Microcontroller-based Underwater Acoustic ECG ... - Semantic Scholar

150

IEEE TRANSACTIONS ON INFORMATION TECHNOLOGY IN BIOMEDICINE, VOL. 1, NO. 2, JUNE 1997

Communications Microcontroller-Based Underwater Acoustic ECG Telemetry System Robert Shukri Habib Istepanian and Bryan Woodward

Abstract— This paper presents a microcontroller-based underwater acoustic telemetry system for digital transmission of the electrocardiogram (ECG). The system is designed for the real time, throughwater transmission of data representing any parameter, and it was used initially for transmitting in multiplexed format the heart rate, breathing rate and depth of a diver using self-contained underwater breathing apparatus (SCUBA). Here, it is used to monitor cardiovascular reflexes during diving and swimming. The programmable capability of the system provides an effective solution to the problem of transmitting data in the presence of multipath interference. An important feature of the paper is a comparative performance analysis of two encoding methods, Pulse Code Modulation (PCM) and Pulse Position Modulation (PPM). Index Terms— Biomedical telemetry, electrocardiogram, pulse code modulation, pulse position modulation, SCUBA diving, underwater acoustic communications.

I. INTRODUCTION Research has been carried out over many years on the development of techniques for the radio frequency (RF) telemetry of biomedical data. These techniques have also been applied for studies of the ecology, behavior and physiology of marine life [1]. Although acoustic telemetry is considered the most practical method for underwater physiological monitoring, few studies on this topic have been reported [2]–[7], although there are some potentially important applications such as deep diver monitoring and stress-related responses. Consequently, the evolution of RF biotelemetry systems has not been parallelled with similar advances in underwater acoustic biotelemetry, especially for human subjects. The main technical problem is to overcome the multipath interference that is generally present to some extent in any underwater channel; this may cause severe distortion of the received signals due to phase and amplitude fluctuations. In the research described here, the aim is to transmit digitally the electrocardiogram (ECG), then subsequently to receive it in a clinically acceptable form. The multipath problem, when compounded with the complex characteristics of the ECG signal itself, have limited the wider use of through-water acoustic monitoring of this type in the past. So far as is known, no study to date has addressed real time, through-water acoustic telemetry of the ECG using programmable concepts, especially in multipath environments such as lakes and swimming pools. This paper presents a performance analysis of a microcontrollerbased underwater acoustic telemetry system capable of real time Manuscript received February 16, 1997; revised August 4, 1997. R. S. H. Istepanian is with the Department of Electrical and Electronic Engineering, University of Portsmouth, Portsmouth PO1 3DJ, U.K. (e-mail: [email protected]). B. Woodward is with the Department of Electronic and Electrical Engineering, Loughborough University, Loughborough, Leics. LE11 3TU, U.K. (e-mail: [email protected]). Publisher Item Identifier S 1089-7771(97)08010-2.

cardiovascular monitoring during SCUBA diving and swimming. The use of programmed intelligence and serial bit transmission allows the acquisition, encoding and telemetry of continuous data, for example an analogue signal such as the ECG, as well as intermittent data such as pulses representing heart rate, breathing rate, and depth. The system can use either of two data encoding schemes, pulse code modulation (PCM) or pulse position modulation (PPM), and employs on-off keying (OOK) burst-mode transmission. The ECG data acquired during diving trials were validated for both digital encoding schemes.

II. MULTICHANNEL UNDERWATER ACOUSTIC TELEMETRY SYSTEM The system shown in Fig. 1 was developed for physiological monitoring of a diver using digital acoustic underwater telemetry techniques. Both the transmitter and receiver contain identical 8-bit single-chip Intel 87C51 microcontrollers; see Fig. 2. While this paper is mainly concerned with the telemetry of the ECG only, the wider application of telemetering several sets of data will first be presented. The transmitter microcontroller controls the digital encoding and sequencing of signals from three sensors (more in principle) by assigning different priority interrupt levels. It also controls the acoustic transmission of these signals by multiplexed on-off keying a separate carrier frequency for each biomedical or physical parameter, such as heart rate, ECG, breathing rate or depth. An identical priorityinterrupt hierarchy is implemented in the receiver microcontroller to achieve maximum timing precision and minimum data errors. All the frequencies are crystal-controlled and may be interchanged by reprogramming the two microcontrollers. Full details of the system design are discussed elsewhere [6]. For the purpose of this paper, attention is focussed on the telemetry of the ECG because of the particular problems associated with transmitting a complex analogue signal; no attempt has been made to satisfy industrial standards for ECG transmission. The modular programmable configuration of the system meets the two main design criteria for underwater acoustic telemetry systems [7]: 1) reliable detection, which is dependent on range, carrier frequency and acoustic power, as formulated from sonar equations; 2) a multipath-immune telemetry link that allows detection of wanted signals (direct path) in the presence of multipath reverberations. The complete transmitter is encapsulated in a pressure-proof housing carried by the diver inside a dry-suit. Different physiological signals are acquired from appropriate sensors; in the case of the ECG, waterproofed skin electrodes are attached to the diver’s chest and the associated connectors are passed through the housing. The positions of the electrodes are selected by trial and error until the largest amplitude QRS complex is found [8]. The amplified and filtered signal is then sampled at 200 Hz using a fast 8-bit analog-to-digital converter interfaced to the transmitting processor. The choices of sampling frequency and quantization level are based on earlier studies [9], [10], although the effect of varying these parameters has also been studied. Software controls the QRS detection and feature extraction process, followed by PCM or PPM data encoding applied to the

