A simultaneous multichannel monophasic action potential electrode array for in vivo epicardial repolarization mapping

IEEE TRANSACTIONS ON BIOMEDICAL ENGINEERING, VOL. 48, NO. 3, MARCH 2001

345

A Simultaneous Multichannel Monophasic Action Potential Electrode Array for In Vivo Epicardial Repolarization Mapping Alan V. Sahakian*, Senior Member, IEEE, Ming-Shing Lee Peterson, Sergio Shkurovich, Student Member, IEEE, Mark Hamer, Timothy Votapka, Tongyou Ji, and Steven Swiryn, Member, IEEE

Abstract—While the recording of extracellular monophasic action potentials (MAPs) from single epicardial or endocardial sites has been performed for over a century, we are unaware of any previous successful attempt to record MAPs simultaneously from a large number of sites in vivo. We report here the design and validation of an array of MAP electrodes which records both depolarization and repolarization simultaneously at up to 16 epicardial sites in a square array on the heart in vivo. The array consists of 16 sintered Ag-AgCl electrodes mounted in a common housing with individual suspensions allowing each electrode to exert a controlled pressure on the epicardial surface. The electrodes are arranged in a square array, with each quadrant of four having an additional recessed sintered Ag-AgCl reference electrode at its center. A saline-soaked sponge establishes ionic contact between the reference electrodes and the tissue. The array was tested on six anesthetized open-chested pigs. Simultaneous diagnostic-quality MAP recordings were obtained from up to 13 out of 16 ventricular sites. Ventricular MAPs had amplitudes of 10–40 mV with uniform morphologies and stable baselines for up to 30 min. MAP duration at 90% repolarization was measured and shown to vary as expected with cycle length during sustained pacing. The relationship between MAP duration and effective refractory period was also confirmed. The ability of the array to detect local differences in repolarization was tested in two ways. Placement of the array straddling the atrioventricular (AV) junction yielded simultaneous atrial or ventricular recordings at corresponding sites during 1:1 and 2:1 AV conduction. Localized ischemia via constriction of a coronary artery branch resulted in shortening of the repolarization phase at the ischemic, but not the nonischemic, sites. In conclusion, these results indicate that the simultaneous multichannel MAP electrode array is a viable method for in vivo epicardial repolarization mapping. The array has the potential to be expanded to increase the number of sites and spatial resolution. Index Terms—Cardiac mapping, electrodes, epicardium, monophasic action potential, refractory period, repolarization. Manuscript received July 22, 1999; revised November 8, 2000. This work was supported in part by a grant from the Dr. Scholl Foundation and scholarships from the Medtronic Scholars Program and from Fulbright-Conacyt-IIE. Asterisk indicates corresponding author. *A. V. Sahakian is with the Departments of Electrical and Computer Engineering, and Biomedical Engineering, Northwestern University, 2145 Sheridan Rd., Evanston, IL 60208 USA. He is also with Evanston Northwestern Healthcare, Northwestern University, Evanston, IL 60208 USA (e-mail: [email protected]). M.-S. L. Peterson and S. Shkurovich are with the Biomedical Engineering Department, Northwestern University, Evanston, IL 60208 USA. M. Hamer, T. Votapka, and S. Swiryn are with the Medical School, Northwestern University, Evanston, IL 60208 USA. They are also with Evanston Northwestern Healthcare, Northwestern University, Evanston, IL 60208 USA. T. Ji is with the Northwestern University, Evanston Northwestern Healthcare, Northwestern University, Evanston, IL 60208 USA. Publisher Item Identifier S 0018-9294(01)01575-0.

