Extracting prescriptions from free text field pipeline
Notebooks
There are three notebook files:
1. MedExtractionPipeline
This notebook file is concerned with the extraction of medication
2. ContextPipeline
This notebook file extracts the context according to the rules & targets defined in the corpus directory. The EMR data is merged with Meteor and Medicator information and finally processed.
3. AnalyzeMedicationTrajectory
This notebook was used for performing statistical analysis
Functions
EMR_functions.py
All of the notebooks utilize functions from this file. Every function is documented as well.
Other files
acronyms.py
File that is utilized to convert the acronyms to full medication names.
Visualize.py
Some of the plots are made with this class
Graphs directory
The medication trajectory graphs (HTML) are featured in this map
Bokeh plots directory
The Damerau-Levenshtein distribution is featured in this map
modules
The majority of necessary modules are featured here
ra_diagnosis
Features the notebook that was used for the diagnosis extraction