You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
In the paper about UNETR, it says "Additional data from the same cohort was used for the Free Competition increasing the number of training cases to 80 volumes". I downloaded the BTCV dataset, there are 30 training samples and 20 testing samples. Could you please tell me where the additional data can be downloaded to increase the treaning cases to 80 volumes?
The text was updated successfully, but these errors were encountered:
I got the same question. Using the current 30 training samples with the provided pretrained SSL weights, the best Dice score is 0.796, which is far from 81.86, as reported in the BTCV repository.
In the paper about UNETR, it says "Additional data from the same cohort was used for the Free Competition increasing the number of training cases to 80 volumes". I downloaded the BTCV dataset, there are 30 training samples and 20 testing samples. Could you please tell me where the additional data can be downloaded to increase the treaning cases to 80 volumes?
The text was updated successfully, but these errors were encountered: