experimental and theoretical investigation of ... - Springer Link

Journal of Clinical Monitoring and Computing (2009) 23:63–73 DOI: 10.1007/s10877-008-9157-5

Springer 2008

EXPERIMENTAL AND THEORETICAL INVESTIGATION OF IMPLANTABLE CARDIAC PACEMAKER EXPOSED TO LOW FREQUENCY MAGNETIC FIELD

Babouri A, Hedjeidj A, Guendouz L. Experimental and theoretical investigation of implantable cardiac pacemaker exposed to low frequency magnetic field. J Clin Monit Comput 2009; 23:63–73

A. Babouri, PhD1,2, A. Hedjeidj, PhD1 and L. Guendouz, PhD1

implantable cardiac pacemaker exposed to low frequency magnetic fields. The method used in this study is based on the interaction by inductive coupling through the loop formed by the pacemaker and its loads and the surrounding medium. This interaction results in an induced electromotive force between the terminals of the pacemaker, which can potentially disturb its operation. The studied frequencies are 50/60 Hz and 10/25 kHz. The experimental tests were carried out on several cardiac pacemakers, single chamber, and dual chamber. The results show a window effect of the detection circuits of cardiac pacemakers for the four studied frequencies. The modelling of the test bed requires studying the effects of the induced currents generated by the application of a magnetic field. Analytical calculations and Numerical simulations were carried out. We modelled the interactions of the magnetic field with a simplified representation of pacemaker embedded in the medium. The comparison of the results in the air and in vitro enabled us to make an equivalent electric model. The results obtained in experimental and theoretical studies allowed us to validate the test bed. The method applied is valid for other medical implants such as cardiac defibrillators, implant hearing aids system…etc.

ABSTRACT. This paper presents in vitro investigation of an

KEY WORDS. pacemakers electromagnetic compatibility, frequency interferences, eddy current.

low

INTRODUCTION

From the 1Electronic Laboratory of Instrumentation Electronic of Nancy, Faculty of Science and Technology, UHP Nancy 1, Nancy Cedex, France; 2Laboratory of Electrical Engineering LGEG, Faculty of Science and Engineering, University of Guelma, Universite´ de 8 mai 1945 a` Guelma, B.P 401-24000, Guelma, Alegria. Received 19 October 2008. Accepted for publication 4 December 2008. Address correspondence to A. Babouri, Laboratory of Electrical Engineering LGEG, Faculty of Science and Engineering, University of Guelma, Universite´ de 8 mai 1945 a` Guelma, B.P 401-24000, Guelma, Alegria. E-mail: [email protected]

An important number of pacemaker’s carriers are exposed daily to numerous sources of electromagnetic interferences that can create some influences on the activities of the implanted devices. Numerous, studies dealing with interferences between low frequency electromagnetic fields and these implants have been published [1, 2]. These studies originally were concerned by interferences caused by the power distribution network [3–5]. More recently, studies about the interferences between electronic article surveillance systems and stimulators have also been published [6–8]. Besides, studies have been published and concern with the mobile telecommunication ground, which has incredibly scared over the last few years [9–12]. This ground, a.k.a. intermediate frequencies, is made more complicated by the diversity of technical data. In the mobile telecommunication ground, a common denominator exists between the different manufacturers

64

Journal of Clinical Monitoring and Computing

Fig. 1. Test bed instrumentation.

(frequencies, sources strengths, beaming elements, etc.). On the contrary, for systems implementing low frequency electromagnetic sources that a carrier may be exposed to, in everyday life or in industry (e.g., electrical appliances or induction heating, electronic article surveillance systems), the main difficulty essentially linked with the great diversity of situations which makes comparing results difficult. The induced currents generated by the exposure of the human body to a magnetic field are very considerable. Studies were carried out to calculate these currents analytically [13–15] or numerically. Among the methods used are FEM [16, 17], method of impedance [18], and FDTD [12]. An experimental investigation in free space and in vitro studied the behavior of the cardiac pacemakers exposed to conducted and radiated disturbances [19, 20]. This article presents a study of the effects of eddy currents induced on medical implant by magnetic coupling. The finite element method is used in this study.

EXPERIMENTAL INVESTIGATION

The instrumentation of experimental measurement is composed by a disturbance source. It consists of two 1.5 m loops separated by a 0.75 m gap. The loops are formed by six jointed turns of 16 mm2 wire; this structure is powered by a programmable generator and is used to put a cardiac pacemaker in a controlled magnetic field. The current crossing the Helmholtz structure is measured via a current probe whose sensitivity is 100 mV/A.

