A method and system for classification and detection of sleep snoring based on deep learning

Abstract

The invention discloses a sleep snoring sound classification and detection method based on deep learning. The method mainly includes: collecting sleep sound signals of a patient under test throughout the night through a sensor, and detecting sound segments in the sleep sound signals to obtain sleep sound. The sound segment map in the acoustic signal; use deep learning to classify the sound segment map of snoring and non-snoring sounds, and retain the recognition results of pure snoring sounds; then use deep learning to classify the recognition results of pure snoring sounds into four types of snoring sounds to complete the classification. Automatic recognition and detection of snoring in patients with apnea hypopnea syndrome (OSAHS); according to the recognition and detection results of snoring, count the number of various types of snoring of the tested patient throughout the night, and obtain the AHI index of the tested patient throughout the night. The invention also discloses a detection system of a sleep snoring sound classification detection method based on deep learning. The method and system of the present invention can effectively and accurately evaluate whether the snoring object is ill and the degree of the disease, and provide data reference for OSAHS patients.

Description

technical field [0001] The invention relates to the technical field of disease detection, in particular to a method and system for classifying and detecting sleep snoring based on deep learning. Background technique [0002] Obstructive sleep apnea-hypopnea syndrome (OSAHS) is a relatively serious sleep-disordered breathing. Sometimes snoring accompanied by apnea or low respiratory flow. Apnea refers to the situation where the patient's breathing airflow disappears for more than 10 seconds while sleeping, and hypopnea refers to the situation where the patient's respiratory airflow intensity is less than 50% of the basic value while sleeping, and the blood oxygen concentration drops below 96% of the normal level. . [0003] For the detection of OSAHS, the traditional method is to monitor and measure the patient's sleep for 6 to 7 hours through a polysomnography device, which can record and analyze EEG (electroencephalogram), ECG (electrocardiogram), EOG (electrooculogram) ...

Images