BIOELECTRIC ACTIVITY RECORDING BASED ON A SINGLE

ELECTRODE FOR USE ON WEARABLE DEVICES

M. S. Fernandes, C. M. Pereira, J. H. Correia and P. M. Mendes

Algotimi Center, University of Minho, Campus de Gualtar, 4710-057 Braga, Portugal

Keywords: Bioelectric signals, Wearable systems, Biotelemetry, Contactless measurement.

Abstract: Wearable devices are used to unobtrusively record several physiological signals. Bioelectric signals are one

of the most important variables monitored. Despite the available techniques, including capacitive coupling,

it is still lacking a contactless solution that can be integrated into wearable devices. We propose a new

approach where an instrumentation amplifier is directly driven by a bioelectric signal. In this way, the

voltage drop on the capacitive electrodes is avoided. In this paper we show the proof of concept, and results

are presented to show how to record an Electrocardiogram (ECG) using this new approach. Measurements

were made using a high-impedance instrumentation amplifier. Results have shown that our approach is

viable for bioelectric signal detection using contactless methods.

1 INTRODUCTION

Advances in electronic, textile and information

technologies have contributed to the design of

wearable monitoring systems, aimed to provide

continuous, unobtrusive and remote monitoring of

physiological signals. An important contribute is

made by novel technologies for detecting bioelectric

signals. Not every sensor can be used in a wearable

context and a set of attributes must be taken into

account. These include physical attributes such as

size and weight, as well as easy placement and an

unobtrusive aspect. In addition, wearable sensors

must ideally produce an electrical output in order to

be digitally processed. Properties such as durability,

reliability and low power consumption are also

demanded (Constantine and Fotiadis, 2005)

(Winters and Wang, 2003). Acquisition devices are

based on contact or contactless measurements.

Focusing on the first type, two main options arise:

dry or wet (require gel) electrodes. There are some

semi-invasive solutions available, where the

electrodes are based on micro spikes that go trough

the skin (Ng et al., 2009). On the other hand, with

contactless measurements, the available solutions

consist in the use of a capacitive or inductive

coupling, and also the use of electro-active materials

(e.g. electro-optic material). However, since they use

electrodes to drive the signal, a significant potential

drop occurs, causing difficulties in the detection of

smaller signals such as brain electric activity.

The focus of this work consists in proposing an

approach towards contactless detection of bioelectric

signals. In this paper we will demonstrate the

concept of the non-contact acquisition of bioelectric

signals, using the bioelectric field to directly drive

an instrumentation amplifier, instead of the

conventional use of big capacitive electrodes. This

will significantly benefit the design of contactless

bioelectric sensors particularly for wearable

applications. The experimental setup used to the

proof of concept of our approach will be described

as well as the results obtained.

2 CONTACTLESS

MEASUREMENT

The need for physical contact has for long been a

problem when envisioning a wearable monitoring

application. Ideally, a biopotential recording system

should draw no real charge current from the body,

allowing to perform non-contact measurements of

biopotentials. However, the present solutions imply

a voltage drop across the electrode, either wet, dry or

capacitive, limiting the sensitivity and ability to

provide efficient contactless measurements. When

looking into the wearable matter, the standard

130

Fernandes M., Pereira C., Correia J. and Mendes P..

BIOELECTRIC ACTIVITY RECORDING BASED ON A SINGLE ELECTRODE FOR USE ON WEARABLE DEVICES.

DOI: 10.5220/0003129101300134

In Proceedings of the International Conference on Biomedical Electronics and Devices (BIODEVICES-2011), pages 130-134

ISBN: 978-989-8425-37-9

 2011 SCITEPRESS (Science and Technology Publications, Lda.)

solutions show two main problems: they are difficult

to integrate with the vests and uncomfortable to

wear. A new generation of sensors is required in

order to surpass the limitations of the present

solutions.

We propose a new approach for contactless

detection of biopotentials that avoids the voltage

drop by directly driving an instrumentation amplifier

with a bioelectric signal. The input impedance of the

instrumentation amplifier should be as high as

possible in order to allow the remote detection of

biopotentials. Moreover, input technologies based on

Complementary Metal–Oxide–Semiconductor

(CMOS) and Field-Effect Transistors (FET) should

be used since they are driven by voltage instead of

current. In our approach there’s only one contact

with the subject skin, responsible for establishing a

reference signal – reference electrode. In this way,

bioelectric signals will be simultaneously measured

without any contact at the recording locations,

sharing the same reference.

