Electromagnetic and Electrophysiological Modeling of Cardiac Stimulation Thresholds in 13 <scp>MRI</scp> Gradient Systems and 56 Body Models

PURPOSE: The MRI safety standard IEC 60601-2-33 limits the maximum dB/dt and E-field induced by MRI gradients to protect patients from cardiac stimulation (CS). Those limits were set in the early 2000s based on a compilation of electrode stimulation measurements in animals as well as electromagnetic calculations in a homogeneous ellipsoid. We re-evaluate CS in MRI using state-of-the-art electromagnetic and electrophysiological modeling in realistic body models and gradient coils. METHODS: We predicted CS thresholds in 56 realistic body models and 13 commercial gradient systems for trapezoidal waveforms with rise times (the time during which the gradient amplitude is ramped from zero to maximum) ranging from 0.1 to 5.0 ms. We used threshold predictions in individual body models to estimate the gradient field amplitude below which the probability of occurrence of a CS adverse event is < 2 ppb, the historically agreed upon probability used in IEC 60601-2-33. RESULTS: We found that the variability of CS thresholds due to differences in anatomy and body habitus was 10%-36%, depending on rise time, while the CS threshold variability across gradient coils was 27%. In the range of typically used MRI rise times (≤ 2.5 ms), the predicted gradient amplitude associated with a CS probability of 2 ppb is > 5.5X higher than the IEC CS limit. CONCLUSION: Detailed simulations in a large population of realistic body models and gradient coils indicate that the IEC CS limit is substantially lower (more conservative) than the predicted gradient amplitude associated with a 2-ppb probability for CS.