Background and Purpose Diffusion-weighted magnetic resonance imaging (
MRI) has been shown to be particularly effective in detecting early (0
to 4 hours) pathophysiological changes in localized brain regions aft
er cerebral ischemia. The present study sought to establish whether di
ffusion-weighted MRI would be similarly effective in predicting outcom
e after traumatic brain injury. Methods Diffusion-weighted MRI images
and T2-weighted MRI images were obtained over 4 hours after either mod
erate fluid percussion-induced traumatic brain injury or unilateral ca
rotid ligation in rats. Results Diffusion-weighted MRI images of traum
atic brain injury demonstrated focal regions of image hypointensity as
early as 1 hour after trauma. The relative diffusion coefficient in t
hese hypointense regions was significantly increased (P<.005) by 4 hou
rs after trauma compared with the noninjured hemisphere, but only in t
he transverse plane in the x direction. In contrast, induction of diff
use, nonfocal ischemia by unilateral carotid ligation resulted in scat
tered regions of hyperintensity with a significant (P<.001) decrease i
n relative diffusion coefficient as early as 1 hour after ligation com
pared with the noninjured hemisphere. This decrease exhibited no direc
tionality. Conclusions We conclude that traumatic brain injury results
in an increased water diffusion distance with the directionality indi
cative of bulk flow of extracellular fluid toward the lateral ventricl
es (vasogenic edema). In contrast, the decreased water diffusion dista
nce with no apparent directionality observed in ischemia is most likel
y indicative of cytotoxic edema. Diffusion-weighted MRI therefore has
the potential to differentiate cases of traumatic brain injury with no
focal ischemia from those instances of traumatic brain injury in whic
h focal ischemia is a complication.