A method and system for predicting OSAS based on hearing aid blood oxygen monitoring and respiratory recording

Abstract

This application discloses an OSAS prediction method and system based on hearing aid blood oxygen monitoring and respiratory recording, relating to the technical field of OSAS prediction. This application includes OSAS prediction model acquisition, user data acquisition, blood oxygen data analysis, respiratory and sleep data analysis, and an execution terminal. First, an OSAS prediction training set is obtained, and then an OSAS prediction model is obtained, so that the OSAS prediction model is deployed in the hearing aid for real-time monitoring and prediction; the user's blood oxygen data is analyzed and divided into multiple condition analyses. If no abnormalities are found, the next step is to analyze the respiratory and sleep data, and predict the user's OSAS, which solves the limitations of the current OSAS prediction development process; when the emergency warning is activated, the user and the user's emergency contacts are notified on the user side, and vibration and ringing instructions are sent to the hearing aid, ensuring the user's safety while predicting OSAS for the user.

Description

Technical Field

[0001] The present application relates to the technical field of OSAS prediction, and in particular to an OSAS prediction method and system based on hearing aid blood oxygen monitoring and breathing recording. Background Art

[0002] With the development of science and technology, more and more people are paying attention to their physical health. OSAS has become well known to everyone, and OSAS prediction technology has become more and more mature.

[0003] Based on the existing technology, the inventors of this application have found that the existing technology has at least the following problems:

[0004] 1. Currently, there is a lack of analysis of the user's blood oxygen data. It is not divided into multiple conditions for analysis, and the next step of breathing and sleep data analysis is not carried out when there are no abnormalities. The prediction of the user's OSAS cannot solve the limitations of the current development of OSAS prediction.

[0005] 2. Currently, there is a lack of analysis of the number of coughs and the duration of each cough, the number of body movements, the number of awakenings and the duration of each awakening, and the use of these and more data to train the model. As a result, it is impossible to truly analyze the user's situation and ensure the user's safety.

[0006] 3. There is currently a lack of detachable versions of hearing aids and blood oxygen sensors, which makes it inconvenient for users to use. Users need to spend more money to purchase blood oxygen sensors separately, which does not increase the applicable age range of users and makes it difficult for the elderly to use them. Summary of the Invention

[0007] In response to the above-mentioned technical deficiencies, the purpose of this application is to provide an OSAS prediction method and system based on hearing aid blood oxygen monitoring and breathing recording.

[0008] To solve the above technical problems, the present application adopts the following technical solution: In a first aspect, the present application provides an OSAS prediction method based on hearing aid blood oxygen monitoring and breathing recording, the method comprising the following steps:

[0009] Step 1: Obtaining an OSAS prediction model: Collect blood oxygen data, respiratory data, and sleep data, clean the sleep respiratory data, and perform feature extraction and feature engineering on the sleep respiratory data to obtain an OSAS prediction training set. This OSAS prediction model is then deployed in hearing aids for real-time monitoring and prediction.

[0010] Step 2: User data acquisition: Obtain the user's sleep recording and blood oxygen data, and then transmit the sleep recording to the OSAS prediction model, thereby transmitting blood oxygen data, respiratory data, and sleep data;

[0011] Step 3: Blood oxygen data analysis: Analyze the user's blood oxygen data to determine the user's blood oxygen assessment coefficient, and then determine whether the user's blood oxygen is abnormal;

[0012] Step 4: Respiratory and sleep data analysis: When the user's blood oxygen levels are normal, the user's respiratory data is analyzed to determine the user's respiratory assessment coefficient and sleep assessment coefficient, and then the user's predicted assessment coefficient is analyzed to determine whether the user's respiratory and sleep conditions are abnormal and predict the user's OSAS;

[0013] Step 5. Execution terminal: When the emergency warning is turned on, the user and the user's emergency contacts are notified on the user side, and a vibration and ringing command is sent to the hearing aid to warn the user of low blood oxygen.

