Basics

Basic Concepts

• Pipeline

  • Everything runs in a pipeline.
  • Ordered components to perform work.
  • Standard components are: Collection Readers, Annotation Engines and Writers

• Collection Reader

  • Obtains notes from files, databases, etc.

• Annotation Engine

  • Detects wanted information.

• Writer

  • Stores information in files, databases, etc.

• JCas

  • Java Common Analysis System.
  • Data container passed through pipeline.

• Annotation

  • Simple data element: Token, Word, Number, etc.

• IdentifiedAnnotation

  • More advanced element, including clinical events.

• CUI

  • Concept Unique Identifier.
  • The most common CUIs can be found in the UMLS.
  • Most clinical events and anatomic sites have an associated CUI.
  • Identifier to normalize synonyms of the same concept.
    C2139817: "CT of left upper limb", "CT of left arm", "CT of left upper extremity"

• UMLS

  • Unified Medical Language System.
  • Brings together many separate biomedical vocabularies and standards.