•  
  •  
 

Turkish Journal of Electrical Engineering and Computer Sciences

Author ORCID Identifier

HAFIDA TIAIBA: 0000-0003-3878-2071

LYAZID SABRI: 0000-0002-9266-567X

OKBA KAZAR: 0000-0003-0522-4954

Abstract

The automatic recognition of medical concepts and temporal expressions in narrative clinical text enhances the utility of electronic health records (EHRs) and supports clinical decision-making and research. However, challenges arise due to the complexity of medical language, ambiguity of terms, and variability in expression. To address these issues, the use of medical ontologies significantly improves data management in healthcare. A novel approach integrates various medical ontologies covering drugs, symptoms, diseases, anatomy, disease drivers, and food, and with convolutional neural networks (CNNs) -including Standard, Transposed, and Separable convolution models (CONSEPTR)- to extract both medical events (e.g., clinical departments, treatments, problems) and temporal expressions (e.g., time, date, duration) from narrative medical texts. This method extends beyond conventional extraction tasks by identifying critical entities such as drugs, diseases, symptoms, anatomy, food materials, and disease drivers. This extraction allows for a deeper analysis of health outcomes and a better understanding of the interactions between anatomical details, dietary factors, and disease mechanisms. The result is improved precision and relevance in health-related data, supporting informed decision-making in clinical and research environments. CNNs, combined with BERT embeddings, offer a robust methodology for identifying novel medical terminology, as they excel at discerning complex data features and capturing semantic relationships. This ability is beneficial for the accurate classification of medical terms. Using the i2b2 datasets (2010, 2012, 2018) and integrating ontologies through the Jena API, a framework for building Semantic Web applications, and SPARQL, a query language for RDF data, enhances the extraction of medical entities. This approach ensures more accurate classification and provides the identification of entities types not explicitly defined in the datasets. For example, the model can recognize "methotrexate" as a drug and "dietary manganese" as a food material, providing a comprehensive and precise understanding of medical narratives.

DOI

10.55730/1300-0632.4114

Keywords

deep learning, medical events, narrative medical text, Ontology, semantic, temporal expression

First Page

65

Last Page

85

Creative Commons License

Creative Commons Attribution 4.0 International License
This work is licensed under a Creative Commons Attribution 4.0 International License.

Share

COinS