Skip to main content

Universal De-identification Platform (UDiP)

Healthcare & Life Sciences

Technical Overview

UDIP is provided as either a Java library, standalone process, or a J2EE application that runs under IBM WebSphere Application Server or other J2EE compliant application servers and can be accessed via the following interfaces:

  • EJBean API
  • Web Services over HTTP and JMS
  • JMS, IBM WebSphere MQ
  • File system
  • UDIP has adaptors to ETL (e.g., IBM-Ascential, Informatica) and to DB2

As a Java library, UDIP can be integrated in a client environment and invoked via Java or C API. UDIP is designed as a flexible framework to which new interfaces and de-identification capabilities can be added. The UDIP core, business flows and services expose JMX interface and can be configured and monitored via the provided Web-based administration console. In addition, UDIP management can be done by external JMX compliant products like IBM Tivoli Monitoring.

At the heart of UDIP lies powerful engines that provide de-identification and anonymization for XML, database tables, DICOM objects, spreadsheets, and CSV files. These engines bind required fields (elements or attributes in an XML-based document or columns of certain database tables) with the desired de-identification algorithm. The binding can be done via a Web-based administration console or configuration file. The actions can be a sophisticated algorithms or a simple (Perl, Python, JavaScript, etc.) scripts.

One advantage of UDIP is that decisions are made in the configuration files. This way, any changes in regulations, standards, de-identification rules, or client demands can be answered by simply changing the framework's configuration file.