Deep Learning Algorithm for Brain Tumor Detection and Classification using MRI Images

Brain tumor classification is beneficial for identifying and diagnosing the tumor’s specific location. According to the medical imaging system, early diagnosis and categorization of a tumor extend a person’s life. Clinical specialists rely heavily on magnetic resonance imaging (MRI) among numerous imaging modalities since it provides contrast information on brain malignancies. The primary purpose of this project is to use a competent automated approach that improves tumour identification accuracy. Several segmentation strategies have been developed throughout the years to achieve and improve the categorization precision of brain tumours. Brain picture segmentation has long been recognised as a difficult and time-consuming aspect of medical image processing. This method for detecting brain tumors Brain pictures are classified using the Full Resolution Convolutional Network (FRCN) classification architecture after pre-processing and segmentation. This study presents a Full Resolution Convolutional Network (FRCN) with Support Vector Machine (SVM) approach for detecting tumors on MRI scans. The procedure is broken down into four steps. In the first phase, the anisotropic filter is utilized to pre-process raw MRI images, followed by segmentation using the Support vector machine (SVM) and skull classification. The singular value decomposition and primary component analysis operations are performed in the third step. Tumors are then detected and classified using the Full Resolution Convolutional Network (FRCN) approach. Simultaneously, the Support Vector Machine (SVM) technique is employed to improve the classification precision of the study model. The experimental results showed an amazing accuracy rate of nearly 100% in detecting both normal and diseased tissues from brain MR images, confirming the efficacy of the suggested technique.