Correction: Confidence Measurement Metrics in Multimodal Large Language Models for Ultrasound-Based Radiology Cases: Comparative Evaluation Study of Self-Reported, Consistency-Based, and Hybrid Methods - Summary - MDSpire
Advertisement
Correction: Confidence Measurement Metrics in Multimodal Large Language Models for Ultrasound-Based Radiology Cases: Comparative Evaluation Study of Self-Reported, Consistency-Based, and Hybrid Methods
To report specific corrections made to the original study regarding confidence measurement metrics in multimodal large language models for ultrasound radiology.
Approach:
Footnote Removal: A footnote from Table 2 has been removed to improve clarity.
Formula Correction: The Shannon entropy formula in the Confidence Measurement Metrics section has been corrected for accuracy.
Table Value Update: The values for 'GPT-5' and 'Majority vote (5), n (%)' in Table 1 have been updated to reflect the latest data.
Key Findings:
A footnote has been removed from Table 2 to enhance clarity.
The Shannon entropy formula has been corrected for accuracy.
Specific values in Table 1 for 'GPT-5' and 'Majority vote (5), n (%)' have been updated to reflect the latest findings.
Interpretation:
The corrections will be reflected in the online version of the paper and have been resubmitted to full-text repositories to ensure accuracy in published literature.
Conclusion:
The corrections are intended to enhance the accuracy and reliability of the published study.
