Automatic Population of Synoptic Reports from Narrative Pathology Reports

This page provides information related to the completed project.

Page last updated: 15 December 2015

2013

Automatic Population of Synoptic Reports from Narrative Pathology Reports

University of Sydney

The preceding two stages of this project have demonstrated the accuracy to which elements of the clinical interest in narrative reports can be identified by human annotators and how specified content can be automatically extracted.

The aim of the third stage of the project was to develop methods for extracting pertinent information from narrative pathology reports by natural language processing to automatically populate synoptic reports for three types of cancer: melanoma, skin and lymphoma. The information is displayed in a synoptic reporting format to make it easier for the requestors to interpret pathology reports and give clearer information to consumers.


Automatic Population of Synoptic Reports from Narrative Pathology Reports (PDF 2456 KB)
Automatic Population of Synoptic Reports from Narrative Pathology Reports (Word 400 KB)