IEEE TRANSACTIONS ON BIOMEDICAL ENGINEERING, VOL. 60, NO.

1, JANUARY 2013 179

Non-Contact ECG Sensing Employing

Gradiometer Electrodes
GuoChen Peng∗, Student Member, IEEE, and Mark F. Bocko, Member, IEEE

Abstract—Noncontact, capacitive electrocardiogram (ECG) (CMRR) of the sensor system [4]. Interfering signals can be re-
measurements are complicated by motion artifacts from the rel- duced by employing a driven right leg (DRL) connection, active
ative movement between the ECG electrodes and the subject. To shielding, and guarding of cables, and by employing a notch
compensate for such motion we propose to employ first and second
order gradiometer electrode designs. A MATLAB-based simula- filter at the output.
tion tool to enable assessment of different electrode configurations 2) Triboelectrically generated static charge caused by rubbing
and placements on human subjects has been developed to guide between the electrodes and the subject’s clothing has been dis-
the refinement of electrode designs. Experimental measurements cussed in [3]. Common-mode electrostatic charges on the body
of the sensitivity, motion artifact cancellation, and common mode were also investigated in [5]. In addition to reducing movement
rejection for various prototype designs were conducted with hu-
man subjects. Second order gradiometer electrode designs appear at the electrode-subject interface, other ways to reduce this ef-
to give the best performance as measured by signal to noise plus fect include careful choice of materials and providing a static
distortion ratio. Finally, both gradiometer designs were compared charge discharge path at the electrode-subject interface.
with standard ECG recording methods and showed less than 1% 3) The signal gain may be a function of the source capacitance
beat detection mismatch employing an open source beat detection and any stray capacitance at the preamplifier input, which can
algorithm.
lead to baseline wandering and gain distortion. This effect can
Index Terms—Body surface potential map (BSPM), capaci- be minimized through employing a voltage mode preamplifier,
tive sensors, charge preamplifier, common mode rejection ratio as opposed to a charge-mode preamplifier; however, preamp
(CMRR), electrocardiography, modeling, motion artifacts.
noise then becomes a consideration.
4) The preamplifier immediately following the ECG electrode
is the major contributor to the overall electronic noise level [1],
I. INTRODUCTION [6] and may be addressed by employing careful preamplifier
ONCONTACT biosensors for cardiac monitoring are of design.
N great interest for a number of long-term health sensing
applications ranging from exercise and fitness monitoring to
Changes in source capacitance due to the relative motion
of the electrodes and the subject leads to modulation of both
management of chronic health conditions. However, the pres- the signals of interest as well as the aforementioned sources of
ence of motion-related artifacts and common-mode interference interference and noise, which in turn may generate interference
remains a challenging problem in practice [1]–[3]. The major within the signal band of interest. This effect may be large
challenges are: 1) common mode electromagnetic interference enough in practical scenarios to completely obscure the ECG
from power mains, 2) triboelectrically generated charge from signal. In [2], the effect of motion artifacts by static common
rubbing of the electrodes on the subject’s clothing, 3) modu- mode voltages has been investigated and the original ECG signal
lation of the bioelectric potential signals from motion-related is reconstructed using known movement parameters measured
source impedance changes, and 4) electronic noise. Methods by a secondary sensor placed near the ECG electrodes.
that have been typically employed to address these challenges We have pursued an alternative approach in which gradiome-
are as follows. ter electrode designs are used to cancel common mode inter-
1) Interference from ac power mains may corrupt electro- ference at the source. To explore the design space we have
cardiograph signals due to low-common mode rejection ratio developed a (2 + 1)D human ECG potential map (2 spatial
dimensions plus time) that allows for assessment of various
electrode designs in terms of their signal sensitivity and their
sensitivity to subject motion. In the following section we present
the gradiometer electrode designs considered in this study. In
Manuscript received March 14, 2012; revised June 26, 2012; accepted Section III we discuss the simulations, including the placement
August 21, 2012. Date of publication September 18, 2012; date of current
version December 14, 2012. This work was supported in part by Blue Highway, of electrodes on the subject employing the (2 + 1)D, ECG poten-
the New York State Foundation for Science, Technology, and Innovation, and tial map simulation tool, we then go on to present experimental
in part by the Eunice Kennedy Shriver National Institute of Child Health De- results with various electrode designs on human subjects.
velopment Grant R01HD060789. Asterisk indicates corresponding author.
∗ G. Peng is with the Electrical and Computer Engineering Department,
University of Rochester, Rochester, NY 14627-0231 USA (e-mail: gupeng@
ece.rochester.edu).
II. GRADIOMETER ELECTRODE DESIGNS
M. F. Bocko is with the Electrical and Computer Engineering Department, In conventional ECG configurations the relative motion of the
University of Rochester, Rochester, NY 14627-0231 USA (e-mail: bocko@
ece.rochester.edu). electrodes is independent [7] and unpredictable, as in planar-
Digital Object Identifier 10.1109/TBME.2012.2219531 fashionable circuit board electrodes [8]. To address this issue,

