Gray matter volume correlates of adolescent posttraumatic stress disorder: A comparison of manual intervention and automated segmentation in FreeSurfer.

ABSTRACT

Exposure to early life trauma is common and confers risk for psychological disorders in adolescence, including posttraumatic stress disorder (PTSD). Trauma exposure and PTSD are also consistently linked to alterations in gray matter volume (GMV). Despite the quantity of structural neuroimaging research in trauma-exposed populations, little consensus exists amongst research groups on best practices for image processing method and manual editing procedures. The purpose of this report is to evaluate the utility of manual editing of magnetic resonance (MR) images for detecting PTSD-related group differences in GMV. Here, T1-weighted MR images from adolescent girls aged 11-17 were obtained and analyzed. Two datasets were created from the FreeSurfer reconall pipeline, one of which was manually edited by trained research assistants. Gray matter regions of interest were selected and total volume estimates were entered into linear mixed effects models with method (manual edits or automated) as a within-subjects factor and group dummy-coded with PTSD as the reference group. Consistent with prior literature, individuals with PTSD demonstrated reduced GMV of the amygdala compared to trauma-exposed and non-trauma exposed controls, independent of editing method. Our results demonstrate that amygdala GMV reductions in PTSD are robust to certain methodological choices and do not suggest a benefit to the time-intensive manual editing pipeline in FreeSurfer for quantifying PTSD-related GMV.