Whole Brain Volume Measured from 1.5T versus 3T MRI in Healthy Subjects and Patients with Multiple Sclerosis

Abstract

BACKGROUNDWhole brain atrophy is a putative outcome measure in monitoring relapsing‐remitting multiple sclerosis (RRMS). With the ongoing MRI transformation from 1.5T to 3T, there is an unmet need to calibrate this change. We evaluated brain parenchymal volumes (BPVs) from 1.5T versus 3T in MS and normal controls (NC).METHODSWe studied MS [n = 26, age (mean, range) 43 (21‐55), 22 (85%) RRMS, Expanded Disability Status Scale (EDSS) 1.98 (0‐6.5), timed 25 foot walk (T25FW) 5.95 (3.2‐33.0 seconds)] and NC [n = 9, age 45 (31‐53)]. Subjects underwent 1.5T (Phillips) and 3T (GE) 3‐dimensional T1‐weighted scans to derive normalized BPV from an automated SIENAX pipeline. Neuropsychological testing was according to consensus panel recommendations.RESULTSBPV‐1.5T was higher than BPV‐3T [mean (95% CI) + 45.7 mL (+35.3, +56.1), P < .00001], most likely due to improved tissue‐CSF contrast at 3T. BPV‐3T showed a larger volume decrease and larger effect size in detecting brain atrophy in MS versus NC [−74.5 mL (−126.5, −22.5), P = .006, d = .92] when compared to BPV‐1.5T [−51.3.1 mL (−99.8, −2.8), P = .04, d = .67]. Correlations between BPV‐1.5T and EDSS (r = −.43, P = .027) and BPV‐3T and EDSS (r = −.49, P = .011) and between BPV‐1.5T and T25FW (r = −.46, P = .018) and BPV‐3T and T25FW (r = −.56, P = .003) slightly favored 3T. BPV‐cognition correlations were significant (P < .05) for 6 of 11 subscales to a similar degree at 1.5T (r range = .44‐.58) and 3T (r range = .43‐.53).CONCLUSIONSField strength may impact whole brain volume measurements in patients with MS though the differences are not too divergent between 1.5T and 3T.

Feasibility of Brain Atrophy Measurement in Clinical Routine without Prior Standardization of the MRI Protocol: Results from MS-MRIUS, a Longitudinal Observational, Multicenter Real-World Outcome Study in Patients with Relapsing-Remitting MS

“…In line with other recent studies, 23,25,26 the MS-MRIUS study showed that scanner changes had an impact on brain volume estimates. While it was previously shown that NeuroSTREAMderived PLVVC is relatively robust to different field strengths when imaged during a short time (approximately 2% coefficient of variation in the 1.5T versus 3T scan-rescan test for 72 hours), 23 the current study showed that PLVVC on T2-FLAIR was significantly different in patients with RRMS with hardware changes, compared with those without.…”

“…24 All outcomes of brain atrophy analyses were assessed by an experienced rater. Because hardware changes can affect longitudinal measurements, [25][26][27] SIENA PBVC and VIENA PLVVC analyses were considered invalid when a patient was imaged on different hardware. In addition, because Neuro-STREAM PLVVC was previously shown to be robust to hardware changes in a study that included 125 patients with MS and 76 healthy controls, 23 we explored the stability of this measure in patients with and without MR imaging hardware changes using the current real-world setting dataset.…”

Whole‐brain atrophy assessed by proportional‐ versus registration‐based pipelines from 3TMRIin multiple sclerosis

“…Although manual skull‐stripping is closer to a gold standard for determining ICV, it is time‐consuming and has been largely replaced with automated techniques such as BET (Smith et al., 2002), SPM's integrated tissue segmentation (Ashburner & Friston, 2005), or FreeSurfer watershed algorithm (Dale et al., 2004). As prior authors have noted, the FSL BET can also be a significant source of error (Popescu et al., 2012; Zivadinov et al., 2005) and we found tissue misclassification in several subjects using the default settings; neck cropping and changing the default parameters (−f 0.2 and −B enabled) allowed an optimal solution for our dataset without any significant misclassification errors (Chu et al., 2016). Without any visually prominent errors, several groups have concluded that brain extraction methods are generally a very small source of variance (Clark, Woods, Rottenberg, Toga, & Mazziotta, 2006; Klauschen, Goldman, Barra, Meyer‐Lindenberg, & Lundervold, 2009) and we feel this preprocessing step is unlikely to be a significant source of variance between methods.…”

“…In the BPV pipeline, raw MDEFT images were resliced to the axial plane, followed by removal of all slices inferior to the cervico‐medullary junction using JIM v7 (http://www.xinapse.com). Images then underwent automated segmentation and template normalization using SIENAX, (Smith et al., 2002) part of FSL (v5.0) (Smith et al., 2004) using a previously optimized brain extraction tool (BET) threshold of 0.2 (Chu et al., 2016). T2‐hyperintense lesion volumes were obtained by expert semiautomated segmentation with an edge‐finding tool based on local image intensity thresholds using JIM (v5) as previously published (Dell'Oglio et al., 2015); manual corrections were applied as needed (Ceccarelli et al., 2012).…”

“…Furthermore, brain volume may vary based on pathophysiological factors, including recent start of immunomodulatory therapy, acute inflammation, hydration status, time of day, tobacco use, genetics, and comorbid conditions (Rocca et al., 2017). As MRI technology evolves and increasingly precise high‐field (e.g., 3T) magnets proliferate in clinical practice, there remains an ongoing need for critical evaluation of the sensitivity and validity of postprocessing software pipelines (Chu et al., 2016; Stankiewicz et al., 2011). …”