TOWARDS WEARABLE AND CONTINUOUS 12-LEAD

ELECTROCARDIOGRAM MONITORING

Synthesis of the 12-lead Electrocardiogram using 3 Wireless Single-lead Sensors

C. P. Figueiredo and P. M. Mendes

Centro Algoritmi, University of Minho, Dep. Industial Electronics, Campus Azurém, 4800- 058 Guimarães, Portugal

Keywords: Health Monitoring Devices, Biotelemetry, Wireless Systems, Wearable Systems, 12-lead ECG Synthesis,

12-lead ECG Reconstruction, Low-Power Design.

Abstract: Wearable health monitoring systems have emerged in the last decade as innovative means for patient

observation and healthcare delivery. Among the physiological signals which can be measured using such

systems, the 12-lead electrocardiogram is arguably the most important. However, continuous monitoring of

the standard 12 lead electrocardiogram is impractical and unattractive for a wearable system, due to the

obtrusiveness and discomfort that the placement and connection of 10 electrodes would cause. In this

regard, the use of reduced lead sets for the synthesis of the 12-lead electrocardiogram is a preferable

solution. This work analyses the suitability of a wireless sensor network prototype for continuous and

simultaneous monitoring of a set of 3 modified electrocardiogram leads, which can be used for synthesis of

the 12-lead electrocardiogram by application of a patient-specific transformation matrix, estimated by

multiple linear regression.

1 INTRODUCTION

The development of wearable systems for health

monitoring has, since the last decade, been an active

research topic with growing attention (Pantepoulos

and Bourbakis, 2010), benefiting from the recent

advances in sensor and communication technologies,

along with their increasing degree of miniaturization

due to microelectronic integration. These systems

have been shown to have the potential to

revolutionise healthcare systems, which are

currently under increasing pressure, due their rising

global costs, ageing of the population and associated

prevalence of chronic disease (Hao and Foster,

2008). In this sense, the promotion of a shift in

healthcare systems from reactive management of

illness towards proactive management of wellness is

desired (Milenković et al., 2006), supported by

innovative, affordable and efficient solutions for

personal health monitoring. Many other benefits are

envisioned through the widespread availability of

these solutions for both patients and caregivers. The

quality of life and mobility of monitored patients is

increased and the means for detection of early signs

of disease are improved by long-term continuous

monitoring (Nield et al., 2004). These allow for

patients to keep an independent lifestyle in their

environment of choice and for caregivers to provide

better treatment and prevent further complications.

Among the different physiological signals which

can be continuously measured with health

monitoring systems, in order to evaluate the health

condition of a person, the 12-lead electrocardiogram

(ECG) is arguably the most important, since cardiac

disease is the leading cause of death and disability in

the world. The 12-lead ECG is considered as the

gold standard for non-invasive detection of

abnormal cardiac rhythms and cardiac diseases such

as myocardial ischemia. However, continuous

monitoring of the full set of 12 leads of the standard

ECG, and the placement of the respective electrodes

at their standard locations (which include the distal

extremities of the limbs), is unattractive and

incompatible with the design requirements of a

wearable system for health monitoring, namely

unobtrusiveness, comfortable use and user-

friendliness. Therefore, most of the wearable health

monitoring systems proposed to date are only able to

monitor a single lead or a lead subset (e.g. 3-lead

ECG composed only of Einthoven leads) of the

standard 12-lead ECG (Pantepoulos and Bourbakis,

2010), and thus only a limited interpretation of the

329

P. Figueiredo C. and M. Mendes P..

TOWARDS WEARABLE AND CONTINUOUS 12-LEAD ELECTROCARDIOGRAM MONITORING - Synthesis of the 12-lead Electrocardiogram using 3

Wireless Single-lead Sensors.

