Synapse
⌘+K
Synapse
PulseExploreClubsResearchersJournals
Instagram
HomeClubsExplore
September 10, 2025

Evaluation of Differential Diagnosis Accuracy Using a Customized GPT-Based Artificial Intelligence Model in the Emergency Department (Preprint)

View Full Paper
Ask AI
Bookmark
Share

Authors

HAH A Z Taha AlCACanan AkmanSYSinan YILDIRIM

Discussion

Loading...

Member takes

Overview

Retrospective analysis highlights custom GPT model's 88.4% diagnostic accuracy in ED, suggesting it may assist clinical decision-making.

Key Points

  • The custom GPT model achieved a primary diagnostic accuracy of 88.4%, indicating its strong potential for use in emergency departments.
  • Accuracy varied significantly, with green triage patients showing the highest accuracy at 93.07%, while red triage patients had the lowest at 75.31%.
  • The model’s performance was statistically analyzed based on demographic factors, showing significant correlation with triage categories but not age or sex.
  • Further caution is advised for its application in high-acuity cases, as lower accuracy was noted in cardiovascular system diagnoses and hospitalized patients.

Cite This Study

Al et al. (2025) studied this question.

synapsesocial.com/papers/68c23b8fb210217d64785a62https://doi.org/10.2196/preprints.82363
View Full Paper
Ask AI
Bookmark
Share

Also Consider

Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context:

  1. 1Diagnostic Accuracy of a Large Language Model (ChatGPT-4) for Patients Admitted to a Community Hospital Medical Intensive Care Unit: A Retrospective Case Study2025 · 2 citations
  2. 2Comparison of Intern Doctors and Chat-GPT in Emergency Cases Assessment2025
  3. 3A Retrospective Cross Sectional Study Comparing the Accuracy of Artificial Intelligence Tools in the Diagnosis of Common Pediatric Emergencies2025
  4. 4AI Based Clinical Decision-Making Tool for Neurologists in the Emergency Department2025
  5. 5Large language models for automatable real-world performance monitoring of diagnostic decision support systems: a comparison to manual doctor panel review in a prospective clinical study2025