Scan-rescan and intra-observer variability of magnetic resonance imaging of carotid atherosclerosis at 1.5 T and 3.0 T.

Carotid atherosclerosis measurements for eight subjects at baseline and 14 +/- 2 days later were examined using 1.5 T and 3.0 T magnetic resonance imaging (MRI). A single observer blinded to field strength, subject and timepoint manually segmented carotid artery wall and lumen boundaries in randomized images in five measurement trials. Mean increases in the signal-to-noise ratios (SNR) for T1-weighted images acquired at 3.0 T compared to 1.5 T were 90% (scan) and 80% (rescan). Despite significantly improved SNR and contrast-to-noise ratios (CNR) for images acquired at 3.0 T, vessel wall volume (VWV) intra-observer variability was not significantly different using coefficients of variation (COV), and intraclass correlation coefficients (ICC). VWV interscan variability and consistency at both field strengths were not statistically different (1.5 T/3.0 T COV = 5.7%/7.8%, R(2) = 0.96 for 1.5 T and R(2) = 0.87 for 3.0 T). A two-way analysis of variance showed a VWV dependence on field strength but not scan timepoint. In addition, a paired t-test showed significant differences in VWV measured at 3.0 T as compared to 1.5 T. These results suggest that although images acquired at 1.5 T have lower SNR and CNR VWV, measurement variability was not significantly different from 3.0 T VWV and that VWV is field-strength dependent which may be an important consideration for longitudinal studies.