NIH VideoCast - Using Informatics and Generative AI to Support Patients’ Understanding of Lab Test Results

Skip Navigation

Official websites use .gov
A .gov website belongs to an official government organization in the United States.

Secure .gov websites use HTTPS
A lock ( ) or https:// means you’ve safely connected to the .gov website. Share sensitive information only on official, secure websites.

Using Informatics and Generative AI to Support Patients’ Understanding of Lab Test Results

Loading... Loading video...

161 Views

Air date:	Thursday, January 11, 2024, 2:00:00 PM Time displayed is Eastern Time, Washington DC Local
Views:	Total views: 161 (73 Live, 88 On-demand)
Category:	Special
Runtime:	00:57:27
Closed captioning:	Check box to display CC outside of the video Note: You can drag the captioning window around and resize it
Description:	Bio: Dr. Zhe He is an Associate Professor in the School of Information (iSchool) at the Florida State University and a Core Affiliate of the Institute for Successful Longevity. He is Director of Biostatistics, Informatics, and Research Design (BIRD) Program in the UF-FSU Clinical and Translational Science Award Hub. His research expertise includes machine learning, natural language processing, and knowledge representation, The overarching goal of his research is to improve population health and advance biomedical research through the application of informatics. Abstract: Viewing laboratory test results is patients’ most frequent activity when accessing patient portals, but lab results can be very confusing for patients. Previous research has explored various ways to present lab results, but few have attempted to provide tailored information support based on individual patient’s medical context. In this project, we aim to gain insights into older adults’ challenges in understanding lab test results and act upon them, and design an informatics tool called LabGenie to support older adults’ understanding of lab results and engagement in their own care. To design such a tool, we conducted three studies. In the first study, we conducted an online survey with 270 patients with one or more chronic conditions to understand the factors influencing lab results understanding. In the second study, we collected and annotated 200+ health articles about laboratory tests from AHealthyMe.com, and evaluated transformer-based large language models for recognizing key terms and their types. In the third study, we selected lab results related questions from a community Q&A website Yahoo! Answers and evaluated the answers to these questions from four generative AI models including GPT-4, MedAlpaca, Llama 2, and ORCA_mini.

NLM Title:	Using informatics and generative AI to support patients' understanding of lab test results / Dr. Zhe He.
Author:	He, Zhe. National Institutes of Health (U.S). Natural Language Processing Special Interest Group,
Publisher:
Abstract:	(CIT): Bio: Dr. Zhe He is an Associate Professor in the School of Information (iSchool) at the Florida State University and a Core Affiliate of the Institute for Successful Longevity. He is Director of Biostatistics, Informatics, and Research Design (BIRD) Program in the UF-FSU Clinical and Translational Science Award Hub. His research expertise includes machine learning, natural language processing, and knowledge representation, The overarching goal of his research is to improve population health and advance biomedical research through the application of informatics. Abstract: Viewing laboratory test results is patients" most frequent activity when accessing patient portals, but lab results can be very confusing for patients. Previous research has explored various ways to present lab results, but few have attempted to provide tailored information support based on individual patient"s medical context. In this project, we aim to gain insights into older adults" challenges in understanding lab test results and act upon them, and design an informatics tool called LabGenie to support older adults" understanding of lab results and engagement in their own care. To design such a tool, we conducted three studies. In the first study, we conducted an online survey with 270 patients with one or more chronic conditions to understand the factors influencing lab results understanding. In the second study, we collected and annotated 200+ health articles about laboratory tests from AHealthyMe.com, and evaluated transformer-based large language models for recognizing key terms and their types. In the third study, we selected lab results related questions from a community Q&A website Yahoo! Answers and evaluated the answers to these questions from four generative AI models including GPT-4, MedAlpaca, Llama 2, and ORCA_mini.
Subjects:	Artificial Intelligence Clinical Laboratory Information Systems Comprehension Health Communication--standards Medical Informatics Applications Patient Access to Records--standards
Publication Types:	Lecture Webcast

NLM Classification:	QY 26.5
NLM ID:	9918768586906676
CIT Live ID:	53956
Permanent link:	https://videocast.nih.gov/watch=53956

Related Videos...