Learn2Reg - The Challenge (2022)

Medical image registration plays a very important role in improving clinical workflows, computer-assisted interventions and diagnosis as well as for research studies involving e.g. morphological analysis. Besides ongoing research into new concepts for optimisation, similarity metrics, domain adaptation and deformation models, deep learning for medical registration is currently starting to show promising advances that could improve the robustness, generalisation, computational speed and accuracy of conventional algorithms to enable better practical translation. Nevertheless, before Learn2Reg there was no commonly used benchmark dataset to compare state-of-the-art learning based registration among another and with their conventional (not trained) counterparts.

This year, we want to further extend the Learn2Reg challenge to new tasks and challenges. We will move from a type 1 to a type 2 challenge meaning that only Docker submissions containing the algorithm are allowed. This facilitates reproducibility and further use of the algorithms in the research community.

Further innovations are introduced in Task 1 and 3.
In task 1, the challenge is divided into two phases: In phase 1, participants train/tune their algorithms locally and submit the algorithms via grand-challenge. The best teams of this phase are invited to participate in phase 2. In phase 2, the participants submit a training docker that will be run by the organizers on a larger dataset that includes additional annotations that are not publically available. The trained networks will be made available via grand-challenge.

Task 3 of Learn2Reg 2022 aims to find the best self-configuring registration algorithm that can automatically be trained and optimised on a variety of hidden datasets. Possible solutions include modular architectural designs that enable adaptive changes in cost function and optimisation strategy or hyper-networks that can be efficiently fine-tuned with few parameters after initial training. Participants are encouraged to upload a single docker container that automatically learns a registration model on any given training/validation dataset (e.g. but not limited to multimodal abdominal fusion, intra-patient lung CT and follow-up as well as inter-subject alignment).