Using ChatGPT to assist in judging the indications for emergency ultrasound: an innovative exploration of optimizing medical resource allocation
IntroductionTo assess the performance of the GPT-4O model in determining the indications for emergency ultrasound and to explore its potential for optimizing medical resource allocation.MethodsThis single-center retrospective observational study included 200 patients who underwent emergency ultrasou...
Saved in:
| Main Authors: | , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Frontiers Media S.A.
2025-06-01
|
| Series: | Frontiers in Medicine |
| Subjects: | |
| Online Access: | https://www.frontiersin.org/articles/10.3389/fmed.2025.1567608/full |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | IntroductionTo assess the performance of the GPT-4O model in determining the indications for emergency ultrasound and to explore its potential for optimizing medical resource allocation.MethodsThis single-center retrospective observational study included 200 patients who underwent emergency ultrasound at the emergency department. Senior clinicians assessed the indications for ultrasound based on guidelines, which served as the gold standard. The medical records were input into the GPT-4O model, which generated binary classification results. The model’s performance was analyzed using confusion matrices and ROC curves.ResultsThe GPT-4O model achieved perfect sensitivity and NPV (1.00), with specificity and PPV of 0.86, and an AUC of 0.93. The model accurately identified 92 emergency cases and 93 non-emergency cases, with only 15 non-emergency cases misclassified as emergency cases.ConclusionThe GPT-4O model showed excellent performance in determining the indications for emergency ultrasound, particularly in terms of sensitivity and negative predictive value. It has the potential to reduce unnecessary examinations and optimize the allocation of medical resources. |
|---|---|
| ISSN: | 2296-858X |