Dear DEEP-PSMA Participants,
We will be opening the debug phase in the coming days and validation phase shortly thereafter. In the meantime, please review example algorithm and docker build scripts available here:
https://github.com/Peter-MacCallum-Cancer-Centre/DEEP-PSMA-Algorithm
For those who have participated in previous challenges, the submission process should be very familiar and we have tried to adapt the example methods provided by the GC team with minimum modification.

For the easiest route to adapting the example to your methods, all of the input image objects, threshold, and spatial registration data are available for processing in the interf0_handler() function of the inference.py script. Essentially swap the functions that produce the "output_psma_ttb_label" and "output_fdg_ttb_label" SITK image objects with your method and include any additional scripts or resources in the Dockerfile configuration.

We're steadily updating the pages including a more detailed submission outline and general grand challenge instructions below:
https://deep-psma.grand-challenge.org/how-to-submit-your-algorithm/
https://grand-challenge.org/documentation/test-and-deploy-your-container/

One other note: there is a 30 minute per-case runtime limit using the Nvidia T4 AWS instances. That is the time for any data pre-processing as well as segmentation of both PSMA and FDG PET/CT images. If processing exceeds that time limit the job won't complete for the case. We have flagged 3 patients from the training dataset (25, 60, and 83) which includes two with large file sizes to evaluate worst-case processing times in the debug environment with access to the processing output information.

We will open the live leaderboard for the validation phase on previously unseen data in the week or two to follow and will open submissions for the final test phase towards the end of July with a close date in the second week of August.

More details to come and in the meantime please test your container scripts and feel free to reach out or post in the forum for any technical issues.

Best wishes and Good Luck to everyone!