[0014] In a second aspect, the present application provides an OSAS prediction system based on an OSAS prediction method using hearing aid blood oxygen monitoring and respiratory recording, comprising the following modules:

[0015] The OSAS prediction model acquisition module is used to collect blood oxygen data, respiratory data, and sleep data, clean the sleep respiratory data, and perform feature extraction and feature engineering on the sleep respiratory data to obtain the OSAS prediction training set and then obtain the OSAS prediction model, so that the OSAS prediction model can be deployed in the hearing aid for real-time monitoring and prediction;

[0016] A user data acquisition module is used to obtain the user's sleep recording and blood oxygen data, and then transmit the sleep recording to the OSAS prediction model, thereby transmitting the blood oxygen data, respiratory data and sleep data;

[0017] The blood oxygen data analysis module is used to analyze the user's blood oxygen data to obtain the user's blood oxygen assessment coefficient and then determine whether the user's blood oxygen is abnormal;

[0018] The breathing and sleep data analysis module is used to analyze the user's breathing data to obtain the user's breathing assessment coefficient and sleep assessment coefficient when the user's blood oxygen level is normal, and then analyze the user's prediction assessment coefficient to determine whether the user's breathing and sleep conditions are abnormal and predict the user's OSAS;

[0019] The execution terminal is used to notify the user and the user's emergency contacts on the user side when the emergency warning is turned on, and to send vibration and ringing instructions to the hearing aid to warn the user of low blood oxygen.

[0020] The beneficial effects of this application are:

[0021] 1. This application provides an OSAS prediction method and system based on hearing aid blood oxygen monitoring and breathing recording. First, an OSAS prediction training set is obtained, and then an OSAS prediction model is obtained, so that the OSAS prediction model is deployed in the hearing aid for real-time monitoring and prediction. The user's blood oxygen data is analyzed and divided into multiple conditions for analysis. If there is no abnormality, the next step is to analyze the breathing and sleep data, and predict the user's OSAS, which solves the limitations of the current development of OSAS prediction. When the emergency warning is activated, the user and the user's emergency contact are notified on the user side, and vibration and ringing commands are sent to the hearing aid, ensuring the user's safety while predicting OSAS for the user.

[0022] 2. This application completes the collection of blood oxygen data, respiratory data and sleep data, and the data is more comprehensive. Blood oxygen data includes blood oxygen saturation, pulse rate and perfusion index; respiratory data includes sleep recordings and the corresponding respiratory rate, respiratory sound intensity, number of coughs and duration of each cough in the sleep recordings; sleep data includes the number of body movements, number of awakenings and duration of each awakening, which more effectively creates data sets and obtains clear training sets and test sets, laying a solid foundation for subsequent model training.

[0023] 3. The hearing aid and blood oxygen sensor of this application are detachable, which is more convenient for users to use. There is no need to spend more money to purchase a blood oxygen sensor separately. The installation is very convenient. The connection can be completed by connecting the data cable port on the blood oxygen sensor to the charging port of the hearing aid. The data is transmitted by the Bluetooth in the hearing aid. No extra operation is required, which increases the user's applicable age and makes it easy for the elderly to use.

[0024] 4. In this application, blood oxygen is divided into two situations. When the blood oxygen is low, a warning is issued directly to ensure the safety of the user to the greatest extent. When the blood oxygen is normal, further analysis and judgment are made based on the user's other data. When an abnormality occurs, an emergency warning is issued immediately.

[0025] 5. This application analyzes the user's situation in many aspects based on the user's breathing data and sleep data, and then conducts a combined analysis of the user's data to avoid inconvenience to the user due to minor omissions, thereby improving the user experience to a certain extent. BRIEF DESCRIPTION OF THE DRAWINGS

[0026] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without any creative work.

[0027] Figure 1 This is a flowchart of the steps for implementing the application method.

[0028] Figure 2 This is a schematic diagram of the system structure connection for this application. DETAILED DESCRIPTION

[0029] The following will be combined with the drawings in the embodiments of this application to clearly and completely describe the technical solutions in the embodiments of this application. Obviously, the embodiments described are only part of the embodiments of this application, not all of the embodiments. Based on the embodiments in this application, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of this application.

