Relationship Extraction from Electronic Health Records & Generating Knowledge Graph
Authors:
- 9 February 2024
- conference paper
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 290-295
Abstract
The research delves into the process of extracting relationships from biomedical literature and building a knowledge graph using those relationships. Information retrieval, knowledge discovery, and data analysis are just a few of the uses for named entity extraction in biomedical research, which includes important entities such as Adverse Drug Effects (ADE), doses, drug names, and more. Using BERT's word contextual comprehension, the report investigates the use of the BioBERT model for named entity recognition and relationship extraction. When calibrated on Electronic Health Records (EHR) data, BioBERT— a biological text-specific model—helps with clinical decision-making and research by recognizing links between elements such as diseases, symptoms, and therapies. The use of Neo4j for knowledge graph generation, which permits a thorough representation of biomedical concepts and their relationships, is also highlighted in the paper. Information retrieval and analysis are made easier by this organized knowledge. This strategy has the potential to revolutionize biomedical research and healthcare by providing fresh perspectives and better results, despite ongoing challenges. More developments in this field are anticipated as a result of ongoing research and development.Keywords
This publication has 9 references indexed in Scilit:
- Towards electronic health record-based medical knowledge graph construction, completion, and applications: A literature studyJournal of Biomedical Informatics, 2023
- Modeling electronic health record data using an end-to-end knowledge-graph-informed topic modelScientific Reports, 2022
- Comparison of biomedical relationship extraction methods and models for knowledge graph creationJournal of Web Semantics, 2022
- EHR-Oriented Knowledge Graph System: Toward Efficient Utilization of Non-Used Information Buried in Routine Clinical PracticeIEEE Journal of Biomedical and Health Informatics, 2021
- Experience Simple Transformer library in solving Mojaz Multi-Topic Labelling Task2021
- BERT based clinical knowledge extraction for biomedical knowledge graph construction and analysisComputer Methods and Programs in Biomedicine Update, 2021
- Relation Extraction for Chinese Clinical Records Using Multi-View Graph LearningIEEE Access, 2020
- BioBERT: a pre-trained biomedical language representation model for biomedical text miningBioinformatics, 2019
- Adverse drug events and medication relation extraction in electronic health records with ensemble deep learning methodsJournal of the American Medical Informatics Association, 2019