Brain Tumor: Hybrid Feature Extraction Based on UNet and 3DCNN

Automated segmentation of brain tumors using Magnetic Resonance Imaging (MRI) data is critical in the analysis and monitoring disease development. As a result, gliomas are aggressive diverse that may be split into intra-tumoral groups by effective accurate methods. It intended to extract characteristics from an image Gray Level Co-occurrence (GLC) matrix feature extraction method described proposed work. Using Convolutional Neural Networks (CNNs), which commonly used biomedical segmentation, CNNs have significantly improved precision state-of-the-art tumor. two networks, U-Net 3D CNN, we present major yet easy combinative technique results more precise estimates. The CNN together this study get better estimates what going on. dataset, models were developed assessed provide maps differed fundamentally terms segmented tumour sub-region. Then, was made separate put produce final prediction. In comparison current designs, (percentage) 98.35, 98.5, 99.4 on validation set for tumor core, enhanced tumor, whole respectively.