电子工程师手册（英文版六）：CH115.PDF

Bronzino, J.D., Berbari, E.J., Johnson, P.L., Smith, W.M. “Bioelectronics and Instruments” The Electrical Engineering Handbook Ed. Richard C. Dorf Boca Raton: CRC Press LLC, 2000 115 Bioelectronics and Instruments 115.1 The Electroencephalogram The Language of the Brain ? Historical Perspective ? EEG Recording Techniques ? Frequency Analysis of the EEG ? Nonlinear Analysis of the EEG ? Topographic Mapping 115.2 The Electrocardiograph Physiology ? Instrumentation ? Conclusions 115.3 Pacemakers/Implantable Defibrillators Pacemakers ? Implantable Cardioverter Defibrillators 115.1 The Electroencephalogram Joseph D. Bronzino Electroencephalograms (EEGs) are recordings of the minute (generally less than 300 μV) electrical potentials produced by the brain. Since 1924, when Hans Berger reported the measurements of rhythmic electrical activity on the human scalp, it has been suggested that these patterns of bioelectrical origin may provide clues regarding the neuronal bases for specific behaviors and has offered great promise to reveal correlations between patho- logical processes and the electrical activity of specific regions of the brain. Over the years, EEG analyses have been conducted primarily in clinical settings, to detect gross organic pathologies and the epilepsies, and in research facilities to quantify the central effect of new pharmacological agents. As a result of these efforts, cortical EEG patterns have been shown to be modified by a wide variety of variables including biochemical, metabolic, circulatory, hormonal, neuroelectric, and behavioral factors. In the past, interpretation of the EEG was limited to visual inspection by a trained electroencephalographer capable of distinguishing normal activity from localized or generalized abnormalities of particular types from relatively long EEG records. This approach has left clinicians and researchers alike lost in a sea of EEG paper records. Computer technology has permitted the application of a host of methods to quantify EEG changes. With this in mind, this section provides an introduction to some of the basic concepts underlying the generation of the EEG, a review of the basic approaches used in quantifying alterations in the EEG, and some insights regarding quantitative electrophysiology techniques. The Language of the Brain The mass of brain tissue is composed of bundles of nerve cells (neurons) which constitute the fundamental building blocks of the nervous system. Figure 115.1 is a schematic drawing of just such a cell. It consists of three major components: the cell body (or soma), the receptor zone (or dendrites), and the axon, which carries electrical signals from the soma to target sites such as muscles, glands, or other neurons. Numbering approx- imately 20 billion in each human being, these tiny cells come in a variety of sizes and shapes. Although neurons are anatomically distinct units having no physical continuity between their processes, the axon ends on the soma and the dendrites of other cells in what is called a synapse. Under the microscope this often stands out as a spherical enlargement at the end of the axon to which various names have been given, for example, boutons, Joseph D. Bronzino Trinity College/Biomedical Allience for Central Connecticut (BOACON) Edward J. Berbari Purdue University Philip L. Johnson University of Alabama at Birmingham William M. Smith University of Alabama at Birmingham ? 2000 by CRC Press LLC end-plate, or synaptic terminals. This ending does not actually make physical contact with the soma or dendrite but is separated by a narrow cleft (gap) of approximately 100 to 200 ? (10 –9 m) wide. This is known as the synaptic cleft. Each of these synaptic endings contains a large num- ber of submicroscopic spherical structures (synaptic vesi- cles) that can be detected only under an electron microscope. These synaptic vesicles, in turn, are essentially “chemical carriers” containing transmitter substance that is released into the synaptic cleft on excitation. When an individual neuron is excited, an electrical sig- nal is transmitted along its axon to many tiny branching, diverging fibers near its far end. These axonal terminals end as synapse on a large number of other neurons. When an electrical pulse arrives at the synapse, it triggers the release of a tiny amount of transmitter substance which crosses the synaptic cleft thereby altering the membrane potential of the receiving neuron. If the change is above a certain threshold value, the neuron is activated and generates an action potential of its own which is propagated along its axon, and the process is repeated. Neurons are involved in every conceivable action taken by the body, whether it is to control its own internal environment or to respond to changes in the external world. As a result, they are responsible for such essential functions as: ? Accepting and converting sensory information into a form that can be processed within the nervous system by other neurons. ? Processing and analyzing this information so that an “integrated portrait” of the incoming data can be obtained. ? Translating the final outcome or “decision” of this analysis process into appropriate electrical or chemical form needed to stimulate glands or activate muscles. Evolution has played a role in the development of these unique neurons and in the arrangement and development of interconnections between nerve cells in the various parts of the brain. Since the brain is a most complex organ, it contains numerous regions designed for specific tasks. One might, in fact, consider it to be a collection of organs arranged together to act in the harmony of activity we recognize as the individual’s state of consciousness or as life itself. Over the years, anatomists and physiologists have identified and named most pathways (tracts), most groups of neurons (nuclei), and most of the major parts of the human brain. Such attention to detail is certainly not necessary here. It will serve our purpose to simply provide a broad overview of the organization of the brain and speak of three general regions: the brainstem, cerebellum, and the cerebral cortex. The brainstem, or old brain, is really an extension and elaboration of the spinal chord. This section of the brain evolved first and is the location of all the centers that control the regulatory systems, such as respiration, necessary for physical survival of the organism. In addition, all sensory pathways find their way into the brainstem, thereby permitting the integration of complex input patterns to take place within its domain. Above the brainstem is a spherical mass of neuronal tissue called the cerebellum. This remarkable structure is a complex monitor and modifier of body movements. The cerebellum does not initiate movements, but only modifies motor control activated in other areas. Cerebellar operation is not only dependent on evolutionary development, but relies heavily on actual use and patterns of learned motor behavior acquired throughout life. It is for this reason that the movements of a gymnast are smooth and seemingly effortless. The most conspicuous part of all in the human brain is the cerebral cortex. Compared to most mammals, it is so large in man that it becomes a covering that surrounds and hides most of the other regions of the brain. Wrinkled and folded, the cerebral tissue is literally pressed into the limited space allocated to it. Although it has been possible to ascertain that certain cortical areas such as visual cortex, the sensory projection area, and the motor strip are associated with specific functions, the overall operation of this complex structure is still FIGURE 115.1 Basic structure of the neuron. ? 2000 by CRC Press LLC not completely understood. However, for the sake of convenience, it has been arbitrarily divided (based primarily on anatomical considerations) into the following areas: frontal lobe, parietal lobe, temporal lobe, and occipital lobe (Fig. 115.2). Each of these segments of the cortex, which is the source of intellectual and imaginative capacities, includes millions of neurons and a host of interconnections. It is generally agreed that brain function is based on the organization of the activity of large numbers of neurons into coherent patterns. Since the primary mode of activity of these nerve cells is electrical in nature, it is not surprising that a composite of this activity can be detected in the form of electrical signals. Of extreme interest, then, are the actual oscillations, rhythms, and patterns seen in the cryptic flow of electrical energy coming from the brain itself, i.e., in the EEG. Historical Perspective In 1875, Caton published the initial account of the recording of the spontaneous electrical activity of the brain from the cerebral cortex of an experimental animal. The amplitude of these electrical oscillations was so low, that is, on the order of microvolts, that Caton’s discovery is all the more amazing because it was made 50 years before suitable electronic amplifiers became available. In 1924, Hans Berger, of the University of Jena in Austria, carried out the first human EEG recordings using electrical metal strips pasted to the scalps of his subjects as electrodes and a sensitive galvanometer as the recording instrument. Berger was able to measure the irregular, relatively small electrical potentials (i.e., 50 to 100 mV) coming from the brain. By studying the successive positions of the moving element of the galvanometer recorded on a continuous roll of paper, he was able to observe the resultant patterns in these brain waves as they varied with time. From 1924 to 1938, Berger laid the foundation for many of the present applications of electroencephalography. He was the first to use the word electroencephalogram in describing these brain potentials in man. Berger noted that these brain waves were not entirely random, but instead displayed certain periodicities and regularities. For example, he observed that although these brain waves were slow (i.e., exhibited a synchronized patter of high amplitude and low frequency, <3 Hz) in sleep and states of depressed function, they were faster (i.e., exhibited a desynchronized pattern of low amplitude and high frequency, 15–25 Hz) during waking behavior. He suggested, quite correctly, that the brain’s activity changed in a consistent and recognizable fashion when the general status of the subject changed, as from relaxation to alertness. Berger also concluded that these brain waves could be greatly affected by certain pathological conditions after noting the marked increase in the amplitude of these brain waves brought about by convulsive seizures. However, in spite of the insights provided by these studies, Berger’s original paper published in 1929 did not excite much attention. In essence, the efforts of this most remarkable pioneer were largely ignored until similar investigations were carried out and verified by British investigators. It was not until 1934 when Adrian and Matthews published their classic paper verifying Berger’s findings that the reality of human brain waves was accepted and EEG studies were put on a firmly established basis. One of their primary contributions was the identification of certain rhythms in the EEG, regular oscillations at approximately 10–12 Hz in the occipital lobes of the cerebral cortex. They found that this alpha rhythm in the EEG would disappear when the brain displayed any type of attention or alertness or focused on objects in the visual field. The physiological basis for these results, the “arousing influence” of external stimuli on the FIGURE 115.2 Major divisions of the cerebral cortex. ? 2000 by CRC Press LLC cortex, was not formulated until 1949 when Moruzzi and Magoun demonstrated the existence of widely spread pathways through the central reticular core of the brainstem capable of exerting a diffuse activating influence on the cerebral cortex. This reticular activating system has been called the brain’s response selector because it alerts the cortex to focus on certain incoming information while ignoring other. It is for this reason that a sleeping mother will immediately be awakened by her crying baby or the smell of smoke, and yet ignore the traffic outside her window or the television still playing in the next room. An in-depth discussion of these early studies is beyond the scope of this presentation; however, for the interested reader an excellent historical review of this early era in brain research has been recorded in a fascinating text by Brazier [1968]. EEG Recording Techniques Scalp recordings of spontaneous neuronal activity in the brain, identified as the EEG, allow measurement of potential changes over time between a signal electrode and a reference electrode [Kondraski, 1986]. Compared to other biopotentials, such as the electrocardiogram, the EEG is extremely difficult for an untrained observer to interpret. As might be expected, partially as a result of the spatial mapping of functions onto different regions of the brain, correspondingly different waveforms are visible, depending on electrode placement. Recognizing that some standardization was necessary for comparison of research as well as clinical EEG records, the International Federation in Electroencephalography and Clinical Neurophysiology adopted the 10–20 electrode placement system, [Jasper, 1958]. Additional electrodes to monitor extracerebral contaminants of the EEG such as eye movement, EKG, and muscle activity are essential. The acquisition of EEG for quantitative analysis should also require the ability to view the EEG during collection on a polygraph or high-resolution video display. Since amplification, filtering, and digitization determine the frequency characteristics of the EEG and the source of potential artifacts, the acquisition parameters must be chosen with an understanding of their effects on signal acquisition and subsequent analysis. Amplification, for example, increases the amplitude range (volts) of the analog-to-digital (A/D) converter. The resolution of the A/D converter is determined by the smallest amplitude of steps that can be sampled. This is calculated by dividing the voltage range of the A/D converter by 2 to the power of the number of bits of the A/D converter. For example, an A/D converter with a range of ±5 V with 12-bit resolution can resolve samples as small as ±2.4 mV. Appropriate matching of amplification and A/D converter sensitivity permits resolution of the smallest signal while preventing clipping of the largest signal amplitudes. The bandwidth of the filters and the rate of digitization determine the frequency components of interest that are passed, while other frequencies outside the band of interest that may represent potential artifacts, such as aliasing, are rejected. A filter’s characteristics are determined by the rate of the amplitude decrease at the bandwidth’s upper and lower edges. Proper digital representation of the analog signal depends on the rate of data sampling, which is governed by the Nyquist theorem that states that data sampling should be at least twice the highest frequency of interest. In addition to the information available from spontaneous electrical activity of the EEG, the brain’s electrical response to sensory stimulation can contribute data as to the status of cortical and subcortical regions activated by sensory input. Due to the relatively small amplitude of a stimulus-evoked potential as compared to the spontaneous EEG potentials, the technique of signal averaging is used to enhance the stimulus-evoked response. Stimulus averaging takes advantage of the fact that the brain’s electrical response is time-locked to the onset of the stimulus and the nonevoked background potentials are randomly distributed in time. Consequently, the average of multiple stimulus responses will result in the enhancement of the time-locked activity, while the averaged random background activity will approach zero. The result is an evoked response that consists of a number of discrete and replicable peaks that occur, depending upon the stimulus and the recording parameters, at predicted latencies from the onset of stimulation. The spatial localization of maximum peak amplitudes has been associated with cortical generators in primary sensory cortex. Instrumentation required for EEG recordings can be simple or elaborate [Kondraski, 1986]. (Note: Although the discussion presented in this section is for a single-channel system it can be extended to simultaneous multichannel recordings simply by multiplying the hardware by the number of channels required. In cases that do not require true simultaneous recordings, special electrode selector panels can minimize hardware require- ments.) Any EEG system consists of electrodes, amplifiers (with appropriate filters) and a recording device. ? 