In general, signals resulting from physiological activity have very small amplitudes and must therefore be amplified before their processing and display can be accomplished. The specifications and lists of characteristics of biopotential amplifiers can be as long and confusing as those for any other amplifier. However, for most typical medical applications, the most relevant amplifier characterizing parameters are the seven described below.
1. Gain. The signals resulting from electrophysiological activity usually have amplitudes on the order of a few microvolts to a few millivolts. The voltage of such signals must be amplified to levels suitable for driving display and recording equipment. Thus, most biopotential amplifiers must have gains of 1000 or greater. Most often the gain of an amplifier is measured in decibels (dB). Linear gain can be translated into its decibel form through the use of
Gain(dB) = 20 log10(linear gain)
2. Frequency response. The frequency bandwidth of a biopotential amplifier should be such as to amplify, without attenuation, all frequencies present in the electrophysiological signal of interest. The bandwidth of any amplifier, as shown in Figure 1.1, is the difference between the upper cutoff frequency f2 and the lower cutoff frequency f1. The gain at these cutoff frequencies is 0.707 of the gain in the midfrequency plateau. If the percentile gain is normalized to that of the midfrequency gain, the gain at the cutoff frequencies has decreased to 70.7%. The cutoff points are also referred to as the half-power points, due to the fact that at 70.7% of the signal the power will be (0.707)2 = 0.5. These are also known as the -3-dB points, since the gain at the cutoff points is lower by 3 dB than the gain in the midfrequency plateau: -3dB = 20 log10(0.707).
3. Common-mode rejection. The human body is a good conductor and thus will act as an antenna to pick up electromagnetic radiation present in the environment. As shown in Figure 1.2, one common type of electromagnetic radiation is the 50/60-Hz wave and its harmonics coming from the power line and radiated by power cords. In addition, other spectral components are added by fluorescent lighting, electrical machinery, computers,
Design and Development of Medical Electronic Instrumentation By David Prutchi and Michael Norris ISBN 0-471-67623-3 Copyright © 2005 John Wiley & Sons, Inc. '
and so on. The resulting interference on a single-ended bioelectrode is so large that it often obscures the underlying electrophysiological signals.
The common-mode rejection ratio (CMRR) of a biopotential amplifier is measurement of its capability to reject common-mode signals (e.g., power line interference), and it is defined as the ratio between the amplitude of the common-mode signal to the amplitude of an equivalent differential signal (the biopotential signal under investigation) that would produce the same output from the amplifier. Common-mode rejection is often expressed in decibels according to the relationship
Common-mode rejection (CMR) (dB) = 20 log10CMRR
4. Noise and drift. Noise and drift are additional unwanted signals that contaminate a biopotential signal under measurement. Both noise and drift are generated within the amplifier circuitry. The former generally refers to undesirable signals with spectral components above 0.1 Hz, while the latter generally refers to slow changes in the baseline at frequencies below 0.1 Hz.
The noise produced within amplifier circuitry is usually measured either in microvolts peak to peak (^Vp.p) or microvolts root mean square (RMS) (^VRMS), and applies as if it were a differential input voltage. Drift is usually measured, as noise is measured, in microvolts and again, applies as if it were a differential input voltage. Because of its intrinsic low-frequency character, drift is most often described as peak-to-peak variation of the baseline.
5. Recovery. Certain conditions, such as high offset voltages at the electrodes caused by movement, stimulation currents, defibrillation pulses, and so on, cause transient interruptions of operation in a biopotential amplifier. This is due to saturation of the amplifier caused by high-amplitude input transient signals. The amplifier remains in saturation for a finite period of time and then drifts back to the original baseline. The time required for the return of normal operational conditions of the biopotential amplifier after the end of the saturating stimulus is known as recovery time.
6. Input impedance. The input impedance of a biopotential amplifier must be sufficiently high so as not to attenuate considerably the electrophysiological signal under measurement. Figure 1.3a presents the general case for the recording of biopotentials. Each electrode-tissue interface has a finite impedance that depends on many factors, such as the type of interface layer (e.g., fat, prepared or unprepared skin), area of electrode surface, or temperature of the electrolyte interface.
In Figure 1.3b, the electrode-tissue has been replaced by an equivalent resistance network. This is an oversimplification, especially because the electrode-tissue interface is not merely a resistive impedance but has very important reactive components. A more correct representation of the situation is presented in Figure 1.3c, where the final signal recorded as the output of a biopotential amplifier is the result of a series of transformations among the parameters of voltage, impedance, and current at each stage of the signal transfer. As shown in the figure, the electrophysiological activity is a current source that causes current flow ie in the extracellular fluid and other conductive paths through the tissue. As these extracellular currents act against the small but nonzero resistance of the extracellular fluids Re, they produce a potential Ve, which in turn induces a small current flow iin in the circuit made up of the reactive impedance of the electrode surface XCe and the mostly resistive impedance of the amplifier Zin. After amplification in the first stage, the currents from each of the bipolar contacts produce voltage drops across input resistors Rin in the summing amplifier, where their difference is computed and amplified to finally produce an output voltage Vout.
The skin between the potential source and the electrode can be modeled as a series impedance, split between the outer (epidermis) and the inner (dermis) layers. The outer layer of the epidermis—the stratum corneum—consists primarily of dead, dried-up cells which have a high resistance and capacitance. For a 1-cm2 area, the impedance of the stratum corneum varies from 200 kQ at 1 Hz down to 200 Q at 1 MHz. Mechanical abrasion will reduce skin resistance to between 1 and 10kQ at 1 Hz.
7. Electrode polarization. Electrodes are usually made of metal and are in contact with an electrolyte, which may be electrode paste or simply perspiration under the electrode. Ion-electron exchange occurs between the electrode and the electrolyte, which results in voltage known as the half-cell potential. The front end of a biopotential amplifier must be able to deal with extremely weak signals in the presence of such dc polarization components. These dc potentials must be considered in the selection of a biopotential amplifier gain, since they can saturate the amplifier, preventing the detection of low-level ac components. International standards regulating the specific performance of biopotential recording systems
Figure 1.3 (a) Simplified view of the recording of biopotentials; (b) equivalent circuit; (c) generalized equivalent circuit.
usually specify the electrode offsets that are commonly present for the application covered by the standard. For example, the standards issued by the Association for the Advancement of Medical Instrumentation (AAMI) specify that electrocardiography (ECG) amplifiers must tolerate a dc component of up to ±300mV resulting from electrode-skin contact.
Commercial ECG electrodes have electrode offsets that are usually low enough, ensuring little danger of exceeding the maximum allowable dc input offset specifications of the standards. However, the design of a biopotential amplifier must consider that there are times when the dc offset may be much larger. For example, neonatal ECG monitoring applications often use sets of stainless-steel needle electrodes, whose offsets are much higher than those of commercial self-adhesive surface ECG electrodes. In addition, many physicians still prefer to use nondisposable suction cup electrodes (which have a rubber squeeze bulb attached to a silver-plated brass hemispherical cup). After the silver plating wears off, these brass cup electrodes can introduce very large offsets.
Was this article helpful?