3260 papers • 126 benchmarks • 313 datasets
This task has no description! Would you like to contribute one?
(Image credit: Papersgraph)
These leaderboards are used to track progress in infant-brain-mri-segmentation-20
Use these libraries to find infant-brain-mri-segmentation-20 models and implementations
No datasets available.
No subtasks available.
A novel very deep network architecture based on a densely convolutional network for volumetric brain segmentation that provides a dense connection between layers that aims to improve the information flow in the network.
This work is the first ensemble of 3D CNNs for suggesting annotations within images, and allows the efficient propagation of gradients during training, while limiting the number of parameters, requiring one order of magnitude less parameters than popular medical image segmentation networks such as 3D U-Net.
Adding a benchmark result helps the community track progress.