Snore site discrimination using statistical moments of free field snoring sounds recorded during sleep nasendoscopy

The non-invasive measurement of free field snoring sounds to estimate the site of snoring is an important development in the diagnosis, treatment and management of sleep-related breathing disorders. We investigated characteristics of the probability density function by testing the sensitivity of the statistical moments to amplitude patterns in the snoring acoustic signal. Snoring sounds from 15 patients were recorded whilst performing sleep (under anaesthetic) nasendoscopy evaluation. We demonstrated, using a fuzzy 2-means clustering method, that a combination of the statistical dimensionless moment coefficients of skewness and kurtosis could discriminate between pure palatal and non-palatal snoring subject groups.