Patient-Specific RF Safety Assessment in MRI: Progress in Creating Surface-Based Human Head and Shoulder Models

Abstract

The interaction of electromagnetic (EM) fields with the human body during magnetic resonance imaging (MRI) is complex and subject specific. MRI radiofrequency (RF) coil performance and safety assessment typically includes numerical EM simulations with a set of human body models. The dimensions of mesh elements used for discretization of the EM simulation domain must be adequate for correct representation of the MRI coil elements, different types of human tissue, and wires and electrodes of additional devices. Examples of such devices include those used during electroencephalography, transcranial magnetic stimulation, and transcranial direct current stimulation, which record complementary information or manipulate brain states during MRI measurement. The electrical contact within and between tissues, as well as between an electrode and the skin, must also be preserved. These requirements can be fulfilled with anatomically correct surface-based human models and EM solvers based on unstructured meshes. Here, we report (i) our workflow used to generate the surface meshes of a head and torso model from the segmented AustinMan dataset, (ii) head and torso model mesh optimization for three-dimensional EM simulation in ANSYS HFSS, and (iii) several case studies of MRI RF coil performance and safety assessment.