Optimal Pacemaker Sensing with Respect toAmplitude and Slew Rate of IntracardiacElectrograms: Theoretical Analysis hyComputer Simulation

YUTAKA ARM. MASARU YAMAZOE, TETSURO TOEDA. YUMIKOHOSHINO. AKIRA SHIBATA, HIDEO MAKINO. and YOSHIAKI SAITOHFrom The First Department of Medicine, School of Medicine and The Department of InformationEngineering, Faculty of Engineering. Niigata University, Niigata, Japan

ARAI, Y., ET Al..: Optimal pacemaker sensing with respect to amplitude and slew rate of intracardiacelectrograms: Theoretical analysis by computer simulation. Tho oplimai timp/i(ude and sJeiv rate ofinlrucurdiuc elecliugrams for pacemaker sensing Here examined on a Iheorel/cal basis by computersimulation. The simulation was based on (he concept that it is Ihe vo/loge at the position of (he pacingeiectrode in «n electrical field of a moving eiectricai dipole. By changing the dislance ix;liveen theeiecfrode and the myocardium and the moving velocity of the electrical dipolo, siniuJaletJ ECGs ivilharbitrary amplitudes and slew rates were generated hy the computer and fed to a bandpass filter. Thisfilter was equivalent to those assembled in some models of permanent pacemakers and had a centerpuss-band freqaency of 50 Hz and a Q of 1.0. The outputs of the filter were measured. The resulls showedthat, for pacemaker sensing, simulated intracardiac eleclrograms iW(h high amplilude should have highslew rates and those with low amplitudes should have low slew rates, althuugli (lie absolute valuesdepend on the characteristics of the bandpass filter and the sensing threshold of (lie pacemaker. (PACE,Vol. 7, September-October, 1984)

pacemaker sensing, slew rate, intracardioc electrogram

Pacemakers should sense intracardiac electro-grams lo avoid competition witb spontaneous heartbeats or to synchronize P-waves in VDD, etc.' Todiscriminate intracardiac R- or P-waves from T-waves, or noise sucb as electromyopotentials,pacemakers employ sensing amplifier filters witba pass-band of 25 to 50 Hz. Sensing characteristit:sshould be appropriate to tbe aniplitudu and theslew rate of the intracardiac electrograms.^-^

The purpose of our study was to examine op-timal paciMnaker sensing with respect to ampli-tude and slew rate of intracardiac electrograms ona theoretical basis by computer simulation.

Address for reprints: Yutaka Aral, M.D.. 2-12. 5 chome, Toraei.Niigata City, 'JSO. |apaii.

Received March 10. 1983; revision received June 22, 1983;September 2G, 1983.

Method

A simulated waveform was used for intracardiacelectrograms. This was composed according !o To-gawa's equation.•'•^ Tbis simulated waveform is tbevoltage at the point A in tbe electrical field of anoloclrical dipole (Fig, 1). Suppose tbis electricaldipole moves, i.e. it first approaches tbe point A.passes it and tben goes away; the voltage at thispoint A is determined by tho following equation(see appendix):

V(t) = C(b - (b - 15]

Where V(t) = Voltage at tbe point A;t = time:

d - the smallest distance between poitilA and the electrical dipole;

b = half of tho distant:^ botwooii tbestarting point and tbe end point of

778 September-October 1984. Part V PACE, Vol. 7

AMPLITUDE, SLEW RATE AND SENSING

-q

Figure 1. The electric potcntiol at the point A in ane)ot:(ric/ieJd generated hy on electric dipole vvlfh )eng(h1 and with charges ±q as represented by the followingformula: V =

the movement of the electrical di-pole;

C = constant;V = movingvelocityoftheelectrical di-

pole.

In this equation, q, 1 and e,, [dielectric constant)were assumed to be constant and included in C.The peak-to-peak amplitude of this equation (a),which corresponds to the R-wave amplitude, hadthe foUowing relationship with d [see appendix):

d = [14]

The maximum value of dV(t)/dt, which corre-sponds to the slew rate [S), has the following re-lationship with d and v (see appendix);

V = -d^S/C [10]

Utilizing equations [5], [14] and [10] of the ap-pendix, simulated R-waves with arbitrary ampli-tudes and slew rates were composed using a dig-ital computer and generated through the digital-to-analog converter every second (Fig. 2). This wavewas fed to a bandpass filter with a center frequencyof 50 Hz and a Q of 1.0, which is equivalent to

Figure 2. The bJock diagram of the system used forthe experiment.

those assembled in a model of implantable pace-makers (Fig. 3).

