Font Size: a A A

Snore Detection For Sleep Apnea Hypopnea Syndrome Diagnosis

Posted on:2011-02-05Degree:MasterType:Thesis
Country:ChinaCandidate:W W ChenFull Text:PDF
GTID:2154330332960804Subject:Signal and Information Processing
Abstract/Summary:PDF Full Text Request
Sleep apnea hypopnea syndrome is a high incidence disease with serious hazard and potential danger. As hypoxia and hypercapnia often happen during sleep, ultimately it leads to multi-system and multi-organ progressive damage, the health of patient is greatly affected.The polysomnography has become the gold standard to diagnose SAHS. Because of the limited number of PSG, vast majority of patients can not get timely treatment. A low-cost and more convenient diagnostic method is favored by many researchers.This paper is based on snore detection for screening SAHS. Firstly, based on the method of voice processing, snoring signal is preprocessed. Secondly, after discussing the traditional short-term energy, short-term window-limited rate and dual threshold endpoint detection, this artice proposes the adaptive short-term energy endpoint detection based on the characteristics of snore signal. And compared to the traditional fixed threshold endpoint detection, the new algorithm can dynamically adjust the energy threshold in order to effectively detect the snore segment. Taking into account the discontinuity of the snoring, this paper make reasonable amendment:if the interval of two adjacent snores is less than 0.5 seconds, they will be combined one snore; if the interval of two adjacent snores is more than 60 seconds, they will be excluded.This paper proposes three methods for SAHS diagnosis:(1) The interval greater than 10 seconds is defined as snore apnea. If the number of apnea greater than 5 times per hour, it will be juded as SAHS, and vice versa, juded as disease-free.(2) According to the difference between male and female, different criteria is executed for male and female cases. The threshold of apnea is 8 for SAHS dianosis for male cases, while the threshold is 6 for female.(3) In order to further reduce the impact of individual differences, the interval are classified as normal interval and the apnea interval based on K means. Then according to the distance between two centers and the number of apnea, the case has been detemined whether is SAHS.The experimental results show that three methods can screen SAHS, with senxitivity and specificity are 100% and 85.7%.
Keywords/Search Tags:snore, apnea, hypopnea, K means, PSG
PDF Full Text Request
Related items