search

nainye / RadiomicsFill

[MICCAI 2024] RadiomicsFill-Mammo: Synthetic Mammogram Mass Manipulation with Radiomics Features
Apache License 2.0
3 stars 0 forks source link

[Inquiry] how to prepare a MET model #1

Open YeongHyeon opened 1 month ago

YeongHyeon commented 1 month ago

Hi there Congratulations on being accepted to MICCAI 2024!

I would like to try this provided source code. However, at line 955, it seems that loading the pre-trained MET model is needed. (Please confirm the attached code block below.) Can I ask how to prepare the ''MET_VinDr-Mammo_Mass_enc6_dec3_head1_embedding32_maxpmask90_best_val.pth''?

Thank you for sharing this great work!

https://github.com/nainye/RadiomicsFill/blob/3b4e0e236429ec3d52164c61fe4c86c286f01741/source/train_RadiomicsFill-MET32_VinDr-Mammo.py#L944-L955

nainye commented 1 month ago

Thank you for your interest in our RadiomicsFill project!

I have uploaded the MET (tabular encoder) model training code and usage instructions. Please check the README for details.

I am still organizing the repository, so more detailed information will be uploaded gradually. :)

YeongHyeon commented 1 month ago

Thank you for your response. I am following your updates!

One more! I haven't found any part of the code that utilizes the 'InBreast' dataset yet. Could you please help me to find where it is? And... also for 'jsonl' files...! Sorry for bothering you!

https://github.com/nainye/RadiomicsFill/blob/75635e8930e8ff67423a90100ba54359adc46bc8/scripts/train_RadiomicsFill-MET32_VinDr-Mammo.sh#L3-L5

YeongHyeon commented 1 month ago

Is there anyone here?