Contour detection of hippocampus using dynamic contour model and region growing

In hippocampal morphology, abnormalities including unilateral or bilateral volume loss are known to cause epilepsy, Alzheimer's disease, and certain amnestic syndromes. The accurate segmentation of the hippocampal contour is critical for the diagnosis. However, it is very difficult to extract the contour that matches exactly the hippocampal region in magnetic resonance images (MRI). The brightness of the hippocampal region varies linearly and it's size is small. For the accurate and consistent detection of the hippocampal region, a method which combines region growing and a dynamic contour model to detect the hippocampus from MRI brain images is presented. The segmentation process is performed in two steps. First, region growing with a seed point is performed in the hippocampal region and its output is used for the initial contour of the dynamic contour model. Second, the initial contour is modified on the basis of criteria which integrate energy with contour smoothness and image gradient along the contour. As a result, this method improves the sensitivity of the choice of the initial seed point and precisely extracts the hippocampus from the MRI brain image.