Share This Author
Multimodal PET/CT Tumour Segmentation and Prediction of Progression-Free Survival using a Full-Scale UNet with Attention
- Emmanuelle Bourigault, D. McGowan, A. Mehranian, Bartlomiej W. Papie.z
- Computer ScienceHECKTOR@MICCAI
- 6 November 2021
A new network based on an encoder-decoder architecture with full interand intra-skip connections is proposed to take advantage of low-level and high-level semantics at full scales for segmentation of head and neck tumours and prediction of patient outcome.
Scoliosis Measurement on DXA Scans Using a Combined Deep Learning and Spinal Geometry Approach
Improvements to an automated method for scoliosis measurement are proposed, using a spline to better model the curve of the spine, and pseudo-labelling is employed to re-train the segmentation step to mitigate the domain gap when adapting to a new dataset.