SAR Reduction in Parallel Transmission by Allowing Spatial Phase Variation

Introduction Using multiple coils for transmission, RF pulse duration and/or specific absorption rate (SAR) can be reduced [1,2]. On the other hand, in the SAR reduction method developed in [1], SAR of shortened pulses can not be reduced significantly. In this work an SAR optimization method based on releasing phase constraints of excitation samples is developed. The new method is developed for pulses designed based on small-angle approximation. This method benefits the fact that phase variations causes less error in the final image than magnitude variations (i.e. there is a higher degree of freedom in phase variations). This method allows minimization of SAR in time reduced pulses (time reduction factor (R) up to number of coils). Phase optimization problem is non-convex, therefore a search algorithm, particle swarm optimization (PSO), is used [4]. The method is tested with electromagnetic simulations of a 6-Coil configuration. Effect of phase variations on excitation error is tested with Bloch simulations. Impact of magnitude variations on SAR is also tested. Methods Formulation developed in [5] is used for calculation of current waveforms. Formulation uses small-angle approximation and it is very similar to Transmit SENSE [2]. Relation between current waveforms and desired excitation profile is given by = m Ab (1) where m is a Mx1 column vector that contains the desired magnetization values of the corresponding samples from region of interest and b is a column vector that contains the current samples of all coils. Size of vector b is (M/R)x1 where R is the time reduction factor. (M/R)xM size A matrix consists of the sensitivity profiles of coils and k-space trajectory samples. Relation between the current samples and the whole-body SAR is given by the following formula