Brain Tumor Classification Using UNet Deep Neural Networks from 3D MRI Images

A patient's therapy options and prognosis are both improved by detecting and diagnosing a brain tumor early. Brain tumors can be located and diagnosed with the help of magnetic resonance imaging (MRI). In real practice, however, medical personnel rely entirely on their own time and expertise when manually identifying brain tumors from a vast number of MRI scans. These days, medical diagnosis and treatment recommendation systems that use computer-aided expertise are all the rage. Brain tumor detection makes use of a plethora of frameworks based on machine-learning(ML) &deep-learning(DL). The purpose of this research is to present a DL-based approach for effective brain tumor segmentation and classification. The research uses convolutional-neural-networks(CNNs) for tumor classification after volumetric segmentation of MRI-images using the 3D-UNet model. To demonstrate the models' validity, loss and precision plots are provided. We compare the outcomes of several methods described in the literature with the measured performance of the models we offer. When compared to the most recent cutting-edge methods, the proposed work comes out on top.