search

uni-medical / SAM-Med3D

SAM-Med3D: An Efficient General-purpose Promptable Segmentation Model for 3D Volumetric Medical Image
Apache License 2.0
484 stars 64 forks source link

Request for validation set list #47

Closed cikitime closed 6 months ago

cikitime commented 6 months ago

I hope you're doing well. I recently read your paper and found it to be very informative and valuable.

I noticed that the validation set list was not provided in the paper or the supplementary materials. Would it be possible for you to share the list of samples used in the validation set? Knowing this information helps me conduct a comparative experiment.

Thank you for your time.

blueyo0 commented 6 months ago

As for the result we report in the readme, the case list is here: case_list.txt.

cikitime commented 6 months ago

thank you！

joseph-bae commented 3 months ago

Hello,

I have a data leakage concern. The case list provided contains the entirety of the BraTS 2021 patient list, despite BraTS 2021 also being listed in the train set. Is this an error? Do you have the list of BraTS patients used for training and testing independently?

blueyo0 commented 3 months ago

Hello,

I have a data leakage concern. The case list provided contains the entirety of the BraTS 2021 patient list, despite BraTS 2021 also being listed in the train set. Is this an error? Do you have the list of BraTS patients used for training and testing independently?

Hi, @joseph-bae Mayde here's some misunderstanding. We used the train/val set of BraTS 2021 for our training and validation, just like many other works that trained and validated on BraTS. I think there is no data leakage issue.