Efficient clustering based segmentation and machine learning classification approaches for mri brain tumor prediction

Abstract

Magnetic Resonance Imaging (MRI) is a widely accepted modality for providing anatomical information. Current research focuses on extending MRI methods to provide adequate information regarding biological function in addition to the concomitant anatomical information. The brain possess anatomically distinct processing regions. A complete understanding of brain function requires determination of where these sites are located, what operations are performed, how distributed processing is organized. Changes in neuronal activity are accompanied by focal changes in Cerebral Blood Flow (CBF), Cerebral Blood Volume (CBV), blood oxygenation, and metabolism. These physiological changes can be used to produce functional maps of component mental operations. Magnetic Resonance Imaging techniques sensitive to changes in blood flow and blood oxygenation are developed. Neurodegenerative disorders, psychiatric disorders, and aging are all frequently associated with structural changes in the brain. Brain abnormalities are wide ranging and can be classified as both organic and developmental. Brain tumor is found to be the most predominant abnormality that affects the brain. A brain tumor occurs when abnormal cell forms within the region of brain. It is one of the leading cause of cancer mortality amongst people. An accurate detection of tumor from the Magnetic Resonance Images (MRI) has acquired significance in the field of medical image processing. A stroke occurs when blood supply to the part of brain interrupted or reduced, depriving brain tissue of oxygen and nutrients. Brain tumor tissue detection is a challenging concept in the area of medical image processing. It is a highly complex task due to the structure of White Matter (WM), Gray Matter (GM), and Cerebrospinal Fluid (CSF) tissues. So, various imaging modalities are established, which includes Computed Tomography (CT), Magnetic Resonance Imaging (MRI), and Positron Emission Tomography (PET). Amongst them, the MRI is mostly recommended by the medical experts, because it offers the more details about the brain. Moreover, the MRI is highly suitable for extracting the tumor region and other tissues that is internally affected. Based on the origin of tumor, it is categorized into two categories such as benign, and malignant. An exact detection of tumor from the Magnetic Resonance Images (MRIs) is a critical and demanding task in medical image processing, due to the varying shape and structure of brain. So, different segmentation methods like manual, semi-automatic, and fully automatic are developed in the conventional works. Among them, the fully automatic segmentation methods are increasingly applied by the medical experts for an efficient disease diagnosis. But, it has the drawbacks of over segmentation, increased complexity, and time consumption. In order to solve these problems, this research aims to develop an efficient segmentation and classification system by incorporating a novel image processing techniques. Here, the Distribution based Adaptive Median Filtering (DMAF) technique is employed for preprocessing the image. Then, skull removal is performed to extract the tumor portion from the filtered image. Further, the Neighborhood Differential Edge Detection (NDED) technique is implemented to cluster the tumor affected pixels, and it is segmented by the use of Intensity Variation Pattern Analysis (IVPA) technique. Finally, the normal and abnormal images are classified newline