Coronal slice segmentation using a watershed method for early identification of people with Alzheimer's

Abstract

One physical sign of a person who has Alzheimer's is the diminution of the area of the hippocampus and ventricles. A good quality magnetic resonance imaging (MRI) will provide a high-quality image so that the doctor will quickly analyze the abnormalities of the hippocampus and ventricle area. However, for low-quality MRI, this is difficult to do. This condition will be a significant problem for some regions in developing countries including Indonesia, where many hospitals have only low-quality MRI, and many hospitals do not have them at all. The primary purpose of this research is to develop simple tools to analyze morphological characteristics in Alzheimer's patients. In this paper, we focus only on coronal slice analysis. We will use watershed method segmentation, because of this method able to segment the boundaries automatically, so that parts of the hippocampus and ventricles can be identified in an MRI image. Analysis of morphological characteristics is also classified by age and gender. Then by referring to the value of the clinical dementia rating (CDR), the process of identifying between images with Alzheimer's disease (AD) and healthy models is done based on the morphological analysis that has been done. The results show this method has a better performance compared to the previously work.