A New Method for Monitoring of the Ballistocardiogram

Motonobu Hoshino

Citizen Watch Co.,Ltd.,Tokorozawa, Japan

Correspondence: M Hoshino, Citizen Watch Co.,LTD., 840, Shimotomi, Tokorozawa, 359-8511, Japan.
E-mail: hoshinom@citizen.co.jp, phone +81 42 942 7992, fax +81 42 942 9929

1.    Introduction

Ballistocardiography (BCG) is one of the non-invasive methods used for recording the kinetic effects produced on the body by cardiac contraction [Goehard, 1979]. In the past, BCG was not commonly used in medicine, the preferred method having been the static-charge-sensitive bed (SCSB) [Alihanka et al., 1981; Lindqvist, 1996]. The SCSB is an excellent method that can monitor the BCG, respiratory and body motion during sleeping. However, its application has been limited to the sleeping state. In an effort to expand the application range of BCG, a novel method is required that would allow for the determination of the BCG signal in a broad range of situations other than the sleeping state. Our system measures the change in electrostatic potential of the human body as a result of cardiac contraction and blood flow. Since we employed a single sensor electrode and an isolated capacitor, we were unable to use a differential amplifier to eliminate common mode noise such as the 50Hz AC noise. Consequently, an alternative noise reduction method was employed. The single electrode can easily be fitted to an ambulant device such as a wristwatch. This paper describes the newly developed method to monitor the BCG signal and details the preliminary results obtained.

2.    Material and Methods

A schematic block diagram of the system is shown in Fig.1. The sensor electrode was attached to the body and connected to the ground of the electric circuit. An isolated capacitor with size 3cm x 3cm was connected to the non-inverting input of a preamplifier, that consisted of a JFET-type operational amplifier AD549 (Analog Devices). The input impedance was 10¹³ Ohm.

Figure 1.     Schematic block diagram of the BCG Monitoring system.

As there is friction between the body and peripheral substances such as clothes or a chair, body movements induce electrostatic charge between these objects and itself. When the electric potential of the body changes because of this electrostatic charge, a potential difference exists between the isolated capacitor and the ground connected to the body. The electric potential of human body can be measured with a single sensor electrode. Unlike the metal plate used in the SCSB method, the human body itself can act as a sensor plate. The 50Hz AC noise was eliminated using an active notch filter in an effort to prevent saturation of the electronics circuit. Noise greater than 200Hz and lower than 0.5Hz was removed with a band-pass filter. The output signal was digitally sampled using a 12-bit ADC with a sampling interval of 0.5ms. The residue of the 50Hz AC noise was averaged over one 20 ms cycle using software. The ECG signal was simultaneously determined and compared to the corresponding BCG signal. Measurements were conducted at rest in the sitting state.

3.    Results

Typical BCG and ECG recordings are shown in Fig.2 (a). Sharp peaks are clearly visible in the BCG recording, with intervals that correspond well with those of the QRS peaks present in the ECG recordings. With our configuration, BCG amplitudes of several ten mV were obtained. The averaged signal of 10 BCG complexes using the QRS complex of the ECG as a trigger point is shown in Fig2 (b).

Figure 2. a:Typical BCG and ECG recordings for 8 seconds. b:An aAveraged signal of 10 BCG complexes using the QRS complex of the ECG as a trigger point.

4.    Discussion

The BCG peaks appear approximately 0.3 seconds after the QRS peaks of the ECG and coincide with the downslope of the ECG T wave. Given that the electrostatic potential of the body changes as a whole, the BCG signal is not dependent on the position of the sensor electrode. The averaged BCG signal shown in Fig2 (b) may provide for a clearer representation of hemodynamic behavior. It is difficult to determine the cause of the change in electrostatic potential. Given the relatively large delay time of the BCG peaks compared with other literature values [McKay, 1999, de Faire, 1979], it can be assumed that the potential occurs at the contact surface between the human body and the chair. The signal intensity can be varied by the environmental conditions. Although seven healthy volunteers participated in this study, a clear BCG signal could not be determined in two of the subjects. While the reproducibility of this method is currently under study, further investigations will be needed to fully evaluate the method presented here. Although the present system is large-scale in size, we demonstrated the possibility of obtaining a clear BCG signal using ambulant devices such as a wristwatch.

References

Alihanka J, Vaahtoranta K, Saarikivi I. A new method for long-term monitoring of the ballistocardiogram, heart rate, and

respiration. American Journal of Physiology, May;240(5): R384-92, 1981

de Faire U, Theorell T. A simple ballistocardiographic measure (IJ/HI amplitude ratio) in relation to electrocardiographic evidence of ischaemic heart disease. Scandinavian journal of clinical and laboratory investigation, Sep;39(5): 435-40, 1979

Goedhard W.J.A. Ballistocardiography: Past, Present and Future. Bibliotheca cardiologica, Vol.37: 27-45, 1979

Lindqvist A, Pihlajamaki K, Jalonen J, Laaksonen V, Alihanka J. Static-charge-sensitive bed ballistocardiography in

cardiovascular monitoring. Clinical Physiology, Jan;16(1): 23-30, 1996

McKay WP, Gregson PH, McKay BW, Militzer J. Sternal acceleration ballistocardiography and arterial pressure wave analysis to determine stroke volume. Clinical and investigative medicine, Feb;22(1): 4-14, 1999.