ap-5-11.dvi


Acta Polytechnica Vol. 51 No. 5/2011

Unobtrusive Non-Contact Detection of
Arrhythmias using a “Smart” Bed

Ch. Brüser

Abstract

Wepresent an instrumented bed for unobtrusive, non-contactmonitoring of cardiac and respiratory activity. The system
presented here is based on the principle of ballistocardiography (BCG), and measures cardiopulmonary vibrations of the
body by means of an electromechanical foil (EMFi) attached to the mattress. Using our system, a clinical study with
13 participants was conducted to assess the BCG’s ability to distinguish atrial fibrillations from normal sinus rhythms.
By computing a time-frequency representation of the recorded signals based on parametric autoregressive estimators,
we can show clear qualitative differences between normal and arrhythmic BCG episodes. The same distinctive features
could also be observed when applying our method to a simultaneously recorded reference ECG. Our results suggest that
ECG and BCG both contain the same basic information with respect to the presence of atrial fibrillations, and that a
bed-mounted BCG sensor can indeed be used to detect atrial fibrillations.

Keywords: ballistocardiography, arrhythmia, bed, unobtrusive home-monitoring, atrial fibrillation.

1 Introduction
Cardiovascular diseases in general and heart failure
in particular are among the commonest reasons for
hospitalization in the industrialized countries [1]. In
order to deal with the growing number of patients,
there is a need for technical solutions which enable
personalized monitoring and treatment, preferably
at home. In recent years, the bed has emerged as
a promising place for long-term monitoring of car-
diopulmonary activity at home, as virtually every-
one spends a significant portion of the day in bed.
Instrumented beds could also be applied in the gen-
eral wards of hospitals, where fully automatic, un-
obtrusivemonitoring systems could reduce the work-
load of the staff and increase the safety of the pa-
tients.
A promising approach for measuring cardiopul-

monary activity is the integration of highly sensitive
mechanical sensors into the bed-frame or mattress
which record the vibrations of the body caused by
the mechanical activity of the heart and by the res-
piratory movements of the thorax.
This basic principle, known under the term “bal-

listocardiography” (BCG), was first reported in the
late 19th century [2]. Through improvements in sen-
sor technologies and digital signal processing tech-
niques, the field has gained renewed interest in recent
years. Using a variety of different sensors, including
but not limited to strain gauges [3,4], PVDF and
EMFi sensors [5,6], accelerometers [7], hydraulic [8]
and pneumatic sensors [9,10] as well as optical de-
vices [11], BCG systems have been integrated into

objects of daily life, such as beds [4–12], chairs [13]
and even weighing scales [3]. These systems share
the commonadvantagethat theyareunobtrusiveand
do not require direct skin contact, unlike for exam-
ple, a conventional ECG. User errors, unlike, incor-
rectly attached electrodes can mostly be avoided by
bed-based BCG systems, which do not require any
user interaction or professional supervision to per-
form nightly measurements over extended periods of
time. Hence BCG systems are very well suited for
long-term monitoring.
While there have been significant contributions

from various sensor modalities for BCG recording,
as well as a few studies evaluating modernBCG sys-
tems, using healthy subjects, there is currently a lack
of researchonwhether these systemsactuallyprovide
clinically useful informationwhenmeasuring real pa-
tients. After all, the answer to this question will de-
cide whether or not these systems come into use in
clinical practice. Arrhythmias, in particular, have so
far been regarded as undesirable artefacts in many
publications. We therefore devised a clinical study
with the explicit goal of evaluating our BCG sys-
tem’s ability to detect cardiac arrhythmias. For our
study we focused on atrial fibrillation (AF), since it
is the most common type of arrhythmia [14]. Dur-
ing atrial fibrillation, the two upper chambers of the
heart (atria) fibrillate and do not perform coordi-
nated contractions [15]. AF is a marker for other
severe illnesses such as congestive heart failure [16].
Long-term monitoring of AF episodes might enable
an early detection of worsening conditions in heart
failure patients.

18



Acta Polytechnica Vol. 51 No. 5/2011

In the remainder of this paper, we first introduce
our BCG measurement system (Section 2.1). Then
we describe the design of the clinical study that was
performed to acquire BCG data from real arrhyth-
mia patients (Section 2.2). Next, we present the sig-
nal processing techniques that were applied to evalu-
ate the acquired data (Section 2.3). We conclude by
discussing the results and presenting our conclusions
(Sections 3 & 4).

