Low Resolution Technique for Fast Identification of Carotid Artery in Computed Tomography

Abstract

Just as in human experience, algorithms designed to automatically locate a target within a large picture can benefit from a step back that looks at the big picture. The methodology, shown here, greatly reduces the resolution so that large scale orientation markers can be used. With very large pixels, the processing is very fast, and the results of automated cropping provide a much reduced image for further processing.