Effective Segmentation of Brain Tumors through the GOA Algorithm using Deep Learning

Image segmentation is used to isolate aberrant tumor regions inside the brain, which is a necessary step in the diagnostic procedure. The tissue of the brain is required for the Magnetic Resonant Image (MRI) for segmenting the brain tumor. The primary objective of this proposed work is to develop a non-invasive approach to estimate the grade of a Brain Tumor (BT) using MRI image. Tumor area was recovered from postprocessed image employing the fuzzy C means methodology. The texture, local binary configuration (LBP) and fractal characteristics were recovered when the characteristics were extracted. The settings for three alternative classification algorithms, including Random Forest (RF), K-Nearest Neighbor (KNN) and Support Vector Machine (SVM), were then accustomed using the Grasshopper Optimization Algorithm (GOA). Finally, the performance of three graders was assessed before and after optimizing.