2000 by CRC Press LLC Commonly used scalp electrodes consist of Ag-AgCl disks, 1 to 3 mm in diameter, with a very flexible long lead that can be plugged into an amplifier. Although it is desirable to obtain a low-impedance contact at the electrode ski interface (less than 10 kW), this objective is confounded by hair and the difficulty of mechanically stabilizing the electrodes. Conductive electrode paste helps obtain low impedance and keep the electrodes in place. A type of cement (collodion) is used to fix small patches of gauze over electrodes for mechanical stability, and leads are usually taped to the subject to provide some strain relief. Slight abrasion of the skin is sometimes used to obtain better electrode impedances, but this can cause irritation and sometimes infection (as well as pain in sensitive subjects). For long-term recordings, as in seizure monitoring, electrodes present major problems. Needle electrodes, which must be inserted into the tissue between the surface of the scalp and skull, are sometimes useful. However, the danger of infection increases significantly. Electrodes with self-contained miniature amplifiers are somewhat more tolerant because they provide a low-impedance source to interconnecting leads, but they are expensive. Despite numerous attempts to simplify the electrode application process and to guarantee long-term stability, none has been widely accepted. Instruments are available for measuring impedance between electrode pairs. The procedure is recommended strongly as good practice, since high impedance leads to distortions that may be difficult to separate from actual EEG signals. In fact, electrode impedance monitors are built into some commercial devices for recording EEGs. Standard dc ohmmeters should not be used, since they apply a polarizing current that causes build-up of noisy electrode potential at the skin-electrode interface. Commercial devices apply a known-amplitude sinusoidal voltage (typically 1 kHz) to an electrode pair circuit and measure root mean square (rms) current, which is directly related to the magnitude of the impedance. From carefully applied electrodes, signal amplitudes of 1 to 10 mV can be obtained. Considerable amplification (gain = 10 6 ) is required to bring these levels up to an acceptable level for input to recording devices. Because of long electrode leads and the common electrically noisy environment where recordings take place, differential amplifiers with inherently high input impedance and high common mode rejection ratios are essential for high- quality EEG recordings. In some facilities, special electrically shielded rooms minimize environmental electrical noise, particularly 60-Hz alternating current (ac) line noise. Since much of the information of interest in the EEG lies in the frequency bands less than 40 Hz, low-pass filters in the amplifier can be switched into attenuate 60-Hz noise sharply. For attenuating ac noise when the low-pass cutoff is greater than 60 Hz, many EEG amplifiers have notch filters that attenuate only frequencies in a narrow band centered around 60 Hz. Since important signal infor- mation may also be attenuated, notch filtering should be used as a last resort; one should try to identify and eliminate the source of interference instead. In trying to identify 60-Hz sources to eliminate or minimize their effect, it is sometimes useful to use a dummy source, such as a fixed 100-kW resistor attached to the electrodes. An amplifier output represents only contributions from interfering sources. If noise can be reduced to an acceptable level (at least by a factor of 10 less than EEG signals) under this condition, one is likely to obtain uncontaminated EEG records. Different types of recording instruments obtain a temporary or permanent record of the EEG. The most common recording device is a pen or chart recorder (usually multichannel) that is an integral part of most commercially available EEG instruments. The bandwidth of clinical EEGs is relatively low (less than 40 Hz) and therefore within the frequency response capabilities of these devices. Recordings are on a long sheet of continuous paper (from a folded stack), fed past the moving pen at one of several selectable constant speeds. The paper speed translates into distance per unit time or cycles per unit time, to allow EEG interpreters to identify different frequency components or patterns within the EEG. Paper speed is selected according to the monitoring situation at hand: slow speeds (10 mm/s) for observing the spiking characteristically associated with seizures and faster speeds (up to 120 mm/s) for the presence of individual frequency bands in the EEG. In addition to (or instead of) a pen recorder, the EEG may be recorded on a multichannel frequency modulated (FM) analog tape recorder. During such recordings, a visual output device such as an oscilloscope or video display is necessary to allow visual monitoring of signals, so that corrective action (reapplying the electrodes and so on) can take place immediately if necessary. ? 2000 by CRC Press LLC Sophisticated FM cassette recording and playback systems allow clinicians to review long EEG recordings over a greatly reduced time, compared to that required to flip through stacks of paper or observe recordings as they occur in real time. Such systems take advantage of time compensation schemes, whereby a signal recorded at one speed (speed of the tape moving past the recording head of the cassette drive) is played back at a different, faster speed. The ratio of playback to recording speed is known, so the appropriate correction factor can be applied to played-back data to generate a properly scaled video display. A standard ratio of 60:1 is often used. Thus, a trained clinician can review each minute of real-time EEG in 1 s. The display appears to be scrolled at a high rate horizontally across the display screen. Features of these instruments allow the clinician to freeze a segment of EEG on the display and to slow down or accelerate tape speed from the standard playback as needed. A time mark channel is usually displayed as one of the traces as a convenient reference (vertical “tick” mark displayed at periodic intervals across the screen). Computers can also be recording devices, digitizing (converting to digital form) one or several amplified EEG channels at a fixed rate. In such sampled data systems, each channel is repeatedly sampled at a fixed time interval (sample interval) and this sample is converted into a binary number representation by an A/D converter. The A/D converter is interfaced to a computer system so that each sample can be saved in the computer’s memory. A set of such samples, acquired at a sufficient sampling rate (at least two times the highest frequency component in the sampled signal), is sufficient to represent all the information in the waveform. To ensure that the signal is band-limited, a low-pass filter with a cutoff frequency equal to the highest frequency of interest is used. Since physically realizable filters do not have the ideal characteristics, the sampling rate is usually greater than two times the filter’s cutoff frequency. Furthermore, once converted to a digital format, digital filtering techniques can be used. On-line computer recordings are only practical for short-term recordings or for situations in which the EEG is immediately processed. This limitation is primarily due to storage requirements. For example, a typical sampling rate of 128 Hz yields 128 new samples per second that require storage. For an 8-channel recording, 1,024 samples are acquired per second. A 10-minute recording period yields 614,400 data points. Assuming 8- bit resolution per sample, over 0.5 megabyte (MB) of storage is required to save the 10-minute recording. Processing can consist of compression for more efficient storage (with associated loss of total information content), as in data record or epoch averaging associated with evoked responses, or feature extraction and subsequent pattern recognition, as in automated spike detection in seizure monitoring. Frequency Analysis of the EEG In general, the EEG contains information regarding changes in the electrical potential of the brain obtained from a given set of recording electrodes. These data include the characteristic waveform with its variation in amplitude, frequency, phase, etc. and the occurrence of brief electrical patterns, such as spindles. Any analysis procedure cannot simultaneously provide information regarding all of these variables. Consequently, the selec- tion of any analytic technique will emphasize changes in one particular variable at the expense of the others. This observation is extremely important if one is to properly interpret the results obtained by any analytic technique. In this chapter, special attention is given to frequency analysis of the EEG. In early attempts to correlate the EEG with behavior, analog frequency analyzers were used to examine single channels of EEG data. Although disappointing, these initial efforts did introduce the utilization of frequency analysis to study gross brain wave activity. Although, power spectral analysis, i.e., the magnitude square of Fourier transform, provides a quantitative measure of the frequency distribution of the EEG, it does so as mentioned above, at the expense of other details in the EEG such as the amplitude distribution, as well as the presence of specific patterns in the EEG. The first systematic application of power spectral analysis by general-purpose computers was reported in 1963 by Walter; however, it was not until the introduction of the fast Fourier transform (FFT) by Cooley and Tukey in the early 1970s that machine computation of the EEG became commonplace. Although an individual FFT is ordinarily calculated for a short section of EEG data (e.g., from 1 to 8 s epoch), such segmentation of a signal with subsequent averaging over individual modified periodograms has been shown to provide a consistent estimator of the power spectrum, and an extension of this technique, the compressed spectral array, has been particularly useful for computing EEG spectra over long periods of time. A detailed review of the ? 2000 by CRC Press LLC development and use of various methods to analyze the EEG is provided by Givens and Redmond [1987]. Figure 115.3 provides an overview of the computational processes involved in performing spectral analysis of the EEG, i.e., including computation of auto and cross spectra [Bronzino, 1984]. It is to be noted that the power spectrum is the autocorrellogram, i.e., the correlation of the signal with itself. As a result, the power spectrum provides only magnitude information in the frequency domain; it does not provide any data regarding phase. The power spectrum is computed by: P(f) = Re 2 [X(f)] + Im 2 [X(f)] (115.1) where X(f) is the Fourier transform of the EEG. Power spectral analysis not only provides a summary of the EEG in a convenient graphic form, but also facilitates statistical analysis of EEG changes which may not be evident on simple inspection of the records. In addition to absolute power derived directly from the power spectrum, other measures calculated from absolute power have been demonstrated to be of value in quantifying various aspects of the EEG. Relative power expresses the percent contribution of each frequency band to the total power and is calculated by dividing the power within a band by the total power across all bands. Relative power has the benefit of reducing the intersubject variance associated with absolute power that arises from intersubject differences in skull and scalp conductance. The disadvantage of relative power is that an increase in one frequency band will be reflected in the calculation by a decrease in other bands; for example, it has been reported that directional shifts between high and low frequencies are associated with changes in cerebral blood flow and metab- olism. Power ratios between low (0–7 Hz) and high (10–20 Hz) frequency bands have been demonstrated to be an accurate estimator of changes in cerebral activity during these metabolic changes. Although the power spectrum quantifies activity at each electrode, other vari- ables derivable from FFT offer a measure of the relationship between activity recorded at distinct electrode sites. Coherence (which is a complex number), cal- culated from the cross-spectrum analysis of two signals, is similar to cross-corre- lation in the time domain. The magnitude squared coherence (MSC) values range from 1 to 0, indicating maximum or no synchrony, respectively, and are independent of power. The temporal relationship between two signals is expressed by phase, which is a measure of the lag between two signals for common frequency components or bands. Phase is expressed in units of degrees, 0° indicating no time lag between signals or 180° if the signals are of opposite polarity. Phase can also be transformed into the time domain, giving a measure of the time difference between two frequencies. Cross spectrum is computed by: Cross spectrum = X(f) Y*(f) (115.2) where X(f), Y(f) are Fourier transforms and * indicates complex conjugates and coherence is calculated by (115.3) Since coherence is a complex number, the phase is simply the angle associated with the polar expression of that number. MSC and phase represent measures that can be employed to investigate the cortical interactions of cerebral activity. For example, short (intracortical) and long (cortico-cortical) pathways have been proposed Coherence= Cross spectrum PXf PYf() ()- FIGURE 115.3 Block dia- gram of measures determined from spectral analysis. ? 2000 by CRC Press LLC as the anatomic substrate underlying the spatial frequency and patterns of coherence. Therefore, discrete cortical regions linked by such fiber systems should demonstrate a relatively high degree of synchrony, whereas the time lag between signals, as represented by phase, quantifies the extent to which one signal leads another. Nonlinear Analysis of the EEG As mentioned earlier, the EEG has been studied extensively using signal-processing schemes, most of which are based on the assumption that the EEG is a linear, gaussian process. Although linear analysis schemes are computationally efficient and useful, they only utilize information retained in the autocorrelation function (i.e., the second-order cumulant). Additional information stored in higher-order cumulants is therefore ignored by linear analysis of the EEG. Thus, while the power spectrum provides the energy distribution of a stationary process in the frequency domain, it cannot distinguish nonlinearly coupled frequencies from spontaneously generated signals with the same resonance condition [Nikias and Raghvveer, 1987]. There is evidence showing that the amplitude distribution of the EEG often deviates from gaussian behavior. It has been reported, for example, that the EEG of humans involved in the performance of mental arithmetic task exhibits significant nongaussian behavior. In addition, the degree of deviation from gaussian behavior of the EEG has been shown to depend to the behavioral state, with the state of slow-wave sleep showing less gaussian behavior than quiet waking, which is less gaussian than rapid eye movement (REM) sleep [Ning and Bronzino, 1989a,b]. Nonlinear signal-processing algorithms such as bispectral analysis are therefore necessary to address nongaussian and nonlinear behavior of the EEG in order to better describe it in the frequency domain. But what exactly is the bispectrum? For a zero-mean, stationary process {X(k)}, the bispectrum, by definition, is the Fourier transform of its third-order cumulant (TOC) sequence: (115.4) The TOC sequence {C(m, n)} is defined as the expected value of the triple product (115.5) If process X(k) is purely gaussian, then its third-order cumulant C(m, n) is zero for each (m, n), and consequently, its Fourier transform, the bispectrum, B(w 1 , w 2 ) is also zero. This property makes the estimated bispectrum an immediate measure describing the degree of deviation from gaussian behavior. In our studies [Ning and Bronzino, 1989a,b], the sum of magnitude of the estimated bispectrum was used as a measure to describe the EEG’s deviation from gaussian behavior, that is, (115.6) Using bispectral analysis, the existence of significant quadratic phase coupling (QPC) in the hippocampal EEG obtained during REM sleep in the adult rat was demonstrated [Ning and Bronzino, 1989a,b, 1990]. The result of this nonlinear coupling is the appearance, in the frequency spectrum, of a small peck centered at approximately 13 to 14 Hz (beta range) that reflects the summation of the two theta frequency (i.e., in the 6- to 7-Hz range) waves. Conventional power spectral (linear) approaches are incapable of distinguishing the fact that this peak results from the interaction of these two generators and is not intrinsic to either. To examine the phase relationship between nonlinear signals collected at different sites, the cross-bispectrum is also a useful tool. For example, given three zero-mean, stationary processes ”x j (n)j = 1, 2, 3}, there are two conventional methods for determining the cross-bispectral relationship, direct and indirect. Both methods first divide these three processes into M segments of shorter but equal length. The direct method computes the BCmne jwm wn n a m a ww aa 12 12 ,, ( ) = ( ) -+() =-=- ?? Cmn E XkXk mXk n, ( ) = ( ) + ( ) + ( ){} DB= ( ) () ? ww ww 12 12 , ? 