the cardiogenic movements are measured. Wearable BCG, measured with an accelerometer affixed to the body, may enable continuous, or more regular, monitoring during the day; however, the signals from such wearable BCGs represent local or distal accelerations of skin and tissue rather than the whole body. In this paper, we propose a novel method to reconstruct the BCG measured with a weighing scale (WS BCG) from a wearable sensor via a training step to remove these local effects. Preliminary validation of this method was performed with 15 subjects: the wearable sensor was placed at three locations on the surface of the body while WS BCG measurements were recorded simultaneously. A regularized system identification approach was used to reconstruct the WS BCG from the wearable BCG. Preliminary results suggest that the relationship between local and central disturbances is highly dependent on both the individual and the location where the accelerometer is placed on the body and that these differences can be resolved via calibration to accurately measure changes in cardiac output and contractility from a wearable sensor. Such measurements could be highly effective, for example, for improved monitoring of heart failure patients at home.

25 citations


Cites background from "A comparison between ultralow-frequ..."

  • ...The issue of comparing BCG displacements, velocities, and accelerations has also been raised: Starr and Noordergraaf remarked on the similarity of the HF BCG (displacement) to the second derivative (acceleration) of the ULF BCG [14]....

    [...]


Journal ArticleDOI
TL;DR: Five of the 5 men with initially abnormal ballistocardiograms had normal records at the end of the program, and significant increases occurred in mean I, J, GI, HI and IJ forces.
Abstract: Fifteen middle-aged men participated in a program of endurance exercise and running for six months. Changes in cardiovascular function were evaluated by using an air-supported ultralow-frequency ballistocardiograph. Significant increases occurred in mean I, J, GI, HI and IJ forces. Four of the 5 men with initially abnormal ballistocardiograms had normal records at the end of the program.

21 citations



Dissertation
01 Jan 1978
TL;DR: The validity of the model was tested by comparing the predicted sway based on cardiorespiratory events with actual sway behaviour, which confirmed the widely held hypothesis that sway is a direct outcome of the dynamic equilibrium that exists between gravitatiqnal forces and the myotatic reflex responses.
Abstract: In an attempt to determine the influence of cardiorespiratory events on sway behaviour. a series of four experiments were undertake.n on a total of 95 subjects, all young healthy adults. Sway tiehaviour, defined as the corrective force recorded between the soles of the feet and the surface of a biomechanical measuring platform (Kistler, 9261A), was first examined to determine the extent to which it is a function.of sex and physique. Height, weight and obesity measurements were taken from 58 subjects (29 male, 29 female) and their influence on sway behaviour analysed. The second experiment was an extended ideographic study designed to test the constancy of sway behaviour over a six-week period for ten subjects (six male, four female) in an attempt to identify the personal characteristics of postural sway. This led to the formulation of a dynamic model of postural sway behaviour based on cardiorespiratory events. In the third experiment the magnitude of the cardiac forces and stroke volume,by transcutaneous aortovelography, were measured on 18 subjects (eight male, ten female), and used to establish the direct effect of cardiac action on sway behaviour. In the final experiment the role of 18 antigravity muscles of the lower limbs and trunk in postural maintenance was examined in nine subjects (five male, four female) to test the widely held hypothesis that sway is a direct outcome of the dynamic equilibrium that exists between gravitatiqnal forces and the myotatic reflex responses. The validity of the model was tested by comparing the predicted sway based on cardiorespiratory events with actual sway behaviour.

15 citations


Journal ArticleDOI
TL;DR: In this study, ultra low frequency force ballistocardiograms were recorded throughout the course of various types of acute cardiomyopathy, suggesting that this type of recording appears to offer a useful means of diagnosing and following the Course of such disease entities.
Abstract: In this study, ultra low frequency force ballistocardiograms were recorded throughout the course of various types of acute cardiomyopathy. Conditions studied included rheumatic carditis, lupus myocarditis, sarcoid carditis, viral myocarditis, acute glomerulonephritis, idiopathic myocarditis, and familial fibrous disease of the myocardium. The instrument used was characterized by an unusually light platform and a very high performance accelerometer. In 14 of the 15 cases studied, tracings were abnormal initially; the recorded force pattern subsequently manifested progressive change which appeared to parallel the clinical course of the disease. Changes included appearance of abnormal forces in early ventricular systole, progressive change in amplitude of acceleration and deceleration forces, appearance of abnormal high frequency components in various portions of the complex, and appearance of abnormal footward forces in late systole. The alterations thus recorded appear to offer a useful means of diagnosing and following the course of such disease entities. In some cases, this type of recording appears to provide information not available through any other conventional means.

12 citations


References
More filters


Journal ArticleDOI
TL;DR: A ballistocardiograph is described which is designed to give a record of the force acting upon the "total body mass" and typical and calibrated records are shown and discussed.
Abstract: The influence of the body mechanical system on the ballistocardiogram is discussed and a ballistocardiograph is described which is designed to give a record of the force acting upon the “total body mass.” Typical and calibrated records are shown and discussed.

48 citations


Journal ArticleDOI
TL;DR: The ballistocardiograms of young normal subjects as recorded with a wide frequency range technic are described and an analysis of these components in terms of known cardiovascular events strongly suggests that they have definite dynamic significance.
Abstract: The ballistocardiograms of young normal subjects as recorded with a wide frequency range technic are described. These records contain high frequency forces not previously described. An analysis of these components in terms of known cardiovascular events strongly suggests that they have definite dynamic significance. In addition, these correlations offer evidence as to the genesis of some forces in the ballistocardiogram.

38 citations


Journal ArticleDOI
TL;DR: Two methods for calculating the amplitude characteristic of the high-frequency BCG, the difference in movement of subject and BCG taken into account, are compared and discussed.
Abstract: Two methods for calculating the amplitude characteristic of the high-frequency BCG, the difference in movement of subject and BCG taken into account, are compared and discussed. Some remarks are made upon the calculation of the amplitude characteristic of the middle-frequency BCG according to Nickerson.

37 citations


Journal ArticleDOI
TL;DR: The nature and magnitude of force artefacts in the ballistocardiogram (BCG) arising from coupling the body to ground was discussed, and the size of these errors and their bearing on multilateral recording were discussed.
Abstract: In part I was discussed the nature and magnitude of force artefacts in the ballistocardiogram (BCG) arising from coupling the body to ground. The effect of this grounding on the reading of cardiovascular motion, momentum and force was explained, using the simple one-mass dynamics. Part II dealt with the effects on the ballistocardiogram of a second mass, a platform supporting the body. The body itself is a cutoff filter (of the resonant type) by its springy supporting tissues which couple the body and platform masses. In the displacement ballistocardiogram this attenuates the upper frequencies recorded from stiffly sprung platforms, or from the direct-body whatever its coupling to earth. In the acceleration record, the body mass and spring cut off the lower ballistocardiographic frequencies, while platform mass and body-spring cut off the higher frequencies. The size of these errors and their bearing on multilateral recording were discussed.)

34 citations


Frequently Asked Questions (1)
Q1. What are the contributions in this paper?

With this experience before us the instrument used in this study was constructed by Mr. George Peirce. The authors expected that a study of the differences between the two force records would provide important information, because each instrument approached the problem from a different direction, and neither method seemed altogether free of error.