Gray Matter and White Matter Segmentation from MRI Brain Images Using Clustering Methods

Brain is the most complex and master organ of the human body. Quantitative analysis of anatomical brain tissues such as GM and WM is important for the clinical diagnosis and therapy of neurological diseases. Brain imaging is a widely used method by the doctors and clinicians for the representation of human brain. Magnetic Resonance Imaging (MRI) is a recently emerged method for brain imaging. Accurate segmentation of the brain tissues from the MR images is useful in neuro-diagnosis and neurosurgery. It will help the doctors for the diagnosing of some complex diseases such as Epilepsy, Stroke, Alzheimers disease, brain tumor, brain infection and multiple sclerosis. This paper proposes an efficient method for the automatic segmentation gray matter and white matter regions from MRI brain images using Fuzzy c-Means (FCM) and Kohonen means (k-Means) clustering methods. We implemented the clustering of gray matter and white matter using intensity values and statistical feature based values. Finally the results are compared with the manually marked ground truths using some standard accuracy measurement coefficients. The experimental result shows that statistical feature based clustering produces more prominent result than intensity based clustering.