I. INTRODUCTION

M

APPING of cardiac electrical activity is an important research and clinical procedure. Mapping is commonly based on recording extracellular electrical signals from multiple simultaneous sites, but may also be accomplished optically on in vitro preparations [1], [2]. Extracellular electrical recordings tend to be biphasic and yield accurate information about the time of local activation (depolarization), but do not accurately show other features of the action potential such as the action potential duration and, therefore, repolarization. Transmembrane (intracellular) recordings do show this information but are experimentally difficult to make on the intact, beating heart. In many important physiologic and clinical situations, it cannot be assumed that changes in repolarization merely mirror changes in depolarization. For example, the sequence of depolarization in the ventricle is endocardial to epicardial, but for repolarization the sequence is reversed, resulting in T-waves with the same polarity as the QRS complex. Electrophysiologic remodeling is a clinically important phenomenon exemplified by the finding that, after a period of altered activation even if depolarization returns to normal, repolarization may remain altered for some time [3]. Therefore, mapping studies which have focused on patterns of depolarization lack important information. Monophasic action potential (MAP) extracellular recordings were first made in the late 19th century by Burdon-Sanderson and Page [4]. In MAP recording, an electrode is pressed against the myocardium with a controlled pressure. This pressure creates an injury current which allows a local (composite) intracellular potential to be sampled, in a somewhat attenuated form, in the extracellular space under the electrode. This potential is measured relative to a reference element in the blood or saline pool a short distance away from the measurement site, yielding a MAP recording. If the reference is relatively near to the recording site then a close bipole is formed and distant potentials are cancelled [5]. MAP recordings exhibit many features seen in transmembrane recordings of the action potential, even during interventions which change the transmembrane action potential [6]. MAP recordings are made clinically using either an endocardial catheter or an epicardial wand [7]. In both cases, the recording probe only has a single active electrode element and a single reference, so only a single site may be examined at a time. Franz [8] reviews the current status and history of MAP recording.

0018–9294/01$10.00 © 2001 IEEE

346

IEEE TRANSACTIONS ON BIOMEDICAL ENGINEERING, VOL. 48, NO. 3, MARCH 2001

Fig. 1. Face view of the electrode array showing the 16 active (recording) elements arranged into four quadrants. Each quadrant of four electrodes has a single reference electrode which serves as the negative input for the MAPs in that quadrant. All measurements are in millimeters.

Although mapping is possible using a single-element probe by moving it from site to site and making serial recordings, there are several limitations to this method. First, there is the implicit assumption that activations repeat in the same way from cycle to cycle at each site. While this assumption may be valid in well-organized rhythms with stable rates (such as sinus rhythm, a paced rhythm or a stable tachycardia), it may not be valid in more complex rhythms (such as polymorphic tachycardia or fibrillation) or if the rate is not stable. Other rapidly changing physiological states, such as ischemia, may not be well studied this way. A second limitation is that the electrode takes time to stabilize at each site. Thus, serial MAP mapping is a very time-consuming procedure. Aside from simple inconvenience, this is a fundamental limitation when designing protocols using serial mapping since the tissue may change with time. There are also practical and ethical limits to the duration of a surgical procedure, especially in humans. To address these limitations we designed, fabricated and validated a multiple-electrode array which acquires simultaneous, rather than serial, MAP recordings from 16 sites. While previous work by others has recorded MAPs from up to three sites simultaneously in vivo [9] and up to ten sites in vitro [10], we are unaware of any reports of MAPs recorded simultaneously from a large number of sites in vivo. II. METHODS A. Design of the Electrode Array The array consists of 16 active electrode elements (model 4 matrix E205A, In Vivo Metric, Healdsburg, CA) in a 4 with an inter-electrode distance of 5 mm [11] as shown in Fig. 1 [12], [13]. Each electrode element contains a cylindrical sin-