0018-9294/$31.00 © 2012 IEEE

180 IEEE TRANSACTIONS ON BIOMEDICAL ENGINEERING, VOL. 60, NO. 1, JANUARY 2013

Fig. 1. Two types of gradiometer electrode configurations were investigated.

On the left a first order gradiometer dual electrode design is shown and on
the right a second order quad electrode design is illustrated. The areas of the
electrodes in a given configuration are equal but the electrode aspect ratio a/b
and spacing are variable.
Fig. 3. Total source capacitance is the series combination of the capacitance
of the polymer overlayer C O L , the air gap, C A , and the subject’s clothing C C .

In (1) A is the area of each subelectrode, A(θx , θy ) is the pro-

jected area of the electrode, which is a function of θx and θy and
ε0 , εOL , and εC are respectively the permittivity of free space,
the relative permittivity of the electrode overlayer and the rela-
tive permittivity of the subject’s clothing. Finally, the effective
gaps of the three capacitors are also indicated in Fig. 3.
To compare the performance of the first and second order gra-
diometer electrode configurations of various designs we com-
pute the signal strength, the electronic noise, and the common
Fig. 2. Definition of the orientation variables, where the rotation angles around mode distortion artifacts.
the x- and y- axes are denoted θx and θy and the distance from the center of the
electrode array to the subject’s clothing is given by h.
A. Electronic Output Noise
Each electrode is coupled to an independent charge preampli-
we considered using first and second order gradiometer designs fier, which contributes independent electronic noise. The output
(see Fig. 1) in which two or four closely spaced electrodes are noise power, Vn2,out (θx , θy , h), is given by the amplifier input
employed. Such gradiometric measurement schemes may en-
noise power, Vn2,in (θx , θy , h), multiplied by the noise gain of the
able cancellation of common mode signals at the point of sens-
amplifier G2n
ing and thus, to a first approximation, are insensitive to com-
mon mode changes in the electrode array to subject distance. Vn2,out = Vn2,in (θx , θy , h) · G2n (Cs (θx , θy , h)) (2)
The first order sensor has two equally sized rectangular subelec-
trodes. The second order electrode array employs four equally where the noise gain Gn is defined as (Cs + Cf + Cin )/Cf
sized subelectrodes. In the following we discuss the effects above the low frequency roll-off frequency of 1/Rf Cf rad/sec.
of the sizes and spacing between the electrodes on the sensor Both terms on the right of (2) are affected by the electrode
performance. orientation variables θx , θy , and h. It is assumed that the input
The output signal of the dual electrode configuration is given noise current contribution of the preamplifier is negligible when
by taking the difference of the voltages of the two electrodes using an ultralow input leakage current amplifier [9], thus the
(marked + and −); in our implementation we employ two sep- additive noise of the preamplifier dominates.
arate charge preamplifiers [9]. In the second order electrode
configuration, the signals from diagonal pairs are added and the B. Common Mode Noise (CMRR Sensitivity)
two sums are then subtracted, for which we employ four pream- The sensor common mode output is determined by the com-
plifier channels. As illustrated in Fig. 2(a) the distance from the mon mode gain, which is minimized by matching the source
center of the electrode array to the subjects clothing is given by capacitances. The CMRR is therefore affected by the differ-
h and the rotation angles about the x- and y-axes (defined in the ential schemes in both gradiometer types and the smaller the
figure) are given by θx and θy . Changes in h, θx, or θy modulate differences of the source capacitances the less the sensitivity
the source capacitance by changing the gap between the sensor of the CMRR to relative motion. The motion-induced common
and the subject and to a lesser extent the projected area of the mode distortion is defined as the difference between the sensor
electrodes A(θx , θy ). The total source capacitance is the series common mode output when there is electrode–subject relative
combination of the three capacitances COL , CA , and CC from motion and the output with no electrode motion. This may be
the overlayer, air gap, and clothing respectively. See Fig. 3. expressed in (3) as
  Vn o C M (θx , θy , h, t) = Vout (t)|(θ x ,θ y ,h)
ε0 A(θx , θy ) ε0 εOL A ε0 εC A(θx , θy )
Cs = series , , (1)
h tOL tC − Vout (t)|(θ x =0,θ y =0,h=fixed) . (3)
IEEE TRANSACTIONS ON BIOMEDICAL ENGINEERING, VOL. 60, NO. 1, JANUARY 2013 181

