Prediction of subsequent hemorrhage in acute ischemic stroke using permeability CT imaging and a distributed parameter tracer kinetic model

J Neuroradiol. 2007 May;34(2):101-8. doi: 10.1016/j.neurad.2007.02.003.


Purpose: Hemorrhagic transformation (HT) is a common consequence of infarction independent of thrombolytic therapy. Our purpose was to examine if permeability imaging in admission perfusion CT data of patients with acute stroke might indicate a subsequent HT by imaging the disrupted permeability barriers between blood and brain.

Materials and methods: A distributed parameter model analysis of the perfusion data were used to analyze the admission perfusion surveys of eight patients with HT of the initial infarct without thrombolysis. The perfusion findings of these patients were compared with those of eight age- and gender-matched patients from the initial group that did not present with HT.

Results: The applied statistics for comparing the ischemic voxels with the contralateral healthy tissue showed significantly higher permeability-surface product (PS), extraction ratio (E), and extracellular extravascular space volume (V(EES)) in the ischemic voxels (P range, 0.05-0.0001). In the patients without HT, the PS, E and V(EES) values in the ischemic voxels were not significantly different from those in the normal region.

Conclusion: Our findings indicate that early perfusion CT physiological imaging in stroke is a promising tool for identifying patients with higher risk of HT and, thus, may serve to guide therapeutic options.

MeSH terms

  • Aged
  • Aged, 80 and over
  • Brain Ischemia / complications
  • Brain Ischemia / diagnostic imaging
  • Case-Control Studies
  • Cerebral Hemorrhage / diagnostic imaging*
  • Cerebral Hemorrhage / etiology*
  • Cerebrovascular Circulation
  • Contrast Media
  • Female
  • Humans
  • Male
  • Middle Aged
  • Predictive Value of Tests
  • Radiographic Image Interpretation, Computer-Assisted
  • Stroke / complications*
  • Stroke / diagnostic imaging
  • Tomography, Spiral Computed / methods*


  • Contrast Media