MENELAS: an access system for medical records using natural language

Area of Science:

• Medical Informatics
• Natural Language Processing

Background:

• Accessing information in patient discharge summaries (PDSs) is challenging.
• Current methods for PDS analysis are often manual and time-consuming.

Purpose of the Study:

• To design and implement a pilot system for accessing medical reports via natural language.
• To improve information retrieval from French patient discharge summaries.

Main Methods:

• Developed the MENELAS (Medical ELectronic data Access System) indexing prototype.
• Encoded free-text PDSs into an internal representation and ICD-9-CM codes.
• Conducted a preliminary evaluation of the prototype's coverage and precision.

Main Results:

• An experimental version of the MENELAS indexing prototype for French was assembled.
• Preliminary evaluation indicated potential for reasonable coverage and precision in medical report encoding.
• The system demonstrated the feasibility of natural language access to PDS information.

Conclusions:

• The MENELAS prototype shows promise for enhancing access to patient discharge summary data.
• Further enhancement and extension into a pilot system for hospital testing are planned.
• This approach could significantly improve the efficiency of medical data analysis and retrieval.