2 Materials and Methods

2.1 BCG Acquisition System

The BCG acquisition system used in this study con-
sists of a single electromechanical-film (EMFi) sen-
sor [17] (30 cm ×60 cm, thickness < 1 mm) and the
acquisition electronics for amplifying and digitizing
the analog sensor signal. Mechanical deformation of
the electromechanical film generates a charge, ΔQ,
which is proportional to the dynamic force,ΔF , act-
ing along the thickness direction of the sensor

ΔQ = kΔF (1)

where k is the sensitivity coefficient. The result-
ing charge is amplified by a charge amplifier, and is
then digitized with 12 bits at a sampling frequency
of 128 Hz.
The EMFi foil is mounted on the underside of

a thin foam overlay, which is then placed on top of
the mattress of a regular bed. Due to its thinness
and its flexible properties, the presence of the sen-
sor is almost imperceptible to the person lying in the
instrumented bed. However, owing to the sensitiv-
ity of the EMFi foil, cardiopulmonary movements of
the person lying in bed can be recorded. In order
to obtain optimal signals, the sensor is mounted in a
fixed position under the thorax region (see Figure 1).
A short segment of a signal recorded by our BCG
system containing three heart beats is shown in Fi-
gure 2. Vertical dotted lines indicate the time points
at which R peaks occurred in the reference ECG.

Fig. 1: Picture of the EMFi BCG sensor attached to the
bottom of the thin foam overlay

2.2 Measurement Scenario

In order to assess whether atrial fibrillations and si-
nus rhythm can be distinguished in a BCG record-
ing, the following study was performed at the Uni-
versity Hospital in Aachen, Germany. The study
was approved by the ethics board of the Univer-
sity Hospital Aachen (ref. number: EK075/10, date:
05.05.2010). A total of 13patients (3 female, 10male,

age: 63.6±16.3 years, BMI: 28.6±4.1
kg
m2
) whowere

visiting the hospital to undergo ambulatory treat-
ment for atrial fibrillation gave their informed writ-
ten consent and were included in our study. To re-
turn the patients’ hearts to a regular sinus rhythm, a
routine procedure called synchronized electrical car-
dioversion [18] was performed on each patient. Dur-
ing this procedure, an electrical current is adminis-
tered to the heart. Unlike defibrillation, the initial
current dose is smaller and the shock is triggered by
the R-peak in the ECG in order to reduce the risk of
induced ventricular fibrillation.
For the entire duration of their treatment, the

subjects were placed in a hospital bed instrumented
with an EMFi foil sensor, as described above. In
addition to the BCG, a 3-lead reference ECG was
recorded with a sampling rate of 500 Hz. BCG and
ECG data was continuously acquired before, during,
and after the procedure. The mean length of the in-
dividual BCG recordings is 45 minutes.
This measurement scenario has the major advan-

tage that it allows theBCGof the samepatient to be
recordedwhile exhibiting thepathology(i.e. arrhyth-
mia/atrial fibrillations) as well as when the patient’s
heart is returned to a normal sinus rhythm. Hence
an inter-personal as well as an intra-personal com-
parison of the BCG signal during arrhythmias and
during normal rhythms is possible.

Fig. 2: Exemplary trace of three heart beats recorded by
the bed sensor. Vertical lines indicate the occurrence of
R peaks in the simultaneously recorded reference ECG

2.3 Signal Analysis

PSD Estimation using AR Models

Autoregressive (AR) models are a common choice
for parametric estimation of the power spectral den-

19



Acta Polytechnica Vol. 51 No. 5/2011

sity (PSD) of a signal [19]. First, a given signal,
x(n), n ∈ [0, N − 1], is modelled as the output of a
discrete, all-pole, infinite impulse response (IIR) fil-
ter whose input is white noise, w(n), of variance σ2:

x(n)= −
p∑

k=1

akx(n − k)+ w(n). (2)

Thus, for an AR(p) model of p-th order, only the
filter coefficients a1, . . . , ap and the noise variance σ

2

need to be estimated to fully describe this process.
After estimating these parameters, the PSD, P(f),
of the modelled process can be computed as:

P(f)=
σ2∣∣∣∣1+ p∑

k=1
ake−j2πf p

∣∣∣∣
2 . (3)

A number of different AR parameter estimation
methods are known in the literature [19]. These are
typically based on minimizing the estimate of the
prediction error power. Popular estimators include
the Yule-Walker method and Burg’s method. For
our analysiswhich follows, we have chosen to use the
Burg estimator [20].
Autoregressive spectral estimation can provide

better modelling of the peaks in the PSD than non-
parametric methods, especially when dealing with
short signal lengths [19]. This improvement comes
at the cost of a less accurate description of the val-
leys of the PSD. When dealing with quasi-periodic
signals measuring cardiac activity, however, this is a
valid trade-off, as we are primarily interested in the
peaks of the PSD. However, this advantage of AR
spectral estimators exists only when the assumption
of an underlying AR process is indeed valid for the
given signal. Furthermore, the model order needs to
be carefully chosen to achieve high quality estimates.