tered Ag-AgCl pellet (1-mm diameter 2.5-mm length) encapsulated in a polycarbonate tube up to but not including the active face of the pellet as shown in Fig. 2. A silver wire (0.25-mm diameter) is sintered into the pellet and passed through the rear of the encapsulated electrode. This wire was insulated with Teflon tubing (1.07-mm outer diameter) and both were fed through beryllium copper springs (10 mm long) such that the spring sat beneath the element. The elements were then mounted so they could move freely through the base and protruded 7–8 mm, during a relaxed state, out of a cylindrical base of PVC plastic (26.6-mm diameter 16.8 mm high). The array was held manually against the epicardium with enough pressure to partially depress the recording elements into the housing. The springs and electrode protrusions were chosen to yield a pressure at the face of each electrode element of between 10 and 30 g/mm , which is large enough to create the necessary injury currents but small enough to avoid damaging the epicardium [7]. An additional unencapsulated Ag-AgCl electrode pellet protruded 1 mm out of the base at the middle of each quadrant (group of four) of active elements and served as the reference for that quadrant. A sponge with holes for the active elements to pass through was placed over the array and saturated with isotonic saline. This served as the medium for conduction from the tissue to the reference element. The wires from all active and reference electrodes passed through the rear of the base and were soldered individually to small-gage multistrand Teflon-insulated wires which led to the input amplifiers. All solder joints were insulated and supported with conformal tubing. For some of our early experiments a second plastic housing was placed around the electrode array to hold it at a fixed dis-

SAHAKIAN et al.: SIMULTANEOUS MULTICHANNEL MONOPHASIC ACTION POTENTIAL ELECTRODE ARRAY

347

Fig. 2. A cross section through the electrode array showing the method of independently suspending the active electrode elements to maintain a controlled pressure at each electrode face. The reference elements are mounted close to the housing, recessed from the epicardial surface, but within a saline-soaked sponge which maintains ionic continuity to the epicardium. All measurements are in millimeters.

Fig. 3. The complete array. The major divisions on the scale are centimeters.

tance from the tissue. This was found to be unnecessary and was abandoned. Fig. 3 shows a photograph of the completed array. Our experiments followed a protocol that was approved by the Institutional Animal Care and Use Committee of Evanston Northwestern Healthcare. MAPs were recorded in six Yorkshire 4.43 kg pigs (four female and two male) weighing 40.48 SD). Prior to each experiment the pigs were fasted (mean overnight. The pigs were anesthetized with intramuscular ke-

tamine (20–30 mg/kg) and sodium pentobarbital (2.5–5 mg/kg to effect) was given intravenously. The pigs were intubated with a cuffed endotracheal tube and placed on a ventilator. Additional inhalation anesthesia consisting of isofluorane (1%–5%), nitrous oxide, and oxygen was administered continuously. Arterial pressure was continuously monitored via a catheter in the femoral artery and the ECG was continuously recorded. At the end of the procedures, the animals were euthanized with sodium pentobarbital. The MAP signals, the ECG and the arterial pressure signal were recorded on a digital electrophysiologic recording system (GE Medical Systems, Milwaukee, WI). The recording system acquired with a bandwidth of dc to 500 Hz, and with a sample rate of 977 Hz. A median sternotomy was performed and the heart was exposed and suspended in a pericardial cradle. Bipolar pacing electrodes were sewn on the right ventricular apex and in the right atrial appendage. For two of the experiments, ablation of the atrioventricular (AV) node was performed to reduce the ventricular rate. We validated the ability of the array to record accurate MAPs in several experiments as described below: B. Evaluation of Acceptable MAP Waveforms To ensure that accurate MAPs were being recorded by the array we rejected signals which did not meet the following criteria [14]: 1) ventricular MAP amplitudes must be at least

348

IEEE TRANSACTIONS ON BIOMEDICAL ENGINEERING, VOL. 48, NO. 3, MARCH 2001

Fig. 4. (Left) Example of 13 acceptable right ventricular MAP recordings. Site numbers appear above the waveforms. RVA is a right ventricular electrogram. BP is arterial blood pressure from 0 to 200 mmHg. (Right) Example of four acceptable right atrial MAP recordings, with surface lead II at the top and BP at the bottom.