2000 by CRC Press LLC Fourier transform of each segment for all three processes and then estimates the cross-bispectrum by taking the average of triple products of Fourier coefficients over M segments, that is, (115.7) where X j m (w) is the Fourier transform of the mth segment of {x j (n)}, and * indicates the complex conjugate. The indirect method computes the third-order cross-cumulant sequence for all segments: (115.8) where t is the admissible set for argument n. The cross-cumulant sequences of all segments will be averaged to give a resultant estimate: (115.9) The cross-bispectrum is then estimated by taking the Fourier transform of the third-order cross-cumulant sequence: (115.10) Since the variance of the estimated cross-bispectrum is inversely proportional to the length of each segment, computation of the cross-bispectrum for processes of finite data length requires careful consideration of both the length of individual segments and the total number of segments to be used. The cross-bispectrum can be applied to determine the level of cross-QPC occurring between {x 1 (n)} and {x 2 (n)} and its effects on {x 3 (n)}. For example, a peak at Bx 1 x 2 x 3 (w 1 , w 2 ) suggests that the energy component at frequency w 1 + w 2 of {x 3 (n)} is generated due to the QPC between frequency w 1 of {x 1 (n)} and frequency w 2 of {x 2 (n)}. In theory, the absence of QPC will generate a flat cross-bispectrum. However, due to the finite data length encountered in practice, peaks may appear in the cross-bispectrum at locations where there is no significant cross-QPC. To avoid improper interpretation, the cross-bicoherence index, which indicates the significance level of cross-QPC, can be computed as follows: (115.11) where P xj (w) is the power spectrum of process {x j (n)}. The theoretical value of the bicoherence index ranges between 0 and 1, i.e., from nonsignificant to highly significant. In situations where the interest is the presence of QPC and its effects on {x(n)}, the cross-bispectrum equations can be modified by replacing {x 1 (n)} and {x 3 (n)} with {x(n)} and {x 2 (n)} with {y(n)}, that is, (115.12) B M XXX xxx m m M mm 123 12 1 1 12 23 1 2 1 ww w w w w, * ( ) = ( ) ( ) + ( ) = ? Ckl xnxnkxnl xxx mm n mm 123 12 3 , ( ) = ( ) + ( ) + ( ) ? et Ckl M Ckl xxx xxx m m M 123 123 1 1 ,, ( ) = ( ) = ? BCkl xxx xxx jk l lk 123 123 12 12 ww ww a a a a ,, ( ) = ( ) -+() =-=- ?? bic B PP P xxx xxx xx x 123 123 12 3 12 12 1212 ww ww wwww , , ( ) = ( ) ( ) ( ) + ( ) B M XY X xyz m m M mm ww w w w w 12 1 1 212 1 , * ( ) = ( ) ( ) + ( ) = ? ? 2000 by CRC Press LLC In theory, both methods will lead to the same cross-bispectrum when data length is infinite. However, with finite data records, direct and indirect methods generally lead to cross-bispectrum estimates with different shapes (Fig. 115.4). Therefore, like power spectrum estimation, users have to choose an appropriate method to extract the information desired. Topographic Mapping Computerized tomography (CT) and magnetic resonance imaging (MRI) have demonstrated the impact of spatial displays on data interpretation and analysis. Similarly, mapping techniques have been applied to elec- trophysiologic data to depict the spatial information available from multielectrode recordings. This effort has been assisted by the development and implementation of low-cost, high-resolution graphic displays on micro- computer systems. The data are frequently presented as two-dimensional topographic color maps [Zappulla, 1991]. In the time domain, color values depict the changes in potential across the scalp at each time point. This is exemplified by mapping peaks of an evoked potential or the spatial distribution of an epileptic spike. Temporal changes in the spatial distribution of voltage can be presented graphically as a series of maps constructed at adjacent time points or by cartooning the topographic maps over the time interval of interest. In the frequency domain, color coding can be used to spatially map power, covariance, and phase values. These maps may be constructed for the broadband activity or for selective frequency components. Unlike CT and MRI displays where each picture element or pixel value represents real data, most of the pixels comprising an EEG and ER topographic map consist of interpolated values. This is because the activity FIGURE 115.4A and B represent the averaged power spectra of 80 4-s epochs of REM sleep (sampling rate = 128 Hz) obtained from hippocampal CA1 and the dentate gyrus, respectively. Note that both spectra exhibit clear power peaks at 7- Hz (theta) and 14-Hz (beta) frequencies. C and D represent the bispectrum of these same epochs from CA1 and the dentate gyrus, respectively. Computation of the bicoherence index at 7 Hz shows significant quadratic phase coupling at this frequency, indicating that the 14-Hz peak is not spontaneously generated, but results from quadratic phase coupling. ? 2000 by CRC Press LLC from a finite number of electrodes represents a sampling of the spatial activity over the scalp. Consequently, the remaining values of the map located outside the electrode positions must be estimated from this sampled activity. One technique for deriving these values is linear interpolation. In the case of a four-point interpolation, the map is divided into boxes whose corners are defined by real data. The interpolated points within the boxes are calculated by the weighted sum of the four real data points, based on their distance from the interpolated point. Although linear interpolation is the most popular technique, polynomial regression and surface spline interpolation have been employed as alternative procedures. These methods reduce the discontinuities inherent in linear interpolation and offer better estimates of extreme values. Polynomial regression has the additional advantage of permitting quantitative comparisons between maps by taking into account the topographic information represented in the map. Maps can be presented in any of several projections to assist in interpretation [Zappulla, 1991]. The most common projection is the top view which presents the spatial distribution of variables from all leads simulta- neously. Lateral, posterior, and anterior projections highlight focal areas of interest. Although mapping presents a method by which spatial information can be efficiently communicated, it is important to be alert to the artifacts that can arise from map construction and manipulation. Topographic spatial artifacts that can lead to misinterpretation include ring enhancement around a spike using source-derivation references, spatial aliasing arising from linear interpolation which causes maximal activity to be mapped at electrode sites, the enhancement of activity away from the midline, and the attenuation of midline activity on amplitude asymmetry maps (centrifugal effect). The quality of the spatial information derivable from EEG recordings depends upon the number of recording electrodes, the choice of the reference electrode, and the conductive properties of intracranial and extracranial structures. The localization of cortical activity from scalp recordings assumes that the potentials recorded from the scalp reflect cortical activity generated in proximity to the recording electrode. Therefore, the greater the density of recording electrodes, the more accurate the estimate of the spatial distribution of scalp potentials and the localization of cortical generators. However, since the distance between the cortical source and recording electrode, as well as the low conductivity of the skull, results in a selective attenuation of small dipole fields, most available EEG information can be obtained with an average scalp-electrode spacing of 2 cm. Topographic maps are constructed from monopolar electrodes referenced to a common cephalic (linked ears or mandible, chin and nose) or noncephalic (linked clavicles or a balanced sternum-vertebra) electrode. Although the reference electrode should be free of any EEG activity, in practice most cephalic electrodes contain some EEG activity, while noncephalic electrodes are a potential source of EKG or muscle activity. Differential amplification of an EEG-contaminated reference electrode can decrease or cancel similar activity in neighboring electrodes, while at electrodes distant from the reference, the injected activity will be present as a potential of opposite polarity. Similarly, noncerebral potentials can be injected into scalp electrodes and misinterpreted as cerebral activity. Therefore, a nonneutral reference electrode can result in misleading map configurations. Several techniques have been applied to circumvent this problem. The construction of multiple maps using several different references can sometimes assist in differentiating active and reference electrode activity. This can be accomplished by acquiring serial EEG records using different references. Alternatively, various references can be acquired simultaneously during acquisition, and various montages can be digitally reconstructed, post hoc. A more computationally intensive method for localizing a source at an electrode involves calculating the local source activity at any one electrode based on the average activity of its neighbors, weighted by their distance from the source. The technique has the advantage of suppressing potentials that originate outside the measure- ment area and weighing factors for implementing source deviation techniques for each of the electrodes in the 10–20 system are available. Another reference technique, the average head reference, uses the average activity of all active electrodes as the common reference. In this approach, the activity at any one electrode will vary depending upon the activity at the site of the reference electrode, which can be anywhere on the recording montage. Therefore, for N number of recording electrodes, each being a potential reference, there are N – 1 possible voltage measurements at each instant of time for each electrode. Maps constructed using the average head reference represent a unique solution to the problem of active reference electrodes in that the average reference produces an amplitude-weighted reference-free map of maximal and minimal field potentials. Power maps constructed from the average reference ? 2000 by CRC Press LLC best depict the spatial orientation of the generating field, and the areas with extreme values are closest to the generating processes [Zappulla, 1991]. Topographical maps represent an efficient format for displaying the extensive amount of data generated by quantitative analysis. However, for reasons discussed above, the researcher and clinician must be cautious in deriving spatial and functional conclusions from mapped data. Although the replicability of map configurations across subjects or experimental conditions may represent a useful basis for experimental and diagnostic clas- sification, judgments concerning the localization of cortical generators or functional localization of cerebral activity are less certain and more controversial. Research continues on defining models and validating assump- tions that relate scalp potentials to cortical generators in an attempt to arrive at accurate mathematical solutions that can be applied to mapping functions. Defining Terms Bispectra: Computation of the frequency distribution of the EEG exhibiting nonlinear behavior. Cross spectra: Computation of the energy in the frequency distribution of two different electrical signals. Electroencephalogram (EEG): Recordings of the electrical potentials produced by the brain. Fast Fourier transform (FFT): Algorithms that permit rapid computation of the Fourier transform of an electrical signal, thereby representing it in the frequency domain. Magnitude squared coherence (MSC): A measure of the degree of synchrony between two electrical signals at specific frequencies. Power spectral analysis: Computation of the energy in the frequency distribution of an electrical signal. Quadratic phase coupling: A measure of the degree to which specific frequencies interact to produce a third frequency. Related Topic 108.1 Introduction References M. Brazier, Electrical Activity of the Nervous System, 3rd ed., Baltimore: Williams and Wilkins, 1968. J.D. Bronzino, M. Kelly, C. Cordova, “Utilization of amplitude histograms to quantify the EEG: Effects of systemic administration of morphine in the chronically implanted rat,” IEEE Trans. Biomed. Eng., 28(10), 673, 1981. J.D. Bronzino, “Quantitative analysis of the EEG: General concepts and animal studies,” IEEE Trans. Biomed. Eng., 31(12), 850, 1984. J.W. Cooley and J.S. Tukey, “An algorithm for the machine calculatio of complex Fourier series,” Math Comput., 19, 267, 1965. A.S. Givens and A. Remond, Eds., “Methods of analysis of brain electrical and magnetic signals,” in EEG Handbook, vol. 1, Amsterdam: Elsevier, 1987. S.M. Kay and S.L. Maple, “Spectrum analysis—A modern perspective,” Proc. IEEE. 69, 1380, 1981. G.V. Kondraski, “Neurophysiological measurements,” in Biomedical Engineering and Instrumentation, J.D. Bronzino, Ed., Boston: PWS Publishing, pp. 138–179, 1986. C.L. Nikias and M.R. Raghuveer, “Bispectrum estimation: A digital signal processing framework,” Proc. IEEE, 75, 869, 1987. T. Ning and J.D. Bronzino, “Bispectral analysis of the rat EEG during different vigilance states,” IEEE Trans. Biomed. Eng., 36(4), 497, 1989a. T. Ning and J.D. Bronzino, “Bispectral analysis of the EEG in developing rats,” in Proc. Workshop Higher-Order Spectral Anal., Vail, Colo.: 1989b, pp. 235–238. T. Ning and J.D. Bronzino, “Autoregressive and bispectral analysis techniques: EEG applications,” Special Issue on Biomedical Signal Processing, IEEE Eng. Med. Biol. Mag., 9, 47, 1990. J.R. Smith, “Automated analysis of sleep EEG data,” in Clinical Applications of Computer Analysis of EEG and Other Neurophysiological Signals, EEG Handbook, revised series, vol. 2, Amsterdam: Elsevier, 1986, pp. 93–130. ? 2000 by CRC Press LLC Further Information The Biomedical Engineering Handbook, J.D. Bronzino, Ed., Boca Raton, Fla.: CRC Press, 1995. The Electroencephalogram: Its Patterns and Origins, by J.S. Barlow (Cambridge, Mass., MIT Press, 1993). See also the journals, IEEE Transactions in Biomedical Engineering and Electroencephalography and Clinical Neurophysiology. 