i2b2 Hive

The i2b2 (Informatics for Integrating Biology and the Bedside) Hive is a set of microservices aimed at genomics researchers. The microservices, known as cells, communicate via stuctured messages sent over the Internet. They may be administered using a Web interface, and used with either a simple Web client or a rich desktop client to construct a research workflow.

The following are core server-side cells:

  • Project management (user, group and role management; cell co-ordination)
  • Ontology management
  • Identity management (handling of sensitive patient data)
  • Workflow framework
  • File repository (for large files such as radiological images and genetic sequences)
  • Data repository ('clinical research chart', for genotypic and phenotypic data)

The following are some of the optional server side cells available:

  • Natural language processing
  • Pulmonary function test processing
  • Patient counts (Data repository plug-in)

The desktop client, i2b2 Workbench, is based on Eclipse and may be extended with the following plug-ins:

  • High performance computing
  • Text analyzer (for decrypting notes in the data repository)
  • Export data (from dimension and mapping tables)
  • Import data (to dimension and mapping tables)
  • Correlation analysis
  • Table view
  • Annotator (to assist with natural language processing)

Provider

The i2b2 Center is an NIH-funded National Center for Biomedical Computing, established in 2004.

Licensing and cost

The software is free to download and use. The source code is released under the i2b2 Open Source Licence.

Development activity

Version 1.7 was released in December 2013, followed every few months by a bug fix release or minor feature enhancement.

Platform and interoperability

The server-side components may be run on either Windows or Linux servers, while the rich desktop client may be run on Windows or Mac OSX. The Web client explicitly supports Internet Explorer 6-9, Firefox 2-3 and Safari 3; other browsers may also work.

Each cell defines its own API, but they all respond to RESTful HTTP requests and some support alternative protocols such as SOAP. They also share a common XML format for exchanging messages.

Documentation and user support

Documentation is provided for each individual component, and for the server-side Hive as a whole. In addition to the PDF manuals on the i2b2 website there is also a community wiki, and the Academic User Group meets annually.

Usability

Two interfaces are offered. The Web client has limited functionality, but as it is hosted from the server it can be used widely with no installation required on the part of the user. The desktop client is much more powerful, but needs to be set up on each user's machine.

Expertise required

Although the process is well documented, installation of the server-side components is non-trivial so some system administrator expertise would be a distinct advantage. Installation of the Workbench is straightforward. Use of the client software should be straightforward for anyone used to the query–results paradigm of systems like SPSS.

Standards compliance

The i2b2 Hive is based on common Web technologies but uses its own XML format for communicating messages. Data may be imported and exported in CSV format.

Influence and takeup

The i2b2 website lists around 90 installations worldwide, including universities, hospitals and commercial organisations.

Last reviewed: 
17 November, 2014