For any given nucleus, it is possible to calculate its net binding energy for various configurations (positions) of the protons and neutrons in it, including which way all their spins are organized. The most likely packing of the nucleons is the one which maximizes the binding energy, resulting in the most stable nucleus.
You can then perturb the nucleus in your binding energy model by pulling it out of round or re-arranging the spins and recalculating the binding energy for any weird configuration you want. the difference in energy between the most stable or "ground" state of the nucleus and what it has in the deformed state is the energy which must be put into the nucleus to get it deformed, and will be the amount released when this excited state decays back into the ground state.
Now, the most likely lifetime of the excited state can itself be calculated knowing the size of that energy difference and a bunch of other quantum considerations (including, for example, the tunneling probability for nucleons to leak through a potential barrier that traps them in a metastable state) and from that you get a half-life for the nuclear isomer.
Then you discover that the theoretically-calculated half-life for certain isomers is really, really long while for others it is extremely short.
Since the energy release in a nuclear isomer decay process is of the order ~gamma rays, the expected decay product will be a gamma-ray photon- which led some physicists to conclude that if you had a large number of metastable isomers and beamed them with gamma rays, you could trigger their simultaneous decay and thereby make a bomb far more powerful than any chemical bomb yet less powerful than a fission bomb- and that bomb would be extremely compact and easy to carry around.
But no conclusive evidence exists that you can in fact trigger that all-at-once decay with a gamma ray beam- so the isomer bomb remains a science-fiction weapon only.