Investigative Ophthalmology & Visual Science Cover Image for Volume 65, Issue 7
June 2024
Volume 65, Issue 7
Open Access
ARVO Annual Meeting Abstract  |   June 2024
Diagnosing Glaucoma Based on a Large Language Model Chatbot
Author Affiliations & Notes
  • Hina Raja
    The University of Tennessee Health Science Center Department of Ophthalmology Hamilton Eye Institute, Memphis, Tennessee, United States
  • Xiaoqin Huang
    The University of Tennessee Health Science Center Department of Ophthalmology Hamilton Eye Institute, Memphis, Tennessee, United States
  • Mohammad Delsoz
    The University of Tennessee Health Science Center Department of Ophthalmology Hamilton Eye Institute, Memphis, Tennessee, United States
  • Yeganeh Madadi
    The University of Tennessee Health Science Center Department of Ophthalmology Hamilton Eye Institute, Memphis, Tennessee, United States
  • Asma Poursoroush
    University of Memphis Department of Biomedical Engineering, Tennessee, United States
    The University of Tennessee Health Science Center Department of Ophthalmology Hamilton Eye Institute, Memphis, Tennessee, United States
  • Asim Munawar
    International Business Machines Corp, Armonk, New York, United States
  • Malik Kahook
    University of Colorado Anschutz Medical Campus School of Medicine, Aurora, Colorado, United States
  • Siamak Yousefi
    The University of Tennessee Health Science Center Department of Ophthalmology Hamilton Eye Institute, Memphis, Tennessee, United States
  • Footnotes
    Commercial Relationships   Hina Raja None; Xiaoqin Huang None; Mohammad Delsoz None; Yeganeh Madadi None; Asma Poursoroush None; Asim Munawar None; Malik Kahook None; Siamak Yousefi NIH, Code F (Financial Support), Enolink, Code R (Recipient), Remidio, Code R (Recipient), M&S Technologies, Code R (Recipient), InsightAEye, Code R (Recipient)
  • Footnotes
    Support  EY030142
Investigative Ophthalmology & Visual Science June 2024, Vol.65, 1636. doi:
  • Views
  • Share
  • Tools
    • Alerts
      ×
      This feature is available to authenticated users only.
      Sign In or Create an Account ×
    • Get Citation

      Hina Raja, Xiaoqin Huang, Mohammad Delsoz, Yeganeh Madadi, Asma Poursoroush, Asim Munawar, Malik Kahook, Siamak Yousefi; Diagnosing Glaucoma Based on a Large Language Model Chatbot. Invest. Ophthalmol. Vis. Sci. 2024;65(7):1636.

      Download citation file:


      © ARVO (1962-2015); The Authors (2016-present)

      ×
  • Supplements
Abstract

Purpose : To evaluate the capabilities of a large language model (LLM) based chatbot for diagnosing glaucoma using the Ocular Hypertension Treatment Study (OHTS) dataset.

Methods : A total of 3170 eyes of 1585 subjects from the OHTS were included in this study. We selected demographic, clinical, ocular, visual field, optic nerve head photo, and history of disease parameters of each subject. We automated the process of case report generation by converting tabular data into textual format based on information from both eyes of all subjects using application program interface (API) of ChatGPT (Fig 1). We randomly selected subsets of patients and tested different final questions to engineer a prompt with the highest accuracy (Fig 2A). We then tested different combinations of parameters to assess the diagnostic performance and selected the best performing subset for the downstream analysis. We subsequently developed a procedure using API of ChatGPT, to automatically input prompts into the chat box followed by querying ChatGPT (3.5 and 4.0) regarding the underlying diagnosis of each subject based on the onset and last visits.

Results : ChatGPT3.5 achieved AUC of 0.74, accuracy of 66%, specificity of 64%, sensitivity of 85%, and F1 score of 0.72. ChatGPT4.0 obtained AUC of 0.76, accuracy of 87%, specificity of 90%, sensitivity of 61%, and F1 score of 0.92 based on the last visit (Fig 2B).

Conclusions : The accuracy of ChatGPT4.0 in diagnosing glaucoma based on OHTS was promising. The overall accuracy of ChatGPT4.0 was higher than ChatGPT3.5. However, ChatGPT3.5 was found to be more sensitive than ChatGPT4.0. Currently, ChatGPT may serve as a useful tool in exploring disease status of ocular hypertensive eyes with clinical parameters. In the future, leveraging LLM with multi-modal capabilities for integration of demographic, clinical, and imaging data, may further enhance diagnostic capabilities.

This abstract was presented at the 2024 ARVO Annual Meeting, held in Seattle, WA, May 5-9, 2024.

 

Figure 1. A)Diagram of prompt generation and diagnosis. Demographic, clinical, ocular, visual field, optic nerve head, and history parameters inform case reports generated with ChatGPT 3.5. Reports were then analyzed using ChatGPT 3.5 and 4.0 for diagnosis. B)Case report design based on ChatGPT web and API. C)Prompt for diagnosis based on ChatGPT web and API.

Figure 1. A)Diagram of prompt generation and diagnosis. Demographic, clinical, ocular, visual field, optic nerve head, and history parameters inform case reports generated with ChatGPT 3.5. Reports were then analyzed using ChatGPT 3.5 and 4.0 for diagnosis. B)Case report design based on ChatGPT web and API. C)Prompt for diagnosis based on ChatGPT web and API.

 

Figure 2. A)Comparative analysis of prompt engineering based on the final question. B)Evaluation of ChatGPT versions for diagnosing glaucoma based on information from the glaucoma onset and final visits.

Figure 2. A)Comparative analysis of prompt engineering based on the final question. B)Evaluation of ChatGPT versions for diagnosing glaucoma based on information from the glaucoma onset and final visits.

×
×

This PDF is available to Subscribers Only

Sign in or purchase a subscription to access this content. ×

You must be signed into an individual account to use this feature.

×