[0030] See also Figure 1 As shown, in a first aspect, the present application provides an OSAS prediction method based on hearing aid blood oxygen monitoring and respiratory recording, comprising the following steps: Step 1, OSAS prediction model acquisition: collecting blood oxygen data, respiratory data, and sleep data, cleaning the sleep respiratory data, and performing feature extraction and feature engineering on the sleep respiratory data to obtain an OSAS prediction training set, and then obtaining an OSAS prediction model, thereby deploying the OSAS prediction model in the hearing aid for real-time monitoring and prediction;

[0031] In a specific example, the OSAS prediction training set is obtained, and the specific acquisition process is as follows:

[0032] Blood oxygen data, respiratory data, and sleep data were collected. Blood oxygen data included blood oxygen saturation, pulse rate, and perfusion index. Respiratory data included sleep recordings and the corresponding respiratory rate, respiratory sound intensity, number of coughs, and duration of each cough. Sleep data included the number of body movements, number of awakenings, and duration of each awakening. The collected blood oxygen data, respiratory data, and sleep data were cleaned, missing values processed, and smoothed. Feature extraction and feature engineering were performed on the blood oxygen data, respiratory data, and sleep data. The blood oxygen data, respiratory data, and sleep data were converted into feature vectors for training, thereby obtaining an OSAS prediction training set.

[0033] In a specific example, the OSAS prediction model is obtained, and thus deployed in a hearing aid for real-time monitoring and prediction. The specific acquisition process is as follows:

[0034] Based on whether the patient has an OSAS tendency, the samples in the OSAS prediction training set are labeled as OSAS positive or negative. The labeled data set is divided into a training set and a test set. The training set is used to train the selected model, and the trained OSAS prediction model is evaluated using the test set. The OSAS prediction model parameters are adjusted based on the evaluation results, and the OSAS prediction model is optimized to obtain the OSAS prediction model, which is then deployed in hearing aids for real-time monitoring and prediction.

[0035] It should be noted that blood oxygen data, respiratory data and sleep data are monitored and recorded in real time through the built-in sensors of the hearing aid; blood oxygen saturation, pulse rate and perfusion index are obtained by connecting the charging port of the hearing aid to the connecting cable of the blood oxygen sensor; respiratory rate, respiratory sound intensity, number of coughs and duration of each cough are obtained by the sound sensor in the hearing aid; the number of awakenings, duration of each awakening and number of body movements are obtained by the gravity sensor.

[0036] It should be noted that using the training set to train the selected model means optimizing the model parameters so that it can better fit the training data set and learn the patterns and rules of the data.

[0037] It should be noted that the test set is used to evaluate the trained model, and indicators such as accuracy, recall rate and F1 value are usually used to evaluate the performance of the model.

[0038] It should be noted that model optimization may involve operations such as parameter adjustment, feature selection, and model fusion to improve model performance and generalization capabilities.

[0039] This application completes the collection of blood oxygen data, respiratory data and sleep data, and the data is more comprehensive. Blood oxygen data includes blood oxygen saturation, pulse rate and perfusion index; respiratory data includes sleep recordings and the corresponding respiratory rate, respiratory sound intensity, number of coughs and duration of each cough in the sleep recordings; sleep data includes the number of body movements, number of awakenings and duration of each awakening, which more effectively creates data sets and obtains clear training sets and test sets, laying a solid foundation for subsequent model training.

[0040] Step 2: User data acquisition: Obtain the user's sleep recording and blood oxygen data, and then transmit the sleep recording to the OSAS prediction model, thereby transmitting blood oxygen data, respiratory data, and sleep data;

[0041] In a specific example, the sleep recording and blood oxygen data of the user are obtained in the following specific acquisition process:

[0042] The user selects the OSAS prediction option in the personal center, and connects the blood oxygen sensor cable to the charging port of the hearing aid according to the text and picture prompts, and clips the blood oxygen sensor clip on the earlobe. After successful activation, OSAS prediction is turned on to obtain the user's sleep recording and blood oxygen data.

[0043] The hearing aid and blood oxygen sensor of this application are detachable, which is more convenient for users to use. There is no need to spend more money to purchase a blood oxygen sensor separately. The installation is very convenient. The connection can be completed by connecting the data cable port on the blood oxygen sensor to the charging port of the hearing aid. The data is transmitted by the Bluetooth in the hearing aid. No extra operation is required, which increases the user's applicable age and makes it easy for the elderly to use.

