Using U-Net network for efficient brain tumor segmentation in MRI images

Magnetic Resonance Imaging (MRI) is the most commonly used non-intrusive technique for medical image acquisition. Brain tumor segmentation process of algorithmically identifying tumors in brain MRI scans. While many approaches have been proposed literature segmentation, this paper proposes a lightweight implementation U-Net. Apart from providing real-time scans, architecture does not need large amount data to train Moreover, no additional augmentation step required. The U-Net shows very promising results on BITE dataset and it achieves mean intersection-over-union (IoU) 89% while outperforming standard benchmark algorithms. Additionally, work demonstrates an effective use three perspective planes, instead original three-dimensional volumetric images, simplified segmentation. • A An accurate deep learning-based approach. Demonstration planes 3D images. neural network with other