A NEW METHOD FOR THE EVALUATION OF THE SIGNAL
ACQUIRED FROM QUANTITATIVE SEISMOCARDIOGRAPH
Hardware and Software Solution for the New Field of Monitoring Heart Activity
Z. Trefny, J. Svacinka, S. Trojan, J. Slavicek, P. Smrcka, K. Hana
Faculty of Biomedical Engineering, Czech Technical University in Prague, Studnickova 7, 2028 Prague 2, Czech Republic
Keywords: Time-domain segmentation of the seismocardiogram, J-wave recognition.
Abstract: The device for quantitative seismocardiography (QSCG) detects cardiac vibrations as they affect the entire
body; the measuring sensors (solid-state accelerometers) are usually placed in the plate of the chair –
additional instruments applied on the proband’s body are not required. The results of the QSCG analysis are
usable in various clinical fields. The first and most important step in the process of detection of significant
characteristics of measured QSCG curves is to detect pseudo-periods in the signal regardless of the initial
pseudo-period position. Other characteristics can be acquired by a relatively simple process over the
appointed pseudo-period. We have developed the experimental equipment for the QSCG measuring and
analysis. We have also developed special algorithms for preprocessing, segmentation and interactive
analysis of the QSCG signal. In this contribution we will introduce technical principles of the quantitative
seismocardiography and then we will focus on the original method for the basic segmentation of the QSCG
signal in time-domain; the method is easy, robust and is appropriate for real-time QSCG processing.
1 INTRODUCTION
Ballistocardiography (BCG): In 1936, Starr took
part in a conference held by the American Society of
Physiology which dealt with methods of determining
cardiac output. Thus began the era of high-
frequency ballistocardiography, which lasted
approximately 15 years. Other types of instruments
were developed, on which the displacement, velocity
or acceleration of a body lying on a bed was
measured. Later studies showed that there are
difficulties when comparing records registered on
different apparatuses. This is mainly caused by two
factors: (a) the instrument’s natural frequency, (b)
the instrument’s damping.
Figure. 1: Records registered using the old BCG
instrument with a frequency of 2Hz and critical damping.
The lower curve depicts the effect of force applied, which
is still of the same intensity but differs in the duration of
its effect. The upper curve is a record, from which one
cannot determine either size or duration of the acting
force.
Quantitative ballistocardiography (QBCG):
Following the critical evaluation of all these facts,
we began in 1952 our own experiments related to the
construction of an apparatus which would lack the
aforementioned shortcomings. Thus, over the years,
we constructed an apparatus whose advantages lie
not only in the simplicity of its design, but also in its
important functional qualities. To achieve a minimal
distortion caused by the transmission from the origin
61
Trefny Z., Svacinka J., Trojan S., Slavicek J., Smrcka P. and Hana K. (2006).
A NEW METHOD FOR THE EVALUATION OF THE SIGNAL ACQUIRED FROM QUANTITATIVE SEISMOCARDIOGRAPH - Hardware and Software
Solution for the New Field of Monitoring Heart Activity.
In Proceedings of the Third International Conference on Informatics in Control, Automation and Robotics, pages 61-66
DOI: 10.5220/0001220000610066
Copyright
c
SciTePress
of the force to the recorder it is necessary that the
natural frequencies of the transmission systems lie as
far as possible from the mentioned frequency range.
The cardiovascular activity is manifested by a
force acting on the human body which represents a
mechanical vibratory system transmitting the force
to the balistocardiographic apparatus.
The basic part of our portable quantitative
balistocardiograph is a very rigid piezoelectric force
transducer resting on a rigid steel chair. The
examined person sits (Figure 2) on the light seat
placed on the transducer and the force caused by the
cardiovascular activity is measured in this way. The
output of the piezoelectric pick-up is fed into an
operational amplifier.
Figure 2: Position of the examined person during the
QBCG session.
The advantage of the piezoelectric transducer is
in very low compliance together with a very high
natural frequency of the apparatus. Another
advantage of the rigid pick-up is the fact that it can
be preloaded with a substantial static force – the
weight of the examined person, and it is still
possible to measure the alternating forces of the
magnitude of g+ (gram as weight). The simple push
button is used to dispose of the static charge caused
by the weight of the person. The measured force is
registered (REG).