The amplitudes of the waveforms at the inputand output terminals of the bandpass filter weremeasured using a 314 storage oscilloscope.*

11.0-

10.0

9.0-

e.o-

7,0-

6.0-

5.0-

*.a-

3.0-

Z.O-

1.0-

6.25 12.5 25 SO 100 200 400

Frequency (Herz)

Figure 3. The characteristic of tlm bandpass filter whichis equivaleni to those assembled in some modeis ofimpiantab/e pacemakers. The center pass-band fre-quency is 50 Hz and the Q is 1.0.

'Sony Tektronix. 5-9-13, Kitashinagawa. Shinagawa Ku, To-kyo. 141.japan

PACE, Vol. 7 September-October 1984. Part V 779

ARAI. ET AL.

SimulatedR -wave

Filter out

>Eo

100 msecFigure 4. The upper waveform is an example of the simulated intracardiac electrogram dis-played on the oscilloscope. The lower trace is the filter output of the above curve.

Results

In Figure 4, examples of the simulated R-waveand tho output waveform of the bandpass filter areshown.

Table I shows the output voltage of the bandpassfilter of the simulated R-wave with various am-

plitudes and slew rates. In some cases, the outputis larger than the input, because the bandpass filterused was an active one with a gain of 2.75 about50 Hz.

The matrix of Table i is illustrated in contourlines in Figure 5, which illustrates the following:the range of the slew rate that can be sensed is

Table I.

The Filter Outputs (mV) of the Simulated Intracardiac Electrograms with Various Amplitudes (mV) and SlewRates (V;s)

R-waveamplitude

(mV)

28.020.014.010.07.05.03.52.51.8

O.S

2.162.933.364.004.746.605.974.002.74

0.7

3.564.535.607.007.747.406.353.832.28

1.0

6.678.53

10.111.710.58.405.973.131.71

1.4

10.714.916.114.011.57.805.232.431.25

Slew Rate (V/s)

2.0

20.022.120.715.510.76.403.731.760.98

2.8

30.729.322.414.59.064.802.521.360.65

4.0

41.332.721.012.06.593.201.491.130.53

5.6

45.331 317.99.504.742.740.930.750.36

8.0

40.123.412.76.453.181.850.880.500.28

780 September-October 1984, Part V PACE. Vol. 7

AMPLITUDE. SLEW RATE AND SENSING

28.0

Slew rate of the simulated R-wave (V/sec)Figure 5. The data in Tnble 1 are displayed in confoLir lines. The unit of /igures on confourJines is in mV, From these conlour Jines. the/oiJowing can be read; suppose (he sensing thresholdat the filter output (erminaJ is 8.0 mV and the amplitude of an R-wave is 7.0 mV, the range ofthe slew rate that can be sensed is 0.75-3.20 V/s. In addition, suppose the slew rale of an R-wave is I.O V/s, the range of R-wave amplitude that cun be sensed is 5.0-20.0 mV al thn samesensing threshold.

0.75-3.2 V/s, if a bandpass is given with a centerfrequency of 50 Hz and a gain of 2.75, the sensingthreshold at the filter output terminal is 8.0 mV,and tho amplitude of the simulated R-wave is 7.0mV. In addition, if the slew rate of the R-wave is1.0 V/s, the range of the simulated R-wave am-plitude that can be sensed is 5.0-20.0 mV at thesame sensing threshold.

Discussion

Excluding sensing failure responsible for pace-makers, there still remains a demand failure rateof 2-4%.*'^ To improve this situation, the ampli-tudes and slew rates of intracardiac electrogramsshould be checked to see if they match with the

characteristics of the bandpass filter.- In otherwords, the center frequency of the bandpass filtershould be appropriate to the amplitude and thefrequency component ol the electrograms.

The sine wave is less appropriate as a simulatedelectrogram because its frequency component issingle. So in this study, a more theoretically rea-sonable waveform was employed as the simulatedintracardiac electrogram.

Togawa attempted a more elaborate simulationof the intracardiac electrogram. taking the numberof myofibrils and the conducting direction of ex-citation into consideration, and composed QRScomplexes of R, RS and QS patterns.'*'' To simplifythe experiment, in this study only the diphasicpattern was employed. By changing the nearest

PACE, Vol. 7 September-October 1984, Part V 781

ARAL ET AL.

position ofpacing electrode

movingelectricdipole

starting pointof excitation

end ofexcitation

b —-^ ^ bFigure 6. The electric dipoie illustrated in Figure 1 was supposed to move according to (heconduction of" excitation in the heart. The changes of eiectn't: polentiai at (he position of thtipacing eiecfrode were considered lo be the simulated intracardiac eleclrograni, v = conductionvelocity of excitation; b = haJ/Jength of myocardium, t = time; d = distance between electrodeand myocardium. See the appendix for details.

distance between point A and the electrical dipole.which corresponds to the distance between thepacing electrode and the myocardium, and bychanging the moving velocity of the electrical di-pole, which corresponds to the conduction veloc-ity of excitation, simulated intracardiac electro-grams with arbitrary amplitudes and slew rateswere generated.