1089–7771/97$10.00  1997 IEEE


151

Fig. 1. Multichannel underwater acoustic telemetry system.

sampled signals. The coded signals are then transmitted by OOK transmission using a class-E power amplifier to drive the acoustic projector. The main function of the receiver is to scan the acoustic frequencies picked up by the hydrophone and to provide the logical input to the receiver processor, which is synchronized with the interrupt telemetry protocol mentioned above and provides the timing control and digital pulse decoding for each individual parameter. The receiver is equipped with a digital-to-analog converter, automatic gain control and low pass filter for reconstuction of the ECG signal, and is interfaced to a portable multichannel data logger (Grant/Elteck, Model 1200, Cambridge, U.K.) via an RS-232 to a portable laptop computer.

III. UNDERWATER ECG DETECTION AND TELEMETRY The software development included signal encoding/decoding, communication protocols and memory handling, and programming was carried out using Assembly code, then tested using an InCircuit Emulator (ICM) for the Intel 87C51 microcontrollers. Fig. 3 illustrates the QRS detection and the digital encoding process and Fig. 4 shows the flow chart for the real time QRS detection algorithm. While illustrated here for PCM, the process is equally applicable to PPM. Figs 5 and 6 show, respectively, the ECG transmitter and receiver programs that allow the encoding, transmission, reception and decoding of the ECG. An example of the reconstruction process performed by the receiver is shown in Fig. 7. Fig. 8 shows the timing diagram for digital PPM encoding. Ten PQRST samples are selected for encoding and transmission; this number of samples was found by experiment to be adequate to reconstruct clinically acceptable ECG signals at the receiver, as judged by a cardiologist and against “true” tape-recorded ECG signals. In the PPM method, each PQRST sample frame is divided into 2N possible time slots. The transmit pulse (pulse ) occupies just one of these time slots, whose width slot is fractionally more than pulse ; its position indicating the value of the particular sample.

Fig. 2. Transmitter showing single-chip Intel 87C51 microcontroller.

A similar procedure is followed for the PCM method except that each sample is represented by several pulses representing a binary code instead of just one pulse. A block of 11 time slots (the additional slot is for the frame synchronization) of binary coded digits is mapped per transmission frame in accordance with the encoding format for each PQRST sample block; this indicates the presence of the incoming successive PQRST frames to the receiver processor. A specific reverberation decay time, based on the nature and geometry of the underwater channel, is allowed between each transmitted burst to allow unambiguous discrimination of the direct path signal from any reverberation signals present. Clearly, the more sample time slots allocated per frame, the higher the information rate achievable for a given frame rate, but it also becomes more difficult to discriminate

152


Fig. 3. Real-time QRS detector and DPCM encoder.

Fig. 4. Flow chart of the QRS detection algorithm.

the individual frame boundaries to achieve correct decoding of the data from the incoming signals. In this work, the ECG transmission time for the PCM and PPM schemes is governed by the following equations, where N is the

number of bits: frame(PCM) = sync + 10N (slot + decay )

N frame(PPM) = sync + 10(2 slot + decay ):

(1) (2)


153

Fig. 6. Flow chart of the ECG receiver program. Fig. 5. Flow chart of the ECG transmitter program.

IV. EXPERIMENTAL RESULTS The maximum bit rate selected for ten ECG quantized samples for each transmission frame is given for PCM and PPM respectively as: fECG

(PCM)

fECG

(PPM)

10N sync + 10N (slot + decay ) 10N = : sync + 10(2N slot + decay ) =

(3) (4)

The heart rates for the two modulation schemes are as follows: fheart0rate(PCM) =

60

sync + 10N (slot + decay ) + QRS 1

fheart0rate(PPM) =

beats per minute

(5)

60

sync + 10(2N slot + decay ) + QRS 1 beats per minute:

These equations are the basis for programming the system.