[0044] Step 3: Blood oxygen data analysis: Analyze the user's blood oxygen data to determine the user's blood oxygen assessment coefficient, and then determine whether the user's blood oxygen is abnormal;

[0045] In a specific example, the analysis obtains the user's blood oxygen assessment coefficient, and the specific analysis process is as follows:

[0046] Denote blood oxygen saturation, pulse rate, and perfusion index as X, M, and G, respectively, and substitute them into the calculation formula The user's blood oxygen assessment coefficient β is obtained, where X′, M′, and G′ represent the reference values of blood oxygen saturation, pulse rate, and perfusion index in the database, respectively. σ1 represents the weight factors corresponding to the blood oxygen saturation and perfusion index in the database, and σ2 represents the weight factor corresponding to the pulse rate in the database.

[0047] It should be noted that 0<σ1≤1, 0<σ2≤1, σ1+σ2=1.

[0048] It should be noted that the weight factors corresponding to blood oxygen saturation and perfusion index and the weight factors corresponding to pulse rate were obtained by factor analysis. First, the information of blood oxygen saturation, perfusion index and pulse rate was condensed, and then the variance explanation rate after rotation was obtained. The weights were obtained by dividing the cumulative variance explanation rate.

[0049] It should be noted that factor analysis is a well-known technology. It is a multivariate statistical analysis method that starts from studying the internal dependencies of variables and reduces some variables with complex relationships to a few comprehensive factors; information concentration is expressed as calculating the median; the variance explanation rate is the amount of information extracted by the factor, and the variance explanation rate = characteristic root / total number of analysis items; the variance explanation rate after rotation is expressed as the variance explanation rate of the factor after maximum variance rotation.

[0050] In a specific example, the process of determining whether the user's blood oxygen level is abnormal is as follows:

[0051] A1. When the user's blood oxygen assessment coefficient is 0, an emergency warning will be issued immediately;

[0052] A2. When the user's blood oxygen assessment coefficient is not 0, the user's blood oxygen assessment coefficient is compared with the user's blood oxygen assessment coefficient threshold in the database. When the user's blood oxygen assessment coefficient is less than or equal to the user's blood oxygen assessment coefficient threshold in the database, it is determined that the user's blood oxygen is abnormal and an emergency warning is immediately issued; when the user's blood oxygen assessment coefficient is greater than the user's blood oxygen assessment coefficient threshold in the database, it is determined that the user's blood oxygen is normal and the next step is performed.

[0053] In this application, blood oxygen is divided into two situations. When the blood oxygen is low, a warning is issued directly to ensure the safety of the user to the greatest extent. When there is no abnormality in the blood oxygen, further analysis and judgment are made based on the rest of the user's data. When an abnormality occurs, an emergency warning is issued immediately.

[0054] Step 4: Respiratory and sleep data analysis: When the user's blood oxygen levels are normal, the user's respiratory data is analyzed to determine the user's respiratory assessment coefficient and sleep assessment coefficient, and then the user's predicted assessment coefficient is analyzed to determine whether the user's respiratory and sleep conditions are abnormal and predict the user's OSAS;

[0055] In a specific example, the analysis obtains the user's breathing assessment coefficient, and the specific analysis process is as follows:

[0056] The corresponding respiratory rate, respiratory sound intensity, number of coughs and duration of each cough in the sleep recording were recorded as HP, HQ, KC and KS respectively. i ; Where i represents the number corresponding to each cough, i = 0, 1, 2, ... KC, KC is an integer greater than or equal to 0;

[0057] According to the calculation formula The user's breathing assessment coefficient λ is obtained, where KS0=0, HP′, HQ′ and KS′ represent the reference values of breathing frequency, breathing sound intensity and cough duration in the database, respectively, ΔKS represents the allowed floating value of cough duration in the database, ω1 and ω2 represent the weight factors corresponding to breathing frequency and breathing sound intensity and the weight factor corresponding to cough duration in the database, respectively.

[0058] It should be noted that 0<ω1≤1, 0<ω2≤1, ω1+ω2=1.

[0059] It should be noted that the weight factors corresponding to the respiratory rate and respiratory sound intensity and the weight factor corresponding to the duration of cough were obtained through factor analysis. First, the information of the respiratory rate, respiratory sound intensity and cough duration was condensed, and then the variance explanation rate after rotation was obtained. The weight was obtained by dividing the cumulative variance explanation rate.

