mediation for more than two mediators or a latent variable as mediator

[Mahnaz-Olfati]

I want to do a mediation analysis which none of my variables are not fMRI or sequential (just simple variables). How should I make a latent variable from like 8 parameters and use it as mediator?