Local SAR management by RF Shimming: a simulation study with multiple human body models

Abstract

Parallel transmission facilitates a relatively direct control of the RF transmit field. This is usually applied to improve the RF field homogeneity but might also allow a reduction of the specific absorption rate (SAR) to increase freedom in sequence design for high-field MRI. However, predicting the local SAR is challenging as it depends not only on the multi-channel drive but also on the individual patient. The potential of RF shimming for SAR management is investigated for a 3 T body coil with eight independent transmit elements, based on Finite-Difference Time-Domain (FDTD) simulations. To address the patient-dependency of the SAR, nine human body models were generated from volunteer MR data and used in the simulations. A novel approach to RF shimming that enforces local SAR constraints is proposed. RF shimming substantially reduced the local SAR, consistently for all volunteers. Using SAR constraints, a further SAR reduction could be achieved with only minor compromises in RF performance. Parallel transmission can become an important tool to control and manage the local SAR in the human body. The practical use of local SAR constraints is feasible with consistent results for a variety of body models.