[0060] It should be noted that factor analysis is a well-known technology. It is a multivariate statistical analysis method that starts from studying the internal dependencies of variables and reduces some variables with complex relationships to a few comprehensive factors; information concentration is expressed as calculating the median; the variance explanation rate is the amount of information extracted by the factor, and the variance explanation rate = characteristic root / total number of analysis items; the variance explanation rate after rotation is expressed as the variance explanation rate of the factor after maximum variance rotation.

[0061] In a specific example, the analysis obtains the user's sleep assessment coefficient, and the specific analysis process is as follows:

[0062] The number of body movements, number of awakenings, and duration of each awakening in the sleep data are recorded as TD, QC, and QX respectively. j , where j represents the number corresponding to each awakening, j = 0, 1, 2, ... QC, QC is an integer greater than or equal to 0;

[0063] According to the calculation formula Get the user's sleep assessment coefficient Where QX0=0, TD represents the reference value of the number of body movements in the database, QX′ represents the reference value corresponding to the duration of wakefulness in the database, ΔQX represents the allowable floating value corresponding to the duration of wakefulness in the database, υ1 and υ2 represent the weight factors corresponding to the number of body movements and the duration of wakefulness in the database, respectively.

[0064] It should be noted that 1<υ1≤1, 1<υ2≤1, υ1+υ2=1.

[0065] It should be noted that the weight factors corresponding to the number of body movements and the weight factors corresponding to the duration of wakefulness were obtained through factor analysis. First, the information of the number of body movements and the duration of wakefulness was condensed, and then the variance explanation rate after rotation was obtained. The weight was obtained by dividing the cumulative variance explanation rate.

[0066] It should be noted that factor analysis is a well-known technology. It is a multivariate statistical analysis method that starts from studying the internal dependencies of variables and reduces some variables with complex relationships to a few comprehensive factors; information concentration is expressed as calculating the median; the variance explanation rate is the amount of information extracted by the factor, and the variance explanation rate = characteristic root / total number of analysis items; the variance explanation rate after rotation is expressed as the variance explanation rate of the factor after maximum variance rotation.

[0067] In a specific example, the analysis obtains the user's breathing assessment coefficient and sleep assessment coefficient, and then further analyzes the user's prediction assessment coefficient, thereby determining whether the user's breathing and sleep conditions are abnormal and predicting the user's OSAS. The specific analysis and judgment process is as follows:

[0068] B1. According to the calculation formula The user's prediction evaluation coefficient α is obtained through analysis, where μ1 and μ2 represent the weight factors corresponding to the user's breathing evaluation coefficient and sleep evaluation coefficient in the database, respectively.

[0069] B2. Compare the user's prediction evaluation coefficient with the lower limit and upper limit of the prediction evaluation coefficients of users in the database. When the user's prediction evaluation coefficient is less than the lower limit of the prediction evaluation coefficient of users in the database, it is judged that the user's breathing and sleep conditions are abnormal, and the user is predicted to develop OSAS, and an emergency warning is immediately issued; when the user's prediction evaluation coefficient is greater than the upper limit of the prediction evaluation coefficient of users in the database, it is judged that the user's breathing and sleep conditions are not abnormal, and the user is predicted not to develop OSAS; when the user's prediction evaluation coefficient is less than the lower limit of the prediction evaluation coefficient of users in the database and greater than the upper limit, it is judged that the user's breathing and sleep conditions are about to become abnormal, and the user is predicted to develop OSAS.

[0070] It should be noted that 0<μ1≤1, 0<μ2≤1, μ1+μ2=1.

[0071] It should be noted that the weight factors corresponding to the user's breathing assessment coefficient and the sleep assessment coefficient are calculated in the same way as the weight factors corresponding to the number of body movements and the length of time awake.

[0072] In this application, after analyzing the user's situation in many aspects based on the user's breathing data and sleep data, the user's data is combined and analyzed to avoid inconvenience to the user due to minor omissions, thereby improving the user experience to a certain extent.

[0073] Step 5. Execution terminal: When the emergency warning is turned on, the user and the user's emergency contacts are notified on the user side, and a vibration and ringing command is sent to the hearing aid to warn the user of low blood oxygen.

