Tier-based formalism for safety assessment of custom-built RF transmit coils.

The purpose of this work is to propose a tier-based formalism for safety assessment of custom-built RF coils that balances validation effort with the determination of more or less conservative safety factors. The formalism has 3 tier levels. Higher tiers require increased effort when validating EM simulation results but allow for less conservative safety factors. In addition, we propose a new method to calculate modeling uncertainty between simulations and measurements and a new method to propagate uncertainties in the simulation into a safety factor which minimizes the risk of underestimating peak SAR. The new safety assessment procedure was completed for all tier levels for an 8-channel dipole array for prostate imaging at 7T and an 8-channel dipole array for head imaging at 10.5T, using data from two different research sites. For the 7T body array, the validation procedure resulted in a modeling uncertainty of 77% between measured and simulated local SAR distributions. For a situation where RF shimming is performed on the prostate, average power limits of 2.4 and 4.5 W/channel were found for tier 2 and 3. When the worst-case peak SAR among all phase settings was calculated, power limits of 1.4 and 2.7 W/channel were found for tier 2 and 3. For the 10.5T head array, a modeling uncertainty of 21% was found based on B 1 + mapping. For the tier 2 validation, a power limit of 2.6 W/channel was calculated. The demonstrated tier system provides a strategy for evaluating modeling inaccuracy, allowing for the rapid translation of novel coil designs with conservative safety factors and the implementation of less conservative safety factors for frequently used coil arrays at the expense of increased validation effort.