Back to the list
Congress: ECR24
Poster Number: C-21819
Type: EPOS Radiographer (scientific)
Authorblock: A. Yagahara; Sapporo/JP
Disclosures:
Ayako Yagahara: Nothing to disclose
Keywords: Artificial Intelligence, MR, Safety, Education and training
Purpose The purpose of this study was to calculate the similarity in outputs between the same model, different models, and within and between multiple languages, in order to evaluate whether the responses generated by ChatGPT and GPT-4 prompts in both Japanese and English regarding MRI-related queries maintain consistency.
Read more Methods and materials BackgroundWith its superior soft tissue contrast, magnetic resonance imaging (MRI) is one of the useful tools in diagnostic imaging. However, due to the generation of strong magnetic fields, it carries inherent risks. To prevent accidents, not only is it essential for medical professionals like radiologists and radiologic technologists to receive safety training, but it is also crucial to provide patients with appropriate explanations.The accuracy of ChatGPT [1] has had a significant impact in the radiology field, yielding promising results for...
Read more Results Summary of generated sentencesTable 1 presents a summary of the sentences generated using four different models. Representative examples of these generated sentences are illustrated in Table 2. Cosine similarityThe results of the similarity analysis in question 1 are shown in Figure 1. Notably, a high degree of similarity was observed in the English outputs. Regardless of language or model differences, a generally high similarity trend was evident compared to the other two query texts.The results of the similarity analysis in question...
Read more Conclusion This study highlighted the output variability in MRI examination explanations based on model versions and languages.
Read more References [1] OpenAI. ChatGPT. https://openai.com/chatgpt (Accecced 7/1/2024)[2] Fink MA, Bischoff A, Fink CA, Moll M, Kroschke J, Dulz L, Heußel CP, Kauczor HU, Weber TF. Potential of ChatGPT and GPT-4 for Data Mining of Free-Text CT Reports on Lung Cancer. Radiology. 2023 Sep;308(3):e231362. [3] Chung EM, Zhang SC, Nguyen AT, Atkins KM, Sandler HM, Kamrava M. Feasibility and acceptability of ChatGPT generated radiology report summaries for cancer patients. Digit Health. 2023 Dec 19;9:20552076231221620.[4] Gilson A, Safranek CW, Huang T, Socrates V, Chi...
Read more
GALLERY