(6)

AND

DISCUSSION

The performance of the ECG telemetry system was tested on different SCUBA divers in a large indoor tank (9 2 5 2 2 m). The main objective of the tests was to verify the functioning of the system by acquiring clinically acceptable ECG signals and to validate its performance in a severe multipath environment. System performance is characterized by the bit rates and corresponding heart rates for PCM and PPM, as shown in Table I for various values of quantization level (number of bits, N ); channel reverberation decay time (decay ); maximum bit rate (fbit ) and maximum heart rate (fheart0rate ): These results demonstrate that the performance of the system depends on the digital encoding method used and the reverberation decay time inherent in the relevant underwater channel. It can also be seen that for the same quantization level and reverberation delay, PPM encoding provides ECG signals at higher bit rates and consequently can operate at higher heart rates compared to the PCM

154


TABLE I COMPARATIVE PERFORMANCE OF PCM AND PPM FOR DIFFERENT QUANTIZATION LEVELS AND REVERBERATION DECAY TIMES

(a)

(b)

(c)

Fig. 7. Received PCM pulses and reconstructed ECG at the receiver; (a) received PCM pulses, including 7-bit synchronization word (V: 5 V/div; t: 100 ms/div); (b) reconstructed ECG before filtering (V: 0.5 V/div; t: 200 ms/div); (c) reconstructed ECG after filtering (V: 0.5 V/div; t: 200 ms/div).

V. CONCLUSIONS A novel microcontroller-based programmable underwater acoustic telemetry system for monitoring human cardiovascular activity and other physiological reflexes during SCUBA diving is presented. A comparative performance analysis using real-time digital PCM and PPM methods, together with programmable ECG detection and transmission, is presented. The results obtained illustrate that PPM is superior to PCM in achieving higher heart rates with clinically acceptable ECG signals under the same multipath reverberation conditions. The programmable methodology employed in the design of the system also allows other biomedical signals, such as the electroencephalogram (EEG), blood pressure and temperature to be telemetered. REFERENCES

Fig. 8. ECG digital encoding and transmission format.

method. This is particularly useful when transmitting the ECG at higher heart rates, such as during hard underwater work or fast swimming. Although a lower quantization level is associated with much higher bit rates and heart rates it results in poor quality, distorted ECG signals [9], [10]. The received ECG signals obtained using the PPM encoding method presented in this paper have been validated by a cardiologist and judged to be clinically acceptable. For comparison purposes, ECG signals were also recorded directly on an instrumentation tape recorder carried by a diver. The system hence provides a noninvasive means of studying cardiovascular reflexes during diving, such as tachycardia and bradycardia together with the associated heart rate variability [11], [12].

[1] D. C. Jeutter, “Overview of biomedical telemetry techniques,” IEEE Eng. Med. Biol. Mag., vol. 2, pp. 17–24, 1983. [2] J. Kanwisher, K. Lawson, and R. Strauss, “Acoustic telemetry from human divers,” Undersea Biomed. Res., vol. 1, pp. 99–109, 1974. [3] A. B. Baggeroer, “Acoustic telemetry—An overview,” IEEE J. Oceanic Eng., vol. OE-9, pp. 229–235, 1984. [4] B. Woodward and R. Sh. Habib, “Physiological monitoring by underwater ultrasonic biotelemetry,” J. Soc. Underwater Technol., vol. 18, pp. 24–44, 1992. [5] B. Woodward and R. Sh. Habib Istepanian, “Acoustic biotelemetry of data from divers,” in Proc. IEEE-EMBS ’93, San Diego, CA, pp. 1000–1002, Oct. 1993. [6] R. S. H. Istepanian, “Use of microcontrollers for diver monitoring by underwater acoustic biotelemetry in multipath environments,” Ph.D. dissertation, Loughborough Univ., U.K., 1994. [7] B. Woodward and R. Sh. Habib Istepanian, “The use of underwater biotelemetry for monitoring the ECG of swimming patient,” Proc. 1st. Regional IEEE-EMBS Conf., New Delhi, India, vol. 4, pp. 107–108, Feb. 1995. [8] N. Utsuyama, H. Yamaguchi, S. Obara, H. Tanaka, S. Fukuta, J. Nakahira, S. Tanube, E. Bando, and H. Miyamoto, “Telemetry of human electrocardiograms in aerial and aquatic environments,” IEEE Trans. Biomed. Eng., vol. 35, pp. 881–884, 1988. [9] R. S. Andrews and L. F. Turner, “On the performance of underwater data transmission system using amplitude shift keying techniques,” IEEE Trans. Sonics Ultrason., vol. 23, pp. 64–71, 1977. [10] J. Wartak, J. Milliken, and D. Lywood, “Theoretical and practical aspects of analog-to-digital conversion of the electrocardiogram,” Med. Res. Eng., pp. 21–23, Sept. 1970. [11] G. Pizzuti, S. Cifaldi, and G. Nalfe, “Digital sampling rate and ECG analysis,” IEEE J. Biomed. Eng., vol. 7, pp. 247–251, 1985. [12] R. Sh. Habib Istepanian and B. Woodward, “Spectral analysis of heart rate variability during SCUBA diving,” in Proc. IEEE-EMBS ’96, Amsterdam, The Netherlands, Oct. 1996, pp. 430–432.