DOI: 10.5220/0003771203290332

In Proceedings of the International Conference on Biomedical Electronics and Devices (BIODEVICES-2012), pages 329-332

ISBN: 978-989-8425-91-1

 2012 SCITEPRESS (Science and Technology Publications, Lda.)

electrical activity of the heart is obtained. This

limitation can be critical for the sensibility of the

ECG monitoring system on the detection of certain

cardiac abnormalities, e.g., ischemic episodes may

be only visible in part of the 12 standard ECG leads

and may be overlooked (Klootwijk et al., 1998),

even during continuous monitoring. Moreover,

caregivers could perform more efficiently if the

health monitoring system could provide them with

the ECG signal in the same fashion in which they

were trained to analyse, i.e., the standard 12-lead

ECG.

Due to the conflicting design requirements

concerning wearable and continuous monitoring of

the 12-lead ECG, solutions are needed to allow

monitoring the full set of 12 ECG leads, while

simultaneously minimising the obtrusiveness of the

monitoring system. On this subject, several authors

have proposed and studied the synthesis of the 12-

lead ECG using reduced lead sets, based on the

application a generic or patient-specific

transformation matrix, estimated through multiple

linear regression, to the lead signals of the reduced

lead set (Dower et al., 1988; Nelwan et al., 2004).

The synthesis of the 12-lead ECG using wireless

technology was proposed by Trobec and Tomašić

(2011), based on a set of 3 differential leads formed

by pair of proximal electrodes. However, the impact

caused by the use of wireless communications in the

performance of the synthesis algorithm was not

analysed. In this regard, this work aims to analyse

the suitability of a developed wireless network

prototype, consisting of 3 wearable single-lead ECG

sensors, for application in 12-lead ECG synthesis.

The quality of the acquired lead signals is evaluated,

as well as the issues caused by the wireless

transmission of these independent signals, such as

reliability (transmission errors leading to loss of

information) and synchronization between nodes.

2 SYSTEM OVERVIEW

The currently proposed system is illustrated in

Figure 1, and it has the aim of enabling continuous

monitoring of a synthesised 12-lead ECG. It can be

seen that the proposed solution is based on a

wireless network of three wearable single-lead ECG

sensors, placed at different locations and orientations

on the chest of the monitored patient. In addition, a

coordinating node is used to receive the signals

acquired by each of the 3 single-lead ECG sensors,

and to forward them via USB connection to a

personal computer where the 12-lead ECG is

synthesized and displayed. The use of these

Figure 1: Proposed wireless network of wearable sensors

for continuous monitoring of a synthesised 12-lead ECG.

wearable single-lead ECG sensors, combined with

the use of wireless technology and an algorithm for

12-lead ECG synthesis based on independent local

lead signals, has allowed the design of a wearable

12-lead ECG monitoring system featuring total

absence of wired connections, such as electrode lead

wires, and which also does not require the monitored

leads to be referenced to the same ground.

The synthesized 12-lead ECG is calculated on a

personal computer, from the independent single-lead

ECG signals obtained by the 3 sensors in the

network, using the following linear equation:

= M

+ M

node 1

+ M

node 2

+ M

node 3

(1)

where L

indicates the amplitude of synthesized lead

i (i = 1, …, 12 or {I, II, III, aVR, aVL, aVF, V1, …,

V6}), l

node j

indicates the amplitude of the lead signal

at node j (j = 1, 2, 3) and M

is the transformation

matrix filled with patient-specific coefficients,

estimated using multiple linear regression to fit the

model shown in (1) to a dataset of a previously

measured standard 12-lead ECG (dependent

variable) and the three single-lead ECG signals

(independent variables) (i = 1, …, 12 and j = 0, ...,

3).

Concerning the currently presented approach, it

can be noticed from Figure 1 that the wearable

single-lead ECG sensors are centred on V2 and that

their monitored local leads are oriented towards the

proximal extremities of the limbs. This configuration

was selected based on existing evidence that the 12-

lead ECG can be synthesized with a fair degree of

accuracy (cross correlation coefficients greater than

0.9) using a 3-lead subset composed of leads I, II

and V2 (Atoui et al., 2010; Nelwan et al., 2004),

even if the Mason-Likar modification of the standard

12-lead ECG is used for the reduction of motion

artefacts. The position of each single-lead ECG

sensor on the chest of the monitored patient can be

moved along the directions indicated by the dashed

lines shown in Figure 1, to increase the accuracy of

the synthesised ECG or to increase user comfort, as

long as these positions are kept constant between

BIODEVICES 2012 - International Conference on Biomedical Electronics and Devices

330

different monitoring sessions of each particular user.

