Keyword extraction and structuralization of medical reports

Area of Science:

• Medical Informatics
• Natural Language Processing
• Computational Pathology

Background:

• Medical examination reports often contain unstructured text data.
• Advances in medical technology necessitate efficient data organization for patient care.
• Structured data facilitates better understanding and analysis of patient health status.

Purpose of the Study:

• To develop a system for automatically structuring unstructured medical examination reports.
• To enhance the efficiency of doctors in understanding patient examination statuses.
• To enable further association analysis for identifying disease-affecting factors.

Main Methods:

• Applied Part-of-Speech (POS) tagging for keyword phrase extraction from renal disease pathology reports.
• Established a medical dictionary for examination items to retrieve terms.
• Utilized topical probability modeling to discover candidate keyword phrases.
• Implemented a system to generate structured reports based on the medical dictionary.

Main Results:

• Demonstrated effective construction of structured examination reports.
• Successfully extracted keywords for popular examination items.
• Validated the system's capability in organizing unstructured medical text.

Conclusions:

• The proposed methods effectively structure medical examination reports.
• Accurate extraction of keywords from popular examination items was achieved.
• The techniques support automatic processing and analysis of medical text reports.