Synopsis

Keywords: Safety, Safety

The influence of the thickness, loss tangent value of the lead insulation material along with lead length effects on the heating of an active implantable medical device (AIMD) lead electrode is analyzed using full-wave EM simulations. Decreasing the thickness of the insulation decreases heating of the electrode significantly when lead lengths are less than or equal to wavelengths in the medium. Increasing the loss tangent of the insulation material has a moderate effect in reducing heating at the electrode. Standing wave distortions affect lead electrode heating at lead lengths greater than wavelengths in the medium.

Introduction

Induced currents on the leads of the AIMDs under MRI RF transmit fields results in heating on the leads, electrodes and implantable pulse generators (IPG) posing risks to patients with AIMDs. The transfer function (TF) method is a state-of-the-art approach to analyze AIMD heating under MRI. The TF method overcomes the massive computational challenge of modeling highly intricate geometry of AIMDs placed inside heterogenous dielectric human body models. AIMD heating in the MRI environment is predicted in a two-step process: (i) by either modeling or measuring the TF; (ii) by numerically calculating the electric (E) fields in human body models without the presence of AIMD¹. In this study the influences of the thickness, lossy property of the insulation material – i.e., loss tangent, tan δ (σ/(ωε)), and the lead length relative to the wavelengths (λ_64MHz, λ_{128 MHz}) in the medium on the heating of straight wire AIMD leads are explored using full-wave electromagnetic (EM) modeling of the heating TFs² employing the reciprocity principle.

Methods

AIMDs with a 10 mm long electrode section connected to straight wire leads (lengths 45cm ~ λ_64MHz, 65cm > λ_64MHz at 64 MHz and 25 cm ~ λ_{128 MHz}, 45 cm > λ_{128 MHz} at 128 MHz) were modeled and excited with 1 A sources at 64 MHz and 128 MHz placed at 2 mm distance from the electrode tip to determine heating TFs. For insulation thickness analysis Polyethylene Terephathalate (PET) material (ε = 2.8) was used and for lossy insulation loss tangent values (tan δ = 0.005 to 0.1, typical to that of silicon, neoprene rubber) were explored. Simulations and experiments were performed in a rectangular ASTM phantom containing dielectric medium (σ = 0.47 S/m; ε = 78). Current distributions on the leads were numerically evaluated using the Method of Moments with surface equivalence principle (FEKO, Altair, USA). An experimental validation of the numerical simulation was done with a straight wire lead of 65 cm length with 64 cm insulation and 1 cm bare wire at the excitation end. Straight lead was excited 2 mm away from the bare wire end and current measurements were performed with a current sensor (Pearson Electronics, Palo Alto, CA), using a vector network analyzer as shown in Fig. 1

Results

TF shapes in Fig.2 (a) and (c) show that decreasing the insulation thickness significantly (up to 50%) decreases heating of the electrodes and along the leads for lead lengths shorter than or equal to the wavelengths in the medium. Electrode heating is reduced on thinly insulated leads due to an increase in the leakage currents along the lead length³. For lead lengths significantly higher than wavelengths in the medium (Fig.2 (b) and (d)), TF shapes indicate standing wave distortions. In Fig.3, TF shapes predict decreased heating at the electrodes as the loss tangent of the insulation material is increased. Electrode heating is only moderately reduced by increasing the leakage currents along the lossy insulation material. Also, standing wave distortions occur for lead lengths greater than the wavelengths in the medium. Fig.4 show experimental validation of simulation findings (see Fig. 2 (b) and (d)) on the influence of decreasing insulation thickness and wavelength effects on reducing heating on lead and electrodes of AIMDs. Current values at excitation points vary between simulation to experiment due to variability in the source location from the bare electrode section.

Discussion

The impedance mismatch conditions at the distal ends of leads where IPGs attach cause standing wave distortions. By appropriately mating the IPGs’ input impedances to the leads’ impedances, the AIMD heating could be minimized by reducing the thickness and increasing the loss tangent of the insulation material. We observe good correlation between experimental and simulation results.

Conclusion

In this study we demonstrated the potential utility of EM simulations to investigate the influences of lead thickness, insulation material loss tangent, and lead lengths on AIMD heating. Further work will be done involving simulation and experimental validations with E-field calculations, current measurements, and temperature measurements in an appropriate phantom under MRI body coil exposure.