Taking into account the limits of the source amplifier, the maximal values of the field vary from 157 lT for a 50 Hz frequency signal to about 8 lT for 25 kHz frequency (Figure 1). Tracking and programming of the pacemaker housing is achieved with the telemetry system. In this study, the devices have all been configured in inhibited stimulation (SSI or VVI mode according to the international codification), which is the most widespread configuration. The device under test (DUT) is composed by the leadequipped pacemaker and forms a loop. This last is completed by a coplanar rectangular loop made of 30 or 90 jointed turns. The areas are, respectively, A 180 cm2 and Aapp 160 cm2; the total effective area is about 4,980 cm2 (for the case of 30 turns) and using this additional loop allows the increase of the effective loop area formed by the pacemaker and its probe. For dual chamber pacemaker, each lead is completed by an additional loop. This device is studied in two configurations (Figure 2). For the first one, consisting of the ‘‘free space’’ tests where the device is completed by a resistance R representing the load of the pacemaker. In this case the DUT is placed on a directional support at the centre of the Helmholtz source. The second case corresponds to in vitro tests; the probeequipped pacemaker is inserted in the gelatine based model [21]. All the experiments were carried out in a Faraday cage to ensure neutral electromagnetic conditions. The tests were carried out on 11 pacemakers’ manufacturered by Ela Medical, six of which are single chamber (A–F) and five dual chamber (G–K). Table 1 gives the references devices tested.

Babouri et al.: Experimental and Theoretical Investigation of Implantable Cardiac Pacemaker

65

Fig. 2. The device under test (DUT).

Table 1. References of tested pacemakers Single chamber pacemakers

Dual chamber pacemakers

OPUS 4023 (A) Talent SR 113 (B) OPUS G4624 (C) OPUS S 4124 (D) Opus 4003 (E) Regency SR+ (F)

Talent SN 009UF081(G) Talent SN 952UF013 (H) Talent SN030UFI84 (I) Brio DR SN 02UK092 (J) CHORUS 6033 (K)

Characterization of the disturbance source The experimental characterization of the source is the measurement of the magnetic field generated in the volume between the loops. These tests were performed using a probe to take measurement in the three directions. Figure 3 shows for example changes in the magnetic field at the center of the structure as a function of the effective value of the current through the coils (Helmholtz coils) for a frequency of 50 Hz. The power of the amplifier is 500 VA, the maximum values of the field are 157 lT for a frequency of 50 Hz and 7.92 lT for 25 kHz. The homogeneity of the field has been verified in a volume equal to the tank containing Table 2. The characteristics of the source for the four frequencies studied Frequency

50 60 10 25

Hz Hz kHz kHz

k (lT/A)

6.40 6.35 5.72 5.62

Maximum field in the center

157.6 153.1 8.5 7.92

Maximum field recommended in the European recommendation (lT) 100 83.3 6.25 6.25

Fig. 3. Magnetic field at the center of the Helmholtz coils.

the pacemaker. Table 2 shows the characteristics of the source for the four frequencies studied. K is the factor of proportionality between the magnetic field and the electrical current. The last column represents the maximum values allowed by the European recommendation of the public exposure to electromagnetic fields.

THEORY AND MODELING

In free space We present the analytical calculation in free space of the electrical voltage generated by the magnetic field through the loop formed by the pacemaker and its probe. From the Faraday law: The electromotive force induced through a loop is:

66


e¼

@U ! ! and U ¼ B S @t

In quasi-static, time-harmonic analysis: ! @B ! ¼ jx B @t

ð1Þ

ð2Þ

By inserting Eq. 2 in Eq. 1 the electromotive force (emf): ð3Þ e j ¼ S 2p f B j! 0

which S0 is the loop area formed by the DUT and f is the frequency signal applied (the magnetic field). The induced voltage in the terminals of the cardiac pacemakers is: V0 ¼

R0 S0 2p f B R0 þ RS

ð4Þ

where R0 and Rs represent, respectively, the internal resistance and the load of the cardiac pacemakers. In this case the electrical model of the system subjected to a disturbance radiated in free space me be represented by an EMF generated by the magnetic field through the loop in series with the internal impedance of the pacemaker Z0 and load impedance Zs (Figure 4).