Spectrogram

Since biosignals, such as ECG and BCG, are also
highly non-stationary in nature (especially in the
presence of arrhythmias), a deeper insight into
the properties of these signals can be achieved by
analysing their time-frequency distributions. A com-
mon approach to obtain a time-varying spectral rep-
resentation of a signal is to divide the signal into
smaller (overlapping) epochs and estimate the PSD
for each of these epochs separately [21]. This so-
called, spectrogram is usually computed bymeans of
the short-time Fourier transform (SFFT). However,
AR-based spectral estimators can equally be used.
Let P Lm(f) denote the estimatedPSDof an epoch

of the signal x(n) which starts at the m-th sample
and has a length of L samples. We can then define
the AR-based spectrogram as

S(f, n)= P Ln (f). (4)

The better distinction of the peaks that can be
obtained through AR estimators means that smaller
epoch lengths can be chosen. This allows an increase
in time resolutionwhile at the same timemaintaining
a similar resolution in the frequency domain.

ECG and BCG Signal Analysis

Both the BCG signals and the lead II ECG signals
recordedduring our studywere first low-pass filtered
to 15 Hz and then downsampled to 30 Hz. Then
the signals were split into 5 s long epochs with 4 s
of overlap, thus resulting in one epoch every second.
For each epoch, the PSDwas estimated using anAR
model of order 50. Spectrogramswere then obtained
for each signal from the sets of estimated PSDs.

3 Results and Discussion
Figure 3 shows the BCG and ECG spectrograms of
a healthy reference subject. Both spectrograms show
a very similar image containing clear bright lines re-
lated to the heart frequency and its harmonics. From
a purely visual standpoint, one could conclude that
both signals, i.e. the non-contact bed measurement
and the reference ECG, contain similar information
about the current heart rate of the subject.
Much to our surprise, the spectrograms also

showedapparent visual similarities duringpathologic
episodes of atrial fibrillation. Figure 4 shows the
time-frequencyanalysisof the signals recordedduring
the treatment sessionwith one of the patients in our
study. Before the cardioversion was performed, the

Fig. 3: Spectrograms of simultaneously recorded BCG
and ECG signals of a healthy subject, respectively. Both
images show clearly visible lines corresponding to the
heart rate of the subject and its harmonics. The higher
power densities in the lower frequencies of theBCG spec-
trogramare related to respiratory-inducedmotions in the
BCG signal

20



Acta Polytechnica Vol. 51 No. 5/2011

Fig. 4: Time-frequency analysis of the BCG and the reference ECG of patient 11 before and after the cardioversion
is performed. Following the cardioversion, both spectrograms change from a noise-like appearance into clearly visible
lines representing a base frequency and its harmonics. (From top to bottom: 1. BCG signal recorded by the bed-sensor,
2. beat-to-beat heart rates computed from the reference ECG, 3. spectrogram of the BCG signal, and 4. spectrogram of
the reference ECG signal)

patient suffered from atrial fibrillation, which caused
strong and rapid fluctuations in the ECG-derived
heart rates, as shown in the second plot of the figure.
After the cardioversion event, the patient’s heart re-
turned to a sinus rhythm and the heart rate stabi-
lized.
When inspecting the spectrograms of the ECG

signal and the BCG signal, respectively, the change
of state induced by the cardioversion is also immedi-
atelyvisible. While the spectrogramduringatrialfib-
rillationhas a smeared, almostnoise-like appearance,
the spectograms change into the previously seen pat-
tern of distinct lineswhen the subject’s cardiac activ-
ity returns to a sinus rhythm. This pattern conforms
with what we observed earlier for healthy subjects.
In addition to the example shown here, we observed
the samedifferences in spectrogrampatternsbetween
arrhythmia and normal periods for all patients that
took part in our study.
These preliminary results seem to support our

