Simultaneous Unipolar and Bipolar Recording of Cardiac Electrical Activity ELIAS SEVAPTSIDIS,* STEPHANE MASSE,* IAN D. PARSON,** EUGENE DOWNAR,t and SHANE KIMBERt From the *Department of Medicine. **Ryerson Polytechnical Institute and Institute of Biomedical Engineering, and the tDepartment of Medicine, Toronto General Hospital and Women's College Hospital, University of Toronto, Toronto, Canada

SEVAPTSIDIS, E., ETAL.: Simultaneous Unipolar and Bipolar Recording of Cardiac Electrical Activity. An analog mapping system using a true bipoJar left ventricuJar balloon electrode array is described, which enables simultaneous unipoJar and bipoJar recordings. It is an adaptation of a previous clinical analog mapping system used in the investigation o/ventricuJar arrhythmias. The bipolar ba]Joon array consists of 112 electrode pairs, each having a 2-mm separation. The signals from the electrodes are sensed in paraiJel by separate unipolar and bipolar amplifier units, which then drive a common multiplexer bus. The bipolar recording unit consists of high quaJity instrumentation amplifiers with adjustable gain and exhibits a full bandwidth minimum common mode rejection of 78 dB. Using this combination, it is possible to record local cardiac micropotentiaJs while still retaining the advantages of unipolar electrograms to track overall cardiac activation. {PACE, Vol. 15, January 1992} activation mapping systems, arrhythmias, electrophysioJogy, bipolar recording

Introduction There is controversy in cardiac mapping literature as to whether unipolar or bipolar cardiac electrograms are hetter for detecting cardiac activation.^'^ Bipolar recordings are best suited for detecting local activation^ and are less sensitive to distant activation fronts."* Noise reduction is greater and high gains can he used for increased sensitivity when dealing with microelectrograms. Disadvantages include directional sensitivity (determined by the axis of the hipole relative to the activation front), which complicates the interpretation of the hipolar signal.^ Furthermore, unless a hipolar electrode makes close contact with the recording surface, no useful activation data is oh-

tained. On the other hand, unipolar recordings have the advantage of a well-established criterion for detecting activation, do not need intimate contact with adjacent myocardium, have an omnidirectional recording field, and a greater signal strength. They can also he used for isopotential mapping. However, since the amplitude of a unipolar recording is determined primarily hy remote activity,^ this recording mode is less sensitive to discrete local events. Recognizing that there are distinct advantages to each recording mode, we developed a cardiac mapping system that, for the first time, allows simultaneous acquisition of unipolar and hipolar electrograms from the same electrode array. Recording Electrodes

Supported by the Canadian Heart Foundation and the Heart and Stroke Foundation of Ontario. Address for reprints; Elias Sevaptsidis, Medical Sciences Building, Room 7363, University of Toronto, Toronto. Ontario M5S 1A8, Canada. Fax: (416) 978-8765, Received July 5, 1991; accepted September 9, 1991.

PACE, VoL 15

For left ventricular endocardial recording, an array of silver head electrodes is used where the electrodes are sutured to an expandahle nylon mesh stretched over a douhle latex balloon. The array consists of 112 electrodes arranged as 14 ra-

January 1992

45

SEVAPTSIDIS, ET AL.

dial rows of 8 electrode pairs. This arrangement is similar to that described in a previous report^ except that at each recording site, a bipolar configuration is used. Each bipolar electrode is made of two 2-mm diameter beads separated by 2 mm (center to center). Both beads are drilled to enable a secure attachment to the nylon mesh. Similar to the unipolar array, stainless steel wires (25 strand. 40 gauge. Teflon insulated) are soldered to the beads and provide the electrical connection to the instrumentation. This left ventricular balloon array is used in conjunction with the unipolar epicardial sock^ and right ventricular balloon, giving a total of 336 unipolar electrodes and 112 bipolar electrodes. System Description The unipolar and bipolar electrograms from the balloon and sock arrays are amplified by instrumentation amplifiers, multiplexed, and stored permanently on their respective video cassette recorders. Radiofrequency transformers ensure patient isolation while video cassette recorders allow immediate off-line playback of the electrograms.^ Each 112 electrode array consisting of sock, right ventricular, and left ventricular balloon is depicted on a video monitor by low intensity dots. The arrays are displayed on a two-dimensional plane composed of 14 rows of 8 dots [electrodes] arranged in a radial pattern, where the center represents the apex of the heart. The occurrence of tissue activation at each electrode is indicated by a high intensity flash of its corresponding dot on the video monitor. Tbe sequence of activation, being a series of video images, is recorded by another video cassette recorder, which allows slow motion or individual frame-by-frame sequencing during replay.^ For analysis purposes, the demultiplexed electrograms are available in any desired group of 8 or 16 on a standard strip chart recorder. The electrograms can be printed on the chart recorder during on-line (signal acquisition) and off-line (playback) modes. Time code annotation is provided on the video display as well as the chart recorder to facilitate necessary referencing of timed events. Instrumentation Amplifiers Recent requirements to detect micropotentials have necessitated high quality instrumenta-