The characteristic responses of the bandpass fil-ter to continuous vi aves and single pulses likeelectrocardiograms are quite different. The selec-tivity of the latter is poor, particularly at higherfrequency part. Therefore, although the slew rate(or frequency component) is too high, sensing maybe possible. This is also apparent from our data ifthe amplitudes of the R-waves are high enough.Whon tbe amplitude is low, the range of slew ratethat can be sensed becomes narrow and low.

Ohm reported the conditions of R-wave ampli-tude and slew rate uuder wbich sensing is possible

in various pacemakers using mainly triangularwaves." The result was different from ours as tothe lower limit of the optimal slew rate. This isprobably due to the differences in tho simulatedwaveforms.

In conclusion, the present study indicated tbatintracardiac electrograms with high amplitudes,but not high slew rates, may result in underson-sing. In addition, intracardiac electrograms witblow amplitudes, but not low slew rate;, may aisoresult in undersensing.

In multiprogrammable pacemakers, undersen-sing can be avoided by lowering tho sensingthreshold. But tbis. on the other hand, induces thehazard of oversensing noisos sucb as oluctromy-ograms. If the center frequency of tbe bandpassfilter can be changed by an external controller, itcan be another effective means of avoiding de-mand failures.

References

1. Parsonnet, V.. Funnan, S.. Smyth. N.P.D.: Implant-able cardiac pacemakers: Status report and resourceguideline. Am. /. Cardiol.. 35:497. 1974.

2. Furman. S.. Hurzeler, P.. DeCaprio. V.: Cardiac pac-ing and pacemakers II. Sensing the cardiac eleclro-gram. Am. Heart /.. 93:794. 1977.

3. BarDld. S.S.. Ong. L.S.. Heinle, K.A.: Stimulationand sensing thresholds for cardiac pacing: Electro-physiologic and technical aspects. Prog. Cardio-vasc. Dis., 24:1. 1981.

Togawa. T.: A method of computation of endocar-dial surface electrogram. K. Matsutia (Ed.): Outlineoj i'hy.si()Jngy (Supplement). Tokyo. Igiikii Shoiu.1971. p. 118 (in Japantt.se).Togawa. T.; Technological problems of pacemaker:Particularly about sensing. Cardiac Pacing, 1980. p.1 (in Japanese).Ohm, O.-J.: The interdepondencH between electro-gram, total electrode impedance aiitl pat:emaker in-put impedance necessary to obtain adequate func-

782 September-October 1984. Part V PACE. Vol. 7

AMPLITUDE, SLEW RATE AND SENSING

tioning of demand [Hicemakers. PACE. 2:465, 1979.7. Furman. S.. Hurzeler. P.. De Caprio, V.: The ven-

tricular endocardial ehictrograni and pacemakersensing. /. Thorac. Curdiovasc. Surg.. 73:258, 1977.

8. Ohm, O.-f.: Inhibition/filter characteristics and in-put impedances of QRS-inhibited demand pace-makers by in vitro studies. PACE, 3:318, 1980.

APPENDIX

From Figure 1 :

V = qlcose/4ireor^

r = (b - vt) + d

From Figure 6:

cose = (b - vt)/{d^ + [b - vt)^} '

C = ql/4'iTe,,

substituting [3| and [4| in ]1|:

V(t) = C[b - vt)/{d^ + {b - vtl^P' '

differentiating [5] with respect to time:

dV[t)/dt = -Cv/{d^ + (b - vt) }*' + 3Cv(b - \{d + [b — vt)^) '

Tbe maximum absoiute va]ue of dV(t)/dt is the slevi

(S) and is obtained by the following condition:

b - vt = 0

ort = b/v

substituting 18] in [6]:

S = -Cv /d '

rearranging [9]:

V - -d-'S/C

[1][2]

[3]

[4]

[5]

[6]

/ rate

[7]

[8]

[91

[10]The maximum value of V(t) if a half of its amplitude(a) and is obtained with the condition below

dV(t]/dt = 0

applying this condition to 16]

b - vt = d/V2substituting [12] in [5]

rearranging [13]d = \/2C''V3-'^a"^

[11]

[121

[131

[14]

Simulated intracardiac electrogram with arbitrary amplitudeand siew rate could be obtained using equations [14], [10], and[5].

PACE, Vol. 7 September-October 1984, Part V 783