In∞uence of a Dielectric Insert of High Permittivity on the Transmit Performance of a 300MHz Multi-channel MRI Loop Array

Several recent reports have suggested that the transmit performance of a given MRI RF coil can be enhanced by incorporating suitable dielectric materials. To analyse this proposed approach more deeply, we investigated the efiect of a dielectric material insert on loop array power balance, transmit performance and slice homogeneity. To obtain a useful database in a reasonable time, we varied only the insert permittivity, the load diameter, the separation between array and load, and the length of loop elements. For a range of parameter combinations, use of the insert increases the mean transmit RF fleld over the central transverse slice B1+s, and the homogeneity of this slice, but not the transmit performance, expressed as ratio of the mean RF fleld over the fleld of view to the square root of power deposited in the FOV, B1 + A = p Pl. An array with the insert cannot provide better B1+s and homogeneity than an array without the insert in which the geometry is optimized for a given target region. We have found no general recipe for achieving the best array performance and homogeneity using the insert in an array with arbitrary geometry. 1. INTRODUCTION Recently several reports (1,2) have suggested that the transmit performance of a given MRI RF coil can be enhanced by incorporating suitable dielectric materials. However, these reports, which are mostly based on experimental results, do not provide detailed analysis of the power balance of the coil with and without an insert composed of such high permittivity dielectric materials, nor data regarding the transmit fleld homogeneity for difierent slices of the object. To analyse this proposed approach more deeply, we investigated the efiect of the insert on loop array power balance, transmit performance and slice homogeneity. To obtain a useful database in a reasonable time, we varied only the insert permittivity, the load diameter, the separation between array and load, and the length of loop elements. 2. METHOD We investigated 8-channel loop-type array coils with loop angular width of 37.5 degrees, having a range of diameters (from 200mm to 280mm) and lengths (80, 100 and 120mm), mounted on a cylindrical acrylic former (not shown on Fig. 1) and loaded by cylinder with diameters 80, 120, 155 and 190mm and length 375mm. The latter minimizes the in∞uence of load length on transverse slice homogeneity for arrays with difierent lengths. For all loads, electrical properties were close to those of average human tissue at 300MHz | permittivity 52 and conductivity 0.55S/m. A high permittivity hollow dielectric insert (in red color on Fig. 1), with permittivity ranging from 10 to 79, and with cylindrical geometry, was placed between the coil and load. In all cases, the thickness of the cylindrical acrylic former (5mm) separated the array and the insert, and there was no air gap between the insert and load. The fleld of view (FOV) was taken as the part of the load located inside the array. In this arrangement, the cylindrical symmetry of the array, the insert and the load obviates performance optimization by RF shimming (adjustment of amplitude and phase for excitation signals). All arrays were excited in circular polarization mode, applying 1W power to each port (array transmit power | Ptransmit = 8W), with a sequential 45 degree phase increment.