Automatic sleep stage determination for sleep apnea syndrome patients: Multivariate probability distribution based on expert knowledge

In this study, the multivariate probability distribution was investigated to develop the expert knowledge-based automatic sleep stage determination technique. The ultimate purpose is to develop adaptive automatic sleep stage determination algorithm for clinical practice. Gaussian distribution was adopted to realize automatic parameter selection while Cauchy distribution was adopted to estimate the parameter distribution. The different usage of multivariate probability distribution of Gaussian and Cauchy distribution presented effectiveness for automatic sleep stage determination.