[Information extraction methodology used in electronic medical records]

Area of Science:

• Natural Language Processing
• Medical Informatics
• Computational Linguistics

Context:

• Medical records contain vast amounts of unstructured text data.
• Accurate extraction of disease information is crucial for clinical research and patient care.
• Existing methods for information extraction from clinical notes require refinement.

Purpose:

• To develop and evaluate a dictionary and rule-based approach for named entity recognition in medical records.
• To extract disease information from unstructured clinical text.
• To build experience in complete information extraction from medical records.

Summary:

• This research applies information extraction techniques to medical records, focusing on disease entity recognition.
• The methodology employs dictionary lookups, rule-based systems, and pattern sentence matching.
• A 3-level finite state automaton supports shallow parsing for information extraction.

Impact:

• This work contributes to advancing automated disease information extraction from clinical notes.
• Improved information extraction can enhance clinical decision support systems.
• The findings facilitate the accumulation of experience for more comprehensive medical record analysis.