The Impact of the Siemens Trio to Prisma Upgrade and Volumetric Navigators on MRI Indices: A Reliability Study with Implications for Longitudinal Study Designs

Abstract

Many magnetic resonance imaging (MRI) indices are being studied longitudinally to explore topics such as biomarker detection and clinical staging. A pertinent concern to longitudinal work is MRI scanner upgrades. When upgrades occur during the course of a longitudinal MRI neuroimaging investigation, there may be an impact on the compatibility of pre-and postupgrade measures. Similarly, subject motion is another issue that may be detrimental to longitudinal MRI work; however, embedding volumetric navigators (vNavs) within acquisition sequences has emerged as a technique that allows for prospective motion correction. Our research group recently implemented an upgrade from a Siemens MAGNETOM 3T Trio system to a Siemens MAGNETOM 3T Prisma Fit system. The goals of the current work were to: 1) investigate the impact of this upgrade on commonly used structural imaging measures and proton magnetic resonance spectroscopy indices ("Prisma Upgrade protocol") and 2) examine structural imaging measures in a sequence with vNavs alongside a standard acquisition sequence ("vNav protocol"). In both protocols, while high reliability was observed for most of the investigated MRI outputs, suboptimal reliability was observed for certain indices. Across the scanner upgrade, increases in frontal, temporal, and cingulate cortical thickness (CT) and thalamus volume, along with decreases in parietal CT, amygdala, globus pallidus, hippocampus, and striatum volumes were observed across the Prisma upgrade, and were linked to increases in signal-to-noise ratios. No significant impact of the upgrade was found in 1 H-MRS analyses. Further, CT estimates were found to be larger in MPRAGE acquisitions compared to vNav-MPRAGE acquisitions mainly within temporal areas, while the opposite was found mostly in parietal brain regions. The results from this work should be considered in longitudinal study designs and comparable prospective motion correction investigations are warranted in cases of marked head movement.

Estimating the effect of a scanner upgrade on measures of grey matter structure for longitudinal designs

“…The results of our study are in line with previous findings which have indicated systematic effects of scanner upgrade on GM imaging outcomes (Lee et al 2019;Han et al 2006;Jovicich et al 2009;Brunton et al 2015). Notably, in two recent studies, an upgrade from Magnetom Trio to Prismafit induced a significant increase in cortical thickness (CT) and volume as well as differences in subcortical volumes (Potvin et al 2019;Plitman et al 2020). Similar to our findings, ICC values for cortical measures were good to excellent in both studies.…”

“…CNR and other measures of image quality indicated higher quality on the (earlier) Verio scanner. In contrast, studies investigating the effects of real upgrades showed increased image quality (quantified as signal-to-noise ratio (SNR) or CNR) after the upgrade (Potvin et al 2019;Plitman et al 2020). This may contribute to the observed bias pattern in GMV, yet increased SNR/CNR was also associated with higher CT in our study, independent of scanner-dependent regional CT differences (Shuter et al 2008;Potvin et al 2019).…”

“…We found a pattern of frontal to occipital differences, with frontal-precentral regions biased towards higher CT and GM volume values in Skyra compared to Verio, and occipital regions biased towards higher CT and GM volume in Verio. Yet, in (Potvin et al 2019;Plitman et al 2020) the biased regions were located in prefrontal and temporal regions where CT and GM volume consistently increased with the upgrade. As shown above, there is a mismatch in location and direction of the upgrade effect between our study and (Potvin et al 2019;Plitman et al 2020).…”

“…Yet, in (Potvin et al 2019;Plitman et al 2020) the biased regions were located in prefrontal and temporal regions where CT and GM volume consistently increased with the upgrade. As shown above, there is a mismatch in location and direction of the upgrade effect between our study and (Potvin et al 2019;Plitman et al 2020). Thus, additional factors in our studies probably led to the observed frontal-occipital bias, and contributed to higher CT and GM volumes in Skyra compared to Verio.…”

Adolescent to young adult longitudinal development of subcortical volumes in two European sites with four waves

“…Similarly, sequence-related differences in subcortical structures have recently been found using CAT-12 (personal communication with J.-L. Martinot). Further, a scanner update was completed during wave three in Paris (Siemens Trio to Prisma) which may have introduced additional bias in results as Medawar et al (2020) and Plitman et al (2020) reported very good ICC but also percent volume differences in several cortical and subcortical regions when comparing Siemens Scanners (Trio versus Prisma / Verio versus Skyra). Since many recent longitudinal studies on brain development included several samples with differences in imaging sequence and scanner type (Dennison et al, 2013; Mills et al, 2021; Tamnes et al, 2017; Wierenga, Sexton, et al, 2018), results should be interpreted with caution.…”