Handling box deformation during NEB

Hello LAMMPS developers and users,

I want to do an NEB calculation in which the simulation box of the initial and final images are not the same. The number of atoms in the initial and final are still same but the density changes since the crystal transforms from one phase to another.

The NEB documentation says that LAMMPS assumes that the simulation box stays the same in all the images. Is there a way to handle the simulation box parameters (lx,ly,lz,xy,xz,yz) as a part of the reaction coordinates? (similar to how cell parameters are included in reaction coordinates in solid state NEB https://aip.scitation.org/doi/10.1063/1.3684549).

Any help with this would be much appreciated