We are particularly envisioning the application of

this approach in brain biopotential wearable systems.

The existent solutions are not suitable for use in

fully wearable devices. With our approach, we hope

to contribute to the design of appropriate wearable

sensors for further applications in relevant areas

such as Brain-Computer Interface (BCI). Figure 1

shows an example of application of our contactless

approach in Electroencephalogram (EEG) wearable

recordings, where the sensor would be placed close

to the scalp. One of the inputs is connected to the

contact reference electrode. The other is floating in

order to drive the instrumentation amplifier with the

bioelectric signal.

Figure 1: Contactless sensor for biopotential acquisition.

Dashed lines are used to indicate that the sensor may be

placed in several recording locations.

The standard setups use the ear lobes as a

reference signal, since there’s minimal influence of

temporal lobe or muscle electrical activity.

However, from a user perspective, it would be

preferable to place the reference in a more discreet

place, since the main objective of wearable systems

consists in providing ubiquitous monitoring in daily

life activities. Therefore, we propose to place the

reference electrode on the back neck, establishing

the only contact with the skin. The recording sensors

are placed near the target locations, without any kind

of contact with the scalp.

3 BIOELECTRIC SIGNALS

3.1 Signal Characterization

Bioelectric signals are recorded as voltages,

potentials and electric field strengths, at very low

levels, with high source impedances and overlaid

interference signals and noise. They can be

classified according to the type of cells and tissue

they are originated from. Table 1 lists the most

relevant biopotentials along with its properties

(Webster, 1988).

Table 1: Properties of Biopotentials (ECG –

Electrocardiogram; EOG – Electrooculogram; EMG –

Electromyogram).

Biopotential

Tissue Amplitude

Frequency

ECG Heart 1-5 mV 0.05-100 Hz

EEG Brain

10-200 μV

0.5-40 Hz

EOG Retina 0.01-0.1 mV dc-10 Hz

EMG Muscles 1-10 mV 20-200 Hz

It’s important to note that due to several

constrains resultant from travel and propagation (e.g.

tissue resistivity), detected properties such as

amplitude and shape will be very distinct from those

detected inside the specific tissue that originates the

bioelectric event.

3.2 Standard Readout

The three essential components required to measure

a bioelectric signal are: bioelectrodes,

instrumentation amplifiers and filtering components

(Neuman, 1998). On the process of sensing a

biopotential, it is required to provide some interface

between the body and the measuring device. This

interface is carried out by bioelectrodes that convert

the ionic current within the body into electronic

current in metal connecting leads (Neuman, 1998).

Bioelectrodes should have low impedance.

Otherwise the currents driving the subsequent

amplifier will lead to a biopotential drop, leading to

BIOELECTRIC ACTIVITY RECORDING BASED ON A SINGLE ELECTRODE FOR USE ON WEARABLE

DEVICES

131

more difficult readouts. Three types of interface

between the electrode and the skin can be applied:

wet, dry/insulated and capacitive coupled. The first

one makes use of an electrolytic gel that helps to

promote the reduction of the contact impedance,

minimizing the risk of signal loss. This carries time-

consuming and complex procedures. The most

commonly used wet bioelectrode is the gel type

silver/silver chloride (Ag/AgCl), which can be found

both in reusable or disposable form. Dry and

insulated electrodes eliminate the need for an

electrolytic paste. The first type consist of a

biocompatible metal in direct contact with the skin,

being the coupling between them made by the user’s

sweat produced after it’s placement. On the other

hand, insulated electrodes are based on a dielectric

surface layer between the metal or semiconductor

and the skin. In this case, the bioelectric signal is

capacitively coupled between the skin and electrode,

without requiring electrical contact with the skin.

Some examples of dry/insulated electrodes and their

application can be found in (Baek et al., 2008; Ryu

et al., 2005). The third type of interface requires no

physical contact with the skin and it’s based on

capacitive pick-up electrodes. Basically, the

biopotential is obtained by capacitive coupling

between the body and the electrode, working both as

plates of a capacitor (Harland et al., 2002).

Bioelectric signals need to be amplified in order

to make them compatible with a variety of devices

such as A/D converters or display equipments. The

instrumentation amplifier is commonly used to

record biopotentials since it fulfils the basic

requirements for biopotential amplifiers, being

designed to have extremely large input impedance

and a small bias current. It works as a differential

amplifier, by applying high gain amplification

between signals at the positive and negative inputs.

