Fig. 1: Demonstration of multiple stains being virtually generated using a class-conditional neural network and images in two autofluorescence channels (DAPI and Texas Red) of a label-free tissue sample. | Light: Science & Applications

Fig. 1: Demonstration of multiple stains being virtually generated using a class-conditional neural network and images in two autofluorescence channels (DAPI and Texas Red) of a label-free tissue sample.

From: Digital synthesis of histological stains using micro-structured and multiplexed virtual staining of label-free tissue

Fig. 1

a Steps involved in virtually creating the various stains. By adding a class condition to the network using a digital staining matrix, a single network can be used to generate multiple stains or a blending of stains in the same tissue cross section on demand. b A second field of view demonstrating the three digital stains generated using a single trained network. Contrast-enhanced unstained tissue images are provided for visual guidance; unprocessed raw versions of these images were used as the input to the neural network. N/A (not available) refers to the fact that once a tissue section has been histochemically stained with one type of stain, we cannot subsequently stain it with other stains; therefore, the comparison includes N/A entries

Back to article page