10 mV, while atrial MAP amplitudes must be greater than 3 mV (atrial recordings were acquired during one experiment, described below); 2) MAP recordings must depict a sharp upstroke, with a smooth repolarization phase; 3) The upstroke spike must be no greater than the MAP amplitude; 4) The MAPs must exhibit uniform morphology; 5) The baseline must be undisturbed and free of artifacts. C. Correlation Between MAPd90 and Cycle Length We investigated the ability of the electrode array to record the cycle-length dependency of MAP duration in the following experiment. The right ventricular bipolar pacing electrodes were connected to a Bloom DTU 215 stimulator (Fischer Imaging Corp., Denver, CO). The diastolic pacing threshold at a pulse width of 2 ms was determined and the stimulator was then set to pace at this width continuously at twice diastolic threshold. The ventricles were paced for up to 2 min at cycle lengths ranging from 550 to 300 ms in decrements of 50 ms and from 300 to 220 ms in decrements of 20 ms while recording using the MAP array. The MAP duration to 90% repolarization (MAPd90) was measured on these recordings using an algorithm similar to that described by Kanaan et al. [15].

D. Correlation Between MAPd90 and Ventricular Effective Refractory Period (VERP) To validate the ability to record MAPs which accurately reflect changes in the true refractory period, both the MAPd90 and VERP were determined at the same site. An active electrode from the array and an electrode on the skin were connected to the stimulator for unipolar cardiac pacing. Using a pacing pulse width of 2 ms and twice diastolic threshold, the stimulator was set to deliver a train of eight stimuli at a steady (S1-S1) cycle length followed by an extrastimulus through the active electrode. The VERP was determined by reducing the S1-S2 interval in 10-ms decrements until loss of capture. This experiment was repeated for S1-S1 cycle lengths of between 400 and 250 ms in decrements of 50 ms. The MAPd90 was determined as described above. E. Recording Localized MAPs To validate the ability of the electrode array to record localized MAPs, we placed the array over the AV groove so that part of the array was over atrial tissue and part over ventricular tissue. We then paced in the atria at a rate high enough to cause 2:1 AV block. In addition, we ligated a small coronary artery while recording both from tissue distal (affected tissue) and proximal (unaffected tissue) to the site of ligation.

SAHAKIAN et al.: SIMULTANEOUS MULTICHANNEL MONOPHASIC ACTION POTENTIAL ELECTRODE ARRAY

Fig. 5.

349

Steady-state MAPd90 versus cycle length for four different electrodes.

III. RESULTS A. Evaluation of Acceptable MAP Waveforms Multiple simultaneous MAP recordings meeting the criteria described above were obtained from up to 13 out of 16 sites, see Fig. 4 (left). Ventricular MAP recordings exhibited sharp upstrokes, distinct plateau phases, and a relatively steep phase of final repolarization. Ventricular MAPs had amplitudes ranging from 10 to 40 mV, uniform morphology, and stable baselines for up to 30 min. Atrial recordings were more triangular in shape, exhibiting a sharp upstroke, a short or absent plateau phase, and a slower repolarization phase, and had amplitudes ranging from 4 to 21 mV as shown in Fig. 4 (right). B. Correlation Between MAPd90 and Cycle Length During sustained pacing, MAPd90 measurements showed an initial sharp decrease and then slowly leveled out to steady-state values over a period of 0.5 to 2 min. MAPd90 decreased with decrease in cycle length. Fig. 5 shows typical pacing rate dependencies for steady-state MAPd90 values averaged over 30 s on the left ventricle. The average MAPd90 increased linearly for values ranging from cycle lengths from 200 ms to 450 ms 0.968 to 0.993). Above 450 ms, steady-state MAPd90 measurements reached a plateau state. Premature ventricular contractions and noncapturing stimuli caused outliers in this group and