46

tion amplifiers to precisely extract their temporal characteristics. Micropotentials have high frequency components with low energy characteristics, which become obscured by noise originating in the input stages of the unipolar operational amplifiers. High quality instrumentation amplifiers operated in the bipolar mode provide good common mode rejection (CMR), signal-to-noise ratio, and low channel cross-talk and can be used to overcome the detection problem. Since most electronic circuitry is concerned with the elimination of the ubiquitous 60-Hz power frequency, amplifiers traditionally have their CMR quoted at this frequency. However, in cardiac mapping the micropotentials of interest have their significant power spectrum in the region 100 Hz to 1 kHz. so the CMR of an amplifier at 500 Hz is much more important than that quoted by manufacturers at 60 Hz. Unfortunately, CMR decreases with increasing frequency (all amplifiers look good at 60 Hz) and it requires special attention to amplifier characteristics at 500 Hz. as well as careful circuit board layout to ensure the optimal CMR at the frequencies characteristic of the micropotentials. For this reason, all the CMRs quoted here are at 500 Hz and will be significantly smaller than those of individual amplifiers quoted at 60 Hz. Since the current trend is to digitize signals to a resolution of 12 bits." a true bipolar signal will then require at least a 72-dB CMR to be considered "hipolar." Therefore, a design requirement for the bipolar system was to have a CMR >72 dB at 500 Hz.

Construction and Measurement The instrumentation amplifiers chosen for the mapping system have a specified CMR of 90 dB and high input impedance (10^" il]. The amplifier chosen for the bipolar design was the INAlOlHP (Burr-Brown Corp.. Tucson. AZ, USA). The circuit components were assembled on 160 x 100 mm Eurocard boards and tested. Tbe following measurements were performed with a frequency sweep signal from DC to 500 Hz. Results yielded a worst case CMR of 78 dB (84 dB average). Crosstalk was 103 dB (average), while signal-to-noise ratio gave 108 dB minimum (average 123 dB). These recordings were made at 500 Hz and mea-

January 1992

PACE, Vol. 15

SIMULTANEOUS UNIPOLAR AND BIPOLAR RECORDING

sured at the output of the final amplifier stage before multiplexing. The bipolar instrumentation assembly uses on-board regulators for each 8-channel amplifier card. This aids in reducing switching noise and other such disturbances from communicating between the amplifier cards via a common bus. Cmos (complementary metal-oxide semiconductor) circuitry was used throughout (with the exception of the instrumentation amplifier) in order to minimize power consumption since the patient-connected amplifier circuitry is battery powered. Simultaneous Unipolar/Bipolar Signal Recording Simultaneous bipolar and unipolar data recording is provided by separate bipolar and unipolar amplifier circuits. Figure 1 shows unipolar and bipolar amplifier circuits connected together at the signal source. The unipolar section uses a fixed gain of 288 and permits a signal bandwidth from 0.036 Hz to 200 Hz (3 dB). The output of the final

unipolar amplifier is handled by a multiplexer operating synchronously with the bipolar multiplexer each having a capacity of 128 channels. Ideally, unipolar signals would be obtained from the instrumentation amplifier's unipolar output. This was not possible, however, since the unipolar front-end amplifiers and multiplexers were the existing equipment at the time of tbe bipolar amplifier design. The bipolar section had to be added in a manner that allowed simultaneous operation. This approach saved time and money. A potential disadvantage could be a sacrifice in signal integrity due to loading effects at the amplifier input stages (see Appendix). Figure 2 shows an example of simultaneously recorded unipolar and bipolar electrograms during sinus rhythm using a left ventricular cavity balloon electrode array. Panel A shows the sequence of electrical activation over the left ventricular endocardium in 12-msec isochrones. The left ventricle is depicted in a polar projection with the apex at the center and the base at the periphery. Activation begins in the ventricular septum at the

gains GI-G4 ADJUSTABLE CAIN CIRCUIT

Za

COMMON MULTIPLEXER BUS

MULTIPLEXEK

(UNIPOLAR) U = Unipolar

(syslcm gain: 288 x)

B = Bipolar

(syslcm gain: lOOOx. 3000x, 6000x, 1 lOOOx)

Figure 1. Amplifier diagram showing inpul electrode configuration for simultaneous unipolar and bipolar recording.

PACE, Vol. 15

January 1992

47

SEVAPTSIDIS. ET AL.

