De-identification of Electronic Health Records
This repository contains the code accompanying my AMIA submission: Benchmarking Modern Named Entity Recognition Techniques for Free-text Health Record Deidentification
Please refer to either testing.ipynb or ccv_test.py for a walkthrough of how to run the code. Note that raw data must be in the i2b2 format. Ensure that isLoading is set to False the first time i2b2 data is processed.