In vitro The test bed is modeled by the Helmholtz coil as a source of disturbance and the tank filled by a gelatin whose electrical conductivity is r = 0.1S/m and its permittivity is e0. ! An excitation by a magnetic field B 0 ; will generate an ! electrical field E 1 according to Faraday-Maxwell law: ! @B0 ! ! r E1 ¼ @t ! ! This al field ð E 1 Þ creates the induced currents J 1 according to Ohm law: ! ! J1 ¼ r E 1

Fig. 4. Equivalent model of the pacemaker circuit in free space.

This current, in turn, gives rise to a secondary magnetic field according to Maxwell–Ampere’s law. ! ! The induced currents J 1 create a magnetic field B 1 according to Maxwell–Ampere law: ! ! ! r H 1 ¼ J 1 ðBy neglecting currents displacementÞ ! ! As B 1 ¼ l H 1 ; This is added to the magnetic field ! generated by the disturbance source B 0 : With the same ! ! ! reasoning: B 1 creates E 2 which creates J 2 . . . etc. We can summarize these phenomena by an organigram composed of three series; we will call it the Triangle of Induction TI (Figure 5). To analyze these phenomena of induced currents generated by the low frequency electromagnetic fields [14], [16], and [13], we have adopted quasi-static formulations. These last consist of writing the Maxwell equations in harmonic regime @t@ , jx ; neglecting the ! displacement currents @@tD and the phenomena of wave propagation. Model of DUT without the pacemaker First of all, we present the analytical calculation of the eddy current generated by an homogeneous magnetic field in a Tank containing a gelatin-based-model simulating the al conductivity of tissues without the pacemaker (Figure 6).

Fig. 5. Triangle of induction.

Fig. 6. In vitro model without the pacemaker.


67

Maxwell equations: Maxwell–Faraday: ! @B ! ! r E ¼ @t

ð5Þ

Maxwell–Ampere: ! ! ! ! @D rH ¼ J þ @t

ð6Þ

! ! r B ¼0

ð7Þ

! ! rD ¼q

ð8Þ

Fig. 7. Mesh 2D of simple model.

From these equations: @ 2 Hx @ 2 Hx ! þ jxrl H ¼ 0 @y2 @z2

RESULTS AND DISCUSSION

ð9Þ

Or ! ! ! ! r r J ¼ jxrl J

ð10Þ

The analytic solution of the Eq. 10, when considering the boundary conditions is as follow: Hx = H0 in the border y = ±a and z = ±h 1 xlrH0 ð2aÞ X 4 ð1Þn 2 p2 0 ð2n þ 1Þ ! sinh ð2nþ1Þp y ð2n þ 1Þp 2a y cos 2a cosh ð2nþ1Þp y 2a

Jy ffi j

Experimental results In the Figures 8 and 9, we present experimental statistical results obtained by telemetry system. On these figures, the x-axis represents the intensity of the applied magnetic field. The vertical axes present the value given by the statistical counters. These figures, present the detection levels of a cardiac pacemaker type single chamber exposed to a 10 kHz magnetic field. Figure 8 corresponds to tests in free space, where the detection levels are between 1.09 and 1.51 lT. Figure 9 corresponds to tests in vitro, where the levels of detection are between 0.69 and 1.04 lT. These experimental results show that the detection levels in free space are low compared to the levels of

1 xlr H0 ð2hÞ X 4 ð1Þm 2 p2 0 ð2m þ 1Þ sinhð2mþ1Þp ! y ð2m þ 1Þp 2h z cos coshð2mþ1Þp 2h a 2h

Jz ffi j

n, m are integers. Model of DUT with pacemaker For the DUT with the pacemaker, it’s difficult to determine analytical effects of eddy currents. But it is possible to do it by numerical solution for a simple geometrical model. This simulation is solved by COMSOL Multiphysics 3.2 software which uses the Finite Element Method (Figure 7).

Fig. 8. Behaviour of pacemaker exposed to a magnetic field (free space).

68


Table 5. Influence of sensitivity on the levels of detection

Fig. 9. Behaviour of pacemaker exposed to a magnetic field (in vitro).

detection in vitro. These results confirm the influence of the currents induced on the levels of detection. For tests in the free space, Tables 3 and 4 summarize the detection ranges that were obtained on pacemakers single and dual chambers. These last are equipped with one or two additional coplanar rectangular loop made of 90 jointed turns according to their types. The Table 5 Gives detection ranges of the pacemaker A in the free space, for magnetic fields of frequencies 10 kHz and 25 kHz. In this case the pacemaker form via

