Purpose: Sequence parameters for most clinical diagnostic imaging protocols have traditionally been designed based upon subjective image interpretive feedback by radiologists. Imaging parameters are typically selected based on compromises between contrast, resolution, scan time, SNR, and motion. While perceived differences in contrast exist for different sequences, the standardization and validation of intrinsic sequence contrast across scanner platforms and field strengths is not addressed by standard QA procedures. Hence longitudinal imaging on disparate scanners can be associated with observed contrast changes between timepoints, negating radiologists’ diagnostic confidence. In this study, we used the ISMRM-NIST system phantom to evaluate contrast of a T2-weighted acquisition protocol used for prostate imaging across several scanners in our institution.
Methods: The system phantom was scanned on three 1.5T and three 3T scanners with a thin slice T2-weighted Turbo Spin Echo PIRADS 2.1 compliant acquisition. Within field strengths, the scanners had identical hardware configurations and executed identical scan parameters. Since the phantom cannot be imaged with an endorectal coil, coronal slices were acquired using the body coil to maximize signal homogeneity. Signal intensity of the T2-sensitive compartments was measured for evaluation of relative contrast, as defined by change in signal intensity as a function of T2. Background water signal near the spheres was also recorded as a measure of overall B1- inhomogeneity, which was applied as a correction factor to minimize residual signal inhomogeneity in the image.
Results: Contrast as defined by signal percent difference between two designated spheres averaged 28.0% at 1.5T and 45.0% at 3T. For T2 ranging from 46.42 to 581.3 msec, contrast curves varied within ±1.3% of average at 1.5T and ±9.4% at 3T.
Conclusion: A standardized system phantom and measurement protocol enables the measurement and validation of T2 contrast between scanners. Notable contrast deviations were detected at 3T, which warrant further investigation.
Funding Support, Disclosures, and Conflict of Interest: This work was supported in part by Siemens Healthineers. Ken-Pin Hwang also receives research funding from GE Healthcare.