B1-Gradient Based MRI Using a Multi-Element Transmit System

Introduction: Spatial signal encoding in MRI is usually performed via B0-gradients. An alternative method is the use of B1-gradients (so-called RF encoding), already suggested in the early days of MR (see, e.g., [1-3]). RF encoding offers the possibility to omit all B0-gradients, which would allow for MR scanning more or less free of acoustic noise. This advantage might be counterbalanced by long scan times, long reconstruction times, reduced image contrast, and/or enhanced SAR. This study investigates the principle possibility of RF encoding using an eight-channel whole-body transmit system [4] for low-resolution imaging of phantoms. Future studies can combine B0 and B1-gradients, alleviating the advantages and disadvantages of RF encoding. Theory: Image acquisition via RF encoding can be divided into five steps. (1) A set of I desired, ideal, complex B1-gradients ideal i B1 is defined (i=1…I). The gradients do not need to be constant, any gradient set leading to the required orthogonality is sufficient. (2) The sensitivities Sn of the N transmit array elements (n=1…N) of the multichannel MR system are measured using the phantom / subject under investigation. (3) The desired gradients are approximated by an optimized superposition of the