A voxel-wise uncertainty-guided framework for glioma segmentation using spherical projection-based U-Net and localized refinement.

Key Points

• This research aims to develop an effective framework for automating glioma segmentation while addressing areas of uncertainty.
• Utilized interpretable uncertainty maps for spatial attention.
• Combined 2D and 3D image processing techniques.
• Implemented spherical projection-based U-Net for enhanced performance.
• Achieved improved segmentation accuracy in anatomically ambiguous regions.
• Demonstrated efficient computational resource allocation.
• Provided a robust solution for automated glioma segmentation tasks.