Spatial generalized linear mixed models are used commonly for modeling discrete spatial responses. In this models the spa
tial correlation of the data is considered as spatial latent variables. For simplicity, it is usually assumed in these models that
 spatial latent variables are normally distributed. An incorrect normality assumption may leads to inaccurate results and is
 therefore erroneous. In this paper we model the spaial latent variables in a general random field, namely the closed skew
 Gaussian random field which is more flexible and includes the Gaussian random field. We propose a new algorithm for
 maximum likelihood estimates of the parameters. A key ingredient in our algorithm is using a Hamiltonian Monte Carlo
 version of the EM algorithm. The performance of the proposed model and algorithm is presented through a simulation study.