Interpret medication prescriptions into a structured format using dependency parses or third-party drug NER pipeline
StructuRX interprets medication prescriptions into a structured format using dependency parses  or third-party drug NER pipeline .
The dependency-parse version uses a hand-crafted set of rules as described in  based on the output of the ClearParser dependency parser, while the alternative method described in  uses the output of the drug NER pipeline from the cTAKES clinical NER system.
Recommended installation method:
$ pip install StructuRX
But of course you can instead do an old-fashioned:
$ python setup.py install
from a VCS checkout or tarball/zip download. In the latter case you will need to ensure that ClearWrap is installed, and in either case you will need a copy of the NICTA fork of ClearParser available to do the dependency parsing. We highly recommend installing into a virtualenv. If you wish to test against cTAKES, that will naturally need to be installed as well.
After installing, run the provided script structurx-run.py and follow the instructions for configuration.
-  A. MacKinlay and K. Verspoor. 2012. Extracting structured information from free-text medication prescriptions using dependencies. DTMBIO 2012.
-  A. MacKinlay and K. Verspoor. 2013. A Comparison of Strategies for Extracting Structured Information from Free-Text Medication Prescriptions. LOUHI 2013.