-
Notifications
You must be signed in to change notification settings - Fork 6
Home
EasyCIE(GUI) is a rule-based clinical information extraction tool for non-NLP(natural language processing) expert users. It a GUI wrapper on top of EasyCIE, an UIMA-based command line version that allows executing on servers.
Although this GUI component is designed to be easily extensible for adapting different databases and wrapping other pipelines without hard coding, this wiki is mainly written as a user guide for EasyCIE(GUI).
- Import: Import documents from files into database. Import eHOST or Brat annotations as the reference standard. Import ontologies (https://github.com/Blulab-Utah/resource_ontologies)
- RunEasyCIE: Execute the EasyCIE pipeline on imported data
- Compare: Compare the pipeline output between different version or against the imported reference standard.
- Debug: Use a snippet of text to tackle the errors
- Export: Export the EasyCIE output into eHOST, Brat or UIMA XMIformat.
EasyCIE is under active development. The components are expanding rapidly. In the current released version, the built-in NLP pipeline includes the following components:
- RuSH: an efficient, reliable, and easy adaptable rule-based sentence segmentation solution.
- FastNER: a speed-optimized dictionary-based (name entity recognition)[https://en.wikipedia.org/wiki/Named-entity_recognition] solution.
- FastContext: an optimized Java implementation of the ConText algorithm (https://www.ncbi.nlm.nih.gov/pubmed/23920642).
- FeatureInferenceAnnotator: allows customizable conversion among UIMA types.
- DocInferenceAnnotator: allows to draw document level conclusion from mention level annotations using customizable rules.
All the components below are configurable through an excel file:
- Configure RuSH
- Configure FastNER
- Configure FastContext
- Configure FeatureInferenceAnnotator
- Configure DocInferenceAnnotator
The most recent compiled releases can be found at: https://drive.google.com/drive/folders/0B0hTn1B4kXcPM2hOZDZwY2NOTG8?usp=sharing