115.2 The Electrocardiograph Edward J. Berbari The electrocardiogram (ECG) is the recording on the body surface of the electrical activity generated by the heart. It was originally observed by Waller in 1889 using his pet bulldog as the signal source and the capillary electrometer as the recording device. In 1903 Einthoven enhanced the technology by using the string galva- nometer as the recording device and using human subjects with a variety of cardiac abnormalities. Einthoven is chiefly responsible for introducing some concepts still in use today including the labeling of the various waves, defining some of the standard recording sites using the arms and legs, and developing the first theoretical construct whereby the heart is modeled as a single time varying dipole. We also owe the “EKG” acronym to Einthoven’s native Dutch language where the root word “cardio” is spelled with a “k”. In order to record an ECG waveform, a differential recording between two points on the body is made. Traditionally each differential recording is referred to as a lead. Einthoven defined three leads numbered with the Roman numerals I, II, and III. They are defined as: I = V LA – V RA (115.13) II = V LL – V RA (115.14) III = V LL – V LA (115.15) where RA = right arm, LA = left arm, and LL = left leg. Because the body is assumed to be purely resistive, at ECG frequencies, the four limbs can be thought of as wires attached to the torso. Hence lead I could be recorded from the respective shoulders without a loss of cardiac information. Note that these are not independent and the following relationship holds: II = I + III. For 30 years the evolution of the ECG proceeded when F. N. Wilson [1934] added concepts of a “unipolar” recording. He created a reference point by tying the three limbs together and averaging their potentials so that individual recording sites on the limbs or chest surface would be differentially recorded with the same reference point. Wilson extended the biophysical models to include the concept of the cardiac source enclosed within the volume conductor of the body. He erroneously thought that the central terminal was a true zero potential. However, from the mid-1930s until today the 12 leads composed of the three limb leads, three leads in which the limb potentials are referenced to a modified Wilson terminal (the augmented leads [Goldberger, 1942]), and six leads placed across the front of the chest and referenced to the Wilson terminal form the basis of the standard 12-lead ECG. Figure 115.5 summarizes the 12-lead set. These sites are historically based, have a built in redundancy, and are not optimal for all cardiac events. The voltage difference from any two sites will record an ECG, but it is these standardized sites with the massive 90-year collection of empirical observations that has firmly established their role as the standard. Figure 115.6 is a typical or stylized ECG recording from lead II. Einthoven chose the letters of the alphabet from P to U to label the waves and to avoid conflict with other physiologic waves being studied at the turn of the century. The ECG signals are typically in the range of ±2 mV and require a recording bandwidth of 0.05–150 Hz. Full technical specification for ECG equipment has been proposed by both the American Heart Association [1984] and the Association for the Advancement of Medical Instrumentation [Bailey et al., 1990]. There have been several attempts to change the approach for recording the ECG. The vectorcardiogram used a weighted set of recording sites to form an orthogonal XYZ lead set. The advantage here was minimum lead set but in practice it gained only a moderate degree of enthusiasm among physicians. Body surface mapping ? 2000 by CRC Press LLC ? 2000 by CRC Press LLC INFRARED CAMERA ASA, teaming with an industry partner, has developed a revolutionary infrared camera that offers important applications not only in aerospace research but in other areas such as air transportation, environment monitoring, and medicine. An innovative feature of the infrared camera shown is its use of highly sensitive quantum-well pho- todetectors of QWIPS. The greater sensitivity of long wavelength QWIPS could allow physicians to detect tumors using thermographic techniques; improve pilots’ night vision to allow better landings; and enable environmental scientists to monitor pollution and weather patterns with enhanced measurement accu- racy. Other possible applications include law enforcement, industrial process control, search and rescue, and military antimissile surveillance. The camera weighs only 9.9 pounds and measures 4.4 inches wide, 10.3 inches deep, and 7.2 inches long. The prototype plugs into a wall socket for power but the camera can be converted readily to battery power for portability. Because infrared light detectors must operate at extremely low temperatures, the camera contains a Stirling cryocooler, a closed-cycle refrigerator about the size of a fist that cools the camera from room temperature to about 343 degrees below zero Fahrenheit in about 10 minutes. The camera was developed by the Center for Space Microelectronics Technology at the Jet Propulsion Laboratory in cooperation with Amber, a Raytheon company. (Courtesy of National Aeronautics and Space Administration.) This revolutionary infrared camera, developed by an industry/government team, has broad applications in medicine, environment monitoring, industrial processing, and law enforcement, as well as in aerospace research. (Photo courtesy of National Aeronautics and Space Administration.) N refers to the use of many recording sites (>64) arranged on the body so that isopotential surfaces could be computed and analyzed over time. This approach still has a role in research investigations. Other subsets of the 12-lead ECG are used in limited mode recording situations such as the tape recorded ambulatory ECG FIGURE 115.5 The 12-lead ECG is formed by the three bipolar surface leads: I, II, and III; the augmented Wilson terminal referenced limb leads: aVR, aVL, aVF; and the Wilson terminal referenced chest leads: V 1 , V 2 , V 3 , V 4 , V 5 , and V 6 . FIGURE 115.6 Stylized version of a normal lead II recording showing the P wave, QRS complex, and the T and U waves. The PR interval and the ST segment are significant time windows. The peak amplitude of the QRS is about 1 mV. The vertical scale is usually 1 mV/cm. The time scale is usually based on mm/s scales with 25 mm/s being the standard form. The small boxes of the ECG are 1 ′ 1 mm. ? 2000 by CRC Press LLC (usually two leads) or in intensive care monitoring at the bedside (usually one or two leads) or telemetered within regions of the hospital from patients who are not confined to bed (one lead). The recording electronics of these ECG systems have followed the typical evolution of modern instrumentation, e.g., vacuum tubes, transistors, ICs, and microprocessors. Application of computers to the ECG for machine interpretation was one of the earliest uses of computers in medicine [Jenkins, 1981]. Of primary interest in the computer-based systems was the replacement of the human reader and the elucidation of the standard waves and intervals. Originally this was performed by linking the ECG machine to a centralized computer via phone lines. The modern ECG machine is completely integrated with an analog front end, a 12- to 16-bit A/D converter, a computational microprocessor, and dedicated I/O processors. These systems compute a measurement matrix derived from the 12 lead signals and analyze this matrix with a set of rules to obtain the final set of interpretive statements. The depiction of the 12 analog signals and this set of interpretive statements form the final output, with an example shown in Fig. 115.7. The physician will over-read each ECG and modify or correct those statements which are deemed inappropriate. The larger hospital-based system will record these corrections and maintain a large database of all ECGs accessible by any combination of parameters, e.g., all males, older than 50, with an inferior myocardial infarction. More recently the high-resolution ECG (HRECG) as been developed whereby the digitized ECG is signal averaged to reduce random noise (Berbari et al., 1973, 1977). This approach, coupled with post averaging high- pass filtering, is used to detect and quantify low-level signals (?1.0 mV) not detectable with standard approaches. This computer-based approach has enabled the recording of events which are predictive of future life-threat- ening cardiac events [Berbari et al., 1978; Simson, 1981]. FIGURE 115.7Example of an interpreted 12-lead ECG. A 2 1?2-second recording is shown for each of the 12 leads. The bottom trace is a continuous 10-second rhythm strip of lead II. Patient information is given in the top area, below which is printed the computerized interpretive statements. (Tracing is courtesy of the Hewlett-Packard Co., Palo Alto, CA.) ? 2000 by CRC Press LLC Physiology The heart has four chambers; the upper two chambers are called the atria and the lower two chambers are called the ventricles. The atria are thin-walled, low-pressure pumps which receive blood from venous circulation. Located in the top right atrium are a group of cells which act as the primary pacemaker of the heart. Through a complex change of ionic concentration across the cell membranes (the current source) an extracellular potential field is established which then excites neighboring cells and a cell-to-cell propagation of electrical events occurs. Because the body acts as a purely resistive medium, these potential fields extend to the body surface [Geselowitz, 1989]. The character of the body surface waves depends upon the amount of tissue activating at one time and the relative speed and direction of the activation wavefront. Therefore the pacemaker potentials which are generated by a small tissue mass are not seen on the ECG. As the activation wavefront encounters the increased mass of atrial muscle, the initiation of electrical activity is observed on the body surface and the first ECG wave of the cardiac cycle is seen. This is the P wave and it represents activation of the atria. Conduction of the cardiac impulse proceeds from the atria through a series of specialized cardiac cells (the A-V node and the His-Purkinje system) which again are too small in total mass to generate a signal large enough to be seen on the standard ECG. There is a short relatively isoelectric segment following the P wave. Once the large muscle mass of the ventricles is excited, a rapid and large deflection is seen on the body surface. The excitation of the ventricles causes them to contract and provides the main force for circulating blood to the organs of the body. This large wave appears to have several components. The initial downward deflection is called the Q wave, the initial upward deflection is the R wave, and the terminal downward deflection is the S wave. The polarity and actual presence of these three components depends upon the position of the leads on the body as well as a multitude of abnormalities that may exist. In general, the large ventricular waveform is generically called the QRS complex regardless of its makeup. Following the QRS complex is another short relatively isoelectric segment. After this short segment the ventricles return to their electrical resting state and a wave of repolarization is seen as a low-frequency signal called the T wave. In some individuals a small peak occurs at the end or after the T wave and is called the U wave. Its origin has never been fully established but is believed to be a repolarization potential. Instrumentation The general instrumentation requirements for the ECG have been addressed by professional societies through the years [American Heart Association, 1984; Bailey et al., 1990]. Briefly, they recommend a system bandwidth 0.05–150 Hz. Of great importance in ECG diagnosis is the low-frequency response of the system because shifts in some of the low-frequency regions, e.g., the ST segment, have critical diagnostic value. While the heart rate may only have a 1-Hz fundamental frequency, the phase response of typical analog high-pass filters is such that the system corner frequency must be much smaller than the 3-dB corner frequency where only the amplitude response is considered. The system gain depends upon the total system design. The typical ECG amplitude is ± 2mV and if A/D conversion is used in a digital system, then enough gain to span the full range of the A/D converter is appropriate. To first obtain an ECG the patient must be physically connected to the amplifier front end. The patient/ampli- fier interface is formed by a special bioelectrode which converts the ionic current flow of the body to the electron flow of the metallic wire. These electrodes typically rely on a chemical paste or gel with a high ionic concen- tration. This acts as the transducer at the tissue-electrode interface. For short-term applications silver-coated suction electrodes or “sticky” metallic foil electrodes are used. Long-term recordings, such as the case for the monitored patient, require a stable electrode/tissue interface and special adhesive tape material surrounds the gel and a Ag + /Ag + Cl electrode. At any given time, the patient may be connected to a variety of devices, e.g., respirator, blood pressure monitor, temporary pacemaker, etc., some of which will invade the body and provide a low resistance pathway to the heart. It is essential that the device not act as a current source and inject the patient with enough current to stimulate the heart and cause it to fibrillate. Some bias currents are unavoidable for the system input stage and recommendations are that these leakage currents be less than 10 mA per device. This not only applies to the normal setting but if a fault condition arises whereby the patient comes in contact with the high voltage side of the ac power lines, then the isolation must be adequate to prevent 10 mA of fault current as well. This mandates that the ECG reference ground not be connected physically to the low side of the ac power line or ? 2000 by CRC Press LLC its third wire ground. For ECG machines the solution has typically been to AM modulate a medium-frequency carrier signal (?400 kHz) and use an isolation transformer with subsequent demodulation. Other methods of signal isolation can be used but the primary reason for the isolation is to keep the patient from being part of the ac circuit in the case of a patient to power line fault. In addition, with many devices connected in a patient monitoring situation it is possible that ground loop currents will be generated. To obviate this potential hazard a low-impedance ground buss is often installed in these rooms and each device chassis will have an external ground wire connected to the buss. Another unique feature of these amplifiers is that they must be able to withstand the high-energy discharge of a cardiac defibrillator. Figure 115.8 shows a three-channel ECG amplifier schematic used in a high-resolution ECG system. The patient is dc coupled to the front end differential, instrumentation amplifier. The first stage of gain is relatively low (?100) because there can be a significant signal drift due to a high static charge on the body or low- frequency offset potentials generated by the electrolyte in the tissue/electrode interface. In this particular amplifier the signal is bandpass filtered prior to the isolation stage. To further limit the high floating potential of the patient and to improve the system common mode rejection a driven ground is usually used. This ground is simply an average of the limb potentials inverted by a single amplifier and connected to the right leg. Older style ECG machines recorded one lead at a time, then evolved to three simultaneous leads. This necessitated the use of switching circuits as well as analog weighting circuits to generate the various 12 leads. This is usually eliminated in modern digital systems by using an individual single-ended amplifier for each electrode on the body. Each potential signal is then digitally converted and all of the ECG leads can be formed mathematically in software. This would necessitate a nine-amplifier system. By performing some of the lead calculations with the analog differential amplifiers this can be reduced to an eight-channel system. Thus only FIGURE 115.8This schematic represents a typical three-lead XYZ amplifier set used in a high-resolution ECG. The instrumentation amplifier (INA101) and bandpass filter (OP400) for each channel are on the isolated side of the power supply. The diode pairs and 4.1 kW resistors on each lead wire provide high-voltage defibrillation protection. The outputs of each differential amplifier are averaged through an amplifier (OPA121) and provide the right leg drive. (Schematic is courtesy of Corazonix Corp., Oklahoma City.) ? 