search

QtacierP / PRIOR

Official repository for the paper "Prototype Representation Joint Learning from Medical Images and Reports, ICCV 2023".
Apache License 2.0
65 stars 6 forks source link

lack the json file #8

Open Lalalalala-l opened 8 months ago

Lalalalala-l commented 8 months ago

hello,mentor.can you provide the annotation.json,I meet some mistake.Please!

QtacierP commented 8 months ago

Which json file do you mean？If you say the json File in MimicCXR dataset, maybe I cannot provide the exact file due to the license limitation. You can apply and download the file from the [homepage]（https://physionet.org/content/mimic-cxr/）, and run the preprocess scripts as shown in README.

Lalalalala-l commented 8 months ago

Which json file do you mean？If you say the json File in MimicCXR dataset, maybe I cannot provide the exact file due to the license limitation. You can apply and download the file from the [homepage]（https://physionet.org/content/mimic-cxr/）, and run the preprocess scripts as shown in README.

Okay, thank you！

QtacierP commented 8 months ago

https://physionet.org/content/mimic-cxr-jpg/2.1.0/ You can download images here

Lalalalala-l commented 8 months ago

https://physionet.org/content/mimic-cxr-jpg/2.1.0/ You can download images here

Can you tell me the format of the data in the Json file? I have some problems here. I plan to modify them directly with the files that have been processed before. Thank you very much.

QtacierP commented 8 months ago

The json file is obtained by 

cd codes/
python prior/data/pretrain/mimiccxr.py build --root_image_path <image_dataset_path> --root_report_path <report_dataset_path> --save_path <dataset_json_path> --meta_csv_path <meta_csv_path>

as presented in the README :)