3D Densely Convolutional Networks for Volumetric Segmentation
This work addresses the problem of low-contrast tissue segmentation in infant brain MRI for medical imaging researchers, representing an incremental advance in network design.
The authors tackled the challenge of accurate volumetric brain segmentation in the isointense stage by proposing a novel densely convolutional network architecture, achieving significant improvements in segmentation accuracy and parameter efficiency on the MICCAI grand challenge dataset.
In the isointense stage, the accurate volumetric image segmentation is a challenging task due to the low contrast between tissues. In this paper, we propose a novel very deep network architecture based on a densely convolutional network for volumetric brain segmentation. The proposed network architecture provides a dense connection between layers that aims to improve the information flow in the network. By concatenating features map of fine and coarse dense blocks, it allows capturing multi-scale contextual information. Experimental results demonstrate significant advantages of the proposed method over existing methods, in terms of both segmentation accuracy and parameter efficiency in MICCAI grand challenge on 6-month infant brain MRI segmentation.