[0074] See also Figure 2As shown, in a second aspect, the present application provides an OSAS prediction system based on an OSAS prediction method using hearing aid blood oxygen monitoring and respiratory recording, comprising the following modules: an OSAS prediction model acquisition module for collecting blood oxygen data, respiratory data, and sleep data, cleaning the sleep respiratory data, and performing feature extraction and feature engineering on the sleep respiratory data to obtain an OSAS prediction training set, thereby obtaining an OSAS prediction model, and deploying the OSAS prediction model in the hearing aid for real-time monitoring and prediction;

[0075] A user data acquisition module is used to obtain the user's sleep recording and blood oxygen data, and then transmit the sleep recording to the OSAS prediction model, thereby transmitting the blood oxygen data, respiratory data and sleep data;

[0076] The blood oxygen data analysis module is used to analyze the user's blood oxygen data to obtain the user's blood oxygen assessment coefficient and then determine whether the user's blood oxygen is abnormal;

[0077] The breathing and sleep data analysis module is used to analyze the user's breathing data to obtain the user's breathing assessment coefficient and sleep assessment coefficient when the user's blood oxygen level is normal, and then analyze the user's prediction assessment coefficient to determine whether the user's breathing and sleep conditions are abnormal and predict the user's OSAS;

[0078] The execution terminal is used to notify the user and the user's emergency contacts on the user side when the emergency warning is turned on, and to send vibration and ringing instructions to the hearing aid to warn the user of low blood oxygen.

[0079] The present application provides an OSAS prediction method and system based on hearing aid blood oxygen monitoring and breathing recording. The method first obtains an OSAS prediction training set, then obtains an OSAS prediction model, and then deploys the OSAS prediction model into the hearing aid for real-time monitoring and prediction. The method then analyzes the user's blood oxygen data and divides it into multiple conditions for analysis. If no abnormalities are found, the method proceeds to analyze the breathing and sleep data and predicts the user's OSAS, thus resolving the limitations of the current development of OSAS prediction. When an emergency warning is activated, the method notifies the user and the user's emergency contacts on the user side, and sends vibration and ringing commands to the hearing aid, ensuring the user's safety while predicting OSAS for the user.

[0080] The above content is merely an example and explanation of the concept of the present application. Technicians in this technical field may make various modifications or additions to the specific embodiments described or replace them in a similar manner. As long as they do not deviate from the concept of the invention or exceed the scope defined in this specification, they should all fall within the scope of protection of the present application.

Claims

1. A method for predicting OSAS based on hearing aid blood oxygen monitoring and breathing recording, characterized in that: include: Step 1: Obtaining an OSAS prediction model: Collect blood oxygen data, respiratory data, and sleep data, clean the sleep respiratory data, and perform feature extraction and feature engineering on the sleep respiratory data to obtain an OSAS prediction training set. This OSAS prediction model is then deployed in hearing aids for real-time monitoring and prediction. Step 2: User data acquisition: Obtain the user's sleep recording and blood oxygen data, and then transmit the sleep recording to the OSAS prediction model, thereby transmitting blood oxygen data, respiratory data, and sleep data; Step 3: Blood oxygen data analysis: Analyze the user's blood oxygen data to determine the user's blood oxygen assessment coefficient, and then determine whether the user's blood oxygen is abnormal; The analysis results in the user's blood oxygen assessment coefficient. The specific analysis process is as follows: Blood oxygen saturation, pulse rate, and perfusion index were recorded as 、 and , substitute into the calculation formula Get the user's blood oxygen assessment coefficient ,in 、 and They are respectively represented by the reference values of blood oxygen saturation, pulse rate and perfusion index in the database. It is expressed as the weight factor corresponding to the blood oxygen saturation and perfusion index in the database, It is represented as the weight factor corresponding to the pulse rate in the database; The specific process of determining whether the user's blood oxygen is abnormal is as follows: A1. When the user's blood oxygen assessment coefficient is 0, an emergency warning will be issued immediately; A2. When the user's blood oxygen assessment coefficient is non-zero, the user's blood oxygen assessment coefficient is compared with the user's blood oxygen assessment coefficient threshold in the database. When the user's blood oxygen assessment coefficient is less than or equal to the user's blood oxygen assessment coefficient threshold in the database, the user's blood oxygen is judged to be abnormal and an emergency warning is immediately issued; When the user's blood oxygen assessment coefficient is greater than the user's blood oxygen assessment coefficient threshold in the database, it is determined that the user's blood oxygen is normal and the next step is performed; Step 4: Respiratory and sleep data analysis: When the user's blood oxygen levels are normal, the user's respiratory data is analyzed to determine the user's respiratory assessment coefficient and sleep assessment coefficient, and then the user's predicted assessment coefficient is analyzed to determine whether the user's respiratory and sleep conditions are abnormal and predict the user's OSAS; Step 5. Execution terminal: When the emergency warning is turned on, the user and the user's emergency contacts are notified on the user side, and a vibration and ringing command is sent to the hearing aid to warn the user of low blood oxygen.

