Comparison of subgroups based on hemorrhagic lesions between SWI and FLAIR in pediatric traumatic brain injury

Purpose: The purpose of this study was to investigate efficient ways to diagnose and predict clinical outcomes for childhood traumatic brain injury. Methods: Hemorrhagic signal intensities in nine brain regions were observed using axial fluid-attenuated inversion recovery (FLAIR) and susceptibility- weighted imaging (SWI). After having divided the subjects into mild presentation (GCS 14-15) and moderate-to-severe presentation groups (GCS ≤13), we divided the patients into three subgroups: Subgroup I, hemorrhagic foci observed only on SWI and not on FLAIR; Subgroup II, hemorrhagic foci observed on both SWI and FLAIR in the same brain regions; and Subgroup III, any cases with additional foci on SWI in other brain regions. We investigated the clinical course and compared lesion numbers and distributions of hemorrhagic lesions on SWI among the subgroups. Results: Three clinical variables (hospitalization period in intensive care unit, total days of hospitalization, and outcome based on Pediatric Cerebral Performance Category Scale score) showed significant relevance to the three subgroups. Subgroup I showed the fewest lesions followed by Subgroups II and III, respectively. In all three subgroups, lesions were most abundant in cortical regions. Lesion in the thalamus, basal ganglia, corpus callosum, and brainstem was least in Subgroup I and gradually increased in Subgroups II and III. Such distinction was more significant in the moderate-to-severe group when compared with the mild group. Conclusions: In cases of pediatric traumatic brain injury, categorizing patients into one of the above three subgroups based on hemorrhagic lesions on SWI and FLAIR is a promising method for predicting patient's clinical outcome.