Dynamic calibration of the QBCG apparatus was
carried out by an electrodynamic exciter (EXC)
acting via a calibrated dynamometer (D), also of a
piezoelectric type, on a pick-up (Q) of the QBCG
(see Figure 3).
Figure 3: A set-up a dynamic calibration of the
quantitative balistocardiographic apparatus.
Figure 4: Records registered using the QBCG instrument.
The lower curve depicts the effect of force applied. The
upper curve is a record, from which we can determine size
and duration of the acting force. Compare with BCG
record on Figure 1.
Quantitative seismocardiography: (QSCG: This
method enables the recording of force applied
without phase or time deformation. Thus, heart rate
may be monitored and analysed using the method of
heart rate variability (statistical and autocorrelation
analysis, spectral analysis, total effect of regulation,
vegetative homeostasis, activity of subcortical
centre, activity of the vasomotor centre and stress
index). The method of QSCG was designated by the
laboratory employees as an absolutely non-invasive,
and the persons examined did not have any
electrodes attached to the body surface and were not
connected by cables to the registering instrument.
This new field of monitoring heart activity, whereby
we determine both amplitude-force and time-
frequency relationships, is termed quantitative
seismocardiography. Thus, one may determine the
force-response of the cardiovascular system to
ICINCO 2006 - SIGNAL PROCESSING, SYSTEMS MODELING AND CONTROL
62
changes in external stimuli, as well as the
autonomous nervous system regulation of the
circulation and the activity of the sympathetic and
parasympathetic systems.
2 METHODS OF QSCG
MEASUREMENT AND
ANALYSIS
2.1 Experimental Equipment
In terms of practical use, a portable telemetric
system for the QSCG measuring has been
developed. This system allows data to be acquired
and assessed using quantitative seismocardiography
(QSCG) and triaxial accelerometric measurements
on the thorax of a patient. It is composed of the three
HW modules that are telemetrically interconnected
with the option of interconnecting through a metallic
line. These are the seismocardiographic, the
accelerometric modules and a module for the data
transfer interconnected with the PC through the USB
interface. Block scheme of the whole system is on
figure 7.
Figure 5: Main sensor of she QSCG measuring equipment
- detail of the solid-state accelerometer between measuring
metal plates.
Sensing mechanical body reactions, which are
induced in response to the cardiovascular dynamics,
is provided by the seismocardiographic module,
which reads the strain coming from the mechanical
deformation of the piezo-electric plate. This sensing
module is mounted on a special device, which works
by transmitting the mechanical body reactions onto
the piezo-electric recorder. The accelerometric
module is applied for measuring thorax acceleration
as induced by movements from the heart activity,
this measurement is made on the three basic
orthogonal axes. The core of the module is the
sensing device composed of the two biaxial
monolithic semiconductor accelerometers. The data
transfer module is designed to transmit the data from
the radio-module into the PC through the USB
interface.
Figure 6: Measuring plates of the proposed QSCG device.
Figure 7: Block scheme of the experimental QSCG device.
2.2 Algorithm for the Time-Domain
Segmentation of the QSCG
We have developed algorithms for preprocessing,
segmentation and interactive analysis of the QSCG
signal. In this contribution we will focuse on the
original method for basic segmentation of the QSCG
Control
unit,
A/D
Converter
Commu-
nication
interface
WIFI
Bluetooth
Q
SCG unit
ECG measuring
unit
Auxiliary
measuring units
PATIENT‘S MEASURING DEVICE
PATIENT
PC
WORKSTATION
FOR
SUPERVISON,
DATA
PROCESSING
WIRELESS CONNECTION
A NEW METHOD FOR THE EVALUATION OF THE SIGNAL ACQUIRED FROM QUANTITATIVE
SEISMOCARDIOGRAPH - Hardware and Software Solution for the new Field of Monitoring Heart Activity
63
signal in time-domain; this first step is crucial for the
successfulness of the whole diagnostic process. Our
method is relatively simple and was developed for
the detection of QSCG pseudo-periods in real time.