2. The OSAS prediction method based on hearing aid blood oxygen monitoring and breathing recording according to claim 1, characterized in that: The OSAS prediction training set is obtained, and the specific acquisition process is as follows: Blood oxygen data, respiratory data, and sleep data were collected. Blood oxygen data included blood oxygen saturation, pulse rate, and perfusion index. Respiratory data included sleep recordings and the corresponding respiratory rate, respiratory sound intensity, number of coughs, and duration of each cough. Sleep data included the number of body movements, number of awakenings, and duration of each awakening. The collected blood oxygen data, respiratory data, and sleep data were cleaned, missing values processed, and smoothed. Feature extraction and feature engineering were performed on the blood oxygen data, respiratory data, and sleep data. The blood oxygen data, respiratory data, and sleep data were converted into feature vectors for training, thereby obtaining an OSAS prediction training set.

3. The OSAS prediction method based on hearing aid blood oxygen monitoring and breathing recording according to claim 2, characterized in that: The OSAS prediction model is obtained, and thus deployed in the hearing aid for real-time monitoring and prediction. The specific acquisition process is as follows: Based on whether the patient has an OSAS tendency, the samples in the OSAS prediction training set are labeled as OSAS positive or negative. The labeled data set is divided into a training set and a test set. The training set is used to train the selected model, and the trained OSAS prediction model is evaluated using the test set. The OSAS prediction model parameters are adjusted based on the evaluation results, and the OSAS prediction model is optimized to obtain the OSAS prediction model, which is then deployed in hearing aids for real-time monitoring and prediction.

4. The OSAS prediction method based on hearing aid blood oxygen monitoring and breathing recording according to claim 1, characterized in that: The specific process of obtaining the user's sleep recording and blood oxygen data is as follows: The user selects the OSAS prediction option in the personal center, and connects the blood oxygen sensor cable to the charging port of the hearing aid according to the text and picture prompts, and clips the blood oxygen sensor clip on the earlobe. After successful activation, OSAS prediction is turned on to obtain the user's sleep recording and blood oxygen data.

5. The OSAS prediction method based on hearing aid blood oxygen monitoring and breathing recording according to claim 4, characterized in that: The analysis results in a user's breathing assessment coefficient, and the specific analysis process is as follows: The corresponding respiratory frequency, respiratory sound intensity, number of coughs and duration of each cough in the sleep recording were recorded as 、 、 and ;in It is represented by the number corresponding to each cough. , is an integer greater than or equal to 0; According to the calculation formula Get the user's breathing assessment coefficient ,in , 、 and They are respectively represented as the reference value of respiratory frequency, the reference value of respiratory sound intensity and the reference value of cough duration in the database, The allowed floating value of the cough duration in the database is expressed as: and They are respectively represented as the weight factors corresponding to the respiratory frequency and respiratory sound intensity in the database and the weight factor corresponding to the duration of coughing.

6. The OSAS prediction method based on hearing aid blood oxygen monitoring and breathing recording according to claim 5, characterized in that: The analysis results in a user's sleep assessment coefficient, and the specific analysis process is as follows: The number of body movements, number of awakenings, and duration of each awakening in the sleep data are recorded as 、 and ,in It is represented by the number corresponding to each awakening. , is an integer greater than or equal to 0; According to the calculation formula Get the user's sleep assessment coefficient ,in , Represents the reference value of the number of body movements in the database, It is represented as the reference value corresponding to the awake time in the database. It is represented by the floating value allowed for the duration of wakefulness in the database. and They are respectively represented as the weight factor corresponding to the number of body movements in the database and the weight factor corresponding to the duration of wakefulness.