B

Figure 2. Comparison befiveen simultaneously recorded unipoiar and bipolar fjiectrograms during sinus rhythm using a left venfricuiur cavity balloon electrode array. (A) Activafion map with 12-ms isochrones. (B, C) Unipolar (LJ) and bipolar (Bj electrograms recorded from rows 2 and 3, respectively. The arrow heuds indicate left bundle branch activity. Note that the unipolar recordings fail to indicate left bundle brunch activity.

isochrone marked with an asterisk. Panels B and C are the unipolar and corresponding bipolar electrograms, which were recorded on electrode rows 2 and 3, respectively. The arrow heads indicate left bundle branch electrograms seen on the hipolar signals. Note that the corresponding unipolar signals obtained from one of each electrode pair fail to give any indication of underlying left bundle branch activation.

48

Input Loading Effect The unipolar signal is obtained by summing the two bipolar inputs of the instrumentation amplifier using two l-MIl resistors, Rg and R4 (Fig. 1). This value is a compromise: a higher value would attenuate significantly the unipolar signal and a lower value would affect the CMR of the bipolar amplifier. These resistors load each input equally

January 1992

PACE, Vol. 15

SIMULTANEOUS UNIPOLAR AND BIPOLAR RECORDING

and, therefore, minimize imbalance of the instrumentation amplifier inputs. Computer simulation has been used to evaluate the CMR degradation (see Appendix). Simulation and measurement results show that the GMR is most sensitive to protection resistor imbalance (Ri and Rz). Using 1% value resistors for R| and R2 can help to maintain the CMR higher than 72 dB for frequencies up to 1 kHz.

Adjustabie Gain Adjustable gain is a departure from the earlier mapping system, which had a single fixed gain. Higher gain is necessary when switching from unipolar to bipolar mode, since the 2-mm spacing of the bipolar eiectrode couplets results in significantly smaller signals. The reason for adjustable gain is to permit measurement of the low electrical activity in the ischemic regions of the endocardium, where the large variations in signal size demand variable gain to derive meaningful signals from ischemic regions compared with those of healthy tissue. The adjustable gain circuit in Figure 1 uses a standard cmos multiplexer chip with each multiplexed line connected to a gain resistor, RQ. Each of these resistors determines a gain setting for the amplifier. Four adjustable gains of 1,000 x , 3,000 X, 6,000 X. and 11,000 X were implemented. The gain-select inputs of all multiplexer chips were tied together so that all channels would have the same gain setting in unison. This arrangement has the advantage of simplicity, while its disadvantage is that it prevents the user from independently adjusting each individual channel gain to remedy occasional clipping of the signai in some channels.

Discussion Initial clinical experience with simultaneous unipolar and hipolar recording has yielded results similar to those illustrated in Figure 2. It is indeed possible to record signals in the \LV range originating from the specialized conducting tissue and from small tracts of surviving muscle fibers embedded in myocardial scar. Such recordings may be vital to understanding arrhythmia mecha-

PACE, Vol. 15

nisms and directing surgical procedures to eradicate arrhythmias. Normally these signals are lost in the background noise of existing mapping systems. Standard local eiectrograms such as those provided by the unipolar amplifier are still required to provide overall activation data such as the isochrone map shown in panel A of Figure 2. On the other hand, while high gain bipolar signals are likely to be clipped and manifest a complexity that is confusing for identification of local activation, they do provide data on micropotentials and endocardial fiber activation. Unipolar recordings in this case will not reject common mode noise and distant activity, which obscures the micropotentials. Also, in such circumstances, the video activation image is confused with false activation flashes related to the excessive noise. Bipolar signals, on the other hand, provide good common mode rejection, signal-to-noise ratio, and have sufficient gain to extract and observe critical microactivations. For these reasons, an ideal cardiac mapping system should offer simultaneously both these sets of complementary data. Until now, such a system has not been available.

Conclusion Bipolar recordings have advantages such as detection of local events, immunity to distant activation fronts, good common mode rejection, and sensitivity. These advantages facilitate the detection of microelectrograms. Unipolar recordings, on the other hand, have advantages such as greater signal strength, omnidirectional recording field, and less dependency on intimate contact with the adjacent myocardium. Simultaneous unipolar and bipolar mapping combines the advantages of both recordings and effectively compensates for disadvantages when recording with only one mode at a time. These unipolar and bipolar modes of recording cardiac activation each has unique properties that can facilitate or inhibit an understanding of cardiac events. Simultaneous recordings of both types of symbols maximizes their advantages and compensates for the limitations inherent in either type alone. Combining complementary data in this system as described is providing important new insights into arrhythmogenic mechanisms in the human heart. In particular, investigations into the

January 1992

49

SEVAPTSIDIS, ET AL.

