Movement during MR Scanning in Children with Autism Spectrum Disorder

Background:  GABA levels, measured with edited MRS, are reduced in children with Autism Spectrum Disorder (ASD). Long scan times are needed to obtain a sufficiently high signal-to-noise ratio (SNR) to quantify GABA reliably. SNR increases with scan duration but suffers from motion and frequency drift. Measurements in neurological pediatric cohorts can be difficult and understanding the pattern of movement, in particular when children move, might benefit the design of future imaging studies and optimize the trade-off between scan time and SNR.

Objectives:  To assess the movement behavior of children with and without ASD during edited MRS of GABA.

Methods:  Volunteer and parental consent was obtained under local IRB approval. Eligibility: Children with ASD met DSM-V diagnostic criteria, confirmed using the Autism Diagnostic Observation Schedule-Version 2 (ADOS-2). All typically developing children (TDC) were free of criteria for psychiatric disorders based on the Diagnostic Interview for Children and Adolescents. Data were acquired in 21 children with ASD and 20 TDC with normal IQ (8–12 yrs). Imaging: GABA-edited MR spectra were acquired from a 27 mL voxel over the right primary sensorimotor area using MEGA‑PRESS on a 3T Philips Achieva scanner (Philips Medical Solutions) (320 transients; scan time = 10 min). Children watched a movie during scanning. Analysis: The frequency of the suppressed water signal in each transient (frame) was extracted for both groups using Gannet 2.0. The step size between each frame was calculated and a frequency step over 2 Hz (0.016 ppm) was defined as a motion event. Differences in movement behavior in the two groups were assessed in terms of (i) size and number of movements and (ii) correlation with time point of movement (i.e., are children more likely to move near the beginning or end of the scan?).

Results:  Figure 1 shows a histogram of movement frequency among the cohort, as well as accumulative percentages. Movement frequency increased significantly throughout the scan (frames binned into groups of 10) across all children (Fig 2; R = 0.56), but more in ASD (R = 0.54) than in TDC (R = 0.3) although this was not significantly different (Fisher R-to-Z, p = 0.38). While no differences were shown in movement frequency, children with ASD had larger movements (4.8 ± 5.1 Hz) compared to TDC (3.4 ± 1.2 Hz), although not significantly (p = 0.3).

Conclusions: These data show that children with ASD do not show more movements than TDC during MRS scanning. These results are important in establishing that previously reported GABA differences are not driven by movement. More children with ASD show no movement behavior (ten without movement versus five for TDC, Fig 1). There were no significant differences in the size of movement. Importantly, children are more likely to move near the end of a 10-min scan so while shorter scans (e.g. 7 minutes instead of 10) may impact SNR, they may benefit from reduced movement related artefacts. Near-end-of-scan motion should be taken into account when planning MRS experiments, as well as prospective frequency and motion-correction.