The method is derived from a well-known and
robust algorithm for QRS complex detection in
traditional electrocardiograms (ECG), originally
developed by Hamilton et al. The algorithm was
based on the first derivative of the input signal and
many thresholds and parameters are automatically
adapted to individual changes in the input signal
using sophisticated empirical rules. The results
(position of the dominant – so-called R - wave) are
obtained with some detection delay (above 200 ms).
For details on the algorithm, see [Hamilton].
For our purposes it is important that the initial
values of many parameters are adjustable and by
modification of these values the original method was
slightly adapted to QSCG’s different curve
morphology. Namely the following parameters were
changed: (1) length of the first derivative from the
original 10 ms to 80 ms, (2) length of the high-pass
pre-filter from 125 ms to 350 ms, (3) length of
moving window integration from 80 ms to 200 ms.
Optimal values were selected experimentally in
order to achieve the best detection results.
Additionally, we developed a special backward
searching process for the precise detection of the
position of the I-wave and J-wave in each QSCG
pseudo-period.
The function of the whole algorithm is as follows:
output of the traditional ECG QRS detector gives the
rough position of the systolic complex inside the
QSCG - candidate X. Then the specific morphology
of the QSCG curve is utilized to backward search
the position of the J-wave – we expect the first big
negative peak in MTI samples (about 100 ms). If the
detection is successful, we assign the position of the
peak as the I-wave; see Figure 8.
6000
7000
8000
9000
10000
11000
12000
13000
1 58 115 172 229 286 343 400 457
time [samples]
Force (quant. le ve ls )
X
I
H
K
L
M
N
I
max
MTI
Figure 8: Backward local I-peak searching in the QSCG
cycle.
Finally we search forward for the position of the
J-wave, which we expect to be the first big positive
peak in maximally MTJ samples (about 160 ms), see
Figure 9.
6000
7000
8000
9000
10000
11000
12000
13000
1 58 115 172 229 286 343 400 457
time [samples]
Force (quant. le ve ls )
I J
H
K
L
M
N
J
max
MTJ
Figure 9: Forward local J-peak searching in the QSCG
cycle.
For the peak-detection we used a very simple
method based on the first difference (length 15 ms):
when the transition from negative to positive value
of the difference occurs, then the sequence is marked
as a negative peak; the transition from a positive to
negative difference means a positive peak. If
searching for the J-wave or the I-wave fails,
candidate “X” is rejected and the algorithm
continues without detection of the QSCG pseudo-
period.
The rejection of “candidate X” is very important
step and it increases robustness of the whole
detection procedure against the artifacts – see
demonstration on the Figure 10.
8800
10800
12800
1 11 21 31 41 51 61 71 81 91 101 111 121 131 141 151 161 171 181 191 201 211 221
time (samples)
Force (quant. levels)
X3
I
max
MTI
Figure 10: Rejection of the false beat detection. We search
backward from “candidate X3” for the first big negative
peak. The I-wave must be recognized in MTI samples
(about 100 ms), so in this case the detection was not
successful.
The false detection of the dominant “candidate
X”, which is not a true QSCG cycle, was corrected
by the proposed simple backward searching
ICINCO 2006 - SIGNAL PROCESSING, SYSTEMS MODELING AND CONTROL
64
algorithm, because the morphology in the nearest
neighborhood of the point X3 does not match the
detection rules – backward searching for the I-wave
in MTI samples was not successful, the false
positive detection of the systolic complex was
correctly rejected.
5000
7000
9000
11000
13000
15000
1 73 145 217 289 361 433 505 577 649 721 793 865 937 1009 1081 1153 1225
time (samples)
Force (quant. levels)
J
J
J
X1
X2 X3 X4
Figure 11: Result of the whole detection: false candidate
X3 was correctly rejected.
The whole software system contains additional
modules for statistical and autocorrelation analysis,
spectral analysis, assessment of the aggregated effect
of the regulation of autonomous functions of
vegetative homeostasis, activity of the vasomotor
centre, activities of the sympathetic cardiovascular
centre and the stress index (SI). Our experimental
software allows also automatic extraction of
classical QSCG hemodynamical parameters,
especially so called systolic force (SF). The current
version of the system is designed for OS Windows
XP and has user-friendly interface. Block scheme of
the system is on figure 12:
Figure 12: Block scheme of the software system.
