Skip to main content
. 2021 Mar 23;12(4):2280–2298. doi: 10.1364/BOE.415962

Fig. 6.

Fig. 6.

Bad quality predictions of patches from the test dataset. Column (A) shows NLM patches, column (B) visualizes contrast inverted NLM patches, column (C) shows histopathologically stained H&E patch, and columns (D) and (E) depict computationally stained H&E patch by the Pix2Pix model and the cycle CGAN model, respectively. The scale bar represents 50 μm. Here, computationally stained H&E patches (columns D and E) are not similar to histopathologically stained H&E patches (column C), as histopathologically stained H&E patches show different structures than the corresponding NLM patch. Furthermore, in dark NLM patches (second and third row), the computationally stained H&E patches fail to generate appropriate color contrast in crypt regions (marked by white arrows).