Test-retest reliability of magnetic resonance imaging data across the Siemens VE11C to XA30 software upgrade

Motivation: The question of test-retest reliability in magnetic resonance imaging has led to difficulties in scientific replication. This may be compounded by changes in scanner software during ongoing longitudinal studies. Goal(s): Our goal was to evaluate the effects of the Siemens VE11C to XA30 software upgrade on longitudinal MRI data. Approach: Eleven participants were each scanned at four sessions on the same 3.0 T scanner: twice using the VE11C software and twice after the upgrade to XA30. Results: Cortical thickness values were consistent across software within 1%, and functional data showed similar variance between software as compared to within software. Impact: We found no meaningful difference between MRI datasets collected using the Siemens VE11C and XA30 software packages. This evidence of consistency in longitudinal MRI data across software lends validity to imaging studies that are affected by unforeseen software upgrades.