Since the input signal of the amplifier consists of the

desired bioelectric signal and unwanted components

(e.g. power line interference signals, noise, etc.), it is

crucial to include a filtering stage. Generally, a

notch filter centered at 50 Hz (60 Hz in USA), and a

bandpass filter are used to remove these unwanted

signal components, that sometimes have higher

amplitudes than the desired bioelectric signal.

4 MEASUREMENTS

Measurements were carried in order to demonstrate

the concept of biopotential contactless recording.

The proof of concept consists into two stages: the

first experiment uses a conventional instrumentation

amplifier with subsequent filtering and amplification

stages; then, the resultant filtered and amplified

ECG is directly and contactless driven into a FET-

input instrumentation amplifier.

4.1 Experimental Setup

The modules used to validate our approach for

contactless detection of bioelectric signals include:

instrumentation amplifiers, notch filters, band-pass

filters and voltage amplifiers. The type of amplifiers

used, instrumentation amplifiers, need to fulfil a

particular set of requirements in order to provide

selective amplification to the biopotential, rejecting

the superimposed noise and interference

components:

- Have high input impedance (at least 10 MΩ)

and electrical isolation in order to inhibit

interference or distortion of the recorded signals.

- High CMRR (>80 dB according to (Neuman,

1998)) in order to separate as much as possible the

relevant signal from noise and interferences;

- Supply enough gain within its bandwidth in

order to reach an output level compatible with the

remaining system.

- Have low output impedance and supply the

amount of current necessary to the load.

- Provide protection to the patient from any

hazard of electrical shock.

Figure 2 shows the first module used in the

carried experiments.

Figure 2: First module comprising a conventional

acquisition circuit for biopotentials. LA corresponds to

Left Arm, and RA to right arm.

At this stage, an electrode was placed on each

arm, according to Lead I of Einthoven’s triangle,

resulting in a differential recording. The electrodes

were connected to both inputs of a precision

instrumentation amplifier (INA129, Texas

Instruments) with an input capacitance of 2 pF

(10

Ω) and a CMRR of 125 dB. The gain of the

amplifier was set to 155, by placing an RG of 320 Ω.

Since electrical circuits are usually interfered by ac

power lines, a notch filter was used to remove this

50 Hz interference, with an attenuation of 33.9 dB.

BIODEVICES 2011 - International Conference on Biomedical Electronics and Devices

132

To remove other unwanted signal components such

as other bioelectric signals or movement artefacts,

we designed a bandpass filter with unity gain and a

bandwidth set according to the frequency

components of interest (see Table 1). After the

filtering components, the ECG signal drives an

amplification stage with a gain of 100. The total gain

of this module is 255.

The second stage implements the contactless

module and consists of directly driving a FET-input

instrumentation amplifier, by placing an isolated

wire loaded with the resultant bioelectric signal from

the first stage. Figure 3 depicts the contactless

module of the experiment.

Figure 3: Contactless module comprising a conventional

acquisition circuit for biopotentials (x = No connection).

The ECG signal is used to directly drive the INA116 with

an isolated probe.

The bioelectric signal was directly and

contactless coupled by placing an isolated wire 1 cm

above a FET-input instrumentation amplifier

(INA116, Texas Instruments) with an input

impedance of 0.2 pF (10

Ω) and a CMRR of 94 dB.

The signal was further amplified with a gain of 20.

4.2 Results

Experiments were carried out in order to test the

recording of bioelectric signals with no contact with

the skin, neither electrical nor mechanical. Fig. 4

shows the bioelectric signals recorded at the end of

the first stage.

A first view of Fig. 4 allows to point some of ECG

components, including the P wave, QRS complex

Figure 4: ECG signal obtained at the end of the first circuit

module. The average amplitude of the signal is 3 V.

and the T-wave. Each one corresponds to a specific

electrical event that occurs during heart activity, for

instance the QRS complex occurs as the ventricles

depolarize. The average signal amplitudes reached a

value of 3 V, after a 255 total gain.

This signal is further used for contactless driving

the INA116, using an isolated probe placed at

different distances. To test the distance influence, we

varied the distance between the probe and the

instrumentation amplifier. Initially, the probe was

placed in external contact with the INA116, and then

went up to a 10cm distance. Figure 5 shows the ECG

signal strength obtained for the different distances

used.

