Leveraging Large Language Models for Simplified Patient Summary Generation, Literature Retrieval and Medical Information Summarization: A Health CASCADE Study
Files
Date
2024-01-03
Contributor
Advisor
Department
Instructor
Depositor
Speaker
Researcher
Consultant
Interviewer
Narrator
Transcriber
Annotator
Journal Title
Journal ISSN
Volume Title
Publisher
Volume
Number/Issue
Starting Page
3255
Ending Page
Alternative Title
Abstract
In the evolving healthcare landscape, integrating advanced technologies such as machine learning and natural language processing has become vital. This paper presents an innovative system that leverages modern Natural Language Processing (NLP) capabilities to extract information from Electronic Health Records (EHRs) and generate simplified patient summaries (SPS). These SPS are subsequently used to provide clinicians with summaries of relevant academic literature, improving their ability to access pertinent information efficiently. The system architecture employs Large Language Models (LLMs) to generate SPSs and summarize relevant information, while dense vector retriever models are used for information retrieval from document corpus, which is created by combining parts of publicly available datasets such as PubMed, the CORD19 dataset, and more. The presented system has the potential to significantly reduce the time and effort required by clinicians to access relevant patient information, allowing them to concentrate more on patient care and contribute to improved patient outcomes.
Description
Keywords
Data Platforms and Ecosystems in Healthcare, electronic health records, fhir, information retrieval, natural language processing, summarization
Citation
Extent
10 pages
Format
Geographic Location
Time Period
Related To
Proceedings of the 57th Hawaii International Conference on System Sciences
Related To (URI)
Table of Contents
Rights
Attribution-NonCommercial-NoDerivatives 4.0 International
Rights Holder
Local Contexts
Collections
Email libraryada-l@lists.hawaii.edu if you need this content in ADA-compliant format.