TABLE I
SIMULATION PARAMETERS FOR THE CAPACITIVE SENSOR

Fig. 5. (a) Gradiometer electrode with ILC3As. It was mounted on a movable

platform. (b) Experimental setup.

the average common mode distortion for the Type B sensor is

always less than type A for all sensor sizes.
The simulation results shown in Fig. 4 averaged over various
electrode sizes gives an average motion-related gain distortion
signal of 31.6 μVrm s for the dual electrode design (type A) and
11.2 μVrm s for the quad electrode design (type B). Thus, the
quad electrode design performs about 2.8 times better than the
dual electrode design.

IV. GRADIOMETER EXPERIMENTS

Fig. 4. Simulation results of the motion-related gain distortion artifacts(V rm s ) A. Experiments on the Common Mode Signal
due to common mode distortion and the electronic noise for the type A and type
B electrodes. Artifact from common mode distortion dominates as the electrode In order to verify the model a gradiometric sensor was
size is increased. Also note that Type B electrodes produce smaller total motion- mounted on a movable platform to give control of h and the
related gain distortion artifact signals.
rotation angles θx and θy . The gradiometer and experimental
setup are shown in Fig. 5. The total sensor area for each type
The power spectral density of this signal is then the Fourier of gradiometer electrode was 4.41 cm2 with 0.5-mm in-plane
transform of its autocorrelation function [10], spacing between electrodes. A 5 V 10 Hz sinusoidal signal was
2
applied to an aluminum sheet covered in textile to serve as a
SN ,C M = |F {R (Vn o C M (θx , θy , h, t))}| (4) surrogate for the subject. The sensor was held at a distance of
4.2 cm from the equipotential surface. Data were recorded over
III. GRADIOMETER SIMULATION a tilt range from 0◦ to 9◦ in both θx and θy with Δθ = 0.9◦ .
The measurements and simulation results are shown in Fig. 6.
To simulate the effects of motion on the sensor output, the
The differences between the measurements and the simulation
(2 + 1)D body surface potential map (BSPM) has proven to
may be due to additional fringe effects and static charge. Both
be a useful source-modeling tool. The BSPM consists of high
simulations and the measurements show that the dual electrode
time-resolution recordings of ECG signals taken at several lo-
configuration has a greater common-mode voltage than the quad
cations on a human subject’s torso [11]. Employing this data in
electrode sensor, thus the quad sensor has improved immunity
simulations enables an investigation of subject to sensor relative
to common-mode signals, which is in agreement with the sim-
motion for various electrode sizes, configurations, and sensor
ulated result shown in Fig. 4.
placement. Table I lists the parameters of the capacitive sensor
modeled in the simulations.
B. Experiments on a Human Subject
For comparison we computed the average of the electronic
noise power in (2) and the motion-related distortion artifacts To demonstrate the proposed measurement schemes with a
induced by common mode distortion in (4) resulting from all human subject we placed gradiometer sensors on a subject’s
possible orientation vectors for the set of electrode sizes and right chest. In order to obtain a fair comparison of the electrode
configurations, Type A and type B. The former are four dual configurations the lateral spacing of the quad electrode was
electrodes, while the later are four quad electrodes. Both designs chosen to be d2 = 0.5 mm (minimum lateral spacing) while its
have electrodes sized b = 1.56, 3.12, 4.68, and 6.24 cm with d1 value was chosen to be the same as d of the dual electrode. The
equal areas (for b = a/2 and d = d1 = d2 = 0.5 mm) as shown adjustable lateral sizes were chosen as follows: d1 = d = 5, 15,
in Fig. 1. We assumed a fixed electrode to subject gap and a 21, and 28 mm. The outputs for both sensor configurations are
fixed electrode centroid location relative to the BSPM. Then computed simultaneously. Fig. 7(a) is the final output following
the output signal was computed for the time-dependent surface 0.5–100 Hz bandpass filtering. We note the following features
potential over a set of possible electrode orientations. Results in comparing the two gradiometer schemes.
of the simulations are shown in Fig. 4. The electronic noise 1) 60 Hz AC Power Mains: In order to compare the sen-
levels of the type A and type B sensors are similar; however, sitivity of both gradiometers to common mode ac power line
182 IEEE TRANSACTIONS ON BIOMEDICAL ENGINEERING, VOL. 60, NO. 1, JANUARY 2013

