Telemetric Monitoring of Foetal ECG
Telemetric Monitoring of Foetal ECG
ABSTRACT
In this paper we propose an implantable telemetric system that meets the clinical needs for
accurate foetal ECG monitoring during labour for all pregnancies where monitoring is
indicated, as well as during pregnancy for a selected number of foetuses undergoing foetal
surgery.
INTRODUCTION
As a consequence of recent developments, one can state that the foetus has become a patient,
accessible for diagnosis and therapy. Foetuses may become critically ill during their in utero
life, and may benefit from timely delivery or foetal therapy. Until now, continuous in utero
foetal monitoring has not been possible, since it requires permanent and direct access to the
foetus. Therefore, information from the unborn patient can only be retrieved by indirect and
often inaccurate means (ultrasound, doppler and cardiotocography). If these findings are
suggestive of foetal distress, a foetal blood sample must be taken for confirmation. Foetal
blood sampling is an invasive procedure with its associated risks and offers only momentary
information on the foetal status. The availability of a device that allows for continuous and
accurate monitoring of foetal well-being will reduce the inappropriate preterm delivery or
therapy of foetus, with good foetal well-being and avoid foetal death or brain damage, where
foetal distress has gone unrecognised.
trocart
METHODS
FIRST EXPERIMENTS
We started this research by investigating the possibility of measuring a foetal ECG by
telemetric means. Therefore we used a commercial available sensor
EA-F20 (DSI, Minnesota, figure 2). This 2 lead miniature sensor measures ECG on a
continuous base when it is magnetically activated.
The signal of the sensor is transmitted and subsequently captured via the maternal abdomen of
the ewe by an antenna. This antenna also provides the necessary electronics to convert the RF
signal into an analog ECG signal. This signal is then further digitalised using a Keithley
Analog-to-Digital PCI/MCIA card. The received digital information is extracted from this
card and stored into the PC using a program developed with Testpoint.
Medical research has shown that the use of ECG to obtain a beat-to-beat analysis (Nijhuis, J.,
et al) can, in combination with ST-analysis, provide the gynaecologist enough information to
detect and prevent foetal stress (Sundström, A., et al). Beat-to-beat analysis consists of the
calculation of the R-R interval of the ECG, whereas ST-analysis consists of determining the
T/QRS ratio.
R RR-interval R
contraction of
the atria QRS-amplitude
T-amplitude
T
P
Q Q
‘ S ST-interval S
QRS-complex
contraction changes during
heart chambers hypoxia
To perform the medical analysis, a program in Visual Basic able to extract automatically
beat-to-beat and ST analysis, has been written. Figure 4 shows a typical screenshot generated
by this program. The upper trace shows the raw received ECG signal. The extraction of the
QRS-complex is calculated by an algorithm based on a level decision criterion. The result of
this algorithm can be found on the second trace. The third trace is a conversion of the second
one into a beat/minute scale. This trace can be used for a medical beat-to-beat evaluation.
Mean, maximum and minimum heart rate is also automatically calculated. The next trace
shows the variation in heart rate and is also useful for medical interpretation. The ST analysis
can be found on the last trace. With the display of these traces the gynaecologist has sufficient
information to decide on the necessary therapy.
‘
Figure 4: Analysis program
The telemetric device was implanted subcutaneously in a foetal lamb at day 119 of
gestation (term = 147 days). (Figure 5)
After the implantation of the sensor on the foetal lamb, telemetric data was obtained until day
11. On the last day of recording, an elevation in the ST-segment and bradycardia were
present, suggesting foetal hypoxia. On day 12 no foetal ECG signal could be obtained and
indeed the lamb was stillborn 2 days later. Figure 6 shows traces of the first and last day of
recording which illustrates clearly the bradycardia and ST changes as described.
1 2 3
1 : Sensor
2 : Intelligent Core
3 : Bi-directional telemetric unit
2: Intelligent core
Instead of transmitting the entire signal, the recorded parameter will be analysed to eliminate
superfluous data. The system will contain algorithms that can automatically detect foetal
health condition. These algorithms will be based on the already described beat-to-beat and ST
analysis. The ultimate goal is to have a system that contains the intelligence to detect any
abnormality and transmit the necessary information.
Prototype
A first prototype has been assembled. The 2 lead ECG amplifier has been developed using
bootstrapped amplifiers to reduce 50 Hz interference, followed by low and high analog filters
to eliminate signals that are no part of the ECG bandwidth. This creates an operating
frequency range between 0.5 and 150 Hz. The ECG is amplified in such a way that it is
adapted to the full range of the A/D converter. The board also provides a bandpass filter
between 5 and 15 Hz that will extract the QRS complex of the ECG signal (Jiapu P., et al).
A Flash PIC microcontroller PIC12F629 is used. It contains a 10 bit A/D converter. This
microcontroller is programmed to record a sample at a sampling frequency of 300Hz. The
program also provides what is necessary to treat the samples in a certain way that they can be
sent to a RS-232 converter. By doing so, data can be directly recorded on a PC. The board
also has a transceiver TR3000 (RFM) with the antenna integrated on the PCB. Through this,
ECG can also be transmitted wirelessly. The PCB is powered with +3V and –3V obtained
through silver oxide button cells.
Figure 8 gives the schematic overview of the board. Figure 9 shows a picture of the developed
circuit.
µcontroller
Measurements
For recording the ECG on a PC, either wireless or direct, a program in Visual Basic has been
written. This program can read from the PC’s serial port, visualise and store the data. Figure
10 shows a screenplot of the program.
The board was tested on a rat. A typical recording can be seen on figure 10.
The obtained signal has an excellent signal to noise ratio and shows no signs of common
mode interference. Data can be transmitted error-free at a maximum distance of 2 metres.
FUTURE WORK
DISCUSSION
At this moment the microcontroller’s program is under permanent improvement. A peak
detection algorithm must be implemented in order to detect the QRS complexes. When these
complexes are recognised, the time interval between 2 successive beats and the different
amplitudes of the ECG can be found. This must allow the system to acquire the necessary
intelligence to carry out the described analysis in situ and in real time. If this goal is achieved
a miniaturised system can be developed.
CONCLUSION
In this paper we have demonstrated that foetus can be treated with use of its ECG signal. We
have also shown that ECG can be telemetrically received from the foetus.
We have set the first steps to a novel telemetric device with integrated analysis. The
experiences gained from the first prototype will be used for the further development of the
miniaturised device based on CMOS electronics.
ACKNOWLEDGEMENTS
The authors which to acknowledge the support for this program “Verkennende Internationale
Samenwerking” (stimulation of research projects within the fifth framework program),
sponsored by the Flemish Regional Government; VIS/00/009, telemetric and non-telemetric
microsensors for perinatalogy.
LITERATURE CITED
DSI. Data Sciences International. www.transomamedical.com. Minnesota
Nijhuis, J., et al. Foetale bewaking. Elsevier/Bunge. 1998
Sundström A., et al.’ Foetale bewaking met STAN’. Neoventa Medical. 2000
Nitish V. Thakor and John G. Webster. ‘Ground-free ECG Recording with Two Electrodes’.
Trans. Biomed. Eng. Vol 12:pp 699-704. IEEE 1980.
Jiapu P. and Willis J. Tompkins.. ‘A Real-Time QRS Detection Algorithm’. Biomed. Eng. 3:
Vol pp 230-236. IEEE 1985 Trans.