A reliable method for calculating RF coil performance

Introduction: Electromagnetic (EM) simulation is a valuable tool for predicting and evaluating the radiofrequency (RF) field inside the human head during an MRI experiment. At high field strength, knowledge of coil performance, defined as B1+ field magnitude per RF voltage applied to coil input, is vital for reliable definition of SAR, monitoring of SAR, and coil optimization. Experimental confirmation of coil performance simulations is essential, using well-characterized and relatively realistic phantoms. Earlier studies [1, 2, 3] have used textbook values for phantom EM properties, and have generally assumed the phantom dimensions. They also compared only scaled B1+ profiles, normalizing the maximum amplitude of simulated and experimental data. For reliable prediction of transmit coil performance, additional factors must be considered. These include the exact position and shape of the phantom within the coil, precise values for the actual phantom EM properties, full details of the RF pulse shape used in experimental B1+ profiling, and measured losses between the coil input and the point where the RF pulse amplitude is monitored by the scanner hardware. Method: The simulations were performed in CST Studio Suite 2008. The coil 3-D EM model includes all construction details for the resonance elements, simulated with realistic dimensions and material electrical properties. The shape of the phantom was obtained using a 3D TurboFLASH scan. The MRI data was segmented with MatLab and exported in an appropriate format. To determine the position of the phantom inside the coil, gel markers were attached to the coil. With the help of these markers, the translation vector from the scanner coordinate system to that of the coil was determined. The phantom tissue EM properties were measured using the reflection technique with the Agilent network analyzer, HB8510 [5]. The simulated B1sim field magnitude was calculated for a voltage Vsim= 20 V applied to the coil input, and coil performance was estimated as Cp_sim= B1 + sim/ Vsim. The magnitude of B1+ was mapped experimentally by applying rectangular RF pulses with amplitude Vtales=33.3 V (measured at the Transmit Antenna Level Sensor (TALES)) and pulse length τ =2.56 ms. Insko’s double angle method [6] was employed, for which B1exp = φ / γτ , where φ is the flip angle, and γ is the gyromagnetic ratio. There is a measured loss of about 2.8 db in total (i.e. attenuation factor Kloss=1.38) between the TALES and the coil input (Fig.1). Including this loss, the actual coil performance could be estimated as Cp_exp= B1 + exp/ (Vtales/Kloss). Results and Discussion: The experiments were performed using a Siemens 7T whole body scanner with a commercially available