However, the method cannot yet be used in hospitals, for example, and the AI does not generate actual histological information, but rather plausible predictions based on the training data.  

In future, there may be a decline in the use of antibody stains, the traditional production of which still relies on animals (e.g., mice for the production of monoclonal antibodies or rabbits for polyclonal antibodies). The antibodies obtained in this way are conjugated with the dye. However, antibodies produced without the use of animals—from bacteria, yeast, algae, or the well-known phage display—already exist.

Original paper:
Almagro-Pérez C, Peruzzi N, Galambos C, Song AH, Brunnström H, Gawlik KI, Stampanoni M, Tran-Lundmark K, Lovric G. (2026). Histology-guided 3D virtual staining of microCT-imaged lung tissue via deep learning. J. R. Soc. Interface 23 : 20251186. https://doi.org/10.1098/rsif.2025.1186

Source:
https://www.bionity.com/

https://www.psi.ch/en/x-ray-tomography-group/people/goran-lovric