Routine testing of magnetic field homogeneity on clinical MRI systems

Poor main magnetic field (B0) homogeneity (H(B)) leads to artifacts and signal losses in magnetic resonance imaging (MRI). The American College of Radiology's MRI quality control manual mandates annual checks of H(B), suggesting tests using spectral linewidth and phase-difference (delta phi) maps. A new method, the bandwidth-difference (deltaBW) method, which compares the distortion for small and large BW acquisitions to determine the HB, is proposed. The deltaBW method has the advantage that it can be used to measure multiple diameters of spherical volumes (DSV) in a single phantom. A phantom has been developed to exploit this method and results obtained with it are compared to those using three standard methods. Small receiver BW in the presence of poor H(B) leads to geometric distortions because gradients are reduced to the level of the B0 inhomogeneities. Data were acquired using seven MRI systems from different manufacturers, ranging in field strength from 0.2 to 3.0 T. Fast gradient echo pulse sequences were scanned twice using small and large BWs. H(B) was measured from the shift of landmarks between the two BW acquisitions. Results were compared with data from the full width at half maximum (FWHM) method, the delta phi method and one manufacturer's resonant frequency mapping data. The FWHM method was available on two systems and the detla phi method was available on one. The deltaBW method could be performed in all scanners investigated. The H(B) measured ranged 0.11-0.32 ppm to 6.7-12.9 ppm for DSV of 13-22.6 cm. Direct comparisons of the data obtained using the deltaBW method showed good agreement with data obtained using the FWHM method. Data obtained using the deltaBW method compared favorably with the manufacturer's resonant frequency map. The deltaBW method produces measurements of H(B) at various DSV values that can be obtained from a single set of phantom images. The accuracy of deltaBW B0 homogeneity measurements are comparable to the other methods tested.