TeTrIS: Template Transformer Networks for Image Segmentation With Shape Priors

This research, published in IEEE Transactions on Medical Imaging (TMI), introduces TeTrIS, a network architecture that integrates shape priors into the segmentation process.

Summary:

Many medical imaging tasks require segmenting objects with consistent global structures. TeTrIS uses a template-based transformation network to ensure that predicted segmentations adhere to realistic shape priors while still being flexible enough to capture local variations.

Published in IEEE TMI, Vol. 38, Issue 11, Nov. 2019.

[Full Paper on IEEE Xplore]