CN-122004785-A - Sleep apnea monitoring method, device, equipment and storage medium

CN122004785ACN 122004785 ACN122004785 ACN 122004785ACN-122004785-A

Abstract

The application provides a sleep apnea monitoring method, a device, equipment and a storage medium, wherein in the application, physiological data of shallow layers and deep layers are captured by utilizing various wavelength signals corresponding to various light sources, blood oxygen signals are obtained by utilizing the obtained physiological data, then global characteristics representing the change time sequence of a sleep breathing event of a user are obtained by analyzing various physiological data, local characteristics of the blood oxygen signals and different weights, whether the user has a low ventilation event or not is judged by analyzing continuous segment characteristics of a first preset duration in the global characteristics, whether a continuous abnormal signal still appears after the low ventilation event appears, and whether the low ventilation event is accurate or not is verified again by a secondary confirmation mode of the abnormal signal.

Inventors

CHEN JIAHAO
LI YUGAO
CAO QINGXIN

Assignees

深圳市玖治科技有限公司

Dates

Publication Date: 20260512
Application Date: 20260331

Claims (10)

1. A method of sleep apnea monitoring, the method comprising: acquiring PPG signals of multiple color light sources acquired by a wearable device at night; obtaining a blood oxygen signal of a user according to a red light PPG signal and an infrared light PPG signal in the PPG signals of the multiple color light sources; According to local features of PPG signals and blood oxygen signals of the light sources with various colors and feature matrixes constructed for weights configured for different signals, global features for representing the change time sequence of the sleep respiratory event of the user are obtained; Classifying continuous segment features with first preset duration in the global features, and determining whether a low ventilation event occurs in sleep breathing of a user corresponding to the continuous segment features; When a hypopnea event occurs, determining whether an abnormal signal exceeding a second preset duration occurs in the first preset duration in the blood oxygen signal; and when an abnormal signal exceeding a second preset duration appears, the low ventilation event is actually generated, and the sleep apnea of the user at night is determined.
2. The sleep apnea monitoring method of claim 1, wherein the obtaining a blood oxygen signal of the user from the red PPG signal and the infrared PPG signal in the PPG signals of the plurality of color light sources includes: According to the time stamps of the red light PPG signal and the infrared light PPG signal, obtaining a red light PPG signal segment and an infrared light PPG signal segment with third preset duration; For a red light PPG signal segment and an infrared light PPG signal segment, respectively determining peak values and valley values of the red light PPG signal segment and the infrared light PPG signal segment in each sliding window in a preset sliding window and a stepping mode; calculating a blood oxygen value corresponding to each sliding window according to the peak value and the valley value of the red light PPG signal segment and the infrared light PPG signal segment in each sliding window; and interpolating all the obtained blood oxygen values according to a preset sampling frequency to obtain the blood oxygen signal.
3. The sleep apnea monitoring method of claim 1, wherein constructing a feature matrix from local features of PPG signals and blood oxygen signals of multiple color light sources, and weights configured for different signals, comprises: performing maximum and minimum normalization processing on the PPG signals of the light sources with various colors and the blood oxygen signals respectively to obtain a first value of the PPG signals of the light sources with various colors and a second value of the blood oxygen signals; Determining a first local feature of each first value and a second local feature of the second value according to the first value and the second value; Calculating a first weight of each first local feature and a second weight of each second local feature by using a global attention mechanism; Vector stitching is carried out on the product of the first local feature and the corresponding first weight and the product of the second local feature and the second weight, so that the feature matrix is obtained.
4. The sleep apnea monitoring method of claim 1, wherein the deriving global features characterizing a timing of a user's sleep respiratory event changes, comprises: and extracting features of the feature matrix through a time sequence feature model to obtain the global features.
5. The sleep apnea monitoring method of claim 1, wherein classifying consecutive segment features of the global features for a first predetermined duration, determining whether a hypopnea event occurs in a user's sleep breath corresponding to the consecutive segment features, comprises: classifying the continuous segment features through a classification model, and determining a first probability of a normal respiratory event and a second probability of a hypopnea event; And when the second probability is larger than the first probability, determining that the sleep breath of the user corresponding to the continuous segment features has a low ventilation event.
6. The sleep apnea monitoring method of claim 1, wherein the method further comprises: The number of times of sleep apnea occurred per unit time is determined as a sleep apnea hypopnea index.
7. The sleep apnea monitoring method of claim 1, wherein the wearable device comprises a microcontroller, a time-sharing drive circuit, a plurality of color light sources, a photosensor, wherein the acquiring PPG signals of the plurality of color light sources acquired by the wearable device at night comprises: The microcontroller controls the time-sharing driving circuit to adopt a time-sharing multiplexing mode, light sources with corresponding colors are sequentially lightened according to a preset circulation sequence, and current signals of the light sources with corresponding colors are obtained through the photoelectric sensor; according to a preset sampling frequency, digitizing current signals of the light sources with various colors to obtain original PPG signals of the light sources with various colors; and performing low-pass filtering processing on the original PPG signals of the light sources of various colors to obtain PPG signals of the light sources of various colors.
8. A sleep apnea monitoring device, the device comprising: The acquisition unit is used for acquiring PPG signals of the multiple color light sources acquired by the wearable equipment at night; The first determining unit is used for obtaining blood oxygen signals of a user according to red light PPG signals and infrared light PPG signals in the PPG signals of the multiple color light sources; The second determining unit is used for obtaining global features for representing the change time sequence of the sleep respiratory event of the user according to the local features of the PPG signals and the blood oxygen signals of the light sources with various colors and the feature matrix constructed for the weights configured for different signals; A third determining unit, configured to classify continuous segment features of a first preset duration in the global features, and determine whether a hypopnea event occurs in sleep breath of a user corresponding to the continuous segment features; A fourth determining unit, configured to determine, when a hypopnea event occurs, whether an abnormal signal exceeding a second preset duration occurs in the first preset duration in the blood oxygen signal; And a fifth determining unit, configured to determine that the user has sleep apnea at night when the hypopnea event is actually occurred when an abnormal signal exceeding a second preset duration occurs.
9. An electronic device comprising a processor, a storage medium storing machine-readable instructions executable by the processor, the processor and the storage medium in communication over the bus when the electronic device is in operation, the processor executing the machine-readable instructions to perform the steps of the method of any one of claims 1 to 7.
10. A computer-readable storage medium, characterized in that it has stored thereon a computer program which, when executed by a processor, performs the steps of the method according to any of claims 1 to 7.

