Robust Non-rigid Registration Challenge for Expansion Microscopy Banner
Join

Robust Non-rigid Registration Challenge for Expansion Microscopy (RnR-ExM)

About  

Expansion microscopy (ExM) is a fast-growing imaging technique for super-resolution fluorescence microscopy through tissue expansion [1]. When ExM is used with multiplexed imaging, it is critical to robustly register high-resolution 3D microscopy image volumes from different sets of staining where the image quality differs. 

Despite the wide adoption of ExM, there are few public benchmarks to evaluate the registration pipeline, which limits the development of robust methods for real-world deployment. To address this issue, we have launched RnR-ExM, a challenge that releases 24 pairs of 3D image volumes from three different species. Participants are asked to align these pairs and submit dense deformation fields for assessment. Half of the volume pairs (the validation and test set) have annotated cell structures (nuclei, blood vessels) as registration landmarks. 

Any means of registration is welcome, but to facilitate the development of learning-based approaches, we provide participants with code to generate a synthetic dataset that has a realistic deformation and appearance when compared to the unlabeled volume pairs.

Why is this challenge important?
  • The results from ExM registration are used in important downstream tasks. One area that relies on ExM registration is transcriptomics, a field of study that has led to developments in personalized medicine, drug discovery, and cancer research.
  • Image registration has been extensively explored in the medical domain, however ExM images present a unique set of challenges such as repetitive small-scale texture, degraded image quality over rounds due to re-staining, highly non-linear deformation of gel, and image size. 

Important Dates  

December 19, 2022. Release of training set, code to generate synthetic dataset, baseline model and tutorial.

January 19, 2023. Release of validation set, code for evaluation and template for write-up.

February 8, 2023. Release of test set.

~~February 15, 2023. Submission of results begins. ~~

March 15, 2023. Submission of results ends.

~~March 16, 2023. Final leaderboard released. ~~

~~April 18, 2023. On-site workshop presentations at ISBI. ~~

April 18, 2023 onwards. Leaderboards open for continuous benchmarking. 

Prizes  

  • The three top-performing teams are eligible to participate in a joint publication with the committee submitting to IEEE Transactions on Medical Imaging (TMI). There is a fixed maximum of two authors per team. The committee will also invite teams that submit particularly novel solutions to join as co-authors.
  • Certificates, an iPad and two ebook readers will be awarded to the challenge’s top three teams (1 winner and 2 runner-ups).
  • Top-performing teams will have the opportunity to present their results at ISBI 2023. 

How to Participate   

1. Create a Grand Challenge account.

2. Read the competition rules. 

3. Join the RnR-ExM challenge using the join button. 

4. Download the dataset under the dataset tab. 

5. Submit test results starting February 15th. 

References   

[1] F. Chen, P. W. Tillberg, and E. S. Boyden, “Expansion microscopy,” Science, vol. 347, no. 6221, pp. 543–548, 2015.