Impact of mixed media on transfer functions with a pacemaker system for estimation of RF heating during MRI scans

Patients with active implantable medical devices (AIMD) may be at risk of harm from RF-induced heating during an MRI scan. The amount of RF-induced lead heating is mainly attributed to the electric field tangential (Etan) to the lead path, which can be predicted by using an electrical field transfer function (TF) associated in different media. The objective of this study is to investigate the difference in estimating RF heating through TFs of a pacemaker system in high conductivity media (HCM), low conductivity media (LCM) and mixed media simulating in-vivo. The ANSYS HFSS was used to calculate the electromagnetic fields in a tank model built from the TF measurement set-up, with a pacer and a cardiac lead model (length=52 cm) placed inside. The tank is filled with homogenous media such as Blood (σ=1.2 S/m, ε<inf>r</inf> = 78) denoted as high conductivity media HCM or low conductivity media LCM (σ=0.47 S/m, ε<inf>r</inf> = 78). For mixed media, blood vessels (HCM) are around a lead with the cardiac dimension. Outside vessels, the space was filled with myocardium (σ =0.678 S/m, ε=106.5). Reciprocal method was used with current excitation at the tip. Currents along lead conductors are obtained by loop integrals of H fields. The TFs are the currents along the lead divided by the excitation, denoted as S(τ). The heating was accessed by integrals of tangential electrical fields (Etan) over 133 clinical pathways inside virtual human family models giving by normalized power PE ∼ σ∗Etip<sup>2</sup> =σ |∫Etan(τ) ∗S(τ)) d τ|<sup>2</sup>. Three curves of TF magnitude S(τ) run closely to each other and TF for the mixed media was in the middle, LCM on the top and HCM in the bottom. Predicted normalized PEs at 99 percentile are 0.974 for HCM, 0.536 for mixed and 0.519 for LCM. For the scenarios evaluated, using TFs with LCM provides a closer in-vivo heating evaluation with a 3% underestimation while HCM overestimates by 80%.