We are pleased to announce that the MICCAI Learn2Reg Image registration challenge will take place again this year. We invite you to the challenge kick-off event on 18 May at 3 pm (CEST) via Zoom (https://us06web.zoom.us/j/82968027226?pwd=dTJ3RHR0RU1LZG11ZFFpNEI3Zm0zdz09). We will present the new tasks and submission formats and you can ask questions. Find more information on our website: https://learn2reg.grand-challenge.org.
Learn2Reg is a multi-task image registration challenge that aims at gaining insight into which methodological components and supervision strategies and offers the chance to benchmark recent DL-registration algorithms on standardised training and test data for a large variety of relevant medical registration tasks, including: Abdomen CT-CT and multimodal CT-MRI, inter-subject Brain MRI, inspiration-exhale Lung CT, follow-up for lung nodules

New this year:
In addition to the possibility to submit task-specific solutions, there will be a new task this year that addresses the development of a universal image registration framework and the privacy-preserving learning from hidden data. This will be run as a type 3 challenge, which means that participants will submit a training Docker Container that will be trained on hidden data sets by the organisers. Please find more information about different types of challenge at https://grand-challenge.org/documentation/challenges/
Teams participating in the development of a universal image registration framework or one that automatically learns from hidden data will be invited to contribute to a challenge overview paper, which will be submitted to a high impact journal (MedIA/TMI).

For any other questions feel free to email us at learn2reg@gmail.com.

Best wishes, Mattias, Alessa Lasse and Christoph