search

liaohaofu / adn

ADN: Artifact Disentanglement Network for Unsupervised Metal Artifact Reduction
Other
164 stars 39 forks source link

code #14

Open strive160 opened 4 years ago

strive160 commented 4 years ago

When you're training U-Net with a real data set, how do you generate a mask。 thank you

strive160 commented 4 years ago

Can you tell me how you preprocessed deeplesion data with MATLAB? I didn't understand why a patient had several images but only one or two, some three or four, and what the selection criteria were.Why are 90 MAT files generated, or several metals added? Are these 90 MAT files 90 images with artifacts?The whole process of data generation is not very clear, can you describe it

strive160 commented 4 years ago

Recently, I would like to use your pre-processed data for other network framework experiments. However, due to the lack of understanding of your data pre-processing, this process is not explained in the article. Please kindly explain