Skip to main content
. 2022 Jun 18;12:10296. doi: 10.1038/s41598-022-14042-y

Figure 3.

Figure 3

Virtual Staining framework. Pairs of TA-PARS and H&E images are co-registered to create the training dataset, which is then used to train the Pix2Pix39 model. Unseen TA-PARS images are then fed into the trained model for virtual staining, producing H&E-like visualizations. Scale bar: 50 µm.