Fig. 8. ECG recordings displaying motion artifacts. Four values of the lateral
spacing of the electrodes in both the dual(in black) and quad (in gray) configu-
rations were measured in 5 s, d1 = d = 5, 15, 21, and 28 mm. The output was
bandpass filtered from 0.01 to 100 Hz.

the quad electrode has better rejection of this common-mode

signal.
2) Baseline Wandering: The effects on the low-frequency
motion signal induced by respiratory activity can be expressed
as the ratio of the QRS complex to the low-frequency signal as
shown in Fig. 7(c). Here, the bandwidth of the motion artifact
was selected to span 0.5–1.1 Hz. The result for this filter shows
that the signal-to-motion artifact ratio is better for the dual elec-
trode configuration at small electrode lateral spacings and the
quad electrode performs better for larger spacing. To observe
Fig. 6. Simulation and experiment for (a) dual electrode and (b) quad elec- the response of the dual and quad electrode designs to even
trode. (a1) and (b1) are simulations of the common mode output for both lower frequency events we employed a 0.01–100 Hz bandpass
electrodes tilted from 0◦ to 9◦ in both θx and θy . (a2) and (b2) are, respectively, filter to record the signal. The result illustrated in Fig. 8 also
the experimental measurements for the electrodes used to obtain the data shown
in Fig. 5(b). (a3) and (b3) are the differences of the experimental data and the indicates that the quad electrode demonstrated a greater reduc-
simulation for the dual and quad electrodes. The result shows that the quad tion of movement-induced artifacts in comparison to the dual
electrode has better immunity to common mode noise. electrode design, even if the signal-to-motion artifact ratio is not
improved. Therefore, the quad electrode design is less likely to
be prone to output saturation. Noting the large, short time scale
events in the first three traces of Fig. 8, we also see that the
quad electrode design provides a greater degree of attenuation.
This may be due to the fact that when both gradiometers were
tilted the dual electrode summed up the triboelectricity signal
sensed by two positive or negative subelectrodes, while the static
charge signal generated on the quad electrodes was effectively
subtracted out.
3) R-Peak Amplitude: As seen in Figs. 7(a) and 8, the R-peak
amplitude does not always become higher when the electrode
lateral spacing increases. As an example consider the dual elec-
trode signal shown in Fig. 7(a), R-peak (measured 0.8 V when
d1 = d = 15 mm) which is larger than the R-peak (measured
Fig. 7. (a) ECG measurements on a subject’s right chest for the dual (in black) 0.5 V when d1 = d = 21 mm.) This is influenced by the precise
and quad (in gray) electrodes with lateral spacing d1 = d = 5, 15, 21, and
28 mm for 5 s recordings. The unfiltered ECG signals all display common mode location of the electrodes relative to the BSPM.
60 Hz power line noise. (b)Ratio of the QRS complex signal (4–40 Hz) to the
60 Hz pickup (c) Ratio of the QRS complex signal power to low frequency
motion signal power(0.5 to 1.1 Hz) is shown. C. Experiments Compared With Standard ECG Setting
The reliability of heart beat monitoring was assessed by em-
ploying gradiometric electrodes of various geometries, as de-
scribed in Section IV-B, placed on a person’s right and left
interference, the notch filter was turned OFF during the mea- chest. The signals were recorded and compared to conventional
surement. In Fig. 7(a), it is seen that signals display significant Ag/AgCl gel electrodes with the standard lead II placement [13].
ac power noise. We computed the QRS complex signal to 60 Hz A validated QRS-detector [12] algorithm was also applied to
noise ratio shown in Fig. 7(b). The bandwidth of QRS complex each recorded signal to track the heart rate. Fig. 9 shows the
signal was chosen to span 4–40 Hz, which captures most of three traces (5 s duration) with beat annotations marked for
the power in the signal spectrum. The data demonstrates that d1 = d = 28 mm. Using the measured R-peak intervals, the
IEEE TRANSACTIONS ON BIOMEDICAL ENGINEERING, VOL. 60, NO. 1, JANUARY 2013 183

 
 bpmDUAL − bpmStandard 

ΔD =   × 100% (5)
bpmStandard 
 
 bpmQUAD − bpmStandard 
ΔQ =   × 100%.
 (6)
bpmStandard
The results indicate that the highest mismatch rate among all
trials is 0.82% which is very good agreement.

