MRI-based Tumour Prediction based on U-Net and VGG-Net

In the detection and treatment of life-threatening disorders, biomedical technology plays a critical role. Brain tumours have recently become one of the most frequent and dangerous disorders. The treatment of brain tumours is dependent on the experience and understanding of doctors. As a result, for radiologists and clinicians to assist in the identification of brain tumours, an automated tumour detection system is critical. To detect brain tumours, accurate analysis of magnetic resonance imaging (MRI) scans is required, and this study provides an effective approach for predicting brain malignancies using U-Net and VGG-Net. Image pre-processing algorithms that find extreme shapes to remove black borders from pictures are examples of data augmentation. Photos are then resized to their proper dimensions. A ReLU Optimizer has indeed been added to our network in order to speed the training process. To get the model as accurate as possible, it is trained and tested on an MRI image dataset. The performance of the maximized system is evaluated with 0.92 accuracies.