Segmentation of cortical gray and white matters from MRI using density based clustering approach

With the availability of high precision MRI scanners helps in automatic clinical analysis of human brain for tumor detection. Cerebral cortex extraction from brain MR images is a challenging task due to its highly convoluted complex structures. In this paper we first extract cerebral cortex from MRI using morphological approach. Next we segment gray and white matters using DenSeg, a Density based Segmentation/ clustering technique. We use publicly available IBSR datasets for experimentations and validation against manually segmented samples. A prototype java tool is developed based on the proposed method.