Deformable Image Registration with Learned Field of Experts – Onur Ali Zeybekoglu

The aim of this work is to create learnable image-registration models that match the speed of deep-learning methods and at the same time maintains the interpretability and controllability of classical algorithms by integrating learned Field of Experts. Unlike "black-box" networks our models remain fully transparent and preserves the explicit optimization structure. These methods will be integrated into adaptive radiotherapy (ART) workflows, which can benefit both interfraction and intrafraction applications.

About Onur Ali Zeybekoglu