Sensitivity

Detection ranges at 10 kHz

Detection ranges at 25 kHz

0.7 mV 2 mV 3 mV

1.48 lT–1.84 lT 1.22 lT–1.55 lT 3.90 lT–4.51 lT 2.85 lT–3.10 lT -3.93 lT–4.48 lT

the probe, the coplanar rectangular loop made of 30 jointed turns and the load resistance, a loop area of S = 4,980 cm2. This table also shows the influence of sensitivity on the levels of detection. For tests in vitro, Tables 6 and 7 summarize the results obtained for a dual chamber and a single chamber pacemaker. The tests in Table 6 were carried out using two coplanar rectangular loop made of 30 jointed turns. For Table 7 tests were carried out using a coplanar rectangular loop made of 90 jointed turns. Evaluation of the equivalent magnetic field The tests of radiated disturbances necessitate, to reach the threshold of detections, the use of an additional loop in order to increase the surface coupling. Also the detection levels obtained by statistical counters do not correspond to a realistic case of implantation. The modelling which is presented below is based on experimental tests (Figures 10

Table 3. Detection ranges, dual chamber pacemakers-in free space tests F

Pacemaker H

Pacemaker G

SA = 0.4 mV et SV = 2.2 mV

SA = SV = 1 mV

SA = 0.4 mV et SV = 2.2 mV

SA = SV = 1 mV

25 kHz 10 kHz

0.76 lT–12.5 lT 1.8 lT–28.8 lT

14.7 lT–24.5 lT 1.85 lT–30.8 lT

0.54 lT–9.7 lT 1.8 lT–22.2 lT

5.6 lT–18.5 lT 0.7 lT–8.9 lT

F

Pacemaker I

25 kHz 10 kHz

Pacemaker J

SA = 0.4 mV et SV = 2.2 mV

SA = SV = 1 mV

SA = 0.4 m V et SV = 2.2 mV

SA = SV = 1 mV

4 lT–32.1 lT 1.8 lT–28.8 lT

6 lT–9.6 lT 0.96 lT–29.6 lT

– 3.8 lT–28.17 lT

– –

Table 4. Detection ranges, single chamber pacemakers-in free space tests F

25 KHz 10 KHz

Pacemaker C

Pacemaker B

S = 0.4 mV

S = 1 mV

S = 0.4 mV

S = 1 mV

2.98 lT–12.1 lT 2.9 lT–19.8 lT

3.1 lT–14.5 lT 4.5 lT–26.24 lT

8.89 lT–31.38 lT 1.8 lT–22.2 lT

9.21 lT–30.64 lT 1.9 lT–24.8 lT


and 11), concerning the case of single chamber pacemaker (OPUS4023). These experimental results show that the ratio of two surfaces is equivalent to the ratio of the slopes for the two configurations: Table 6. Detection ranges, dual chamber pacemaker-in vitro tests F

Pacemaker k SA = 0.4 mV et SV = 2.2 mV

25 10 50 60

kHz kHz Hz Hz

1.56 3.43 4.96 4.61

lT–1.83 lT–3.91 lT–7.80 lT–6.32

lT lT lT lT

SA = 1.5 mV et SV = 1 mV 1.34 lT–1.88 lT 3.18 lT–4.06 lT 4.96 lT–5.44 lT

S 4;980 345:17 , ¼ 0 S 180 12:59

Table 8. Detection ranges for a surface of 200 cm2 Pacemaker k

Table 7. Detection ranges, single chamber pacemakers- in vitro tests

25 KHz 10 KHz

Pacemaker A S = 1 mV

Pacemaker D S = 2 mV

0.41 lT–1.41 lT 0.84 lT–2.5 lT

14.8 lT–15.3 lT 12.2 lT–18.8 lT

ð11Þ

The relation (11) allowed us to recalculate the detection ranges, for the surface of 200 cm2 (the surface used in the standards) without repeating the steps. For example for the detection sensitivity of 0.7 mV and the frequency of 25 kHz for the surface of 180 cm2, the range of detection is between 34 lT–43 lT and for the surface of 200 cm2, the range of detection is between: 30.6 lT–38.7 lT. Table 8 summarizes the detection ranges for the pacemaker K. These values corresponding to the equivalent levels of detection for a coupling surface of 200 cm2.

F

F

69

25 10 50 60

KHz KHz Hz Hz

SA = 0.4 mV et SV = 2.2 mV

SA = 1.5 mV et SV = 1 mV

38.84 lT–45.56 lT 85.40 lT–97.35 lT 123.5 lT–194.22 lT 114.78 lT–157.36 lT

33.36 lT–46.81 lT 79.18 lT–101.09 lT 123.50 lT–135.45 lT –

Fig. 10. Characteristics of transfer function Vbe = f(Isource) without the additional loop.

