Enterprise Knowledge Graphs for Biomedical Data Integration: A Scalable Architecture for Semantic Interoperability
DOI:
https://doi.org/10.15662/IJEETR.2024.0604015Keywords:
Enterprise Knowledge Graphs, Biomedical Data Integration, Semantic Interoperability, Ontology Engineering, RDF and SPARQL, Explainable Artificial Intelligence (XAI)Abstract
The notion of semantic interoperability and integrated analytics is so challenging because of the scandals that surround which biomedical entities can accessibly create variable information with electronic health records, genomics and medical imaging, lab, wearable and with clinical research environments. The suggested article presents a scalable design of Enterprise Knowledge Graph (EKG-BDI) which applies Biomedical Data Integration to unite different biomedical data with the help of semantic model created on the basis of ontological concepts and standard knowledge representation. Six joint layers are proposed to be applied in the proposed framework and they are: Data Acquisition, Semantic Transformation, Ontology Management, Knowledge Graph Construction, AI-Driven Semantic analytics and Interoperability and Decision Support. The general biomedical data of different structured and unstructured forms, the ontologies of domains are transformed into the triples of RDF and the standardized vocabularies are connected to be represented in the same way, discover relations and semantic inferences. Tasks, such as improving the quality of information, removing semantic inconsistencies and an entity discovery in distributed health care repositories, become easier by inferencing, asking questions using a language called SPARQL, and automated entity resolution. This architecture deploys scalable graph databases in the form of micro-services to be capable of both supporting large volumes of biomedical data as well as compliance with the security and privacy policies and regulations. Moreover, knowledge graph is applicable in explainable AI modules that support clinics in decision-making, biomedical research, predicting illnesses,precision medicine. The suggested architecture gives a robust base of the non-volatile healthcare systems by enhancing the efficacy of data unification, semantic consistency, reuse of knowledge and intelligent data scrutiny throughout the enterprise biomedicine levels
References
[1] D. N. Nicholson and C. S. Greene, “Constructing knowledge graphs and their biomedical applications,” Computational and Structural Biotechnology Journal, vol. 18, pp. 1414–1428, 2020, doi: 10.1016/j.csbj.2020.05.017.
[2] J. Schrodt, A. Dudchenko, P. Knaup-Gregori, and M. Ganzinger, “Graph-representation of patient data: A systematic literature review,” Journal of Medical Systems, vol. 44, no. 5, Art. no. 86, 2020, doi: 10.1007/s10916-020-1538-4.
[3] C. A. Nelson, R. Bove, A. J. Butte, and S. E. Baranzini, “Embedding electronic health records onto a knowledge network recognizes prodromal features of multiple sclerosis and predicts diagnosis,” Journal of the American Medical Informatics Association, vol. 29, no. 3, pp. 424–434, 2022, doi: 10.1093/jamia/ocab270.
[4] D. S. Himmelstein et al., “Systematic integration of biomedical knowledge prioritizes drugs for repurposing,” eLife, vol. 6, Art. no. e26726, 2017, doi: 10.7554/eLife.26726.
[5] S. A. C. Bukhari et al., “LinkedImm: A linked data graph database for integrating immunological data,” BMC Bioinformatics, vol. 22, Art. no. 105, 2021, doi: 10.1186/s12859-021-04031-9.
[6] C. A. Nelson, A. J. Butte, and S. E. Baranzini, “Integrating biomedical research and electronic health records to create knowledge-based biologically meaningful machine-readable embeddings,” Nature Communications, vol. 10, Art. no. 3045, 2019, doi: 10.1038/s41467-019-11069-0.
[7] W. S. Campbell, J. Pedersen, J. C. McClay, et al., “An alternative database approach for management of SNOMED CT and improved patient data queries,” Journal of Biomedical Informatics, vol. 57, pp. 350–357, 2015, doi: 10.1016/j.jbi.2015.08.016.
[8] Y. Mao and K. W. Fung, “Use of word and graph embedding to measure semantic relatedness between Unified Medical Language System concepts,” Journal of the American Medical Informatics Association, vol. 27, no. 10, pp. 1538–1546, 2020, doi: 10.1093/jamia/ocaa136.
[9] J. A. M. Stothers and A. Nguyen, “Can Neo4j replace PostgreSQL in healthcare?” AMIA Joint Summits on Translational Science Proceedings, pp. 646–653, 2020.
[10] D. M. Bean et al., “Knowledge graph prediction of unknown adverse drug reactions and validation in electronic health records,” Scientific Reports, vol. 7, Art. no. 16416, 2017, doi: 10.1038/s41598-017-16674-x.
[11] J. Youn, N. Rai, and I. Tagkopoulos, “Knowledge integration and decision support for accelerated discovery of antibiotic resistance genes,” Nature Communications, vol. 13, Art. no. 2360, 2022, doi: 10.1038/s41467-022-29993-z.
[12] K. Luck et al., “A reference map of the human binary protein interactome,” Nature, vol. 580, no. 7803, pp. 402–408, 2020, doi: 10.1038/s41586-020-2188-x.
