E-Medicine - Information Retrieval in Medicine: Overview and Applications

Information Retrieval (IR) is a branch of computer science that is concerned with the processing of collections of documents containing "free text." Examples of such collections are a set of hospital discharge summaries, radiology reports, or surgery notes, or (in a non-medical context) the full text of the complete works of Shakespeare or the Bible. In contrast to a spreadsheet or a database table, which is divided into rows and columns, such documents have no obvious structure: any structure that is imposed is purely artificial, highly variable, and is of little use in retrieving information. Thus, while the Bible is divided into books, chapters and, occasionally, verses, a reader is more interested in locating a section by the phrases it contains (e.g., the name of particular Biblical characters or a quote). Similarly, the "structure" of a Chest X-Ray report, in terms of the headings under which the radiograph is described, is different from the structure of a Barium swallow or Intravenous Pyelogram report.