Fig. 11. Characteristics of transfer function Vbe = f(Isource) with the additional loop (30 turns).

70


They were obtained from the trials, which we recall, have been achieved with an additional loop. It is possible to study the effects of experimentally induced currents from a comparison of the statistical curves that were made in the air and in vitro.

Numerical simulation Model without pacemaker Figure 12 shows the representation of the currents induced in 2D. In both cases the induced currents are maximal in the boundaries of the tank (This last one contains the medium which simulates the electrical characteristics of the tissues). Those currents are null towards the center and in the four corners of the tank. The comparison between the results obtained by analytical calculation and numerical simulation are comparable. For example, in the case of 50 Hz frequency there is a maximum current density of 0.189208 9 10-3 A/m2 by the analytical calculation and 0.1892194 9 10-3 A/m2 (2D model) and 0.19525 9 10 -3 A/m2 (3D model) by numerical simulation.

Table 9 summarizes the results of the mathematical model and numerical simulation. The simulation results show that in 3D there’s a difference of 3% compared to 2D simulation. This discrepancy is due to the fact of not having enough refined meshes with the software used. All simulations were carried out by a PC with a Pentium IV at 3 GHz. The 3D simulation is expensive in terms of memory requirements and also in terms of time simulation. In addition the results of 2D and 3D modelling are almost identical with a difference of 3%. That is why we preferred 2D modelling. With simple model pacemaker The model of pacemaker proposed for the simulation consists of an open loop isolated with a resistance of 3 KX. This resistance is equivalent to the internal resistance of the pacemaker. Only the extremities of this loop are in contact with the coupling medium. Figure 13 shows the distribution of the current induced, in the case of the model with pacemaker. The extract of the current density of the Figure 14 compared to the horizontal axis shows the same profile

Fig. 12. 2D distribution of eddy current-model without pacemaker.

Table 9. Induced currents-comparison between the analytical calculation and numerical simulation F(Hz)

50 60 10,000 25,000

Mathematical model (A/m2)

Numerical simulations (A/m2)

2D model

2D model

3D model

0.189208 9 10-3 0.227049 9 10-3 37.8415 9 10-3 94.6039 9 10-3

0.1892194 9 10-3 0.2270633 9 10-3 37.846 9 10-3 94.646 9 10-3

0.19525 9 10-3 0.234063 9 10-3 39.051 9 10-3 97.628 9 10-3


71

Fig. 13. 2D distribution of eddy current- model with pacemaker (F = 10 kHz).

Fig. 14. Profile excerpt of the current density for y = 0, with simplified model pacemaker.

obtained in Figure 15 (the case of the tank without pacemaker) with two peaks at the location of the loop simulating the cardiac pacemaker. Both peaks give the total value of the current density generated by the magnetic field and the induced currents generated by the medium. The Figure 16 shows a comparison between the cases in free space and in vitro, of the electrical voltage generated by the magnetic field through the loop formed by the pacemaker and its probe. This comparison shows a very important difference between the two cases. For the first time, we present an original equivalent electrical model. This model represents the behavior of

Fig. 15. Profile excerpt of the current density for y = 0, without pacemaker.

pacemaker when exposed to a low frequency radiated disturbance in vitro (Figure 17). Veb ¼ ðFEMAir þ FEMi Þ

Ri Ri þ R

Ri and R represent, respectively, the internal resistance of the cardiac pacemaker and the equivalent resistance of the surrounding medium (equivalent model of tissues). FEMAir is the electromotive force between the terminals of the pacemaker generated by the magnetic coupling (Faraday law) and FEMi is the induced electromotive force generated by the induced currents in the presence of the medium (gelatin-based-model).

72


Fig. 16. Electrical voltage generated: in free space and in vitro.

Fig. 17. Equivalent circuit pacemaker—in vitro model.


CONCLUSION

According to the experimental results and simulation in free space and in vitro, the magnetic field through the DUT generates induced currents in the presence of a medium. Those currents generate an additional induced electromotive force added to the one generated in the free space. The effects of the induced currents on the generated electrical potential during the cardiac activity, either natural or artificial are very considerable. The method that we proposed is very effective for the study of the low frequencies electromagnetic interferences on the medical implants: cardiac pacemaker, defibrillators…etc.

