Hi,
As far as I understood the validation script requires a file in which we store the subject id, laterality and the confidence. Thus, in cases where we have both modalities CC and MLO and/or multiple images, you ignore this aspect and expect only a confidence per laterality.
Did I understand this correctly?
Thanks,
Anamaria
Created by Anamaria Vizitiu Anamaria > Thus, in cases where we have both modalities CC and MLO and/or multiple images, you ignore this aspect and expect only a confidence per laterality. Did I understand this correctly?
Yes, this is correct. We are asking for a prediction at the breast level (and not at the image level).
Thanks!