# Verifying the concordance between motion corrected and conventional MPRAGE for pediatric morphometric analysis

## TL;DR

This study shows that a new motion correction technique called DISORDER produces reliable brain measurements in children, even when motion is present, improving on conventional methods.

## Contribution

The study validates DISORDER as a motion correction technique for pediatric brain morphometry, showing improved reliability over conventional MPRAGE in motion-corrupted scans.

## Key findings

- DISORDER showed good/excellent agreement with conventional MPRAGE for most subcortical GM and regional brain volumes.
- Motion-corrupt conventional MPRAGE had significantly greater differences compared to DISORDER data in 22/58 brain structures.
- DISORDER provides more reliable morphometric measures in motion-degraded pediatric scans than conventional MPRAGE.

## Abstract

This study aimed to validate a retrospective motion correction technique, Distributed and Incoherent Sample Orders for Reconstruction Deblurring using Encoding Redundancy (DISORDER), for pediatric brain morphometry.

Two T1-weighted MPRAGE 3D datasets were acquired at 3 T in thirty-seven children aged 7–8 years: one with conventional linear phase encoding and one using DISORDER. MPRAGE images were scored as motion-free or motion-corrupt. Cortical morphometry and regional brain volumes were measured with FreeSurfer, subcortical grey matter (GM) with FSL-FIRST, and hippocampi with HippUnfold. Intraclass correlation coefficient (ICC) was used to determine agreement. Mann–Whitney U was used to test the difference between measures obtained using DISORDER and (i) motion-free and (ii) motion-corrupt conventional MPRAGE data.

ICC measures between conventional MPRAGE and DISORDER data were good/excellent for most subcortical GM (motion-free, 0.75–0.96; motion-corrupt, 0.62–0.98) and regional brain volumes (motion-free 0.47–0.99; motion-corrupt, 0.54–0.99), except for the amygdala and nucleus accumbens (motion-free, 0.38–0.65; motion-corrupt, 0.1–0.42). These values were less consistent for motion-corrupt conventional MPRAGE data for hippocampal volumes (motion-free 0.65–0.99; motion-corrupt, 0.11–0.91) and cortical measures (motion-free 0.76–0.98; motion-corrupt, 0.09–0.74). Mann–Whitney U showed percentage differences in measures obtained with motion-corrupt conventional MPRAGE compared to DISORDER data were significantly greater than in those obtained using motion-free conventional MPRAGE data in 22/58 structures.

In the absence of motion, morphometric measures obtained using DISORDER are largely consistent with those from conventional MPRAGE data, whereas improved reliability is obtained by DISORDER for motion-degraded scans. This study validates the use of DISORDER for brain morphometric studies in children.