Multi scale based multimodal brain Tumor image segmentation using Neural computing

Abstract

The brain is one of the important and complex organs that controls all the metabolic activities of human body. The abnormalities or injury in the brain cause grievous problems and sometimes they could be lethal. One vital problem that has been most challenging to the physicians and patients is brain tumor in diagnosis and prognosis. Even tiny errors during the diagnosis or any deviation from the target treatment can adversely affect the patient s health. The radiologists and oncologists have utilized Magnetic Resonance Image (MRI) to diagnose brain tumors for their type and location. newlineMoreover, the development and metastasis of the tumor can be monitored and treated appropriately by MRI accurately. MRI can also render information on the surrounding tissues, necrotic tissues, and the active cancer tissues significant in tumor diagnosis, for early therapy and continuous monitoring. Since it uses non-ionizing radiation as the source for acquiring data from the brain, this method is considered safe, making it useful for constant monitoring and speedy diagnosis. newlineBrain image segmentation is one of the most important clinical diagnoses using MRI because of the complexities of the brain s anatomical structure. The brain images normally encompass cumbersome noise, inhomogeneity, and occasionally deviation. Therefore, accurate segmentation of brain images is a very complicated and untiring task. Decades of research have developed numerous segmentation approaches to predict tumors and other abnormalities caused by these tumors. Fully automatic and semi-automatic segmentation approaches have been utilized over manual segmentation to analyze larger data with more details into the MRI s image. Glioblastoma is antithetic to other brain tumors because of its penetrating and fuzzy structure newline