3 WEARABLE SINGLE-LEAD

ECG SENSOR

Each single-lead ECG sensor within the wireless

network is similar and the appearance of their first

prototype (Figueiredo et al., 2010) is shown in

Figure 2.

Figure 2: Top and bottom views of the developed wireless

sensor node for monitoring of a single modified ECG lead.

The wearable sensor shown in Figure 2 consists

of a flexible printed circuit board containing built-in

flexible dry electrodes (Steltenkamp et al., 2009) on

its bottom layer, which is intended to contact the

surface of the chest of the monitored patient. These

electrodes do not require skin preparation before use

and are therefore suitable for a wearable system. The

top layer contains low-power electronics for signal

acquisition, processing and wireless communication,

making use of commercial off-the-shelf components.

The design of these sensors meets the requirements

for a wearable system for health monitoring such as:

small dimensions, light weight, wireless

communication and low power consumption.

The wireless communications between the

single-lead ECG sensors and the coordinating node

are based on proprietary wireless transceiver and

MAC Protocol (Omeni et al., 2008), which operates

at the non-crowded 862-870 MHz radio band, and

uses a master-slave architecture with collision

avoidance and a Time Division Multiple Access

(TDMA) scheme. These features not only provide

error resilient and low power wireless

communication, but also help to keep synchronism

between the data transmitted from the different

sensor nodes, since the network coordinating node

periodically requests data from each sensor node in

succession, and only one sensor node is transmitting

at a time.

Besides monitoring of a local ECG lead, the

wireless sensor nodes also include a triaxial

accelerometer and a thermistor for activity and

temperature monitoring. Table 1 summarizes the

specifications and features of the developed

wearable single-lead ECG sensors (Figueiredo et al.,

2010).

Table 1: Summary of the specifications and features of the

developed wearable single-lead ECG sensors.

Wireless link

Operation frequency 862-870 MHz

Data rate 50 kbps

ECG lead

monitoring

Gain 500

Resolution 11 bits

Sampling Frequency 250 samples/s

Triaxial

accelerometer

monitoring

Range ± 2g or ± 8g

Resolution 8 bits

Sampling Frequency 400 samples/s (max.)

Temperature

monitoring

Range 20 ºC to 40 ºC

Resolution 11 bits

Sampling Frequency 250 samples/s

Power

consumption

Supply voltage 3.3 V (minimal)

Average current

consumption

0.98 mA

(ECG mode)

1.43 mA (ECG +

accelerometer mode)

4 RESULTS

4.1 Measurement of Single-lead ECG

Signals

A five second period of a modified ECG lead,

obtained by the developed single-lead ECG sensor is

shown in Figure 3. The obtained signal has suitable

quality and signal to noise ratio for the intended

ECG monitoring application, allowing different

features of the ECG signal to be detected, such as

the QRS complex, T wave and the ST interval.

4.2 Wireless Link Performance

The reliability of the wireless communications

between the sensor nodes and the network

coordinating node was tested by evaluating the

TOWARDS WEARABLE AND CONTINUOUS 12-LEAD ELECTROCARDIOGRAM MONITORING - Synthesis of the

12-lead Electrocardiogram using 3 Wireless Single-lead Sensors

331

number of communication errors occurred over time.

Figure 3: ECG lead signal acquired by the developed

wearable sensor.

Table 2 shows a list of different statistics obtained

during a testing period of one hour, within which the

3 wireless sensor nodes were performing signal

acquisition and transmitted their monitored data

when requested by the network coordinating node,

following a TDMA communication scheme.

Table 2: Statistics of error performance in wireless

communications within the developed network, during a

testing period of one hour.

