Access to Banerjee et al. 4,351 annotated CTPA reports for NLP labeler fine-tuning

[vanlam2308]

Hi team,

I'd like to reproduce the NLP auto-generator for radiology report pipeline. From the paper, the fine-tuning step was using OMOP-STARR dataset from Banerjee et al. Could you advise on how to get to these data? I wonder if the de-identified OMOP-STARR dataset is available to download or access with training completion.

Many thanks,

[jason-fries]

@vanlam2308

INSPECT as dataset is distributed in 2 parts via redivis:

INSPECT EHR (via the Shah Lab)
INSPECT IMAGE (via the AIMI Center) https://stanford.redivis.com/datasets/2n96-d71hggrbf

All the labels, mappings, and notes are hosted under the AIMI Center dataset