LLM-Augmented Knowledge Graphs for EHR Summarization
Click the Poster to View Full Screen, Right click to save image
Grant: CPC Integrated Health
Tristram Dacayan
College:
The Dorothy and George Hennings College of Science, Mathematics, and Technology
Major:
Computer Science
Faculty Research Advisor(s):
Daehan Kwak
Abstract:
With the introduction of electronic health records (EHR) in the medical field, doctors and nurses are able to examine patients faster and more efficiently as opposed to using paper records. Despite the advancement in patient documentation technology, one of the main drawbacks for EHRs is the inconsistent format of documents among the different medical specialties, specifically psychiatry and behavioral health EHRs, as well as those used by a range of behavioral healthcare professionals, being more anecdotal and text-based. With the recent advancement of large language models (LLM), there is large potential for this technology to become a viable solution as medical professionals could use them to summarize and inquire about patients at record speeds. While LLMs have the potential to revolutionize the medical industry, their issues include their inconsistently formatted responses and their limited knowledge domain. As a consequence, they are currently not applicable in high-stakes medical situations as a single incorrect diagnosis could result in the patient's injury. We propose the use of knowledge graphs with LLMs to both obtain more consistent outputs of important relationships and expand the knowledge domain of pretrain LLMs via cross-validation with existing medical domain graphs. Through prompt-engineering, the LLM is able to generate formatted knowledge graphs based on a set of rules that focus on extracting as many relationships involving the patient, including afflictions and previous addictions. Using these graphs, we are provided with better visualizations on the patient's current and previous issues and reduce the complexity of future inquiries regarding their health.