This work was supported by EDF ‘Electricite´ De France’. I would like to thank Dr. Benzeltout Boubakeur for helping in the translation of some parts of this paper.

REFERENCES 1. Dodinot B, Godenir JP, Costa AB. Electronic article surveillance: a possible danger for pacemakers patients. Pace 1993; 16: 46–53. 2. Hedjiedj A, Goeury C, Nadi M. A methodological approach for the characterization of cardiac pacemaker immunity to low frequency interferences: case of 50 Hz, 60 Hz, 10 kHz and 25 kHz led disruptions. J Med Eng 2002; 26: 223–227. 3. Scholten A, Joosten S, Silny J. Unipolar cardiac pacemakers in electromagnetic fields of high voltage overhead lines (med/ biol). J Med Eng Technol 2005; 29(4): 170–175. 4. Scholten A, Silny J. The interference threshold of cardiac pacemakers in electric 50 Hz fields. J Med Eng Technol 2001; 25(1): 1–11. 5. Trigano A, Blandeau O, Souques M, Gernez JP, Magne I. Clinical study of interference with cardiac pacemakers by a magnetic field at power line frequencies. med./biol. J Am Coll Cardiol 2005; 45(6): 896–900. 6. Kainz W, Casamento JP, Ruggera PS, Chan DD, Witters DM. Implantable cardiac pacemaker electromagnetic compatibility testing in a novel security system simulator. IEEE Trans Biomed Eng 2005; 52(3): 520–530. 7. Lucas EH, Johnson D, McElroy BP. The effects of electronic article surveillance systems on permanent cardiac pacemakers: an in vitro study Part II. Pace 1994; 17: 2021–2026.

73

8. Mc Ivor ME, Reddinger J, Floden E, Sheppard RC. Study of pacemaker and implantable cardioverter defibrillator triggering by electronic Article surveillance devices (SPICED TEAS). Pace 1998; 21: 1847–1861. 9. G. O. COMAR Reports. Radio frequency interference with medical devices: a technical information statement. IEEE Eng Med Biol Magaz 1998; 17(3): 111–114. 10. Hayes DL, Wang PJ, Reynolds DW, Estes NAM III, Griffith JL, Stefens RA, Carlo GL, Findlay GK, Johnson CM. Interference with cardiac pacemakers by cellular telephones. New Eng J Med (New England Journal of Medic) 1997; 336: 1473– 1479. 11. Tan KS, Hinberg I. Can Wireless Communication Systems Affect Implantable Cardiac Pacemakers?. An In Vitro Laboratory Study, Biomedical Instrumentation Technology 1998; 32(no.1): 18–24. 12. Wang J, Ohshima T, Takahashi E, Fujiwara O. Verification for EMI test of cardiac pacemaker by portable telephones with an anatomically based human model. IEEE International symposium on EMC, Istanbul, 2003. 13. Claycomb JR, Traslshawala N, Miller JH Jr. Theoretical investigation of eddy current induction for non-destructive evaluation by superconducting quantum interference devices. IEEE Trans Magn 2000; 36(1): 292–298. 14. Stoll Richard L. The analysis of eddy currents, Clarendon Press: Oxford, 1974. 15. Perez J-P. Electromagne´tisme fondements et applications. Dunod, Paris, 2002. 16. Scorretti R. Caracte´risation nume´rique et expe´rimentale du champ magne´tique B. F. geńe´re´ par des syste`mes e´lectrotechniques en vue de la mode´lisation des courants induits dans le corps humain. The`se doctorat, 2003. 17. Wang W, Eisenberg SR. A three-dimensional finite element method for computing magnetically induced current in tissues. IEEE Trans Magn 1994; 30(6): 5015–5023. 18. Gandhi OM, Kang G. Calculation of induced current densities for humans by magnetic fields from electronic article surveillance devices. Phys Med Biol 2001; 46: 2759–2771. 19. Babouri A, Hedjiedj A, Guendouz L, Andretzko JP. The behavior of dual-chamber pacemakers exposed to a conducted lowfrequency disruptive signal. Physiol Meas 2006; 27: 725–736. 20. Babouri A, Hedjeidj A. In vitro investigation of eddy current effect on pacemaker operation generated by low frequency magnetic field. 29th IEEE EMBS Annual International Conference. Lyon, August 23–26, 2007. 21. Goeury C, Hedjiedj A, Nadi M. Gel conducteur pour la simulation de la conductivite´ e´lectrique des tissus humains entre 50 Hz et 500 kHz. ITBM-RBM 2001; 22(4): 371–377.