Automated Generation of Synoptic Reports from Narrative Pathology Reports in University Malaya Medical Centre Using...

Area of Science:

• Medical Informatics
• Computational Pathology
• Natural Language Processing

Background:

• Pathology reports are crucial for cancer registries, but unstructured narrative formats hinder data extraction.
• University Malaya Medical Centre (UMMC) faces challenges in data analysis and clinical audits due to free-text pathology reports.
• Efficient information extraction is vital for clinical audits and research.

Purpose of the Study:

• To develop an automated natural language processing (NLP) algorithm for summarizing narrative breast pathology reports.
• To convert unstructured UMMC breast pathology reports into a structured synoptic format.
• To create a checklist-style report template for easier pathology report generation.

Main Methods:

• A rule-based NLP algorithm was developed using the R programming language.
• The algorithm utilized 593 pathology specimens from 174 patients at UMMC.
• Pathologist-defined keywords established semantic rules for data element extraction.

Main Results:

• The NLP algorithm achieved high performance with a micro-F1 score of 99.50% and a macro-F1 score of 98.97%.
• The system successfully processed 178 specimens, extracting 25 data elements.
• The developed algorithm demonstrated strong accuracy in converting narrative to synoptic reports.

Conclusions:

• The automated NLP algorithm effectively summarizes breast pathology reports into a structured format.
• This structured data facilitates data mining and generates valuable insights for research.
• The improved data accessibility enhances communication between pathologists and clinicians.