Figure 5: QRS peak amplitude versus distance between the

instrumentation amplifier and the probe. The values

represent peak-to-peak amplitudes. The inset represents

the signal which is directly coupled from the probe to the

INA116.

These results demonstrate a decrease in signal

strength, as the distance between the INA116 and

the probe increases. Above 10 cm, the signal

becomes smaller than the noise components, causing

difficulties to isolate the relevant bioelectric signal

from the noise and interference. In figure 6 we show

an ECG recorded at a distance of 1cm from the

INA116.

Figure 6: ECG signal recorded at a distance of 1cm from

the INA116. An isolated probe was used to carry the

signal from the first module. The average amplitude is 600

mV.

BIOELECTRIC ACTIVITY RECORDING BASED ON A SINGLE ELECTRODE FOR USE ON WEARABLE

DEVICES

133

As shown in fig. 6, we can easily identify the

QRS complex, and part of the T-wave. The periodic

pattern displayed is similar to the conventional ECG

detected in the first module (Fig. 4). In terms of

amplitudes, the signal reaches a maximum of 600

mV. When envisioning wearable devices, ensuring a

good sensor performance at a distance of 1 cm is

desirable.

5 CONCLUSIONS

The developments described here open a new

approach to non-contact recording of bioelectric

signals, with promising applications in wearable

systems. A method for testing non-contact

acquisition of biopotentials by directly driving an

instrumentation amplifier with a previously

amplified and filtered ECG signal was proposed.

Measurements were made using a high-impedance

FET-input amplifier (INA116) and an isolated

probe, varying the distances between them. We

tested the performance of this approach in these

conditions, and results have shown the possibility of

successfully contactless acquire a readable ECG

until 10 cm of distance from the source, suggesting

that these approach could be used in bioelectric

wearable sensors.

Future work is needed towards the improvement

of sensitivity and noise reduction. This can be

achieved by increasing the input impedance and the

CMRR. The first one will result in smaller

attenuation of the electrophysiological signal. A

higher CMRR improves a better separation of the

relevant signal from noise and interferences. The

gain of the system can be easily set to higher values

by changing the feedback resistor of the amplifier.

ACKNOWLEDGEMENTS

We would like to acknowledge the Center

Algoritmi, the Portuguese Foundation for Science

and Technology (Grant SFRH/BD/42705/2007) and

the MIT Portugal Program, for supporting this work.

REFERENCES

Baek, J. Y., An, J. H., Choi, J. M., Park, K. S., and Lee, S.

H. (2008). Flexible polymeric dry electrodes for the

long-term monitoring of ECG. Sensors and Actuators

A, 143 (2), 423-429.

Constantine, G. and Fotiadis, D. I. (2005). Wearable

Devices in Healthcare. In Silverman, B. G., Jain, A.,

Ichalkaranje, A. and Jain, L. C. (Eds.), Intelligent

Paradigms for Healthcare Enterprises (cap.8, pp. 237-

264). Berlin, Germany: Springer Berlin.

Harland, C. J., Clark, T. D. and Prance, R. J. (2002).

Electric potential probes—new directions in the

remote sensing of the human body. Measurement

Science and Technology, 13 (2), 163-169.

Neuman, M. R. (1998). Biopotential Electrodes. In

Webster, J. G. (Ed.), Medical Instrumentation:

application and design, (cap. 5, pp. 183-232). New

York, USA: John Wiley & Sons.

Ng, W. C., Seet, H. L., Lee, K. S., Ning, N., Tai, W. X.,

Sutedja, M., Fuh, J. Y. H. and Li, X. P. (2009). Micro-

spike EEG electrode and the vacuum-casting

technology for mass production. Journal of Materials

Processing Technology, 209 (9), 4434-4438.

Ryu, C. Y., Nam, S. H. and Kim, S. (2005). Conductive

rubber electrode for wearable health monitoring.

Conference Proceedings - IEEE Engineering in

Medicine and Biology Society, 4, 3479-81.

Webster, J. G. (1988). Encyclopedia of Medical Devices

and Instrumentation (2

ed.), New York, USA : Jonh

Wiley & Sons.

Winters, J. M. and Wang, Y. (2003). Wearable sensors

and telerehabilitation. IEEE Engineering in Medicine

and Biology Society, 22 (3), 56-65.

BIODEVICES 2011 - International Conference on Biomedical Electronics and Devices

134