7. The OSAS prediction method based on hearing aid blood oxygen monitoring and breathing recording according to claim 6, characterized in that: The analysis obtains the user's breathing assessment coefficient and sleep assessment coefficient, and then further analyzes the user's prediction assessment coefficient, thereby determining whether the user's breathing and sleep conditions are abnormal and predicting the user's OSAS. The specific analysis and judgment process is as follows: B1. According to the calculation formula Analyze and obtain the user's prediction evaluation coefficient ,in and They are respectively represented as the weight factor corresponding to the breathing assessment coefficient and the weight factor corresponding to the sleep assessment coefficient of the user in the database; B2. Comparing the user's prediction evaluation coefficient with the lower and upper limits of the prediction evaluation coefficients of users in the database. When the user's prediction evaluation coefficient is less than the lower limit of the prediction evaluation coefficient of users in the database, it is determined that the user's breathing and sleeping conditions are abnormal, and the user is predicted to develop OSAS, and an emergency warning is immediately issued; When the user's prediction evaluation coefficient is greater than the upper limit of the prediction evaluation coefficients of users in the database, it is judged that the user's breathing and sleep conditions are normal, and it is predicted that the user will not develop OSAS; when the user's prediction evaluation coefficient is less than the lower limit of the prediction evaluation coefficients of users in the database and greater than the upper limit, it is judged that the user's breathing and sleep conditions are about to become abnormal, and it is predicted that the user is about to develop OSAS.

8. An OSAS prediction system that implements the OSAS prediction method based on hearing aid blood oxygen monitoring and breathing recording according to any one of claims 1 to 7, characterized in that: include: The OSAS prediction model acquisition module is used to collect blood oxygen data, respiratory data, and sleep data, clean the sleep respiratory data, and perform feature extraction and feature engineering on the sleep respiratory data to obtain the OSAS prediction training set and then obtain the OSAS prediction model, so that the OSAS prediction model can be deployed in the hearing aid for real-time monitoring and prediction; A user data acquisition module is used to obtain the user's sleep recording and blood oxygen data, and then transmit the sleep recording to the OSAS prediction model, thereby transmitting the blood oxygen data, respiratory data and sleep data; The blood oxygen data analysis module is used to analyze the user's blood oxygen data to obtain the user's blood oxygen assessment coefficient and then determine whether the user's blood oxygen is abnormal; The analysis results in the user's blood oxygen assessment coefficient. The specific analysis process is as follows: Blood oxygen saturation, pulse rate, and perfusion index were recorded as 、 and , substitute into the calculation formula Get the user's blood oxygen assessment coefficient ,in 、 and They are respectively represented by the reference values of blood oxygen saturation, pulse rate and perfusion index in the database. It is expressed as the weight factor corresponding to the blood oxygen saturation and perfusion index in the database, It is represented as the weight factor corresponding to the pulse rate in the database; The specific process of determining whether the user's blood oxygen is abnormal is as follows: A1. When the user's blood oxygen assessment coefficient is 0, an emergency warning will be issued immediately; A2. When the user's blood oxygen assessment coefficient is non-zero, the user's blood oxygen assessment coefficient is compared with the user's blood oxygen assessment coefficient threshold in the database. When the user's blood oxygen assessment coefficient is less than or equal to the user's blood oxygen assessment coefficient threshold in the database, the user's blood oxygen is judged to be abnormal and an emergency warning is immediately issued; When the user's blood oxygen assessment coefficient is greater than the user's blood oxygen assessment coefficient threshold in the database, it is determined that the user's blood oxygen is normal and the next step is performed; The breathing and sleep data analysis module is used to analyze the user's breathing data to obtain the user's breathing assessment coefficient and sleep assessment coefficient when the user's blood oxygen level is normal, and then analyze the user's prediction assessment coefficient to determine whether the user's breathing and sleep conditions are abnormal and predict the user's OSAS; The execution terminal is used to notify the user and the user's emergency contacts on the user side when the emergency warning is turned on, and to send vibration and ringing instructions to the hearing aid to warn the user of low blood oxygen.

Images