Detection and Classification of Brain Tumor Based on Multilevel Segmentation with Convolutional Neural Network

• DOI: 10.4236/jbise.2020.134004
• Publication Date: 2020-04-29T07:03:00Z
• Authors: Rafiqul Islam; Shah Imran; Md. Ashikuzzaman; Md. Munim Ali Khan
• Keywords: Kernel (algebra); Brain tumor; Feature (linguistics); Contextual image classification

Magnetic Resonance Imaging (MRI) is an important diagnostic technique for early detection of brain Tumor and the classification from MRI image a challenging research work because its different shapes, location intensities. For successful classification, segmentation method required to separate Tumor. Then features are extracted segmented that used classify In this work, efficient multilevel developed combining optimal thresholding watershed followed by morphological operation Convolutional Neural Network (CNN) then applied feature extraction finally, Kernel Support Vector Machine (KSVM) utilized resultant justified our experimental evaluation. Experimental results show proposed effectively detect as cancerous or non-cancerous with promising accuracy.