Question about the data processing phase

[Roypic]

Dear authors,

I'm tring to reimplement the whole process. However when deal with MIMIC-CXR datasets with data_process.sh script, I found that the caption.csv file is almost blanket. Should I rearrange the dataset( dicom to jpg, establish the image-txt croessponding csv) or there are already prepared ones? Thanks.

[rookiemmmm]

Hello, where did data_process.sh script come from? I can't find it.

[xiaoman-zhang]

That's in the google drive/ Baidu Cloud. ./A1_DATA/MIMIC-CXR/data_preprocess

[rookiemmmm]

Okay, thank you very much！

[rookiemmmm]

Hi, I can't find radgraph_entity_graphes.json in ../A1_DATA/MIMIC-CXR/json_preprocess/

[xiaoman-zhang]

Hi, The file was too large and I ran out of space on Google Drive, so I uploaded it to Baidu Netdisk. https://pan.baidu.com/s/1rozuHHsducL2S0NX4Hs4Tg?pwd=7ncb code: 7ncb

[rookiemmmm]

Okay，thank you.

[rookiemmmm]

I'm sorry, I can't find MRSTY.csv in ../A1_DATA/UMLS/.

[Roypic]

Me too, I also can't find MRSTY.csv anywhere.

[xiaoman-zhang]

upload to https://github.com/xiaoman-zhang/KAD/blob/main/A1_DATA/UMLS/MRSTY.csv

[Roypic]

Much appreciate.