Ge signa excite hd 12.0 twin speed 8 channel scanner

To assess inter-scanner stability, a previously collected dataset consisting of 125 MS patients and 52 healthy controls scanned at both 1.5 T and 3 T was used. (Di Perri et al., 2009 (link), Dwyer et al., 2017 (link)) All subjects were examined on both scanners within one week, and the order in which subjects were scanned was randomized. The scanners used were a 1.5 T GE Signa Excite HD 12.0 8-channel scanner and a 3 T GE Signa Excite HD 12.0 Twin Speed 8-channel scanner (General Electric Milwaukee, WI). 3D T1w images and 2D low-resolution T2-FLAIR images were acquired in each scanning session. Sequences were not identical between the two scanners, but rather reflected optimizations for the specific field strengths as would be seen in clinical practice. As with the scan-rescan dataset, DeepGRAI was run on both scans for each subject, and association was again assessed by pairwise correlation and by CoV.

To assess stability across scanners, a previously collected dataset consisting of 125 MS patients and 76 healthy controls scanned at both 1.5T and 3T was used. Subjects were 72% female, with a mean age of 42.5 ± 11.1 years. All subjects were examined on both scanners within 72 h, and the order in which subjects were scanned was randomized. The scanners used were a 1.5T GE Signa Excite HD 12.0 8-channel scanner and a 3T GE Signa Excite HD 12.0 Twin Speed 8-channel scanner (General Electric Milwaukee, WI). 2D 3 mm T2-FLAIR acquisitions were included in each scanning session. Sequences were not identical between the two scanners, but rather reflected optimizations for the specific field strengths as would be seen in clinical practice.
As with the scan-rescan dataset, NeuroSTREAM was run on both scans for each subject, and association was again assessed by pairwise correlation, ICC, Bland-Altman plot, and casewise CoV. As above, longitudinal error was calculated as the median absolute Jacobian-derived percent change, with the true change assumed to be 0%.