Segmenting and grading brain tumors using neuro-fuzzy and fuzzy clustering methods on magnetic resonance images

It is always been the toughest task to segment and grade the brain tumors simultaneously using computational methods. The conventional computational methods such as threshold, edge detection has been utilized in the past to segment the different pathologies of the brain tumors in magnetic resonance imaging; however they were utilizing the T1 and T2 information of the tissue which subsequently limited their ability to grade and identify the higher vascular regions. The techniques handled in the above approaches have inherent inability to overcome the partial volume effects and handle the complexity of the magnetic resonance image data. They also were utilizing limited information provided by conventional T1 and T2 weighted images