V. CONCLUSIONS
Although artifacts from the relative motion of noncontact
Fig. 9. ECG traces from gradiometer electrodes placed on the right chest ECG electrodes and the subject are problematic and have been
with d1 = d = 28 mm and standard ECG lead II contact electrode recordings addressed previously by a number of researchers, we have found
(5 s traces). All three traces were analyzed with an open source beat detection
algorithm [12] and the detected beats are annotated. that motion-related artifacts may be reduced considerably by
employing gradiometric measurement techniques without em-
ploying the DRL method. Although this may lead to a robust
and convenient heart rate monitor the relevance of short-base
TABLE II line ECG measurements in more sophisticated morphological
HEART RATE MEASUREMENT COMPARISON OF THE GRADIOMETRIC analysis of ECG signals remains to be investigated clinically.
ELECTRODE AND STANDARD LEAD II ECG Further exploration and design optimization clearly will be pos-
sible using the BSPM-electrode simulation tool.

REFERENCES
[1] Y. Chi, T.-P. Jung, and G. Cauwenberghs, “Dry-contact and noncontact
biopotential electrodes: Methodological review,” Biomed. Eng. IEEE Rev,
vol. 3, pp. 106–119, Dec. 2010. DOI: 10.1109/RBME.2010.2084078.
[2] J. Ottenbacher and S. Heuer, “Motion artefacts in capacitively cou-
pled ECG electrodes,” in Proc. World Congr. Med. Phys. Biomed. Eng.,
vol. 25/4, Munich, Germany, Sep. 2009, 2010, pp. 1059–1062.
[3] T. Wartzek, T. Lammersen, B. Eilebrecht, M. Walter, and S. Leonhardt,
“Triboelectricity in capacitive biopotential measurements,” IEEE Trans.
Biomed. Eng., vol. 58, no. 5, pp. 1268–1277, May. 2011.
[4] J. G. Webster, Medical Instrumentation, Application and Design. New
York: Wiley, 2010.
[5] O. Casas and R. Pallas-Areny, “Electrostatic interference in contactless
biopotential measurements,” in, Aug. 2007, pp. 2655–2658.
[6] L. Baxter, Capacitive sensors, design and applications. Piscataway, NJ:
IEEE Press, 1997.
[7] S. Heuer, D. Martinez, S. Fuhrhop, and J. Ottenbacher, “Motion arte-
fact correction for capacitive ecg measurement,” in Proc. IEEE Biomed.
Circuits.Syst. Conf., Nov. 2009, pp. 113–116.
[8] J. Yoo, L. Yan, S. Lee, H. Kim, and H.-J. Yoo, “A wearable ECG acqui-
sition system with compact planar-fashionable circuit board-based shirt,”
IEEE Trans. Inf. Technol. Biomed., vol. 13, no. 6, pp. 897–902, Nov. 2009.
DOI: 10.1109/TITB.2009.2033053.
[9] G. Peng and M. Bocko, “A low noise, non-contact capacitive cardiac
sensor,” in Proc. IEEE Annu. Int. Conf. Eng. Med. Biol. Soc. EMBC 2012,
Aug. 28–Sep. 1, 2012, to be publised.
average beats per minute (bpm) can be determined along with [10] J. Peyton and Z. Peebles, Probability, Random Variables, and Random
its standard deviation. Table II lists the results for 25 s recordings Signal Principles. New York: Mcgraw-Hill, 2001.
[11] R. Bond, “Xml-bspm: An xml format for storing body surface potential
for each electrode spacing d. To compare results, two mismatch map recordings,” BMC Med. Informat. Decis. Making, vol. 10, pp. 10–28,
parameters, the mismatch of bpm from the dual electrode and 2010.
the standard setting(ΔD) and the mismatch of bpm from the [12] P. Hamilton, “Open source ecg analysis,” in Proc.Comput. Cardiol., Sep.
2002, pp. 101–104.
quad electrode and the standard setting(ΔQ) were calculated [13] Nursecom Educational Technologies. (2003). “An ECG primer,” [Online].
(as percentages). Available: http://www.nursecom.com/ECGprimer.pdf