Basically the mixture of solvent and epoxy is much less viscous and easily infiltrates the material. There are numerous protocols that describe the process. For example: http://web.path.ox.ac.uk/~bioimaging/bitm/instructions_and_i...

Start where it says "Dehydrate through a series of ethanols or acetones and propylene oxide." basically this is for replacing the water with solvents, if the sample is dry you can just immerse it in your solvent of choice, probably propylene oxide (toxic) or acetonitrile (less toxic), something miscible with your epoxy resin.

Follow the protocol up to the polymerization. I'm not sure if all resins are suitable for this type of infiltration but there are many suitable ones, see: https://www.tedpella.com/chemical_html/chem2.htm