Connection losses 2

Packets lost due to link errors 60

Duplicate packets received 10

Total number of generated packets/TDMA slots 1383

Packet error rate 5.1%

The obtained packet rate of 5.1% is considerable,

although it could be acceptable for continuous long-

term monitoring applications, where loss of data is

less critical.

5 CONCLUSIONS

From a preliminary analysis of the developed

system, it is expected that the use of a wireless

network of single-lead ECG sensors, using a

protocol based on TDMA and operating in a less

crowed band, could provide reliable support for

continuous monitoring of independent single-lead

signals and for continuous 12-lead ECG synthesis.

Future work will involve testing of the accuracy

of the proposed system for 12-lead ECG synthesis in

volunteers and the development of a new

miniaturised version of the single-lead ECG sensors.

Moreover, tests should be performed to evaluate up

to what degree the data losses caused by errors in

wireless communications affect the accuracy and the

performance of the system for continuous

monitoring of a synthesized 12-lead ECG.

ACKNOWLEDGEMENTS

This work was supported by the Portuguese

Foundation for Science and Technology under grant

SFRH/BD/40341/2007.

REFERENCES

Atoui, H., Fayn, J., Rubel, P., 2010. A Novel Neural-

Network Model for Deriving Standard 12-Lead ECGs

From Serial Three-Lead ECGs: Application to Self-

Care. IEEE Trans. Inf. Technol. Biomed., 14(3),

883-890.

Dower, G. E., Yakush, A., Nazzal, S. B., Jutzy, R. V.,

Ruiz, C. E., 1988. Deriving the 12-lead

electrocardiogram from four (EASI) electrodes. J.

Electrocardiol., 21, 182-187.

Figueiredo, C. P., Becher, K., Hoffmann, K.-P., Mendes,

P. M., 2010. Low Power Wireless Acquisition Module

for Wearable Health Monitoring Systems. Proc. 32nd

Ann. Int. IEEE EMBS Conf., 704-707.

Hao, Y., Foster, R., 2008. Wireless body sensor networks

for health-monitoring applications. Physiol. Meas., 29,

R27-R56.

Klootwijk, P. et al., 1998. Comparison of usefulness of

computer assisted continuous 48-h 3-lead with 12-lead

ECG ischaemia monitoring for detection and

quantitation of ischaemia in patients with unstable

angina. Eur. Heart J., 18(6), 931-940.

Milenković, A., Otto, C., Jovanov, E., 2006. Wireless

sensor networks for personal health monitoring: Issues

and an implementation. Comp. Commun., 29,

2521-2533.

Nelwan, S. P., Kors, J. A., Meij, S. H., Bemmel, J. H. v.,

Simoons, M.L., 2004. Reconstruction of the 12-lead

ECG from Reduced Lead Sets. J.Electrocardiol., 37,

11-18.

Nield, I., Heatley, D. T. J., Kalawsky, R. S., Bowmann, P.

A., 2004. Sensors Networks for Continuous Health

Monitoring. BT Technol. J., 22(3), 130-139.

Omeni, O. C., Eljamaly, O., Burdett, A. J., 2007. Energy

Efficient Medium Access Protocol for Wireless

Medical Body Area Sensor Networks. Proc. 4th IEEE-

EMBS Int. Summer School & Symp. Medical Devices

and Biosensors, 29-32.

Pantepoulos, A., Bourbakis, N. G., 2010. A Survey on

Wearable Sensor-Based Systems for Health

Monitoring and Prognosis. IEEE Trans. Syst., Man,

Cybern. C, Appl. Rev., 40(1), 1-12.

Steltenkamp, S., Becher, K., Doerge, T., Ruff, R.,

Hoffmann, K.-P., 2009. Electrode structures for

acquisition and neural stimulation controlling the

cardiovascular system. Proc. 31st Ann. Int. IEEE

EMBS Conf., 5478-5481.

Trobec, R., Tomašić, I., 2011. Synthesis of the 12-Lead

Electrocardiogram From Differential Leads. IEEE

Trans. Inf. Technol. Biomed., 15(4), 615-621.

BIODEVICES 2012 - International Conference on Biomedical Electronics and Devices

332