Closed anothersin closed 1 year ago
I also want to know the preprocessing way for lit and kits dataset.
I'm very interested in your paper, it's a great job. Can you provide the code for pseudo-label generation?
Your work is awesome and I want to reproduce results of the experiment, but the code for processing la dataset is from UAMT and the preprocessing way for lit and kits dataset are not available, too. Can you provide the preprocessing pipeline and explain in a little more detail how to generate pseudo-label?
Thanks, we've uploaded the code for kits & lits preprocessing and pseudo-label generation. As for the pseudo label generation, registration.py is an example of generating pseudo label for LA dataset from transverse plane. To perform label propagation for other two planes, you need to slice the volume from other directions, i.e. replacing image(label)[:,:,i] by image(label)[i,:,:] or image(label)[:,i,:].
请问lits和kits数据怎么处理为与LA数据类型的格式
Thank you very much for your contribution to the community. We especially want to follow your work, but we didn't find the preprocessing code for the lits and kitss dataset. To avoid unfair competition, can you provide the relevant preprocessing pipeline?
LiTS Dataset and KiTS19 Dataset, Have you successfully processed these two data sets?
Thank you very much for your contribution to the community. We especially want to follow your work, but we didn't find the preprocessing code for the lits and kitss dataset. To avoid unfair competition, can you provide the relevant preprocessing pipeline?
LiTS Dataset and KiTS19 Dataset, Have you successfully processed these two data sets?
is there any problem with preprocessing_kits/lits.py?
is there any problem with preprocessing_kits/lits.py?
The data format of KiTS I got is .nii, how to make it be processed into .h5 format which is the same format as the LA dataset. I tried converting but it still doesn't work correctly, can you help me?
you don't have to necessarily convert to h5, the preprocessing code (code/dataloader/preprocessing_kits.py) takes nii.gz as input
you don't have to necessarily convert to h5, the preprocessing code (code/dataloader/preprocessing_kits.py) takes nii.gz as input
I've tried it, but it didn't work
what's the error message
what's the error message
what's the error message
dont run this function, you only need to run preprocessing_kits.py, which takes nii.gz as input and generates processed h5 files as results, and those h5 files can be directly used for training.
no such .py function-- preprocessing_kits.py
how come
how come
The data I downloaded from the official website KITS19 does not exist image.nii
how come
The data I downloaded from the official website KITS19 does not exist image.nii
Is there any way to get
sorry, the download link we used is invalid now, check https://aistudio.baidu.com/aistudio/datasetdetail/24582 or some else resources on the internet
sorry, the download link we used is invalid now, check https://aistudio.baidu.com/aistudio/datasetdetail/24582 or some else resources on the internet
Is it downloading the one with the red arrow?
maybe all of these four
maybe all of these four
Which one did you use before?
maybe all of these four
Appreciate your answer, thank you. I wish you a happy life.
The error occurred when I used the KiTS dataset to train, I tried to solve it but failed, I need your help, thank you
did you generate pseudo label with registration.py
did you generate pseudo label with registration.py
No
did you generate pseudo label with registration.py
You said last time that you don’t need to run this function, so I didn’t run it
sorry for the ambiguity, i meant dont run la_heart_processing.py but preprocessing_kits.py. The general pipeline for these three datasets is 1) preprocessing, 2) generate two direction pseudo labels with registration.py (note that this code is only for transverse plane, you may need to rewrite few lines for coronal plane), 3) train
What is the specific code? Can you send me your complete project code for reference?
sorry for the ambiguity, i meant dont run la_heart_processing.py but preprocessing_kits.py. The general pipeline for these three datasets is 1) preprocessing, 2) generate two direction pseudo labels with registration.py (note that this code is only for transverse plane, you may need to rewrite few lines for coronal plane), 3) train
What is the specific code? Can you send me your complete project code for reference?
Thank you very much for your contribution to the community. We especially want to follow your work, but we didn't find the preprocessing code for the lits and kitss dataset. To avoid unfair competition, can you provide the relevant preprocessing pipeline?