Diffuse damage in pediatric traumatic brain injury: a comparison of automated versus operator-controlled quantification methods

Abstract

This investigation had two main objectives: 1) to assess the comparability of volumes determined by operator-controlled image quantification with automated image analysis in evaluating atrophic brain changes related to traumatic brain injury (TBI) in children, and 2) to assess the extent of diffuse structural changes throughout the brain as determined by reduced volume of a brain structure or region of interest (ROI). Operator-controlled methods used ANALYZE software for segmentation and tracing routines of pre-defined brain structures and ROIs. For automated image analyses, the open-access FreeSurfer program was used. Sixteen children with moderate-to-severe TBI were compared to individually matched, typically developing control children and the volumes of 18 brain structures and/or ROIs were compared between the two methods. Both methods detected atrophic changes but differed in the magnitude of the atrophic effect with the best agreement in subcortical structures. The volumes of all brain structures/ROIs were smaller in the TBI group regardless of method used; overall effect size differences were minimal for caudate and putamen but moderate to large for all other measures. This is reflective of the diffuse nature of TBI and its widespread impact on structural brain integrity, indicating that both FreeSurfer and operator-controlled methods can reliably assess cross-sectional volumetric changes in pediatric TBI.