mimic-cxr-ehr-split.csv

[jshang11]

Great work! Reading the code, but can't find information about how mimic-cxr-ehr-split.csv got generated. Is it possible to share the idea how this file get generated?

[jshang11]

also wondering whether age, gender and ethnicity were used in the training?

[v4nderstruck]

hi, currently working with this in my thesis so I can help (not affiliated with the author tho)

You can create the split using MedFuse create_split.py. (must do the preprocessing step)
Have a look here: https://github.com/nyuad-cai/MedFuse/blob/6f827589afd89562813cc5aa915762d054c29efc/ehr_utils/create_split.py

From my understanding of the code, the author does extract age, gender and ethnicity but the transformer
only considers images and ehr data. So that would be a no, it is not used during training.