Fig. 6

This figure presents the semantic segmentation results obtained from training various U-Net models, including the standard U-Net as well as its ResNet-U-Net, Inception-U-Net, and VGG-U-Net variants, using different MRI image weights such as T1, T2, T1Gd, and T2-FLAIR.1 Each row in the figure corresponds to a specific MRI image modality, while each column represents a different U-Net architecture. The segmentation outputs demonstrate the models’ ability to accurately delineate regions of interest, including edema, Necrotic, and active tumor regions, across diverse MRI image weights and U-Net models.