Quantitative evaluation of Scout Accelerated Motion Estimation and Reduction (SAMER) MPRAGE for morphometric analysis of brain tissue in patients undergoing evaluation for memory loss

•The SAMER-MPRAGE sequence effectively corrects motion in 3D volumetric brain MRI.•SAMER increases the accuracy of cortical volume and cortical thickness estimation.•Validation was performed on healthy volunteers and dementia-workup outpatients.•SAMER-MPRAGE allows accurate morphometry even for severely-motion-corrupted scans.•SAMER-MPRAGE has been translated to be utilized in practical clinical settings. Three-dimensional (3D) T1-weighted MRI sequences such as the magnetization prepared rapid gradient echo (MPRAGE) sequence are important for assessing regional cortical atrophy in the clinical evaluation of dementia but have long acquisition times and are prone to motion artifact. The recently developed Scout Accelerated Motion Estimation and Reduction (SAMER) retrospective motion correction method addresses motion artifact within clinically-acceptable computation times and has been validated through qualitative evaluation in inpatient and emergency settings. We evaluated the quantitative accuracy of morphometric analysis of SAMER motion-corrected compared to non-motion-corrected MPRAGE images by estimating cortical volume and thickness across neuroanatomical regions in two subject groups: (1) healthy volunteers and (2) patients undergoing evaluation for dementia. In part (1), we used a set of 108 MPRAGE reconstructed images derived from 12 healthy volunteers to systematically assess the effectiveness of SAMER in correcting varying degrees of motion corruption, ranging from mild to severe. In part (2), 29 patients who were scheduled for brain MRI with memory loss protocol and had motion corruption on their clinical MPRAGE scans were prospectively enrolled. In part (1), SAMER resulted in effective correction of motion-induced cortical volume and thickness reductions. We observed systematic increases in the estimated cortical volume and thickness across all neuroanatomical regions and a relative reduction in percent error values compared to reference standard scans of up to 66 % for the cerebral white matter volume. In part (2), SAMER resulted in statistically significant volume increases across anatomical regions, with the most pronounced increases seen in the parietal and temporal lobes, and general reductions in percent error relative to reference standard clinical scans. SAMER improves the accuracy of morphometry through systematic increases and recovery of the estimated cortical volume and cortical thickness following motion correction, which may affect