Published on: May 12, 2025
Researchers at Florida State University's (FSU) eHealth Lab, part of the School of Information, have been exploring the potential of AI to support healthcare providers in making more accurate patient diagnoses. This advancement promises to improve treatment methods and enhance patient outcomes.
The research, co-authored by Zhe He, Senior Author and Director of FSU's Institute for Successful Longevity, and Balu Bhasuran, Visiting Assistant Professor, is part of a multi-institutional study that has garnered significant attention, with the paper receiving over 3,000 views since its publication in mid-March. The study, published in npj Digital Medicine, builds upon FSU's LabGenie project, a tool designed to help older adults better understand lab test results.
The team is investigating the feasibility of using large language models (LLMs), a type of AI that learns from vast amounts of text data to answer questions accurately, to assist clinicians in improving the accuracy and efficiency of differential diagnoses. Differential diagnosis (DDx) is a key element in clinical decision-making, enabling healthcare providers to differentiate between conditions with similar symptoms.
The AI-generated differential diagnosis is extremely thorough, covering all potential diagnoses for patients, said Zhe He. “This study demonstrates how AI can be a valuable tool to help healthcare providers make more informed decisions for their patients.
The research involved using LLMs to generate lists of the top one, five, and ten differential diagnoses for clinicians to review. The team evaluated the accuracy and predictive capabilities of the models, examining how lab test results influenced diagnostic accuracy.
When we asked the model to provide the top differential diagnoses, most of the models identified the correct diagnosis, even in rare disease cases, said Bhasuran. This is particularly interesting because it indicates that the model can predict diagnoses in less common conditions.
The research seeks to address challenges faced by both healthcare providers and patients. Accurate diagnoses are essential for effective treatment and patient management, directly influencing treatment choices and outcomes. By reducing diagnostic errors, the study aims to improve patient care by eliminating unnecessary tests and procedures, ultimately lowering healthcare costs through shorter hospital stays and fewer repeat treatments.
This research was supported by a grant from the Agency for Healthcare Research and Quality, with additional backing from the University of Florida-Florida State University Clinical and Translational Science Award and the National Library of Medicine. The study also involved collaborations with Tampa General Hospital, along with coauthors from Emory University, the University of South Florida, and the University of North Texas Health Science Center. FSU's Undergraduate Research Opportunity Program (UROP) students, including Angelique Deville, Hailey Thompson, Maggie Awad, and Yash Alva, contributed to extracting key data from the case reports.
Back to News
Empowering Innovation Through Global Conferences and Events.
© 2025 SciInov. All Rights Reserved.