initial hypothesis that arrhythmic cardiac activity
can indeed be detected using a bed-mounted BCG
system. Furthermore, the observation that ECG

and BCG spectrograms undergo the same qualita-
tive changes leads us to believe that the BCG signal
recorded using our unobtrusive, non-contact sensor
systemdoes indeed contain similar informationas the
ECG with respect to the presence of arrhythmias.
Nevertheless, our experiments also highlight a

major challenge on to road towards a fully auto-
matic BCG-based arrhythmia detector. As shown
in Figure 4, motion artefacts immediately after the
cardioversion (as evident through the increased am-
plitude of the BCG signal) cause distortions in the
BCG spectrogramwhich, at first glance, appear sim-
ilar to arrhythmias. However, an automatic algo-
rithm might still be able to distinguish motion arte-
facts from arrhythmia by taking the BCG signal am-
plitude and the details of the frequency distribution
into account.

4 Conclusion
We have introduced a bed-based sensor system that
can unobtrusively monitor the cardiopulmonary ac-
tivity of a person lying in bed. Unlike previous work

21



Acta Polytechnica Vol. 51 No. 5/2011

in the field, which has mostly treated arrhythmias in
the data sets as undesirable artefacts, we have de-
vised and presented a clinical study dedicated ex-
plicitly to the goal of evaluating the fitness of the
proposed system for detecting cardiac arrhythmias.
Through the analysis of the data acquired during
this study by means of an AR-model-based time-
frequency representation, we have shown that the
proposed systemdoes indeed enable cardiac arrhyth-
mias to be detected.
Our work prepares the way for future research

on fully automatic algorithms for detecting arrhyth-
mias in BCG signals. While the types of arrhyth-
miaswhich canbe detected inBCGrecordings is still
limited to atrial fibrillation, our findingsmight facili-
tate improvements in the long-termmanagementand
treatment of cardiac diseases for which AF episodes
can be an important marker.

Acknowledgement

The research presented in this paper was supervised
by Prof. S. Leonhardt, RWTH Aachen University in
Aachen, Germany, andwas sponsored byPhilips Re-
search, Eindhoven, the Netherlands.
The author also thanks S. deWaele ofPhilipsRe-

search for insightful discussions. Further thanks go
toProf. P. Schauerte andM.Zink of theDepartment
of Cardiology, Medical Clinic I, University Hospital
Aachen, for enabling and supporting the execution of
this study.

References

[1] Organization, W. H.: The World Health Report
2004. 2004. http://www.who.int/whr/2004/
annex/topic/en/annex 2 en.pdf

[2] Gordon, J. W.: Certain Molar Movements of
the Human Body produced by the Circulation
of the Blood, Journal of Anatomy and Physiol-
ogy, 1877, vol. 11, p. 533–536.

[3] Inan, O. T., Etemadi, M., Wiard, R. M., Gio-
vangrandi, L., Kovacs, G. T. A.: Robust Bal-
listocardiogram Acquisition for Home Monitor-
ing, Physiological Measurement, 2009, vol. 30,
pp. 169–185.

[4] Brüser,C., Stadlthanner,K., Brauers,A., Leon-
hardt, S.: Applying Machine Learning to De-
tect Individual Heart Beats in Ballistocardio-
grams, In Proc. 32nd Ann. Int. Conf. of the
IEEE EMBS, Buenos Aires, Argentina, 2010,
pp. 1926–1929.

[5] Kortelainen, J. M., Virkkala, J.: FFT averaging
of multichannel BCG signals from bed mattress

sensor to improve estimation of heart beat in-
terval, Proc. 29th Ann. Int. Conf. of the IEEE
EMBS,Cité Internationale, Lyon, France, 2007,
p. 6685–6688.

[6] Aubert, X. L., Brauers, A.: Estimation of Vi-
tal Signs in Bed from a Single Unobtrusive
Mechanical Sensor: Algorithms and Real-life
Evaluation, Proc. 30th Ann. Int. Conf. of the
IEEE EMBS, British Columbia, Canada, 2008,
pp. 4744–4747.

[7] Chuo, Y., Tavakolian,K., Kaminska, B.: Evalu-
ation of a Novel Integrated Sensor System for
Synchronous Measurement of Cardiac Vibra-
tions andCardiacPotentials,Journal of medical
systems, 2009, pp. 1–11.

