An Enhanced Brain Tumor Segmentation Using an Improved 3d U-Net Topology: Solution to Brats 2020 Challenge

The process of automatically recognizing malignant brain tissues and labeling them according to tumor types is known as brain tumor segmentation. Manually segmenting a tumor from a brain MRI takes a long time and is prone to errors. Thereby, a variety of fast and accurate brain tumor segmentation approaches have been developed. Deep learning algorithms have recently exhibited outstanding segmentation performance. As a result, this research suggested U-Net architecture, which is a deep learning topology extension for segmentation. The BraTS2020 dataset is used in this study to assess the efficiency of the suggested topology. The proposed U-Net model gained an accuracy of 99.8%, which is better than the state of the result obtained so far. In terms of accuracy, the simulation results reveal that the proposed U-Net design outperforms the existing models.