SHARPS DS2 -- Decision Support for Data Segmentation -- is a technical architecture and software prototype to demonstrate how OpenCDS, an open source Clinical Decision Support (CDS) framework, can be used to identify and redact selected conditions from clinical summary documents in an HIE environment. The DS2 software also includes a suite of related tools for:
- creating, importing, and editing Continuity of Care (CCD) documents;
- testing redacted CCDs; converting CCDs to Virtual Medical Record (vMR) format;
- training machine learning classifiers and calling them from OpenCDS;
- checking for drug-drug interactions against non-redacted CCDs; and
- visualizing the effectiveness and the impact of probabilistic redaction through a web-based "inference analyzer."
From here, you can:
- See About to learn more about DS2.
- See Projects to get an overview of the component projects; and Downloads and Installation for information on obtaining and deploying the prototype software.
- See Publications to access white papers describing DS2 in more detail.
- See Resources for demos, additional documentation, and information about software platforms and data sets used by DS2.
- See License and Disclaimer to view DS2's open source license.
- Or, go directly to the source code for DS2 project components by visiting the public repository at https://bitbucket.org/sharps-ds2/.
Our target community includes health informatics researchers and anyone interested in exploring the use of CDS to assist with the segmentation of clinical records by condition.
SHARPS DS2 is part of SHARPS, is a multi-institutional and multidisciplinary research project, supported by the Office of the National Coordinator for Health Information Technology, aimed at reducing security and privacy barriers to the effective use of health information technology.
Try out the Predicate-Reducer and Inference Analyzer on the Demonstrations page.
This work was supported by HHS 90TR0003-01 (SHARPS) and the Illinois Office of Health Information Technology (OHIT). The views expressed are those of the authors only.