BrainNet: CNN-Powered Diagnosis to Detect and Classify Brain Tumor from MRI Imaging Technique

In medical image processing, brain tumor segmentation is a crucial problem.Patients' chances of survival are increased and treatment options are improved when brain tumors are detected early.It is challenging and time-consuming to manually segment brain tumors for cancer diagnosis from the ample number of MRI images produced during clinical routines.Automatic segmentation of brain tumor images is required.The varied image content, crowded objects, occlusion, image noise, non-uniform object texture, and other characteristics make segmentation a difficult challenge even after much research.Although there are numerous algorithms and methods for image segmentation, a quick and effective method for medical image segmentation still has to be developed.MRI brain images were initially subjected to preprocessing and enhancement methods.The damaged brain tumor region was then segmented using a new 2D Convolutional Neural Network (CNN) approach that is created.The proposed method is not only able to segment the affected area but also able to properly classify the type of brain tumor.The proposed technique achieved an overall accuracy of 91.3% and a recall of 88% respectively.