Background/Purpose: Traditionally, hypoxic-ischemic brain injury has been thought to result in infarction in ‘watershed’ zones. However, formal evaluation of the topography of these lesions has not been performed. Knowledge of the extent and topography of infarction in hypoxic-ischemic brain injury will help determine regions at highest risk of infarction, thus providing a clue into mechanisms of infarction. This study aims to create a digital probability map of infarction in hypoxic –ischemic brain injury in patients post cardio-respiratory arrest.
Methods: We included patients with cardio-respiratory arrest between the years 2008 to 2010. The inclusion criteria were: age >18 years and a clinical diagnosis of coma on admission. Infarcts were manually segmented on T2-weighted magnetic resonance images (FLAIR and diffusion weighted sequences) obtained >24 hours after cardio-respiratory arrest. Images were linearly registered into common sterotaxic coordinate space. Segmented images were averaged to yield the probability of infarction at each voxel.
Results: Twenty-nine patients (mean age 49.6 years, range 21- 81 years) were studied. In our digital atlas, the highest frequency of infarction on the diffusion-weighted sequences was within the putamen (probability = 0.2), caudate (probability = 0.2), thalamus (probability = 0.1), and temporo-parietal cortical involvement (probability = 0.1).
Conclusion: We have created a probabilistic digital of hypoxic-ischemic brain infarction. This approach is useful in establishing the spatial distribution of infarction in hypoxic-ischemic brain injury post cardio-respiratory arrest.