Research On Sleep Breathing Monitoring Method Based On Snoring And Blood Oxygen Saturation

Obstructive sleep apnea hypopnea syndrome(OSAHS)is a common sleep disease,it can cause patients to sleep hypoxia at night,seriously affects the quality of sleep and physical health of patients,so the detection and diagnosis of OSAHS are very important.The current gold standard for OSAHS detection is polysomnography,but polysomnography is more expensive,and requires multiple leads to connect to the patient,it affect the patient's sleep.Therefore,this paper proposes a predictive method of apnea hypopnea index based on probabilistic integrated regression model to achieve the detection of OSAHS in patients.The successful implementation of the algorithm can greatly facilitate the detection of OSAHS suspected patients,and reduce the threat of OSAHS to the health of patients.First of all,the innovative point of the OSAHS diagnostic method proposed in this article is to combine snoring and blood oxygen saturation.The characteristics of OSAHS are used to preprocess the blood oxygen saturation,and the median filter is used to remove the noise in the blood oxygen saturation.The blood oxygen saturation lowering threshold value performs the lowering stage processing.According to the time corresponding find the snoring signal to the blood oxygen saturation decline segment.use the Mel cepstrum coefficients for feature extraction,and input the features into the convolutional neural network.Combine the logistic regression model to classify the audio signal of the decline segment into snoring,breathing,and noise.And extract the characteristics of snoring sound and blood oxygen saturation after treatment.Secondly,through the introduction of the principle and structure of the probabilistic integrated regression model,the probabilistic integrated regression model is used to train the probability of the occurrence of apnea hypopnea per hour,and the apnea hypopnea index is predicted.The experimental results show that the correlation between the prediction results of the probabilistic integrated regression model,and the results of polysomnography reaches 0.9838.Finally,this paper designs and implements a sleep monitoring data management system for managing sleep data of OSAHS patients.The implementation of sleep monitoring data management system includes data identification,analysis and storage,snoring and blood oxygen saturation signal import and processing,database design and system functions,etc.Clinical application shows that the system is of great help to doctors' scientific research,and initial diagnosis of patients.