You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Firstly, congratulations on publishing an excellent article on Radiology! However, during the reading process, I had doubts about the calculation of the binary F1 score. Is this calculation aimed at the accuracy of the recommended structured template, or at the accuracy of binary diagnosis for GPT after completing the structuring?
The text was updated successfully, but these errors were encountered:
For the Chest-Xrays we had binary annotations, which allowed to calculate the F1 score (data and annotations will be published in the near future as well).
For the longer reports, we did not provide any scores but did manual verification of correctness.
Firstly, congratulations on publishing an excellent article on Radiology! However, during the reading process, I had doubts about the calculation of the binary F1 score. Is this calculation aimed at the accuracy of the recommended structured template, or at the accuracy of binary diagnosis for GPT after completing the structuring?
The text was updated successfully, but these errors were encountered: