Medical language processing applied to extract clinical information from Dutch medical documents

Area of Science:

• Computational linguistics
• Medical informatics

Background:

• Existing Dutch Medical Language Processor (DMLP) requires adaptation for broader medical NLP applications.
• The Linguistic String Project-Medical Language Processor (LSP-MLP) offers language-independent modules beneficial for specialized language processing.

Purpose of the Study:

• To extend the Dutch Medical Language Processor (DMLP) for compatibility with the language-independent modules of the LSP-MLP system.
• To demonstrate the utility of this integration through the extraction of clinical information from Dutch patient discharge summaries.

Main Methods:

• Morpho-syntactic analysis of Dutch language data.
• Integration of the Dutch Medical Language Processor (DMLP) with the Linguistic String Project-Medical Language Processor (LSP-MLP) system.
• Development of a system for extracting clinical information from patient discharge summaries.

Main Results:

• Successful extension of the Dutch Medical Language Processor (DMLP) to produce LSP-MLP compatible output.
• Demonstrated feasibility of extracting clinical information from Dutch patient discharge summaries.
• The integrated system leverages Dutch-specific linguistic features while benefiting from language-independent NLP advancements.

Conclusions:

• The enhanced Dutch Medical Language Processor (DMLP) facilitates integration with the LSP-MLP system.
• This approach enables effective clinical information extraction from Dutch medical texts.
• The system has potential applications in hospital education, research, and quality control.