Deformable Image Registration Using Cue-aware Deep Regression Network
We propose a novel deformable registration method of using the deep neural network to directly learn the mapping from an image pair to the corresponding deformation field. This highly non-linear mapping is modeled by the novel cue-aware deep regression network, in which we adopt contextual cue to better guide the learning process…
Structured Learning for 3D Perivascular Spaces Segmentation Using Vascular Features
In this study, we propose a structured-learning-based segmentation framework to extract the perivascular spaces (PVSs) from high-resolution 7T MR images. Specifically, we integrate three types of vascular filter responses into a structured random forest for classifying each voxel into two categories.