are primarily responsible for the relatively large standard deviations observed at cycle lengths 400 and 260 ms. C. Correlation Between MAPd90 and Ventricular Effective Refractory Period (VERP) Since simultaneous VERP measurements and MAP recordings could not be made from a single site, MAPd90 measurements were taken from the data collected from the sustained pacing protocol. VERP data at a specific cycle length were paired with the average of three MAPd90 measurements taken immediately after the first eight paced beats to replicate the conditions used to determine the VERP. It was found that both MAPd90 and VERP are linearly related to S1-S1 intervals less than or equal to 400 ms. Furthermore, MAPd90 and VERP seemed to run parallel across S1-S1 intervals. The VERP/MAPd90 ratio ranged between 0.74 and 0.87 in all sites except at one site that exhibited values between 1.03 and 1.12, consistently. D. Detection of Local Myocardial Events Placement of the array on the AV groove resulted in local atrial or ventricular recordings at different sites during sinus rhythm. Based on ECG and MAP morphology and timing, it was possible to distinguish easily between atrial and ventricular MAPs.

350

IEEE TRANSACTIONS ON BIOMEDICAL ENGINEERING, VOL. 48, NO. 3, MARCH 2001

Fig. 6. Simultaneous atrial MAPs (channels 1, 5, and 6) and ventricular MAPs (channels 3, 9–13, and 15) from the AV groove during 2:1 AV block during the first eight atrial/four ventricular activations. RVA is a right ventricular electrogram apex. Atrial cycle length is 240 ms. And Ventricular cycle length is 480 ms. Note that channels 1 and 3 are from the first quadrant where they share the same reference.

In those instances where atrial and ventricular MAP signals were recorded from two sites in the same quadrant and using the same reference, there were notches in atrial MAP recordings at the times of ventricular repolarization and vice versa. This demonstrates that MAP recordings are unipolar signals and that the reference serves only to minimize far field activity [8]. For this reason, having a central reference for each quadrant does not affect the array’s capabilities to study temporal characteristics of depolarization or repolarization. The local and temporal nature of these recordings was further demonstrated during instances of 2:1 AV block, which exhibited two atrial MAPs for each ventricular MAP (see Fig. 6). E. Local Ischemia Local ischemia created significant differences in MAP morphology between ischemic and nonischemic sites, see Fig. 7. Ischemic sites exhibited a decrease of the plateau and the slope of repolarization in phase 2 and 3, respectively, when compared to those obtained from normal sites. While some ischemic regions showed shorter action potential durations primarily through a shortening of phase 2, [Fig. 8(a)], some other ischemic regions did not show shorter MAP durations due to the concurrent decrease in the slope of repolarization during phase 3 [Fig. 8(b)].

IV. DISCUSSION The MAP electrode array successfully recorded multiple local MAPs, demonstrating true multisite, simultaneous repolarization mapping. The MAPs had appropriate morphologies consistent with transmembrane action potentials, acceptable amplitudes, and stable baselines. Further, it was shown that the MAP durations reproduced the action potential duration rate dependencies seen previously with the use of single MAP catheters. Specifically, the sustained pacing protocol resulted in MAPd90 changes characterized by an initial sharp decrease followed by a slower decrease to a steady-state value over a period of several minutes, as has been shown by others [14]. The MAPd90 and VERP versus cycle length showed both similarities and differences with previously published data. MAPd90 and VERP shared a parallel relation across cycle lengths, decreasing linearly with decreasing cycle lengths [6]. Although the VERP/MAPd90 ratios were constant across cycle lengths, the measured values (range 0.7 to 0.8) were slightly less than what has been reported elsewhere (0.9) [14]. This difference can be a result of temporal differences in repolarization at the site level, since the VERP and MAPd90 measurements were not performed simultaneously. To enhance the array’s performance and capabilities, the next generation of MAP arrays could include pacing elements to allow for

SAHAKIAN et al.: SIMULTANEOUS MULTICHANNEL MONOPHASIC ACTION POTENTIAL ELECTRODE ARRAY

Fig. 7. MAP recordings from the electrode array during local ischemia via constriction of a coronary artery. MAP waveforms have been normalized so that amplitudes are the same across all sites. Sites 1–4 and 9–12 are on normal tissue. Sites 5–8 and 13–16 are on ischemic tissue. Sites 2 and 7 did not have MAP recordings.

