Please reply to this thread to report any data issues encountered with the BraTS-MEN and BraTS-MEN-RT datasets. Lead MEN / MEN-RT organizers: @ecalabr @dlabella29

Created by Verena Chung vchung
Hi, thank you for investigating.
@Mmdanesh I reviewed your mask, and unfortunately yours is incorrect. Metrics seem to be producing the right answers.
Hi, I sent you the segmentation mask. Thanks
Hi @Mmdanesh, could you send the segmentation for the 00002-000 that your model generated? You can email it to me via I would like to investigate.
Hi. I hope you are having a great week. After sending my file to challenge, I found that some samples get zero across all metrics. For example sample 00002-000. I checked the segmentation result in CaPTk and everything is fine and according to the documentations. Since all metrics are zero , could there be a problem in evaluation process? Regards
@dlabella29 Thank you for your quick response and helpful answers!
Hi @Qing123, The dataset and the target lesion regions for segmentation in the preoperative BraTS-MEN 2025 are exactly the same as those in the 2023 MEN competition. The segmentation regions are also the same: ET (Enhancing Tumor), TC (ET + NETC ), and WT (ET + NETC + SNFH). For the BraTS-MEN-RT radiotherapy challenge, there is only a single label = 1 = Red = GTV. All the images and labels for the 2025 BraTS-MEN-RT challenge are the same as the 2024 BraTS-MEN-RT challenge. Best of luck. Thanks, Dominic
Hello, I would like to know whether the dataset and the target lesion regions for segmentation in this competition are exactly the same as those in the 2023 MEN competition? are the segmentation regions still three areas: ET (Enhancing Tumor), TC (ET + NETC ), and WT (ET + NETC + SNFH)？
Thanks for your response. When can we send files for validation? According to the timeline we should be able to send it now but the section says more information coming soon.
Hello, Thank you for the inquiry. You are not doing anything wrong. The data is as you discussed. For the Preoperative BraTS-MEN dataset, there are labels per the following: *1 = Red = non-enhancing tumor 2 = Green = surrounding FLAIR hyperintensity 3 = Blue = enhancing tumor *Note that the Red = 1 = non-enhancing tumor is not found in the majority of cases. https://www.nature.com/articles/s41597-024-03350-9/figures/2 For the BraTS-MEN-RT radiotherapy dataset, there is only a single label = 1 = Red = GTV Please let me know if any further questions. Thanks, Dominic
Hi. I have some questions regarding the labeling. Is the label 1 enhancing or non-enhancing tumor? Also, after processing the data, I found that label 1 is very very scarce. It is either non-existent or has the smallest counts of all. Am I doing something wrong or the data is like this? Thanks

BraTS-MEN and BraTS-MEN-RT Data Feedback page is loading…