Electromagnetic (EM) modelling is an effective method for evaluating the gradient safety of magnetic resonance imaging (MRI) for patients with active implantable medical devices (AIMDs). However, the combined effects of multiple factors-including gradient coil design constraints, implanted lead path, gradient strength, and scan configuration-on gradient-induced voltage (GIV) risk has not been systematically investigated. In particular, the magnetic field distribution outside the region of linearity (ROL) of gradient coils cannot be uniquely determined from their nominal gradient profile, and its impact on AIMD gradient safety assessment remains poorly understood. This study presents a multifactorial analysis of MRI gradient safety by integrating gradient coil modelling with anatomical lead path tracing using a reference human body shell. We examine how variations in coil design constraints affect magnetic field distributions and how these, in turn, influence GIV for three representative AIMDs' pathways: deep brain stimulators (DBS), cardiac pacemakers (PM), and sacral nerve stimulators (SNM). Multiple gradient strengths, coil excitation modes, and scanning positions are assessed. Magnetic field distributions vary significantly between coil designs, particularly in the concomitant Bx and By components, with differences reaching up to 53%. These variations result in GIV difference that increases with gradient strength. The maximum GIV differences for DBS, SNM, and PM reach 1.08V, 0.52V, and 0.93V, respectively, under Y-axis excitation. The concomitant field plays a significant role in these differences. Simultaneous excitation of all axes does not always produce the highest GIV due to cancellation effects. Cross-AIMD analysis shows high-risk zones are concentrated in and around the ROL. This work fills a gap by systematically evaluating how coil design, implant characteristics, gradient strength, and scan configurations influence GIV risk, providing a foundation for more comprehensive, individualized MRI gradient safety assessments.
Boya et al. (Mon,) studied this question.