Description

Sleep apnea monitoring method, device, equipment and storage medium Technical Field The present application relates to the field of data processing technologies, and in particular, to a sleep apnea monitoring method, apparatus, device, and storage medium. Background Sleep apnea Syndrome (SLEEP APNEA Syndrome, SAS) is a common sleep disorder disease characterized by recurrent apneas or hypopneas in a patient during sleep. The disease not only affects sleep quality, but also is an important risk factor for inducing various chronic diseases such as hypertension, heart disease, cerebral apoplexy and the like. Therefore, the method has great significance for early screening and diagnosis of sleep apnea. Currently, clinical diagnosis of sleep apnea depends on polysomnography (Polysomnography, PSG), which is required to be performed in a hospital or a professional sleep center, and synchronously records up to tens of physiological parameters such as brain electricity, eye electricity, electrocardio, myoelectricity, oronasal airflow, blood oxygen saturation, chest-abdomen respiratory effort and the like through a plurality of electrodes and sensors connected to the head face, chest abdomen and limbs of a patient. Although the PSG detection precision is high, the process is complex, the cost is high, the patient is required to fall asleep in a strange environment, the comfort is poor, the accessibility is low, and the PSG detection precision is difficult to be used for daily primary screening and long-term monitoring. Therefore, a solution capable of ensuring high detection accuracy while simultaneously achieving miniaturization, low power consumption and low cost of the device is needed in the art. Disclosure of Invention In view of the above, embodiments of the present application provide a sleep apnea monitoring method, apparatus, device and storage medium, so as to provide a solution that ensures high detection accuracy while simultaneously achieving miniaturization, low power consumption and low cost of the device. In a first aspect, an embodiment of the present application provides a sleep apnea monitoring method, the method including: acquiring PPG signals of multiple color light sources acquired by a wearable device at night; obtaining a blood oxygen signal of a user according to a red light PPG signal and an infrared light PPG signal in the PPG signals of the multiple color light sources; According to local features of PPG signals and blood oxygen signals of the light sources with various colors and feature matrixes constructed for weights configured for different signals, global features for representing the change time sequence of the sleep respiratory event of the user are obtained; Classifying continuous segment features with first preset duration in the global features, and determining whether a low ventilation event occurs in sleep breathing of a user corresponding to the continuous segment features; When a hypopnea event occurs, determining whether an abnormal signal exceeding a second preset duration occurs in the first preset duration in the blood oxygen signal; and when an abnormal signal exceeding a second preset duration appears, the low ventilation event is actually generated, and the sleep apnea of the user at night is determined. In a possible embodiment, the obtaining the blood oxygen signal of the user according to the red light PPG signal and the infrared light PPG signal in the PPG signals of the multiple color light sources includes: According to the time stamps of the red light PPG signal and the infrared light PPG signal, obtaining a red light PPG signal segment and an infrared light PPG signal segment with third preset duration; For a red light PPG signal segment and an infrared light PPG signal segment, respectively determining peak values and valley values of the red light PPG signal segment and the infrared light PPG signal segment in each sliding window in a preset sliding window and a stepping mode; calculating a blood oxygen value corresponding to each sliding window according to the peak value and the valley value of the red light PPG signal segment and the infrared light PPG signal segment in each sliding window; and interpolating all the obtained blood oxygen values according to a preset sampling frequency to obtain the blood oxygen signal. In one possible embodiment, constructing a feature matrix from local features of PPG signals and blood oxygen signals of multiple color light sources, and weights configured for different signals, includes: performing maximum and minimum normalization processing on the PPG signals of the light sources with various colors and the blood oxygen signals respectively to obtain a first value of the PPG signals of the light sources with various colors and a second value of the blood oxygen signals; Determining a first local feature of each first value and a second local feature of the second value according to the first value and the second value