| Autor: |
RIYAHI, Abdullah Mahmoud |
| Předmět: |
|
| Zdroj: |
International Journal of Medical Dentistry; Jul-Sep2023, Vol. 27 Issue 3, p463-466, 4p |
| Abstrakt: |
Aim: Evaluation of ChatGPT capabilities for endodontic diagnosis using assumed scenarios. Materials and methods: 10 questions about endodontic diagnosis were asked and answered correctly. The same questions were then given to the ChatGPT. The Artificial Intelligence (AI) responses were then compared to the correct answers. If AI response did not match the correct answer for either pulp or periapical diagnosis, it was considered inaccurate. Furthermore, if different terms other than the one specified by the question are used, AI response is also considered inaccurate.Results and discussion: Only 3 out 10 answers were accurate, therefore the rate of inaccurate endodontic diagnosis to the questions provided was 70%.Conclusions: The free version of ChatGPT 2023 is not a reliable tool for endodontic diagnosis when the questions are presented in this scenario format. Future studies can evaluate the effect of different formats of the question on the accuracy of AI responses [ABSTRACT FROM AUTHOR] |
| Databáze: |
Complementary Index |
| Externí odkaz: |
|