(a)

351

simultaneous pacing and MAP recording adjacent sites as in the recording/pacing MAP catheter described by Franz et al. [16]. For the site that exhibited a VERP/MAPd90 ratio greater than unity the following three explanations are plausible: 1) The time between the sustained pacing protocol and the VERP measurements was longer than 1 h, thus allowing for large differences in repolarization to occur. 2) Between the sustained pacing and VERP measurements, the electrode array was removed briefly from the tissue and the sponge was saturated again with saline to improve the recording characteristics; the array may not have been replaced exactly on the same location. 3) MAPd90 may not correlate well with VERP due to postrepolarization refractoriness [17]. The ability of the array to record local myocardial events was demonstrated by the recording of simultaneous atrial and ventricular MAPs from different sites of the array during sinus rhythm and pacing induced 2:1 AV block. The notches that appeared in the waveforms are due to incomplete far-field cancellation by the reference electrode. Finally, the ischemic protocol provided more evidence of the ability of the array to record local myocardial repolarization changes. Despite these positive results, several difficulties still exist with the use of this array which prevented us from successfully recording diagnostic quality MAPs from all 16 sites. First, it was extremely difficult to position the array such that all probes were in constant contact with the tissue. This may be due to the fact that the array was flat and unable to conform to the curvature of the heart and that the active beating of the heart caused one or more electrodes to lose contact with the cardiac tissue. One fix for this problem would be to use a material for the base that would allow the curvature of the array to fit the area currently being mapped. Second, due to the small size of the heart, when the array was placed over areas of fatty tissue or large coronary arteries, acceptable MAP waveforms were not recorded; this is a known limitation of the contact MAP recording technique. Third, the recording procedure of placing the MAP array and maintaining constant adequate pressure manually for extended periods of time proved to be difficult. A method to mechanically fix the array at a location on the heart would alleviate this problem. Finally, it should be noted that due to the amount of pressure required to obtain consistent tissue contact with as many active electrodes as possible, some damage to the epicardium resulted. Thus, given the invasive nature of the current technique, the array is not yet acceptable for clinical use. This may be resolved if the improvements described above are made to the array.

V. CONCLUSION

(b) Fig. 8. (a) Two MAPs from normal and ischemic tissue with different action potential duration and morphology. (b) Two MAPs from normal and ischemic tissue with little difference in action potential duration despite morphology differences.

Cardiac repolarization mapping can lead to a better understanding of the mechanisms of cardiac arrhythmias. Current methods to study repolarization are limited due to the technical difficulties inherent to the recording technique, such as the number of simultaneous mapping sites. The MAP electrode array described in this paper has been shown to faithfully generate simultaneous, diagnostic quality MAPs at multiple sites reflecting local repolarization activity.

352

IEEE TRANSACTIONS ON BIOMEDICAL ENGINEERING, VOL. 48, NO. 3, MARCH 2001

The development of a nonmanual method to maintain constant pressure with the tissue will facilitate the contact of all electrodes to the in situ beating heart. Additional improvements will provide more stable MAP recordings at more sites and for longer periods of time, both at the ventricular as well as the atrial level [18]. The MAP electrode array is a viable technique for multisite repolarization mapping which can be used toward the better understanding of the role of repolarization, particularly that of dispersion of repolarization, in the initiation, maintenance and termination of cardiac arrhythmias.