2000 by CRC Press LLC the individual chest leads V 1 through V 6 and any 2 of the limb leads, e.g., I and III, are needed to calculate the full 12-lead ECG. Figure 115.9 is a block diagram of a modern digital-based ECG system. This system uses up to 13 single-ended amplifiers and a 16-bit A/D converter, all within a small lead wire manifold or amplifier lead stage. The digital signals are optically isolated and sent via a high-speed serial link to the main ECG instrument. Here the 32-bit CPU and DSP chip perform all of the calculations and a hard copy report is generated (Fig. 115.7). Notice that each functional block has its own controller and the system requires a real- time, multitasking operating system to coordinate all system functions. Concomitant with the data acquisition is the automatic interpretation of the ECG. These programs are quite sophisticated and are continually evolving. It is still a medical/legal requirement that these ECGs be over-read by the physician. High-resolution capability is now a standard feature on most digitally based ECG systems or as a stand- alone microprocessor-based unit [Berbari, 1988]. The most common application of the HRECG is to record very low-level (?1.0 mV) signals which occur after the QRS complex but are not evident on the standard ECG. These “late potentials” are generated from abnormal regions of the ventricles and have been strongly associated with the substrate responsible for a life-threatening rapid heart rate (ventricular tachycardia). The typical HRECG is derived from three bipolar leads configured in an anatomic XYZ coordinate system. These three ECG signals are then digitized at a rate of 1000–2000 Hz/channel, time aligned via a real-time QRS correlator, and summated in the form of a signal average. Signal averaging will theoretically improve the signal-to-noise ratio by the square root of the number of beats averaged. The underlying assumptions are that the signals of interest do not vary, on a beat-to-beat basis, and that the noise is random. Figure 115.10 has four panels depicting the most common sequence for processing the HRECG to measure the late potentials. Panel A depicts a three-second recording of the XYZ leads close to normal resolution. Panel B was obtained after averaging 200 beats and with a sampling frequency of 10 times that shown in panel A. The gain is also five times greater. Panel C is the high-pass filtered signal using a partially time reversed digital filter having a second-order Butterworth response and a 3-dB corner frequency of 40 Hz [Simson, 1981]. Note the appearance of the signals at the terminal portion of the QRS complex. A common method of analysis, but necessarily optimal, is to combine the filtered XYZ leads into a vector magnitude (X 2 + Y 2 + Z 2 ) 1/2 . This waveform is shown in panel D. From this waveform several parameters have been derived such as total QRS duration, including late potentials, the rms voltage value of the terminal 40 ms, and the low-amplitude signal (LAS) duration from the 40-mV level to the end of the late potentials. Abnormal values for these parameters are used to identify patients at high risk of ventricular tachycardia following a heart attack. FIGURE 115.9Block diagram of a microprocessor-based ECG system. It includes all of the elements of a personal computer class system, e.g., 80386 processor, 2 Mbytes of RAM, disk drive, 640 ′ 480 pixel LCD display, and is battery operable. In addition, it includes a DSP56001 chip and multiple controllers which are managed with a real-time, multitasking operating system. (Diagram is courtesy of the Hewlett-Packard Co., Palo Alto, Calif.) ? 2000 by CRC Press LLC Conclusions The ECG is one of the oldest instrument-bound measurements in medicine. It has faithfully followed the progression of instrumentation technology. Its most recent evolutionary step, to the microprocessor-based system, has allowed for an enhanced, high-resolution ECG which has opened new vistas of ECG analysis and interpretation. Defining Terms 12-lead ECG: Twelve traditional ECG leads comprising the standard set. ECG: Abbreviation for the device (electrocardiograph) or the output (electrocardiogram) depicting the body surface recording of the electrical activity of the heart. ECG lead: Differential signal depicting one channel of the ECG record. HRECG: High-resolution ECG used to detect microvolt-level cardiac potentials most commonly by signal averaging. Wilson central terminal: Reference point for forming most of the standard ECG leads. It is the average of the right arm, the left arm, and the left potentials. It is a time-varying reference. Related Topic 108.5 Instrumentation System FIGURE 115.10 The signal processing steps typically performed to obtain a high-resolution ECG are shown in panels A–D. See text for a full description. ? 2000 by CRC Press LLC References J.J. Bailey, A.S. Berson, A. Garson, L.G. Horan, P.W. Macfarlane, D.W. Mortara, and C. Zywietz, “Recommen- dations for standardization and specifications in automated electrocardiography: bandwidth and digital signal processing,” A report for health professionals by an ad hoc writing group of the Committee on Electrocardiography and Cardiac Electrophysiology of the Council on Clinical Cardiology, American Heart Association, Circulation, vol. 81, no. 2, pp. 730–739, 1990. E.J. Berbari, “High resolution electrocardiography,” CRC Crit. Rev. Bioeng., vol. 16, p. 67, 1988. E.J. Berbari, R. Lazzara, P. Samet, and B.J. Scherlag, “Noninvasive technique for detection of electrical activity during the PR segment,” Circulation, vol. 48, p. 1006, 1973. E.J. Berbari, R. Lazzara, and B.J. Scherlag, “A computerized technique to record new components of the electrocardiogram,” Proc. IEEE, vol. 65, p. 799, 1977. E.J. Berbari, B.J. Scherlag, R.R. Hope, and R. Lazzara, “Recording from the body surface of arrhythmogenic ventricular activity during the ST segment,” Am. J. Cardiol., vol. 41, p. 697, 1978. W. Einthoven, “Die galvanometrische Registrirung des menschlichen Elektrokardiogramms, zugleich eine Beur- theilung der Anwendung des Capillar-Elecktrometers in der Physiologie,” Pflugers Arch. Ges. Physiol., vol. 99, p. 472, 1903. D.B. Geselowitz, “On the theory of the electrocardiogram,” Proc. IEEE, vol. 77, p. 857, 1989. E. Goldberger, “A simple, indifferent, electrocardiographic electrode of zero potential and a technique of obtaining augmented, unipolar, extremity leads,” Amer. Heart J., vol. 23, p. 483, 1942. J.M. Jenkins, “Computerized electrocardiography,” CRC Crit. Rev. Bioeng., vol. 6, p. 307, 1981. M.B. Simson, “Use of signals in the terminal QRS complex to identify patients with ventricular tachycardia after myocardial infarction,” Circulation, vol. 64, p. 235, 1981. “Voluntary standard for diagnostic electrocardiographic devices,” ANSI/AAMI EC11a, Arlington, Va.: Associ- ation for the Advancement of Medical Instrumentation, 1984. A.D. Waller, “On the electromotive changes connected with the beat of the mammalian heart, and the human heart in particular,” Phil. Trans. B., vol. 180, p. 169, 1889. F.N. Wilson, F.S. Johnston, and I.G.W. Hill, “The interpretation of the galvanometric curves obtained when one electrode is distant from the heart and the other near or in contact with the ventricular surface,” Amer. Heart J., vol. 10, p. 176, 1934. Further Information Comprehensive Electrocardiology: Theory and Practice in Health and Disease, Volumes 1–3, P. W. Macfarlane and T.D. Veitch Lawrie, Eds., England: Pergamon Press, 1989. High-Resolution Electrocardiography, M.D. Nabil El-Sherif and M.D. Gioia Turitto, Eds., Mount Kisco, N.Y.: Futura Publishing Company, 1992. Medical Instrumentation: Application and Design, 2nd ed., J. G. Webster, Ed., Boston: Houghton Mifflin, 1992. 115.3 Pacemakers/Implantable Defibrillators Philip L. Johnson and William M. Smith The heart is an amazing machine. Throughout an average lifetime, it contracts over 2.5 billion times to pump blood throughout the body. Without its proper function, an individual will die within minutes. The heart consists of four chambers. The upper two chambers, the atria, are used as primers for the lower two chambers, the ventricles, which serve as the main pump. Blood delivery will be inefficient if the atria and ventricles do not pump in mechanical synchrony (AV synchrony). Optimum efficiency occurs when the atria contract slightly before the ventricles. Electrical depolarization waves are responsible for controlling the contractions of the heart and thus maintaining AV synchrony. The depolarization waves originate from a specialized set of cells, known as the sinus node, that are modulated by neural input and are located in the top of the right atrium. The sinus ? 2000 by CRC Press LLC node is the heart’s natural pacemaker. It is part of the atrioventricular (AV) conduction system, which serves to distribute the wavefronts throughout the heart and to connect the otherwise electrically isolated atria and ventricles. A normal depolarization wave spreads across the atria causing them to contract first, and then, after a brief delay while traversing the AV conduction system, across the ventricles causing them to contract shortly thereafter. With such a demanding, complex organ, it is no wonder there are multiple ways by which it can fail. Failures in the electrical system, known as arrhythmias, may impair the contraction sequence and com- promise blood flow. These failures are often the result of some underlying heart disease, but may also have a genetic etiology. While antiarrhythmic drugs have been available for some time, contemporary treatment of arrhythmias relies heavily on two types of implantable medical devices: pacemakers and implantable cardio- verter defibrillators. Bradyarrhythmias Bradyarrhythmias are defined as heart rates that are abnormally slow (<60 b.p.m.) [Katz, 1992]. They are generally caused by either sinus node disease or AV conduction disorders. In the former, disease of the body’s natural pacemaker cells often results in an unnaturally slow heart rate and significant patient discomfort. Also, the heart rate may not increase in response to exercise due to a loss of neural control of the sinus node, which will inhibit the patient from performing strenuous activities; this is known as chronotropic incompetence. AV conduction disease results from a pathology of the cells that electrically connect the atria and the ventricles. This can result in inefficient blood delivery due to a loss of AV synchrony. Pacemakers are commonly used to attempt to restore a natural heart rate, AV synchrony, and chronotropic competence in patients with these and other diseases. Approximately 115,000 pacemakers are implanted in the U.S. every year [Ellenbogen, 1996]. Tachyarrhythmias Tachyarrhythmias are generally defined as heart rates that are abnormally or inappropriately fast (>100 b.p.m.) [Katz, 1992]. There are many different types of tachyarrhythmias (tachycardias). They are caused by “runaway” depolarization wavefronts that may continue to rapidly activate the same tissue over and over again by a process known as reentry. This can result from a number of underlying physiological problems, such as dying tissue with altered conduction properties due to a blocked coronary artery, around which the wavefront can propagate. The atria and the ventricles can both experience tachycardia. Although somewhat debilitating, atrial tachycardias are not immediately life threatening; however, most ventricular tachycardias are life threatening. The most serious ventricular tachycardia, ventricular fibrillation, has been defined as the rapid, disorganized, and asyn- chronous contraction of ventricular muscle [Epstein and Ideker, 1995] during which the heart’s ability to distribute blood to the body is completely compromised. If not immediately treated with a defibrillating electrical shock, loss of life will follow in only minutes. Ventricular fibrillation is the most common cause of sudden cardiac death, of which nearly 400,000 people die annually in the U.S. alone [Gillum, 1989]. The Implantable Cardioverter Defibrillator (ICD) was developed in an attempt to terminate ventricular fibrillation and prevent sudden death from occurring. Pacemakers The pacemaker is a medical device capable of controlling the heart rate through a set of implanted electrode leads (Fig. 115.11). The first devices in 1958 were simple, fixed rate oscillators controlled by two transistors [Elmquist and Senning, 1960]. Weighing more than 180 grams, they had a lifetime of around 3 years and paced only the ventricles [Sanders and Lee, 1996]. More sophisticated dual-chamber pacemakers, which sense and pace both the atria and the ventricles independently, were introduced in the late 1970s [Funke, 1982]. Modern pacemakers have shrunk to less than 15 grams (6 ccs) and evolved into sophisticated, implantable computers capable of complex pacing algorithms, telemetry, extensive diagnostics, data storage, and a lifetime greater than 5 years [Sanders and Lee, 1996]. Clinical Indications Pacemakers are generally indicated for three major disorders, including sinus node disease, AV conduction system disorders, and certain atrial tachyarrhythmias, as well as other less-common pathologies. There are varying degrees of each of these disorders, and the severity of the symptoms and age of the patient may suggest ? 2000 by CRC Press LLC if a pacemaker is warranted. Symptoms may include syncope, dizziness, seizures, heart failure, depression, and dementia. Although bradycardia accounts for the majority of implantations, pacemakers can be used to treat some tachycardias. Antitachycardia pacing is a special type of pacing that may be indicated for atrial or ventricular tachycardias; however, when managing ventricular tachycardias, an ICD is preferable to a pacemaker so that, if the tachycardia degenerates into ventricular fibrillation, it can be halted with a defibrillation shock. Pacemaker implants may be permanent or temporary. Surgery A pacemaker implant is a fairly standard procedure in which the pacing leads are inserted intravenously into the heart and the pacemaker is implanted subcutaneously in a pocket on the chest just below the clavicle. A number of veins have been used to implant the leads, but the cephalic and the subclavian veins are the most common. Access to either of these veins is obtained from the same incision that is used to form the pocket on the chest. A number of tools, such as guidewires, dilator sheaths, and fluoroscopy (real time X-ray), are used to work the lead down the veins into position in the right side of the heart. Depending on the type of pacemaker used, one or two leads may be implanted. In dual-chamber pacemakers, a lead is required in both the ventricles and the atria. Figure 115.12 shows a diagram of a typical dual-chamber system in the body. The right ventricular lead is usually implanted first and is advanced to the right ventricular apex, where it is fixed. The atrial lead is then advanced to the right atrial appendage, where it is also fixed. A pacing system analyzer is used to test for sufficient electrode contact by measuring impedance, pacing thresholds, and sensing thresholds. The pacing threshold is the smallest charge necessary to stimulate cardiac tissue and initiate a depolarizing wavefront. The Figure 115.11 The Guidant/CPI Discovery family of pacemakers. (Courtesy of Guidant/CPI, St. Paul, MN.) ? 2000 by CRC Press LLC sensing threshold is the smallest acceptable amplitude and slew rate of a sensed cardiac signal. Electrode repositioning may be required if efficient pacing and sensing are not realized. The leads are then connected to the pacemaker, sutured in place to prevent significant pacemaker movement, and the pocket is closed. Pacemaker programming can be accomplished through a wireless telemetry system. Design The pacemaker must be able to reliably detect the wavefronts on the atria and the ventricles, determine their rate, determine if the chambers are synchronized, and intervene in the appropriate chamber if these conditions are not met. Some patients may receive continuous pacing, while others may rarely need pacing at all. The pacemaker must perform over a period of greater than 5 years, in all types of environments, with maximum reliability, and using the smallest, most comfortable design possible. In addition, it must be adaptable to various types of patients through remotely programmable parameters and informative diagnostics. A pacemaker system is made up of three major components, including the leads, the pulse generator, and the programmer. The pulse generator (i.e., the pacemaker itself) houses all of the controlling and pacing electronics in a biologically compatible titanium shell. The leads provide the electrical link from the pulse generator to the heart. The programmer allows the physician to remotely program the pacemaker parameters and assess pacemaker func- tion through a telemetry system. Programmer The programmer provides a bi-directional wireless link to the implanted pacemaker through a telemetry wand. The physician can program parameters that customize functions, check battery capacity, send functional commands to the pacemaker, check for past events, and monitor realtime or stored electrograms measured from the implanted electrodes. Diagnostics provide an invaluable tool for the physician to assess patient welfare, and a well-designed programmer will effectively display this information. Because of the overwhelming avail- ability of parameters and functions built into modern pacemakers, the programmer is an essential part of the pacemaker system. The major problem associated with current telemetry systems is the speed of data transfer. Transfer speeds will need to be significantly increased in the future to keep pace with the increasing number of available diagnostics. Figure 115.12 Placement of a dual-chamber pacemaker in the body and the leads in heart. The right atrial (RA) electrode is placed in the right atrial appendage and the right ventricular (RV) electrode is placed in the right ventricular apex. Both electrodes are capable of pacing and sensing. ? 