role of the specialized conducting system and surrounding muscle tracts in the ischemic arrhythmias should be enhanced by this approach. Appendix In the circuit shown in Figure 1, Za and Zb represent the input impedance seen by the voltage source in branches A and B, respectively. Zjn is the instrumentation amplifier input impedance. Rsi and Rs2 represent the tissue source impedance, which is approximately 1 kft. Finally, Ri and R2 are protection resistors used to limit the current flowing to the instrumentation amphfier during defibrillatiou. According to Winter,^ a common mode voltage (Vc) can be measured as a differential voltage (Vi): Vi = vJl/CMR + Zd/ZJ

loading effect on the bipolar channel. If all impedances in branch A are mismatched to the ones in branch B by 10% and assuming a worst case situation, the effective CMR would be 71 dB at 500 Hz (using computer simulation, Fig. 3). Without the unipolar amplifier present, the calculated CMR^ is 74 dB at 500 Hz. Actual CMRe measurements on eight bipolar channels with and without the unipolar circuitry were always greater than the computer simulation with a worst case of 78 dB at 500 Hz. Results show that the unipolar loading causes a small degradation on bipolar channel CMR^. The difference between simulated and measured values demonstrates that a 10% imbalance between each pair of component in branches A and B is unlikely to occur. The mismatch impedance above 500 Hz is mostly due to the second and fourth terms since:

(1)

where Zd = difference between the two electrode impedances, Zc = common mode impedance, and CMR = common mode rejection ratio of the instrumentation amplifier and the effective CMR of the circuit (CMRe) can be defined as: CMRe = 1/(1/CMR + Zd/ZJ

(2)

CO

3 E u

Zd/Zc gives after simplification: 70

- (Rsi/Za + Ri/[Zi« + Ri])

(3) 60

with Rl and R2 = 100 kfl.

1000

Rsi and Rs2 = 1 kH,

Frequency (Hz)

Za and Zb - 1.5 Mii at 500 Hz, and

—•— —I— —A— —o—

Zin = 7 X 10^ a at 500 Hz (Zd - 10« a I 3 pF, Zc The second and fourth terms in Equation 3 represent the impedance mismatch between the inputs without unipolar channel present, whereas the first and third terms are related to the unipolar

50

1200

simulation, with loading simulaiion, without loading measured, without loading measured, with loading

Figure 3. Computed and measured common mode rejection of bipolar amplifiers with and without the presence 0/ unipolar amplifiers. Note that measured common mode results fwith loading) are always >78 dB up to 500 Hz. H-hiie simulated results showed >74 dB for the same range. Cmrr — common mode rejection ratio.

January 1992

PACE, Vol. 15

SIMULTANEOUS UNIPOLAR AND BIPOLAR RECORDING

1 ka/1.5 MO = 6.7 X 10-" is smaller than 100 ka/(ioo ka + 7 X 10' a) ^ 1.42 X 10"^ at 500 Hz.

Therefore, a 10% imbalance on Ri and R2 would decrease the CMRe more than a similar imbalance on Rsi and Rs2 (since Za = Zb).

References 1. Ideker JD, Smith WM, Blanchard SM, et al. The assumptions of isochronal cardiac mapping. PACE 1989; 12:456-478. 2. Page SM, Cardinal R, Savard P, et al. Sinus rhythm mapping in a canine model of ventricular tachycardia. PACE 1988; 11:632-644. 3. Gallagher JJ, Kasell JH, Cox JL, et al. Techniques of intraoperative electrophysiologic mapping. Am J Cardiol 1982; 49:221-240. 4. Durrer D, Van Der Tweel LH. Spread of activation in the left ventricular wall of the dog. II. Am Heart J 1954; 47:192-203. 5. Spach MS, Barr RC. Ventricular intramural and potential distributions during ventricular activation and repolarization in the intact dog. Circ Res 1975; 37:243-257.

PACE, Vol. 15

Downar E, Harris L, Mickleborough LL, et al. Endocardial mapping of ventricular tachycardia in the intact human ventricle: Evidence for reentrant mechanisms. J Am Coll Cardiol 1988; 11:783791. Parson ID, Downar E. Clinical instrumentation for the intra-operative mapping of ventricular arrhythmias. PACE 1984; 7:683-692. Ideker RE, Smith WM, Wolf P, et al. Simultaneous multichannel cardiac mapping systems. PACE 1987; 10:281-292. Winter BB, Webster JG. Reduction of interference due to common mode voltage in biopotential amplifiers. IEEE Trans Biomed Eng 1983; 30:58-62.

January 1992

51

Simultaneous unipolar and bipolar recording of cardiac electrical activity.

An analog mapping system using a true bipolar left ventricular balloon electrode array is described, which enables simultaneous unipolar and bipolar r...
3MB Sizes 0 Downloads 0 Views