Submissions

Submissions will be expected in the form of short reports which explain the method (with sufficient detail for reproduction) on the validation set. The participants will be asked to submit their code as a docker, and the results will be run on the held-out test set by the challenge team.  All submissions will be required to make an open-release of their code, and results on train, test and validation sets will be published on the MLCN website.

Short report format

Participants must submit a method description along with their submission on the final test set. This should be in Springer Lecture Notes in Computer Science format [link to Overleaf template]. The report should highlight the main steps taken to arrive at the method used on the final test set. In particular, the following should be described:

• Data preprocessing and augmentation
• Method description
• Post Processing
• Results
• Link to public code repository
• Performance of the method on the validation set

Each method description paper should be uploaded to a pre-print platform such as arXiv, so that the paper can be easily linked to the final leaderboard. Each final submission needs to be linked to a pre-print paper.

Docker Submission

To submit your solution, you will have to upload it as a Docker Container on the grand-challenge website. We created a template solution, that you as a participant, can reuse and adapt for your own solution: SLCN Algorithm Container Example.  Detailed explanations of the submission process and requirements are provided in the README file. Please read it carefully. 

The set set will be mounted on the servers into /input/images/cortical-surface-mesh/ and the predictions must be written in /output/birth-age.json. The folder /input/images/cortical-surface-mesh/ will contain all the test images in the format <uid>.mha . The algorithm container will read and make the prediction on test images, successively. 

Participants must write birth age prediction simply as a single float value in a json file /output/birth-age.json. More details in SLCN Algorithm Container Example.

In a nutshell, as a participant, you will have to:

1. Create your algorithm container following the template provided and insert your code and trained weights accordingly.
2. Test the algorithm container localy and export it into a .tar file.
3. Create your algorithm page on the grand-challenge website.
4. Upload your Algorithm Docker container and try-it-out with a .mha  image example.
5. Submit your algorithm on the SLCN challenge webpage.