REFERENCES [1] I. R. Efimov, D. T. Huang, J. M. Rendt, and G. Salama, “Optical mapping of repolarization and refractoriness from intact hearts,” Circulation, vol. 90, pp. 1469–1480, 1994. [2] S. D. Girouard, J. M. Pastore, K. R. Laurita, K. W. Gregory, and D. S. Rosenbaum, “Optical mapping in a new guinea pig model of ventricular tachycardia reveals mechanisms for multiple wavelengths in a single reentrant circuit,” Circulation, vol. 93, pp. 603–613, 1996. [3] M. Wijffels, C. Kirchhof, R. Dorland, and M. Allessie, “Atrial fibrillation begets atrial fibrillation. A study in awake chronically intrumented goats,” Circulation, vol. 92, pp. 1954–1968, 1995. [4] J. Burdon-Sanderson and F. J. M. Page, “On the time relations of the excitatory process in the ventricle of the heart of the frog,” J. Physiol., vol. 2, pp. 385–412, 1882. [5] S. B. Olsson, “Monophasic action potentials from right atrial muscle recorded during heart catheterization,” Acta Med. Scand., vol. 190, pp. 369–379, 1971. [6] M. R. Franz, “Method and theory of monophasic action potential recording,” Prog Cardiovasc. Disease, vol. 33, pp. 347–368, 1991. , “Monophasic action potential mapping,” in Cardiac Mapping, M. [7] Shenasa, M. Borggrefe, and G. Breithardt, Eds. Mount Kisco, NY: Futura Publishing Co., Inc., 1993, pp. 565–583. , “Current status of monophasic action potential recording: Theo[8] ries, measurements and interpretations,” Cardiovasc. Res., vol. 41, pp. 25–40, 1999. [9] E. S. Platou, K. Steinnes, and H. Refsum, “A method for simultaneous epicardial monophasic action potential recordings from the dog heart in situ,” Acta Pharmacol. Et Toxicol., vol. 54, pp. 94–103, 1984. [10] M. Zabel, S. H. Hohnloser, S. Behrens, Y. G. Li, R. L. Woosley, and M. R. Franz, “Electrophysiologic features of torsades de pointes: Insights from a new isolated rabbit heart model,” J. Cardiovasc. Electrophysiol., vol. 8, pp. 1148–1158, 1997. [11] M. W. Kay, P. V. Bayly, and R. B. Schuessler, “Spatial sampling requirements for the determination of the spatial distribution of repolarization in atrial tissue,” in Proc. IEEE Computers in Cardiology Conf., vol. 25, 1998, pp. 129–132. [12] M. S. L. Peterson, “A simultaneous multichannel monophasic action potential electrode array for epicardial repolarization mapping,” M.S. thesis, Northwestern Univ., Evanston, IL, 1997. [13] A. V. Sahakian, M. S. L. Peterson, M. Hamer, T. Votapka, T. Ji, and S. Swiryn, “A simultaneous multichannel monophasic action potential electrode array for epicardial repolarization mapping,” in Proc. IEEE Computers in Cardiology Conf., vol. 25, 1998, pp. 125–128. [14] S. Yuan, C. Blomstrom-Lundqvist, and S. B. Olsson, “Monophasic action potentials: Concepts to practical applications,” J. Cardiovasc. Electrophysiol., vol. 5, no. 3, pp. 287–308, 1994. [15] N. Kanaan, J. Jenkins, and A. Kadish, “An automatic microcomputer system for analysis of monophasic action potentials,” PACE, vol. 13, pp. 196–206, 1990. [16] M. R. Franz, M. C. Chin, H. R. Sharkey, J. C. Griffin, and M. M. Scheinman, “A new, single catheter technique for simultaneous measurement of action potential duration and refractory period in situ,” J. Amer. Coll. Cardiol., vol. 16, pp. 878–886, 1990. [17] H. M. Leerssen, M. A. Vos, K. den Dulk, J. van der Zande, and H. J. Wellens, “Rate dependent effects of procainamide on the threshold current for pacing in the setting of postrepolarization refractoriness in dogs,” PACE, vol. 22, pp. 291–301, 1999.

