An Explainable CNN Model for Image Based Tumor Detection Of Brain

Abstract

The emerging topic in many medical diagnostic applications is automated flaw identification in medical imaging. Automated tumor diagnosis in magnetic resonance imaging (MRI) is extremely important since it offers details about aberrant tissues that are required for treatment planning. Human inspection is the standard procedure for flaw identification in magnetic resonance brain pictures. Using this strategy with a lot of data is not practicable. Therefore, automated tumor identification techniques are created in order to free up radiologist time. Due to the intricacy and variety of tumors, detecting brain tumors with an MRI is a challenging undertaking. In this study, machine learning techniques are used to find tumors in brain MRIs. Three sections make up the proposed work: On brain MRI pictures, pre-processing procedures are used, and texture characteristics are retrieved and retrieved with the use of the Gray Level Co-occurrence Matrix (GLCM), followed by categorization using a machine learning technique.