2000 by CRC Press LLC Leads The purpose of the pacemaker lead(s) is to provide a link between the pulse generator and the cardiac tissue in order to efficiently sense and stimulate the heart. The majority of leads are inserted intravenously and attached to the inside of the heart. Modern leads are composed of five major parts, including the connector, conductor(s), insulation, electrode(s), and a fixation mechanism [Kay, 1996]. The electrode design is critical to minimizing current drain during pacing while ensuring reliable sensing. It has been shown that the pacing threshold is a function of the current density at the electrode [Stokes and Bornzin, 1985]. Minimizing the radius of the electrode will maximize current density and therefore reduce the pacing threshold and current drain. In addition, a small radius will increase the electrode resistance, which also helps to reduce current drain. Con- versely, it has been shown that a large electrode surface area decreases sensing impedance and electrode polarization [Kay, 1996]. Electrode polarization is caused by a buildup of charge on the cardiac tissue after a stimulation pulse and can affect the electrode’s ability to sense properly. The ideal electrode would, therefore, minimize the radius while maximizing the surface area [Sinnaeve et al., 1987]. This has been accomplished by building electrodes with a small radius but with complex microscopic mesh or porous structure to maximize surface area [Kay, 1996; Bornzin et al., 1983]. Electrode material varies among manufacturers, but is often comprised of a platinum alloy. The conductor serves as the electrical pathway between the pulse generator and the electrodes. It must have low impedance and be able to withstand repeated flexing due to heart motion. This is generally accomplished using a nickel alloy and coiling the wire to resist stress. There will be one conductor in the lead for unipolar configurations and two conductors for bipolar. Unipolar leads use a single electrode at the tip of the lead with the reference being the metal shell of the pacemaker. Bipolar leads use two closely spaced electrodes near the tip of the lead, which is helpful in rejecting external noise when sensing. The conductor must be insulated by a material that can withstand flexing and is resistant to harsh biological conditions. Silicone rubber and polyurethane are two commonly used insulating materials. The connector pin provides a physical link between the lead and the pulse generator and was standardized by an international meeting of manufacturers to avoid confusion when mixing brands of leads and pacemakers [Calfee and Saulson, 1986]. Finally, the fixation mechanism is responsible for holding the lead in place on the heart. The two major types of fixation are known as active and passive fixation. The most common active mechanism involves a helical screw that is advanced into the tissue [Markewitz et al., 1988]. The most common passive mechanism uses flexible tines that become entrapped in the cardiac tissue [Furman et al., 1979]. Fibrous tissue often grows around the fixation mechanism due to tissue injury. This further stabilizes the lead but can cause pacing thresholds to increase over time. A good fixation mechanism will minimize tissue injury while ensuring a stable anchor. Some designs use a steroid-eluting electrode to minimize and stabilize fibrous tissue growth [Timmis et al., 1983]. Pulse Generator The primary function of the pulse generator is to interpret the information gained from the atrial and ventricular electrodes and other sensors to determine if the patient requires pacing and to deliver the pacing pulses, if necessary. A number of sophisticated algorithms are applied by the pacemaker to determine when pacing is necessary. The pulse generator also performs several secondary functions, such as telemetry and diagnostics. A hermetically sealed titanium shell is used to house the electronics of the pulse generator because titanium is a strong, lightweight metal that is biocompatible with human tissue and does not corrode. Figure 115.13 shows a block diagram of a typical pacemaker pulse generator. A battery supplies the power for the electronics as well as for the pacing pulses. The sensing circuitry is used to amplify the electrical signals that are present on the heart in order to determine the heart rate. If rate-adaptive sensors are used (discussed below), additional sensing circuitry is needed. The output circuitry generates the pacing pulses by storing a charge on a capacitor so it can be delivered to the heart on demand. The backup pacing circuit is capable of asynchronously pacing either the atria or the ventricles at a preprogrammed rate in the event of excess noise. It also serves as a rate-limiting protection circuit to prevent the heart from being paced at an excessive rate due to a main system failure. The pacing control consists of timing circuitry and logic sections; it is responsible for interpreting sensed data and reacting with an appropriate pacing response. A microprocessor can be integrated with the pacing control and is often utilized as an overall system control to allow for flexibility of design. Memory in the form of RAM and ROM is required to store the microprocessor program, pacing parameters, and diagnostics. The telemetry ? 2000 by CRC Press LLC circuit is capable of swapping information with the programmer and is activated by a magnetic reed switch. The above circuits are currently built using CMOS integrated circuits, VLSI design, and hybrid technology. Continuing advances in the electronics industry will allow for further size reductions. Battery The battery for a pacemaker must be a safe and reliable energy source with a high energy density capable of supplying several microamperes for longer than 5 years. In addition, it must be possible to reliably predict the end of its life so that the pulse generator can be replaced before pacing fails. Because the battery accounts for the majority of pacemaker volume, these requirements must be met while ensuring that the battery is as small as possible. Nearly all modern pacemakers use lithium-iodine technology [Sanders and Lee, 1996]. This battery has a high energy density and a low internal self-discharge, which combine to give a longer lifetime than past batteries. Lithium serves as the anode, iodine combined with poly-2-vinyl pyridine serves as the cathode, and a semisolid layer of lithium iodide serves as the electrolyte. The cell is hermetically sealed to prevent corrosion. A new battery produces 2.8 V and declines linearly to 2.4 V near the end of its life. Either multiple batteries in series or voltage multipliers can be used to achieve voltages greater than 2.8 V. The status of the battery can be determined by several methods and can be telemetered out to the physician. The current drain of the device ultimately determines the lifetime of the battery and is dependent on many factors, such as circuit operating current, electrode impedance, and frequency, duration, and amplitude of the output pulses. Significant improve- ments in these areas have enabled the pacemaker battery to shrink in size. Amplifier Sense System The purpose of the amplifier sense system is to reliably detect the rate of electrical activity of the heart so that the pacemaker can determine if there is a need for therapy. The amplifier must have a high input impedance to ensure adequate signal amplitude. The system must reject all forms of external noise so that no inappropriate counting occurs. The electrical activity on the heart is manifested as deflections in an electrogram measured Figure 115.13 Block diagram of typical dual-chamber pacemaker components. ? 2000 by CRC Press LLC by the sense electrodes. There are multiple deflections in an electrogram during one heartbeat due to near- field and far-field cardiac activity. In a ventricular electrogram, the far-field activity is associated with the depolarization of the atria (P waves), and the near-field activity is associated with the depolarization (R waves) and repolarization (T waves) of the ventricles. The R waves, which are the largest and fastest deflections, indicate a depolarization wavefront located directly below the sensing electrode and are, therefore, used as a means to count the heart rate. Circuitry used to sense R waves usually involves voltage comparators and slew rate detectors. Since the R wave deflection normally has a larger amplitude and faster slew rate than the other deflections, the circuitry can use reference thresholds to detect the R waves. The thresholds may be dynamic and determined by a complicated algorithm that has been termed autosensing [Jacobson and Kroiss, 1996; Castro et al., 1996; Kim, 1998]. In addition, there may be a brief amplifier blanking period after an R wave is detected during which all further deflections are ignored (so as to avoid potential inappropriate sensing of following T waves). In dual-chamber pacemakers, it is necessary to measure an atrial electrogram in addition to the ventricular electrogram. Similar detection strategies are employed in the atria; however, due to the large mass of the ventricles compared to the atria, the far-field ventricular activity in an atrial electrogram is more difficult to reject. A sensed cardiac event is marked by a digital pulse that is input into the timing circuitry. The timing circuitry determines if the heart rate is too slow and also controls all amplifier blanking periods and pacing rates. In addition to far-field effects, many other sources of external noise exist in an electrogram, such as motion artifact, electrode polarization, noise from the skeletal muscles, and environmental noise (e.g., 60-Hz noise and cellular phones). Using bandpass filters and closely spaced bipolar electrodes minimizes external noise. In the event of extreme interference, backup pacing circuits can assume control and asynchronously pace the heart until the noise is gone. Output Circuitry The output circuitry is responsible for delivering the pacing pulses through the electrodes to the cardiac tissue in order to artificially control (capture) the heart. Because the amount of energy needed for capture can vary over the lifetime of a patient (e.g., due to changing electrode impedance or position) and between different patients, it is necessary to be able to deliver a controlled amount of energy per pacing pulse. Many pacemakers use an output voltage much higher than the pacing threshold to ensure capture for every pulse; however, this may unnecessarily waste energy. The minimum reliable pulse voltage necessary for capture is desired in order to minimize the current drain on the battery. Autocapture algorithms are capable of monitoring every pulse for capture and adjusting the output voltage to the minimum necessary value on demand [Jones et al., 1999]. This feature will become more prevalent in future pacemakers. Timing circuitry and output amplifiers are used to control the frequency, pulse width, and amplitude of the stimuli. Capacitors controlled by electronic switches physically deliver the energy. The capacitors are charged by the battery up to the desired voltage in between pacing pulses and then discharged into the heart by the switches at the proper timing. Voltage multipliers can be used to double or triple the battery voltage if necessary. Rate Adaptive Pacing An important feature of modern pacemakers is their ability to modulate the heart rate based on the metabolic needs of the body. There are many conditions that call for heart rate modulation, such as exercise, fever, stress, or sleep. Because it is under neural control, the ideal rate modulator is a normally functioning sinus node. In the case that the patient has atrioventricular conduction problems, the sinus node and the atria may still be functional; therefore, atrial sensing may be used by the pacemaker to regulate the ventricular rate. Patients with sinus node disease or atrial arrhythmias, however, require an artificial means of regulating heart rate. Many attempts have been made to design metabolic sensors that can be used to naturally modulate the heart rate. Control variables tested include blood pH [Cammilli, 1977], blood temperature [Alt et al., 1986], venous blood oxygen saturation [Eityzgrlf et al., 1982], respiratory rate [Rossi et al., 1983], minute ventilation [Alt et al., 1987], vibration [Anderson et al., 1983], acceleration [Matula et al., 1992], right ventricular pressure [Yee and Bennett, 1995], and QT interval [Donaldson and Rickards, 1983]. The ideal sensor would be reliable, have low current drain, require no additional surgery, and accurately reflect metabolic needs. All of the above sensors could potentially be used to reflect metabolic needs; however, only a few are currently practical and in use. The most common sensors in pacemakers today are the activity sensors, which attempt to indicate activity level by ? 2000 by CRC Press LLC transducing vibration and acceleration. Piezoelectric crystals functioning as strain gauges can be mounted to the inside of the pacemaker to detect mechanical vibrations [Anderson and Moore, 1986]. In addition, accel- erometers can be mounted directly to the hybrid circuit to detect acceleration [Kay, 1996]. Although both of these sensors are subject to motion that may not be due to exercise, accelerometers have generally proven to be more proportional to exercise than strain gages [Kay, 1996]. Another sensor that has been successfully integrated into pacemakers is the minute ventilation sensor. Minute ventilation is representative of the amount of air a person breathes and can be estimated by measuring the transthoracic impedance over time. This can be accomplished by emitting a train of very small pulses from one pole of a bipolar pacing electrode and measuring the voltage between the other pole and the pacemaker can [Nappholz et al., 1986]. The impedance calculated from the measured voltage rises when a person breathes in and falls when the person breathes out. Combining activity sensors with minute ventilation sensors has proven to be a clinically successful approach to estimating metabolic needs during exercise [Alt et al., 1995]. Future There are many exciting advancements currently under development for pacemakers. Efforts are being concentrated on making devices more sophisticated but less complicated [Jones et al., 1999]. One approach to realizing these goals is to expand the number and ability of the automatic features, such as autosense and autocapture algorithms. Pacemakers are trending toward automatic self-optimization abilities that are based on the individual patient’s needs. Improvements in rate adaptive sensors and algorithms are key to realizing this goal and represent an area of significant research. In the future, integrated circuit technology will allow for much smaller and more efficient designs. This, in addition to improved battery technology, will allow for increased memory, advanced signal processing, and faster telemetry. More memory will permit more extensive patient diagnostics. These diagnostics will be displayed on advanced programmer interfaces that are more clinically relevant. Pacemaker lead research may ultimately yield single-pass leads, capable of pacing and sensing in both the atria and the ventricles. Finally, nontraditional pacemaker uses are currently being explored. A major area of interest is in the treatment of congestive heart failure (CHF). CHF is a debilitating and deadly disease characterized by an enlarged heart that is incapable of pumping adequate blood to the body. It is possible that the size of the heart facilitates an electromechanical asynchrony between the ventricles. Recent studies have shown that appropriately positioned and timed pacing stimuli can help to improve cardiac output by synchronizing the ventricles [Foster et al., 1994; Bakker et al., 1994]. Implantable Cardioverter Defibrillators The Implantable Cardioverter Defibrillator (ICD) was first conceived of by Dr. Michael Mirowski in the mid- 1960s [Mirowski et al., 1970]. He imagined a device that would continuously monitor the hearts of high-risk individuals for life-threatening arrhythmias and intervene by electrical shock to restore normal sinus rhythm. Just over a decade later, the first patient was successfully implanted with an ICD [Mirowski et al., 1980]. From there, rapid development ensued, and FDA approval was obtained in 1985, at which time Cardiac Pacemakers Incorpo- rated (CPI) took over the marketing and development of the ICD. Since then a number of competitors have arisen, and the ICD has evolved into a remarkably sophisticated medical device capable of bradycardia and antitachycardia pacing, low-energy cardioversion, high-energy defibrillation shocks, and extensive diagnostics (Fig. 115.14). Clinical Indications Initially, ICDs were only indicated for certain patients that had survived an episode of cardiac arrest. Time of intervention is critical to survival of cardiac arrest; only 25% of people that have an episode are successfully resuscitated by first responders [Shuster and Keller, 1993]. ICDs have decreased the first-year mortality rate of these survivors from 30 to 2% [Winkle et al., 1991]. More recently, in addition to cardiac arrest survivors, patients deemed at risk for a first arrest due to sustained ventricular tachyarrhythmias are receiving ICDs as well [Saksena et al., 1996]. In the future, patients with more subtle predictors of sudden death may be indicated for ICDs. Surgery The modern ICD is now implanted similarly to a pacemaker. In the past, it was necessary to open the chest (i.e., thoracotomy) in order to suture large patch electrodes directly onto the ventricles and situate the ICD abdominally. Patch electrodes were required to achieve a low defibrillation threshold (DFT), which is a measure ? 