Article Text

  1. G. A. Sigmund,
  2. J. P. Nickerson,
  3. C. J. Wall,
  4. C. Ashwall,
  5. K. Tong
  1. Loma Linda, CA.


This study was a comparison of lesion detection in the most common modalities for evaluating traumatic brain injury and their association with patient Glasgow outcome score (GOS). The modalities compared were Computed tomography (CT), T2 weighted magnetic resonance imaging (MRI), FLAIR MRI, and susceptibility-weighted imaging (SWI) MRI. The goal was to assess which sequences are most valuable in diagnostic radiology for predicting outcome after traumatic brain injury. The patient population included 40 pediatric traumatic brain injury patients admitted between March, 2001 and September, 2002. CT, T2, FLAIR, and SWI were attained for all patients along with GCS and GOS scores. Radiologic data was analyzed for intracranial lesions and confirmed with board certified neuroradiologist. Intraparenchymal lesions were traced using the Image Pro Plus program (Media Cybernetics, Silver Spring, MD), which was then converted into volume based on slice thickness and pixel size. Based on outcome score, patients were grouped into categories: 1. Normal (No neurologic impairment) 2. Mild disability and 3. Moderate/Severe disability. Rank Sum analysis was performed within each modality between (1 + 2) vs. 3, and each category individually. All modalities except CT showed a statistically significant difference between categories (1 + 2) vs. 3. For FLAIR, T2 and SWI, no significant difference was found between categories 1 and 2, but there were significant differences (p ≤ .05) between (1 vs. 3) and (2 vs. 3). CT showed no correlation with and no statistical significance between any of the categories. We have learned from this comparison that SWI, T2 and FLAIR modalities are sensitive for detection of lesions that are associated with significant outcome impairment in pediatric TBI patients. Further study is required to determine which modality is most useful for outcome prediction, or whether combinations of modalities are best. CT was inconsistent at lesion detection and outcome prediction

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.