A Framework for Extracting, and Validating Named-Entities to Integrate Openehr Using the Example of Free Text Molecular Genetic Findings

Le résumé fourni par la source

Processing and extracting information from unstructured texts written by physicians in Hospitals is still an open problem. There is no efficient solution that ensures the reliability of the extracted information without any human intervention. Many factors, like the low availability of documents in the training phase, patient-sensitive information, and the complexity of the written texts can impact the results. Through our scientific journey, to integrate unstructured texts in openEHR, we have developed tools that together provide a complete process to efficiently extract, validate and integrate data in openEHR. As a use case, we demonstrate the free written texts in molecular genetic findings to present our results. The validation of the pipeline resulted in an F-score of 0.98.

Ce résumé expose les affirmations des auteurs. BNTIC ne l’interprète pas comme une validation indépendante des résultats.