Nonaccidental pediatric head injury: Diffusion-weighted imaging findings

OBJECTIVE: Diffusion-weighted imaging (DWI) reveals nonhemorrhagic posttrau matic infarction hours to days before conventional computed tomographic sca nning or magnetic resonance imaging (MRI). We evaluated the diagnostic util ity of DWI in children with nonaccidental head trauma. METHODS: The medical records and imaging examinations obtained between Janu ary 1998 and May 2000 for all children less than 2 years of age with presum ed or suspected nonaccidental head injury were reviewed retrospectively. Tw enty children who had undergone DWI within 5 days of presentation were incl uded in the study. Computed tomographic scans, conventional MRI sequences, and DWI combined with apparent diffusion coefficient (ADC) maps were evalua ted. RESULTS: Eleven girls and nine boys (median age, 5 mo) were studied. Eighte en children had presumed nonaccidental head trauma, and two children had su spected nonaccidental head trauma. Of the 18 children with presumed nonacci dental trauma, 16 (89%) demonstrated abnormalities on DWI/ADC, as compared with neither of the two children with suspected nonaccidental trauma. In 13 (81%) of 16 positive cases, DWI revealed more extensive brain injury than was demonstrated on conventional MRI sequences or showed injuries not obser ved on conventional MRI. DWI combined with ADC maps allowed better delineat ion of the extent of white matter injury. DWI/ADC abnormalities in the nona ccidental head-injured children were likely to involve posterior aspects of the cerebral hemispheres, with relative sparing of the frontal and tempora l poles. Severity on DWI correlated significantly with poor outcome (P < 0. 005). CONCLUSION: DWI has broad applications in the early detection of infarction in children with nonaccidental head injury and enhances the sensitivity of conventional MRI. In the patients in this study, early DWI provided an ind icator of severity that was more complete than any other imaging modality. The use of DWI may help to identify children at high risk for poor outcome and to guide management decisions.