Diagnostic support of obstructive sleep apnea based on snoring analysis

点击翻译按钮获取中文摘要

As an alternative to polysomnography test, we attempted to develop the snore-based OSA diagnosis techniques. The results obtained show that :(i) high diagnosis accuracy is achieved by noise-robust formant detection,(ii) there is a significant difference between OSA and non-OSA snorers in the high frequency region> 5kHz,(iii) apneic snores exhibits more complex behavior with nonlinear dynamic properties compared to non-apenic snores,(iv) neural network-based method developed can detect the low SNR snore-breathing episode.