Dear BraTS Organization Team Regarding Task 4 of BraTS 2026, I have a few questions to consult: I previously uploaded one Docker image prior to the official one-week extension announcement. Do I still have available submission quota for the test-phase Docker upload? If my quota has been exhausted, would it be possible to manually reopen a submission slot for my team? Moreover, is there any possibility to extend the submission window for the validation leaderboard? The original deadline for validation leaderboard submissions is July 23rd. Looking forward to your reply. Sincerely, IntelliSense team

Created by Chengwei Li lcw
@lcw You're welcome! And thank you :) Best Paula
@paulabuzduga Dear Paula and the BraTS 2026 Organization Team, We are sincerely grateful for your kindness and support in reopening the submission portal and granting our team a second submission opportunity. This considerate arrangement gives us a valuable chance to further optimize and validate our model, and we highly appreciate your professional and patient guidance throughout the competition. Our team will carefully polish our Docker image and complete the final submission strictly in accordance with the competition rules. We will actively cooperate with the follow-up work of the competition. We sincerely wish the BraTS 2026 challenge a complete and successful conclusion. May all your subsequent work go smoothly and smoothly! Best regards, IntelliSense Team
@lcw Dear Chengwei Li, the submission portal has been reopened for your team! IntelliSense should be able to submit a second docker image now :) Let me know if you run into any issues! Best Paula
@paulabuzduga Dear Paula, Hope this email finds you well. We are terribly sorry to disturb you again amid your heavy workload for BraTS 2026 challenge. We are writing just to gently follow up on our previous request for an extra Docker submission slot for our team IntelliSense, which we discussed with you on July 24. We fully understand that you need to hold internal discussions with @vchung to make a decision, and we absolutely respect all the official submission regulations designed to ease the computing burden of your server. There is no rush on our end at all; we only send this message to confirm whether there has been any preliminary update from your internal review. As we mentioned before, we have finalized our optimized model and thoroughly tested the complete Docker package locally using the inpainting evaluation tool you recommended. If we are fortunately granted the extra submission opportunity, we will upload the model immediately and promise no further revision requests afterward. Again, we apologize sincerely for any extra trouble this follow-up might bring to you and the BraTS organizing team. Please take all the time you need for your internal discussion, and we will patiently wait for your reply whenever you are available. Should you need any supplementary information from our team for your discussion, please feel free to let us know. Best regards, IntelliSense Team
@lcw Dear Team IntelliSense, thank you very much for your understanding! We appreciate it. As promised, I will update you immediately if there are any news. In the meantime, good luck for the ongoing development of your model! Best Paula
@paulabuzduga Dear BraTS Organization Team, Thank you so much for your detailed and patient reply, as well as your kind consideration of our team’s situation. We sincerely apologize for the extra workload and inconvenience brought to your team due to our careless early Docker submission before the deadline extension. We fully understand the platform’s rule of limiting one Docker submission per team to avoid excessive computation pressure on your servers. We would be extremely grateful if you could grant us this one additional submission slot to upload our refined, final model. We guarantee that we will thoroughly verify the model and Docker package before uploading, and this will be our only revised submission with no further updates requested afterward. We also appreciate your guidance on the local evaluation tool inpainting; we will follow the instructions to complete offline metric testing on our side. We will wait patiently for your internal discussion result regarding the extra submission chance. Looking forward to your favorable reply at your earliest convenience. Sincerely, IntelliSense team
@lcw @vchung Dear Chengwei Li, thank you for reaching out to us with your questions! Indeed, we can see your docker submission for team IntelliSense. **In general, each team is supposed to only upload one final docker image, while multiple submission to the leaderboard are possible in order to locally refine the model and validate it via the leaderboard.** The reason for a single submission is the following: Oftentimes the algorithms run for a very long time and we do not have the capacity to re-run models if teams continue re-submitting new versions of their algorithm. It gets very messy very quickly. That is why each team is currently limited to submitting only a single final docker. We do, however, sometimes make exceptions. We will discuss this internally (@vchung), but I am quite hopeful that we can open the portal for your team again, so you can submit an improved model. **Please note, however, that we cannot grant you a third, fourth,.... attempt. So in the meantime, please make sure that you prepare one final model that you would like to submit. **I will get back to you as soon as there are any updates. Regarding the validation leaderboard: You are right, it has been closed. We have been discussing reopening it, but I cannot at the moment promise anything. **However, there is the option to locally run the same evaluation as is done for the leaderboard. We are using the package 'inpainting', which you can install via pip. It is a PyPI package and can be found here: https://pypi.org/project/inpainting/** We are using the generate_metrics() function to calculate the three metrics you see on the leaderboard, but the package can calculate some additional ones, not displayed on the leaderboard. Calling the function is really straightforward. There is documentation on PyPI and Github. **Please note that it takes four inputs.** Notably, it requires the original voided T1 image as normalization. If this fourth argument is not passed, it can yield invalid results! I will let you know if there are any news regarding the submission portal or in case we decide to reopen the leaderboard! Feel free to reach out again if there are any remaining questions on your side! Best of luck Paula

Test-phase Docker Submission Quota & Validation Leaderboard Deadline for Task 4 page is loading…