[8] Zhu, X., Chen, W., Nemoto, T., Kanemitsu, Y.,
Kitamura, K., Yamakoshi, K., Wei, D.: Real-
Time Monitoring of Respiration Rhythm and
Pulse Rate During Sleep, IEEE Trans. Biomed.
Eng., no. 12, vol. 53, 2006, pp. 2553–2563.

[9] Watanabe, K., Watanabe, T., Watanabe, H.,
Ando, H., Ishikawa, T., Kobayashi, K.: Non-
invasive Measurement of Heartbeat, Respira-
tion, Snoring and Body Movements of a Sub-
ject in Bed via a Pneumatic Method, IEEE
Trans. Biomed. Eng., no. 12, vol. 52, 2005,
pp. 2100–2107.

[10] Chee, Y., Han, J., Youn, J., Park, K.: Air
mattress sensor system with balancing tube for
unconstrained measurement of respiration and
heart beat movements, Physiological Measure-
ment, vol. 26, 2005, pp. 413–422.

[11] Spillman Jr., W. B., Mayer, M., Bennett, J.,
Gong, J., Meissner, K. E., Davis, B., Cla-
us, R. O., Muelenaer Jr., A. A., Xu, X.: A
‘smart’ bed for non-intrusive monitoring of pa-
tient physiological factors,Measurement Science
and Technology, vol. 15, 2004, pp. 1614–1620.

[12] Mack, D. C., Patrie, J. T., Suratt, P. M.,
Felder, R. A., Alwan, M. A.: Develop-
ment and preliminary validation of heart rate
and breathing rate detection using a pas-
sive, ballistocardiography-based sleep monitor-
ing system, IEEE Trans. Inf. Technol. Biomed.,
no. 1, vol. 13, 2009, p. 111–120.

[13] Junnila, S., Akhbardeh, A., Värri, A.: An Elec-
tromechanical Film Sensor Based Wireless Bal-
listocardiographic Chair: Implementation and
Performance, Journal of Signal Processing Sys-
tems, no. 3, vol. 57, 2008, pp. 305–320.

22



Acta Polytechnica Vol. 51 No. 5/2011

[14] Go, A. S., Hylek, E. M., Phillips, K. A.,
Chang, Y., Henault, L. E., Selby, J. V.,
Singer, D. E.: Prevalence of diagnosed atrial
fibrillation in adults: national implications for
rhythmmanagement and stroke prevention: the
AnTicoagulation and Risk Factors in Atrial
Fibrillation (ATRIA) Study, JAMA, no. 18,
vol. 285, 2001, pp. 2370–2375.

[15] Katz, A. M.: Physiology of the Heart. M. Lip-
pincott Williams & Wilkins, 2005.

[16] Maisel, W. H., Stevenson, L. W.: Atrial fib-
rillation in heart failure: epidemiology, patho-
physiology, and rationale for therapy, Ameri-
can Journal of Cardiology, no. 6, vol. 91, 2003,
pp. 2D–8D.

[17] Paajanen, M., Lekkala, J., Kirjavainen, K.:
ElectroMechanical Film (EMFi): a new mul-
tipurpose electret material, Sensors and Ac-
tuators A: Physical, no. 1–2, vol. 84, 2000,
pp. 95–102.

[18] Shea, J. B., Maisel, W. H.: Cardioversion, Cir-
culation, no. 22, vol. 106, 2002, pp. e176–e178.

[19] Kay, S. M.: Modern Spectral Estimation: The-
ory and Application. Prentice Hall, 1999.

[20] Broersen, P.: Finite-sample bias propagation
in autoregressive estimation with the Yule-
Walker method, IEEE Trans. Instrumentation
andMeasurement, no. 5, vol. 58, 2009,pp. 1354–
1360.

[21] Mitra, S. K.: Digital Signal Processing: A
Computer-Based Approach. McGraw-Hill, 2001.

About the author

Christoph BRÜSER was born in Troisdorf, Ger-
many, in 1983. He holds a Dipl.-Ing. degree in Com-
puter Engineering from RWTH Aachen University,
Aachen, Germany. Currently, he is pursuing a Dr.-
Ing. (Ph.D.) degree in the Department of Medical
Information Technology, RWTH Aachen University,
where he is also working as a Research Assistant.
His research interests include biosignal processing
and classification as well as unobtrusive physiologi-
cal measurement techniques.

Christoph Brüser
E-mail: brueser@hia.rwth-aachen.de
Philips Chair for Medical Information Technology
RWTH Aachen University
Pauwelsstrasse 20, 52074 Aachen, Germany

23