2000 by CRC Press LLC of the amount of energy needed to reliably terminate fibrillation (defibrillate). Due to significant advances in the size and energy efficiency of ICD systems, it is now possible to implant the ICD pectorally and insert specially developed electrode leads into the heart intravenously without the need for a thoracotomy. This ensures greater patient comfort and significantly reduces the risk associated with surgery. There are several different lead configurations available, depending on the manufacturer, each with its own advantages. All of these systems place one electrode in the right ventricular apex, while the position of the return electrode varies between systems. Figure 115.15 shows a typical system layout in the body. The entire surgery can be performed through a single incision, using only local anesthetics and heavy sedation, and is similar to that of the pacemaker. Lead positioning is critical to obtaining low DFTs [Lang et al., 1995; Usui et al., 1995]. Once the lead(s) are in place, the ICD is tested by inducing fibrillation by artificial means and then giving a shock of known energy to halt the arrhythmia. The DFT can be determined by a number of different methods, such as by decreasing the energy of each successive shock until the defibrillation attempt is unsuccessful. The DFT must be well below the maximum output energy of the device before a successful implant is declared; a 10-joule safety margin is typically used [Moss et al., 1996]. If an adequate safety margin cannot be obtained through optimal electrode placement, additional electrodes may be required in order to obtain an acceptable DFT. In extreme cases, a thoracotomy may still be required. It is also necessary to thoroughly test the pacing/sensing characteristics as with pacemakers. Figure 115.14 The Guidant/CPI Ventak Mini IV ICD with the Endotak lead system attached (a single-pass lead). The Mini IV is a single chamber defibrillator. (Courtesy of Guidant/CPI, St. Paul, MN.) ? 2000 by CRC Press LLC Design An ICD system consists of three main components: the programmer, the leads, and the pulse generator. The programmer provides a link to the ICD after it has been implanted and is similar to the pacemaker programmer. The leads deliver the energy from the ICD to the heart for defibrillation, as well as provide for pacing and sensing capabilities. The ICD pulse generator is an ever-evolving technology capable of delivering sophisticated cardiac rhythm management to the patient and diagnostics to the physician. As technology advances, ICDs will become smaller, more reliable, and more versatile. Leads The leads provide the means by which to deliver the energy of the defibrillation shock from the pulse generator to the heart, as well as pacing and sensing capabilities. They are insulated with either medical-grade silicone rubber or polyurethane, except at the electrodes. A major disadvantage of these leads is that they deliver current in a largely nonuniform manner as compared to the older patch electrodes, which results in higher energy requirements for defibrillation. Fortunately, advances in battery, capacitor, circuit, and waveform technology have compensated for this greater energy requirement while still allowing the ICDs to become smaller. Because electrode position is important, many different electrode shocking configurations have been attempted. One of the most popular is to place one electrode in the right ventricle (RV) and the return electrode in the superior vena cava (SVC). In addition to defibrillation electrodes, there must be pacing and sensing electrodes as well. Currently, three major configurations exist to accommodate these requirements. The first, known as the single pass lead, integrates two defibrillation coil electrodes (RV, SVC) and ventricular pace/sense electrodes onto a single lead. The second configuration consists of one lead, which contains the RV defibrillation electrode and the ventricular pace/sense electrodes, and a second lead, which contains the return defibrillation electrode. The third configuration consists of one lead containing both defibrillation electrodes and a separate lead containing the ventricular sense/pace electrodes. Some systems use the titanium housing of the ICD as an additional return electrode in the “active can” configuration in an attempt to distribute the current more evenly throughout the ventricles. Also, the most modern ICD systems utilize atrial pace/sense electrodes in addition to ventricular electrodes to achieve dual-chamber pacing and sensing capabilities. Another requirement of the defibrillation Figure 115.15 Placement of an ICD in the body and a single-pass lead system in the heart. The right ventricular (RV) and the superior vena cava (SVC) electrodes serve as the defibrillation electrodes, and there are pace/sense electrodes at the tip of the lead in the right ventricle. ? 2000 by CRC Press LLC lead is that it have very low impedance due to the large defibrillation currents, but can still withstand repeated flexing due to millions of heartbeats. This is achieved using a combination of high-strength, low-impedance metals. The pace/sense electrodes are similar in design to those used by pacemakers. The leads are fixed in place by either a screw-in mechanism or flexible tines. Pulse Generator The pulse generator is the core of the ICD system. It consists of the batteries, capacitors, and accompanying electronics enclosed in a hermetically sealed titanium can. The can may be used as a return electrode. A header, typically made of epoxy, is attached to the can and provides the link from the electronics to the leads via silicone- sealed ceramic feedthroughs. The size of the pulse generator has steadily declined since the introduction of the ICD and is currently around 40 cc. Significant efforts are underway to further reduce the size of the pulse generator in order to increase patient comfort. As research in defibrillation progresses, more efficient defibril- lation strategies will no doubt be developed and allow the size to be decreased further. Currently, the major barriers to size reduction are the battery and capacitor size required to create waveforms capable of ventricular defibrillation. In addition, designing the electronics for a system that can measure cardiac signals on the order of 100 μV and produce high-energy waveforms on the order of 750 V and 40 A, all within the same small space, presents a significant engineering challenge. A problem associated with this includes high-voltage arcing among internal components. To prevent this, nitrogen gas is sealed inside the can because of its high breakdown voltage barrier. Figure 115.16 shows a block diagram of the key components of a typical ICD pulse generator. The brain of the ICD is the microprocessor. Most ICD manufacturers use industry-standard microprocessors, such as the Z80, 6502, or 8852, to control the ICD [Warren et al., 1996]. In order to conserve energy, it is desirable to put the microprocessor into a sleep mode as often as possible and to wake it only when necessary, such as when an arrhythmia is suspected. To accomplish this, many of the monitoring and pacing functions are implemented using analog and digital circuits. Modern ICD designs have reduced the size of the circuitry to a small number of integrated circuits on a hybrid chip [Warren et al., 1996]. Figure 115.16 Block diagram of typical ICD components. ? 2000 by CRC Press LLC Memory is required to store the program and individual patient parameters for the operation of the ICD. Startup code and some of the main program is often stored in ROM, while the remaining program, parameters, patient diagnostics (e.g., electrograms), and event markers are stored in RAM. Clinicians and investigators are increasingly interested in diagnostics obtained from the ICD, which warrants future memory increases. Additional circuitry includes support for the microprocessor, timers, the telemetry interface, low-voltage power supplies, the high-voltage system, the pacing control, the defibrillation control, and isolation and external protection circuits. The telemetry interface is the link to the external programmer and consists of a coil, which serves as the antenna, support circuitry, and a magnetic reed activator switch. The low-voltage supplies power the analog and digital circuitry as well as the pacing pulses. The defibrillation control determines when defibrillation is necessary and controls the process of defibrillation. The high-voltage system is used to generate the defibrillation shocks and consists of high-current batteries, capacitors, a fly-back transformer, and output switching circuits. The pacing control includes the circuitry to deliver pacing pulses, to interpret the signals from the sense amplifiers, and timers that monitor the current heart rate and wake the microprocessor if necessary. The isolation and protection circuits provide protection against external defibrillation attempts and external noise. Amplifier The purpose of the amplifier sense system is to reliably detect the rate of electrical activity on the heart so that the ICD can determine if there is a need for intervention. The amplifier must be immune to noise and be able to quickly respond to a large range of heart rates (30 to 360 b.p.m.) [Warren et al., 1996]. In order to obtain an accurate heart rate, it is desirable to digitally count R waves, while rejecting all other electrical activity and noise. Since the R wave is much larger in amplitude than the other waves in the electrogram, a simple method to detect them would be to use a comparator with a set threshold. Unfortunately, the amplitude of R waves is not constant; therefore, a simple comparator circuit is not reliable. For example, during tachycardias and fibrillation, the amplitude of the signals decreases significantly. One technique commonly used to solve this is to use a dynamically adjusting comparator threshold, in which the threshold level exponentially decreases over time until the next R wave is detected, at which time, the threshold level is reset [Brumwell et al., 1996]. This ensures that low- amplitude signals will be detected. Another technique is to use an automatic gain control to slowly increase the gain between detected R waves, while keeping the threshold constant [Brumwell et al., 1996]. To avoid double counting caused by undesired T-wave detection, there is, typically, a brief period of time after each detected R wave during which sensed signals are ignored. Detection schemes are often implemented with carefully designed analog chips, known as Application Specific Integrated Circuits (ASICs), in order to keep the current drain on the order of 10 μA [Warren et al., 1996]. An advantage of these chips is that they provide near-perfect component matching, which is critical in engineering predictable gain control and frequency response. Battery The design for an ICD battery has many stringent requirements and presents a unique challenge to the engineer. While the pacemaker battery is optimized for high energy density, a defibrillator battery must sacrifice some energy density for high current capability. An average defibrillator battery must be able to supply a steady background current of 10 to 20 μA for at least 5 years for monitoring and pacing functions, as well as, provide around 200, 2-A pulses for 10 to 15 seconds each in order to charge the capacitors for multiple defibrillation shocks [Holmes, 1996]. All of these criteria must be met while minimizing the size of the battery and maximizing its safety and reliability. In addition, it is necessary to be able to reliably predict the end of life of the battery. Nearly all modern ICDs use lithium silver vanadium oxide battery technology to accomplish these requirements [Liang et al., 1982]. Lithium pressed into a nickel current collector serves as the anode, and silver vanadium oxide serves as the active cathode. The electrolyte is generally a lithium salt dissolved in a mixed organic solvent. This produces approximately 3.2 V, but two cells are often connected in series to give around 6 V. Some newer systems are using single battery technology. A large electrode surface area and low internal impedance are required to achieve the high current pulses [Holmes, 1996]. This is accomplished by folding the anode in an accordion-like fashion and placing cathode plates in the folds. A disadvantage of this battery is that it exhibits a phenomenon in its mid-life known as voltage delay, in which the voltage goes low in the first second or two during a charging pulse. This can cause a prolongation of the capacitor charging time, which can be dangerous to the patient. It is due to an initial high resistance caused by a chemical buildup on the cathode in the cell ? 