A breast phantom for quality assurance in a multi- centre screening trial

Tests of system performance. Measurements of the SNR of a Siemens Vision breast coil showed a 35 % variation along the coil's Y axis (bottom to top). The signal uniformity varied by 15 % along the same axis. The average uniformity in the XZ plane was 80 %. The ratio of the signal intensity of a Tl_weighted acquisition to that of a proton density weighted acquisition varied by 3 % along the Y axis for a 3D FLASH sequence. However, these values varied by as much as 10 % for 2D FLASH sequences. These variations arise from factors such as inhomogeneities in the B, field and the coil sensitivity and for the 2D FLASH sequences, from imperfections in the slice profile. The dependence of the signal intensity ratio on slice position has implications for localised quantitative analysis, especially for pharmacokinetic analysis. In order to overcome these variations, each coil used in the trial will be calibrated using the same phantom. The signal intensity ratio, determined using both 3D and 2D FLASH sequences, will be mapped as a function of slice position. These signal intensity ratios may then be converted to T, values using a calibration curve constructed, using reference gels, for each MR system3. The calibration will be incorporated in all software used to analyse quantitative contrast agent uptake.