好色先生

好色先生

Explore the latest content from across our publications
Join The AAN

Log In

Forgot Password?
Create New Account

Loading... please wait

Abstract Details

Exploring the use of ChatGPT in Reporting a Case of HIV, Syphilis, and Lyme Co-infection Presenting with Multiple Cranial Neuropathies
好色先生, Research, and Methodology
P5 - Poster Session 5 (5:30 PM-6:30 PM)
7-003
This case report delves into the utilization of Microsoft Bing Chat Generative Pre-Trained Transformer (ChatGPT) for reporting a unique case involving co-infection with HIV, syphilis, and Lyme disease, presenting with multiple cranial neuropathies and gait abnormalities.
Microsoft Bing ChatGPT is one of the latest AI search engines powered by GPT language model developed by OpenAI and uses the testing version of the GPT-4 model. The chat option provides a broad range of features, including answering questions, writing articles and sifting through scientific literature to provide references and resources

ChatGPT-4 model was used to write the case report based on the input prompt provided and generate background content for the case report; this was then compared to the background content created by ChatGPT based on the information collected by our own research.
The case report itself revealed that both neurosyphilis and Lyme can lead to a similar presentation with multiple cranial neuropathies. When concomitant HIV is present, clinical findings are exacerbated. In the setting of co-infection, it is hard to determine which pathogen is the most offending. ChatGPT allowed us to generate texts faster whether it be purely by AI generated text or based on our own research. However, text generated based on our own research was more accurate and expressed points more clearly. Purely AI generated text did not provide enough citations and, if provided, was not always correct.
Overall, ChatGPT is a great tool to use to decrease time and improve efficiency in producing text, and thereby promote rapid turnover of literature. It can be used to generate text based on your own research but should be used cautiously if relying on ChatGPT solely to generate the material. If this last method is employed, fact checking should be used. 
Authors/Disclosures
Tse Chiang Chen, MD (Tulane School of Medicine)
PRESENTER 		Dr. Chen has nothing to disclose.
Nimmi N. Wickramasuriya, MD (Tulane University Hospital and Clinics) Dr. Wickramasuriya has nothing to disclose.
Eleftheria Vyras, MBBS (Tulane School of Medicine Neurology Residency) Dr. Vyras has nothing to disclose.
Jorie M. Singer Ms. Singer has nothing to disclose.
Martha Robinson, MD (Steward Health Physician Services) Dr. Robinson has nothing to disclose.