[18] S. Shkurovich, A. V. Sahakian, T. Ji, T. Votapka, and S. Swiryn, “A multichannel monophasic action potential electrode array for simultaneous epicardial and endocardial repolarization mapping,” in Proc. IEEE Computers in Cardiology Conf., vol. 26, 1999, pp. 281–284.

Alan V. Sahakian (S’84–M’84–SM’94) received the Ph.D. degree in electrical engineering from the University of Wisconsin, Madison, in 1984. Since then he has been on the faculty of Northwestern University, Evanston, IL, where he is currently Professor of Electrical and Computer Engineering (ECE) and Biomedical Engineering and a member of the Associate Professional Staff of Evanston Hospital. His research is in the general area of medical instrumentation, with a special focus on the electrophysiology and automatic treatment of the atrial cardiac arrhythmias. He also works on problems of aviation systems reliability and microwave/millimeter-wave communications and imaging systems. He is the author or co-author of over 100 papers, abstracts, and book sections. Dr. Sahakian recently served as EMBS Vice President for Publications and Technical Activities and is active on several EMBS committees.

Ming-Shing Lee Peterson was born in 1970 in Taipei, Taiwan. She received the S.B. degree in electrical engineering from the Massachusetts Institute of Technology (MIT), Cambridge, in 1992, and the M.S. degree in biomedical engineering from Northwestern University, Evanston, IL, in 1997. From July 1992 to August 1995, she was a Software Engineer at Medtronic, Inc., Minneapolis, MN. From September 1995 to June 1997, she was a Medtronic Scholar. From July 1997 to September 1998, she worked at Medtronic, Inc., as a Software Systems Engineer.

Sergio Shkurovich (S’90) received the B.S. degree with honors in mechanical and electrical engineering in 1993 from the National University of Mexico, Mexico City. In 1996, he received the M.S. degree in biomedical engineering from Northwestern University, Evanston, IL. He is a Fulbright scholar currently working toward the Ph.D. degree in biomedical engineering at Northwestern University. His research interests are in biomedical digital signal processing with a special focus on cardiac electrophysiology and, in particular, atrial fibrillation.

Mark Hamer is a clinical cardiac electrophysiologist currently practicing in Rochester, NY.

Timothy Votapka attended the University of Kansas, Lawrence, for his undergraduate studies and for medical school. He did his cardiothoracic surgery residency at Northwestern University, Evanston, IL, followed by a one-year fellowship in thoracic transplantation at St. Louis University, St. Louis, MO. His research interests include atrial fibrillation and mitral valve disease.

SAHAKIAN et al.: SIMULTANEOUS MULTICHANNEL MONOPHASIC ACTION POTENTIAL ELECTRODE ARRAY

Tongyou Ji received the M.S. degree in computer and information sciences from Knowledge Systems Institute, Skokie, IL in 1999, the M.S. degree in cardiac electrophysiology from Qingdao Medical School, Qingdao, R.O.C., in 1982, and the M.D. degree from Tianjin Medical School, Tianjin, R.O.C., in 1978. He currently works in Evanston Northwestern Health Care, Evanston Hospital, Evanston, IL.

353

Steven Swiryn (M’89) received the B.A. degree in psychology from the University of Michigan, Ann Arbor, in 1968 and the M.D. degree from the University of Illinois, Urbana, in 1973. He continued his graduate medical training at the University of Illinois Hospitals and trained in Clinical Cardiac Electrophysiology under the late Dr. K. Rosen. He is currently Professor of Medicine, Northwestern University Medical School, and Senior Attending Physician and Director, Cardiac Electrophysiology, Evanston Hospital, Evanston, Illinois. His research interests include cardiac arrhythmia and conduction disease, especially atrial fibrillation mechanisms, treatment, and detection, and techniques of multichannel cardiac mapping. Dr. Swiryn is board certified in internal medicine, cardiovascular disease, and in clinical cardiac electrophysiology.