A Detection Based On The Ratio Of Sub-band Energy For Sleep Apnea Syndrome Diagnosis

Obstructive Sleep Apnea Syndrome is a higher incidence disease with serious hazard and potential danger. PSG has become the gold standard to diagnose OSAS and it has been widely applied to the detection of OSAS. PSG is restricted greatly by its higher require for the test environment and limited number. According to incomplete statistics, there are 80%-90% patients cannot get timely treatment. A little sleeps disturbance in patients, low-cost, more convenient and being applied to large-scale diagnostic method becomes a problem demanding prompt solution.Snoring is the most direct clinical characterization of OSAS which is also known as snoring disease and it contains the information of the upper airway. In addition, it is very easy to get snoring. For the above reasons, using snoring characteristics for OSAS detection has received considerable attention. From the practice application, in this paper, we analyze the robust methods for OSAS diagnosis in the noisy environment and study on different OSAS graded patients, including:1) We proposed a novel method which is used the rate of sub-band energy. This method which is similar to conventional physical examination can diagnose numerous snorers and make the patients with OSAS positive diagnosed by PSG for further.2) Considering the more serious one patient suffers from, the higher frequency band which his energy spectrum concentrated, we studied on the different OSAS graded patients based on the rate of sub-band energy.3) As a screening method, its anti-noise performance is a crucial factor. We evaluate our method's performance in the case of strong noise (SNR=5 dB).The experimental results show that our approach can effectively overcome the influence of environmental noises, so the high accuracy can be achieved even in the home environment. In our experiments, the sensitivity of the proposed method is 92.8% and the specificity is 92.7%. It has basically met the requirements in practical applications of OSAS diagnosis. In addition, our study of different OSAS grading patients provides an effective method which is used snoring characteristics to distinguish different prevalence of patients.