Algorithm for the detection of the mid-brain in B mode ultrasound images

Abstract

This paper presented an algorithm for the detection of the mid-brain in B mode ultrasound images. Template matching is used to detect and mark the mid-brain area. Templates are created by changing scale and orientation of average mid-brain region outline, which is made by using Procrustes analysis by using statistical methods on 43 mid-brain contour marking by an expert. The algorithm process the original image by cropping it and equalizing histogram. The normalized cross-correlation is used to match the mid-brain contour. The results of the algorithm showed that it can be used to automate the process of evaluating the mid-brain area. The created algorithm detects mid-brain area in most of the cases, but due to variable mid-brain shape, not always shows correct scale and orientation.