2000 by CRC Press LLC and can be alleviated by periodic pulsing of the batteries into the capacitors and internally dumping the charge. This shortens the life of the battery slightly but is not a total waste because the capacitors need this type of reforming as well. Several methods can be used to predict the end of life of a battery. Common indicators are the battery’s open-circuit voltage, voltage during charging, and the time it takes to charge the capacitor [Holmes, 1996]. Future advancements in battery technology are critical in reducing defibrillator size and increasing longevity. Charging Circuit In order for a defibrillation shock to occur, it is necessary to convert the 6 V from the battery to an output voltage of up to 750 V to be stored across a capacitor. This is generally done with a dc/dc converter, or inverter. Unique design considerations include the large size of the conversion, a demand for high efficiency, and a minimized transformer and circuit size. The circuit includes the battery and a low-voltage, high-current switch at the input, a fly-back transformer, and a rectifying diode and storage capacitor at the output. A controlling oscillator typically operates the switch between 30 and 60 kHz for high efficiency [Warren et al., 1996]. This creates a simulated ac current, which is converted by the transformer to the higher voltage. The current in the input stage is, typically, around 2 A. The rectifier diode prevents current from flowing back into the secondary winding of the transformer. The voltage on the capacitor increases as a function of the square root of the time that the oscillator is on and typically reaches full capacity in 10 to 15 seconds [Warren et al., 1996]. It is possible for the transformer to be small because of the high-speed switching. In addition, the diameter of the core and the wire windings in the fly-back transformer can be small because the converter is only used intermittently to charge the capacitor, which allows for ease of heat dissipation compared to continuous conversion [Bach and Monroe, 1996]. There is a trade-off between the size of the converter and the efficiency. The high clock rates allow the transformer to be small, but also introduce losses in energy due to hysteresis in the coils. A typical ICD charging circuit achieves about 75% efficiency [Bach and Monroe, 1996]. This plays a major role in the charging time and, therefore, the delay before shocking therapy can be delivered. Capacitor The function of the capacitor is to store the energy generated from the high-voltage charging circuitry and to deliver that energy on demand to the heart over a few milliseconds. The commercially available aluminum electrolytic photoflash capacitor is currently used in the ICD. This is because of its high energy density of 1.7 J/cm 3 made possible due to special etching techniques that maximize surface area in the aluminum foil. Because the capacitors are commercially available and are not custom-designed for defibrillators, there are several shortcomings that ICD manufacturers must deal with. Since the highest energy density aluminum electrolytic capacitors are designed to operate at around 375 V, two of these capacitors must be used in series to attain the necessary 750 V used in most designs; therefore, the capacitors play a major role in determining the size of the defibrillator. Their geometry is not ideal for efficient packaging. The round shape results in wasted space in the ICD. Perhaps the greatest shortcoming is the requirement that the capacitor be reformed after periods of no use to ensure there is no leakage current during charging. This involves automatic application of the rated voltage to the capacitor for a few minutes every few weeks in order to repair damage that has occurred due to aging. The latest generation capacitors have minimized the need for reforming. As mentioned before, because the battery requires reforming as well, the energy is not completely wasted. Current research in capacitors is centered on reducing the size by increasing the energy density and eliminating the need for reforming. Waveform and Output Switching Optimizing the defibrillation waveform has been the subject of much investigation. Because a capacitor is used to deliver the energy, the waveform is some form of a decaying exponential. Still, there are many parameters that can be varied when creating a waveform, such as the pulse width, amplitude, decay, and the polarity and number of phases. In the past, monophasic, truncated, exponential waveforms were common. More recently, biphasic waveforms, in which the direction of current is reversed at some point during the waveform by switching the electrode polarity, have become more popular. This is due to the work of Schuder and others, who have shown that biphasic waveforms defibrillate with less energy than monophasic waveforms [Schuder et al., 1984; Feeser et al., 1990]. ? 2000 by CRC Press LLC Output switching circuitry is needed in order to time and create these waveforms. Since the load impedance of the defibrillation system ranges from 20 to 70 ?, it is possible to have peak currents of 40 A in the output circuit. There is, therefore, a need for high-power electronic switches to carry the current, such as Silicon Controlled Rectifiers (SCRs), Metal Oxide Semiconductor Field Effect Transistors (MOSFETs), and Insulated Gate Bipolar Transistors (IGBTs). In addition, these switches must be mounted so that there are very low junctional resistances to minimize power loss. MOSFETs and IGBTs are often used in bridge circuits to facilitate switching of biphasic waveforms. SCRs were used mostly in generating monophasic waveforms, but are still used in some biphasic waveform circuits. Because these types of switches require around 15 V for the control, it is necessary to use a low-power dc/dc converter to boost the 6 V from the battery. Timing of the switching is either controlled by timing circuitry, or by voltage monitoring circuitry that causes a polarity reversal to occur when the voltage falls to a certain threshold. Future In the future, ICDs will continue to evolve into increasingly sophisticated cardiac rhythm management devices. The ICD is no longer simply a safeguard against ventricular fibrillation; it is, moreover, being called to better manage bradycardias and tachycardias, and may eventually be used to predict and prevent arrhythmias from ever occurring. Management of atrial arrhythmias is also an important emerging frontier. To facilitate these and other demands, a number of advancements are currently being explored [Morris et al., 1999]. New lead systems will be smaller, more reliable, and better designed to manage atrial arrhythmias. Additional sensors may be incorporated, such as pressure transducers to measure hemodynamic stability. More complex rhythm discrimination algorithms will be developed to ensure that appropriate therapy is given at the appropriate time. Advances in electronics, battery, and capacitor technology will allow the ICD to continue to shrink in size. Diagnostic capabilities will be expanded due to increases in memory, and efforts will be made to dramatically decrease the complexity of programming the ICD. Finally, basic research will produce important advances in arrhythmia management, the significance of which cannot yet be imagined. Defining Terms Arrhythmia: A general term referring to a disorder in the electrical system of the heart. Atria: The upper two chambers in the heart that act as primer pumps for the ventricles. AV synchrony: The timing that must be maintained between the atria and the ventricles in order to pump blood efficiently. Bradyarrhythmia: A class of arrhythmia that results in an abnormally slow heart rate. Cardioversion: Termination of a tachyarrythmia, other than ventricular fibrillation, by a low-energy electrical shock. Defibrillation: Termination of fibrillation by an electrical shock. Fibrillation: A type of tachyarrhythmia characterized by a disorganized rhythm that can occur in either the atria or the ventricles and completely compromises their ability to pump blood. Sinus node: Specialized cells in the top of the right atrium, which act as the heart’s natural pacemaker. Tachyarrhythmia: A class of arrhythmia that results in an abnormally fast heart rate. Ventricles: The lower two chambers of the heart, which are responsible for pumping the blood to the body. References Alt, E., Heinz, M., Hirgstetter, C., Emslander, H. P., Daum, S. and Blomer, H., Control of pacemaker rate by impedance-based respiratory minute ventilation, Chest, 92, 247–252, 1987. Alt, E., Hirgstetter, C., Heinz, M. and Blomer, H., Rate control of physiologic pacemakers by central venous blood temperature, Circulation, 73, 1206–1212, 1986. Alt, E., Millerhagen, J. O. and Heemels, J.-P., Accelerometers. Clinical Cardiac Pacing, Ellenbogen, K. A., Kay, G. N. and Wilkoff, B. L., Eds., W.B. Saunders, Philadelphia, PA, 1995, 275. Anderson, K. and et al., A rate-variable pacemaker which automatically adjusts for physical activity, Pacing and Clin. Electrophys., 6, A-12, 1983. Abstract. ? 2000 by CRC Press LLC Anderson, K. M. and Moore, A. A., Sensors in pacing, Pacing Clin. Electrophysiol., 9, 954–9, 1986. Bach, S. M. and Monroe, P., High power circuitry, Implantable Cardioverter Defibrillator Therapy. The Engi- neering-Clinical Interface, Kroll, M. W. and Lermann, M. H., Eds., Kluwer Academic Publishers, Norwell, MA, 1996, 257–274. Bakker, P. J., Meijburg, H., de Jonge, N., van Mechelen, R., Wittkampf, F. H., Mower, M., et al., Beneficial effects of biventricular pacing in congestive heart failure, Pacing and Clin. Electrophys., 17, 820, 1994. Abstract. Bernstein, A. D., Camm, A. J., Fletcher, R. D., Gold, R. D., Rickards, A. F., Smyth, N. P., et al., The NASPE/BPEG generic pacemaker code for antibradyarrhythmia and adaptive-rate pacing and antitachyarrhythmia devices, Pacing Clin Electrophysiol., 10, 794–799, 1987. Bornzin, G. A., Stokes, K. B. and Wiebush, W. A., A low threshold, low polarization, platonized endocardial electrode, Pacing and Clin. Electrophys., 6, A70, 1983. Brinker, J. and Midel, M., Techniques of pacemaker implantation, Cardiac Pacing, Ellenbogen, K. A., Ed., Blackwell Science, Cambridge, MA, 1996, 216–269. Brumwell, D. A., Kroll, K. and Lehmann, M. H., The amplifier: sensing the depolarization, Implantable Car- dioverter Defibrillator Therapy. The Engineering-Clinical Interface, Kroll, M. W. and Lermann, M. H., Eds., Kluwer Academic Publishers, Norwell, MA, 1996, 275–302. Calfee, R. V. and Saulson, S. H., A voluntary standard for 3.2 mm unipolar and bipolar pacemaker leads and connectors, Pacing Clin Electrophysiol., 9, 1181–5, 1986. Cammilli, L., A New Pacemaker Autoregulating the Rate of Pacing in Relation to Metabolic Needs, Excerpta Medica, Amsterdam,1977. Castro, A., Liebold, A., Vincente, J., Dungan, T. and Allen, J. C., Jr., Evaluation of autosensing as an automatic means of maintaining a 2:1 sensing safety margin in an implanted pacemaker. Autosensing Investigation Team, Pacing Clin. Electrophysiol., 19, 1708–13, 1996. Donaldson, R. M. and Rickards, A. F., Rate responsive pacing using the evoked QT principle. A physiological alternative to atrial synchronous pacemakers, Pacing Clin. Electrophysiol., 6, 1344–1349, 1983. Eityzgrlf, A., Goedel-Meinem, L., Bock, T. and et al., Central venous saturation for the control of automatic rate-responsive pacing, Pacing and Clin. Electrophys., 5, 829, 1982. Abstract. Ellenbogen, K. A., Cardiac Pacing, Blackwell Science, 1996. Elmquist, R. and Senning, A. An implantable pacemaker for the heart, Proceedings of the 2nd International Conference on Medical Electronics, London, Iliffe and Sons, 1960. Epstein, A. E. and Ideker, R. E., Ventricular fibrillation, Cardiac Electrophysiology: From Cell to Bedside, Zipes, D. P. and Jalife, J., Eds., W. B. Saunders Company, Philadelphia, PA, 1995, 927–934. Feeser, S. A., Tang, A. S. L., Kavanagh, K. M., Rollins, D. L., Smith, W. M., Wolf, P. D., et al., Strength-duration and probability of success curves for defibrillation with biphasic waveforms, Circulation, 82, 2128–2141, 1990. Foster, A. H., McLaughlin, J. S. and Fisher, M. L., Improved hemodynamics with biventricular pacing, J. Am. Coll. Cardiol., 23, 156A, 1994. Abstract. Funke, H. D., Cardiac pacing with the universal DDD pulse generator: Technological and electrophysioological considerations, The Third Decade of Cardiac Pacing: Advances in Technology and Clinical Applications, Barold, S. and Mugica, J., Eds., Futura Publ. Co., Mt. Kisco, NY, 1982, 191–223. Furman, S., Pannizzo, F. and Campo, I., Comparison of active and passive adhering leads for endocardial pacing, Pacing Clin. Electrophysiol., 2, 417–27, 1979. Gillum, R. F., Sudden coronary death in the United States, Circulation, 79, 756–765, 1989. Holmes, C. F., The Battery. Implantable Cardioverter Defibrillator Therapy. The Engineering-Clinical Interface, Kroll, M. W. and Lermann, M. H., Eds., Kluwer Academic Publishers, Norwell, MA, 1996, 205–222. Jacobson, P. and Kroiss, D. Automatic control of the sensing threshold for monitoring cardiac rhythm in an implantable device: United States Patent 5564430, 1996. Jones, B. R., Kim, J., Zhu, Q., Nelson, J. P., KenKnight, B. H., Lang, D. J., Warren, J.A., Future of bradyarrhythmia therapy systems: Automaticity, Am. J. Cardiol., 83(5B), 192D–201D, 1999 Mar 11. Katz, A. M., The Arrhythmias. I. Introduction and mechanisms, Physiology of the Heart, Raven Press,1992, 515–545. Kay, G. N., Basic Concepts of Pacing, Cardiac Pacing, Ellenbogen, K. A., Ed., Blackwell Science, Cambridge, MA, 1996, 37–106. ? 2000 by CRC Press LLC Kim, J., Zhu, Q. and Lang, D., An autosense algorithm with dual thresholds, Arch. Mal C?ur, 91, 77, 1998. Abstract. Lang, D. J., Heil, J. E., Hahn, S. J., Lindstrom, C. C. and Derfus, D. L., Implantable cardioverter defibrillator lead technology: improved performance and lower defibrillation thresholds, Pacing and Clin. Electrophys., 18, 548–559, 1995. Liang, C. C., Bolster, M. E., Murphy, R. M. and inventors (1982 Jan 12.). Metal oxide composite cathode material for high energy density batteries. U.S. Patent No. 4,310,609, Wilson Greatbatch Ltd., assignees. Markewitz, A., Wenke, K. and Weinhold, C., Reliability of atrial screw-in leads, Pacing Clin. Electrophysiol., 11, 1777–83, 1988. Matula, M., Alt, E., Fotuhi, P. and et al., Influence of varied types of exercise on the rate adaptation of activity pacemakers, Pacing and Clin. Electrophys., 15, 578, 1992. Abstract. Mirowski, M., Mower, M. M., Staewen, W. S., Tabatznik, B. and Mendeloff, A. I., Standby automatic defibrillator. An approach to prevention of sudden coronary death, Arch. Intern. Med., 126, 158–61, 1970. Mirowski, M., Reid, P. R., Mower, M. M., Watkins, L., Gott, V. L., Schauble, J. F., et al., Termination of malignant ventricular arrhythmias with an implanted automatic defibrillator in human beings, N. Engl. J. Med., 303, 322–4, 1980. Morris, M. M., KenKnight, B. H., Warren, J. A. and Lang, D. J., A preview of implantable cordioverter defibrillator systems in the next millennium: An integrative cardiac rhythm management approach, Am. J. Cardiol., 83(5B), 48D–54D, 1999 Mar 11.. Moss, A. J., Hill, W. J., Cannom, D. S., Daubert, J. P., Higgins, S. L., Klein, H., et al., Improved survival with an implanted defibrillator in patients with coronary disease at high risk for ventricular arrhythmia, N. Engl. J. Med., 335, 1933–1940, 1996. Nappholz, T., Valenta, H., Maloney, J. and Simmons, T., Electrode configurations for a respiratory impedance measurement suitable for rate responsive pacing, Pacing and Clin. Electrophys., 9, 960, 1986. Rossi, P., Plicchi, G., Canducci, G., Rognoni, G. and Aina, F., Respiratory rate as a determinant of optimal pacing rate, Pacing Clin. Electrophysiol., 6, 502–10, 1983. Saksena, S., Prakash, A., Hill, M., Krol, R. B., Munsif, A. N., Mathew, P. P., et al., Prevention of recurrent atrial fibrillation with chronic dual-site right atrial pacing, J. Am. Coll. Cardiol., 28, 687–694, 1996. Sanders, R. S. and Lee, M. T., Implantable pacemakers, Proceedings of the IEEE, 84, 480–6, 1996. Schuder, J. C., McDaniel, W. C. and Stoeckle, H., Transthoracic defibrillation of 100 kg calves with bidirectional truncated exponential shocks, Trans. Am. Soc. Artif. Intern. Organs, 30, 520–525, 1984. Shuster, M. and Keller, J. L., Effect of fire department first-responder automated defibrillation, Ann. Emerg. Med., 22, 721–7, 1993. Sinnaeve, A., Willems, R., Backers, J., Holovoet, G. and Stroobandt, R., Pacing and sensing: how can one electrode fulfill both requirements?, Pacing Clin. Electrophysiol., 10, 546–54, 1987. Stokes, K. and Bornzin, G., The electrode-biointerface: Stimulation, Modern Cardiac Pacing, Barold, S. S., Ed., Futura Publ. Co., Mount Kisco, NY, 1985, 37–77. Timmis, G. C., Gordon, S., Westveer, D. C. and et al. (1983). A new steroid-eluting low threshold lead, Proceedings of the Seventh World Symposium on Cardiac Pacing, Darmstadt, Vienna, Steinkopff-Verlag. Usui, M., Walcott, G. P., KenKnight, B. H., Walker, R. G., Rollins, D. L., Smith, W. M., et al., Influence of malpositioned transvenous leads on defibrillation efficacy with and without a subcutaneous array elec- trode, Pacing and Clin. Electrophys., 18, 2008–2016, 1995. Warren, J. A., Dreher, R. D., Jaworski, R. V., Putzke, J. J. and Russie, R. J., Implantable cardioverter defibrillators, Proceedings of the IEEE, 84, 468–79, 1996. Winkle, R., Mead, H., Ruder, M. and et al., Ten year experience with implantable defibrillators, Circulation, 84, II-426, 1991. Abstract. Yee, R. and Bennett, T. D., Rate-adaptive pacing controlled by dynamic right ventricular pressure (dP/dtmax), Clinical Cardiac Pacing, Ellenbogen, K. A., Kay, G. N. and Wilkoff, B. L., Eds., W.B. Saunders Company, Philadelphia, PA, 1995, 212–218. ? 2000 by CRC Press LLC Further Information Cardiac Pacing, second edition, Kenneth A. Ellenbogen, Ed., Blackwell Science, 1996. Clinical Cardiac Pacing, Kenneth A. Ellenbogen, G. Neal Kay, and Bruce L. Wilkof, Eds., W.B. Saunders Company, 1995. Implantable Cardioverter Defibrillator Therapy: The Engineering-Clinical Interface, Mark W. Kroll and Michael H. Lehmann, Eds., Kluwer Academic Publishers, 1996. Implantable Cardioverter-Defibrillators: A Comprehensive Textbook, N.A. Mark Estes III, Antonis S. Manolis, and Paul J. Wang, Eds., Marcel Dekker, 1994. ? 2000 by CRC Press LLC