Infarct pattern analysis using an average CT brain image

Introduction

One of the risk factors for brain infarcts is stenosis of the Internal Carotid Artery (ICA), i.e. (partial) obstruction of the vessel in the neck that provides the brain with blood. The cause of brain infarcts in patients with severly stenosed ICA is thought to be hemodynamic impairment. This means that blood supply to brain areas at the end of the vascular tree is insufficient and causes so called border zone infarcts. Therefore, these border zone infarcts will differ in location from embolic infarcts.

This was examined by collecting CT images of patients with no or only mild ICA stenosis (0-49%), patients with severe stenosis (50-99%), and patients with an occlusion of the ICA (100%).

Brain infarct

CT Slice showing large infarct in the territory of the Medial Cerebral Artery with compression of the ventricles.

Construction of infarct patterns

An infarct pattern for each stenosis category was constructed to compare all infarct locations. Infarct patterns were made by mapping the patient CT images to an average CT brain image. For information on the construction of the average CT image see Construction of Average CT Brain Image. Infarcts in the patient images were manually segmented. Knowing the mapping of the patient image to the average CT brain image, the mapping of the infarcts to the average CT brain image is, of course, also known. The resulting infarct patterns are shown below.

Infarct patterns for mild ICA stenosis (left column), severe ICA stenosis (middle column), and ICA occlusion (right column).

The colors indicate the percentage of cases that infarcted tissue was found at a certain position.

Analysis of infarct patterns

The analysis of differences between the infarct patterns revealed no differences between infarct location in mild and severe ICA stenosis. However, both mild and severe ICA stenosis infarct patterns differed from the ICA occlusion infarct pattern in some small areas. So these findings do not support the hypothesis that severe ICA stenosis leads to a border zone infarct, which would have created differences between the infarct pattern for mild and for severe ICA stenosis.

Statistics for mild ICA stenosis vs. severe ICA stenosis (left column), statistics for mild ICA stenosis vs. ICA occlusion (middle column), and statistics for severe ICA stenosis vs. ICA occlusion (right column).

The colors indicate the probability that the infarct patterns do not differ.

References

This page describes work by Cynthia Jongen, Paul Nederkoorn, Wiro Niessen, and Josien Pluim.