Duke Center for Health Informatics: Interoperability and Reproducibility of Multimorbidities in EHRs: Pathways to Precision Medicine through Machine Learning and AI
Multimorbidity, characterized as the nonrandom co-occurrence of multiple health conditions in an individual, is prevalent and significantly influences health outcomes. Insight into disease-disease interactions and the underlying mechanisms of multimorbidity can inform new opportunities for the development of novel preventative strategies, interventions, and personalized treatments. Large-scale electronic health record (EHR) systems provide an extensive real-world patient data source for evaluating disease multimorbidities, yet concerns about data accuracy, completeness, and interoperability across institutions persist. In our study, we utilized multivariate analysis and network models to investigate disease multimorbidities across different EHR systems. Our findings demonstrate reasonably consistent multimorbidity patterns across systems when using ICD codes, effectively capturing recognized disease clusters, genetic correlations, causal relationships, and potentially new discoveries. To assist a broader range of researchers in performing complex data analysis, we proposed an interactive visualization solution for in-depth exploration of the comprehensive multimorbidity knowledge base constructed across multiple systems. A case study will further illustrate the value of cross-institutional discovery in EHR-based data research, promoting the advancement of personalized medicine.
Dr. Yaomin Xu, an Assistant Professor of Biostatistics and Biomedical Informatics at Vanderbilt University Medical Center, specializes in leveraging machine learning approaches to extract novel insights from large-scale, real-world health systems, including Biobanks and Electronic Health Records (EHRs). His technical expertise encompasses multivariate data analysis, data visualization, and unsupervised learning, all with a concentrated focus on the application and problem-solving within the realms of translational bioinformatics and health informatics.
Zoom Meeting link:
Meeting number (access code) 920 0552 7024, Meeting password: 128631