Presented algorithm is in the box „Unit for time-domain
segmentation of the QSCG curves“.
3 CONCLUSION
For high-quality measurements we can obtain good-
looking signals for which both methods exhibit
excellent results. For disruptive and spurious signals
there is still a good chance of obtaining authentic
information because we first detect the impairments
and remove the particular interval of the signal. It is
true that in using this method we also remove certain
useful information but simultaneously ensure
processing of the remaining signal. We emphasize
that we need not process all consecutive pseudo-
periods in the signal but only a sufficient amount of
pseudo-periods.
4000
6000
8000
10000
12000
14000
1 70 139 208 277 346 415 484 553 622 691 760 829 898 967 1036 1105 1174 1243 1312 1381 1450
Force (quant. levels)
time (samples)
J J
J
J J
I
I
I
I
I
Figure 13: Typical QSCG signal with correctly placed
reference points.
For good-looking and typical signals, the
methods behave very well, achieving nearly
complete success (see Figure 13). The success
decreases with deterioration of the signal. On the
other way, in such signals it is often difficult even
for the human expert to recognize correct pseudo-
period time points (see Figure 14).
0
2000
4000
6000
8000
10000
12000
14000
16000
18000
1 70 139 208 277 346 415 484 553 622 691 760 829 898 967 1036 1105 1174 1243 1312 1381 1450
Force (quant. levels)
time (samples)
I
I I
J
J J
Figure 14: QSCG signal with the motion artifact; it is
difficult to recognize correct positions of some reference
points.
Measuring
software
Visualization,
printing and
archivation
Setup and
calibration
Signal
acquisition
and data
storage
Data processing uni
t
Unit fo
r
time-domain
segmentation of the
QSCG curves
Unit for extraction of
the
hemodynamical
parameters
Unit for extraction of
the HRV parameters
A NEW METHOD FOR THE EVALUATION OF THE SIGNAL ACQUIRED FROM QUANTITATIVE
SEISMOCARDIOGRAPH - Hardware and Software Solution for the new Field of Monitoring Heart Activity
65
The QSCG signal offers specific information
about functional changes of the cardiovascular
system regulation which preceded the structural
changes coming later. The equipment is ready for
use, algorithms for automatic analysis of the QSCG
signal are prepared.
Quantitative seismocardiography probably offers
a more complex view of both inotropic and
chronotropic heart function. It will be suitable for:
examining operators exposed to stress; for assessing
the effect of work, fatigue, mental stress; for
monitoring persons as part of disease prevention; for
determining a person’s ability to carry out their
duties both on the ground and in the air.
ACKNOWLEDGEMENTS
This work has been supported by the Ministry of
Education of the Czech Republic under project No.
MSM6840770012 and by the project UREKA
E!3031.
REFERENCES
Jerosch-Herold, M. et al., 1999. The seismocardiogram as
magnetic-field-compatible alternative to the
electrocardiogram for cardiac stress monitoring, In
International Journal of Cardiac Imaging, 15(6), pp.
523-31
Trefny, Z. et al., 1996: Some physical aspects in
cardiovascular dynamics, In J. Cardiovasc. Diagnosis
and Procedures, 13(2), pp. 141 - 145
Hamilton, P. – Tompkins, W.J. (1987): Quantitative
investigation of QRS detection rules using the
MIT/BIH arrhythmia database, In IEEE Trans.
Biomed.Eng., 33, pp. 1158-65
Trefny Z. – David E. - Bayevsky R.M.: Achievements in
Space Medicine into Healt Care Practice and Industry,
Development of Space Cardiology metods in the
Earth's Health Service, Berlin 2001
Freisen G., Jannet T.: A comparison of the Noise
Sensitivity of Nine QRS Detection Algorithms’, IEEE
Trans. Biomed. Eng., 1990, 85(1)
ICINCO 2006 - SIGNAL PROCESSING, SYSTEMS MODELING AND CONTROL
66