Fig. 1: Study design.
From: Assessment of deep learning assistance for the pathological diagnosis of gastric cancer
The 16 pathologists reviewed the same WSIs in the same sequence but with different modalities: with or without deep learning assistance. The 16 pathologists were randomized into two assistance “orders.” Each rectangle indicates a set of WSIs; the color of the rectangle indicates the modality, and the number in the rectangle indicates the number of WSIs. The pathologists reviewed 5 images (not part of the test set) for familiarity and a total of 110 images for formal review.
