Stress prediction method for hearing aid system based on real-time monitoring of hormone levels

Abstract

The embodiment of the present invention discloses a stress prediction method for a hearing aid system based on real-time monitoring of hormone levels, and relates to the field of hearing aid devices and algorithm technology. The method comprises: identifying the type of stress curve of the user in a specific time period according to the sound signals obtained by the user's hearing aid in multiple historical time periods, and the basic information of the user, wherein the specific time period covers the multiple historical time periods and future time periods; matching the existing pressure curve in the specific time period with the standard pressure curve of the pressure curve type in segments, and using the curve after the segmentation that matches the existing pressure curve in the standard pressure curve as the user's future pressure curve, wherein the existing pressure curve is monitored by the hearing aid system. This embodiment can monitor and predict the user's psychological stress level in real time.

Description

Technical Field

[0001] The embodiments of the present invention relate to the field of hearing aid devices and algorithm technology, and more particularly to a pressure prediction method for a hearing aid system based on real-time monitoring of hormone levels. Background Art

[0002] At present, people pay more attention to mental health while pursuing physical health. Stress control is an important part of mental health. By testing the human hormone level, we can understand our own stress status, relax in time, and seek medical treatment when it is serious. It has a good effect on the control of mental illness.

[0003] However, existing hormone detection devices are inconvenient to carry and cannot achieve real-time monitoring of hormone levels and psychological stress levels; at the same time, they lack records and analysis, making it difficult to predict future stress trends. Summary of the invention

[0004] An embodiment of the present invention provides a pressure prediction method for a hearing aid system based on real-time monitoring of hormone levels to solve at least one of the above problems.

[0005] In a first aspect, an embodiment of the present invention provides a stress prediction method for a hearing aid system based on real-time monitoring of hormone levels, which is applied to a hearing aid system for real-time monitoring of hormone levels, wherein the hearing aid system includes a hearing aid, a flexible sweat sensor patch, and a disease monitoring system, and the cooperation of each part can monitor the hormone level and psychological stress level of the user in real time;

[0006] The method comprises:

[0007] Identifying the type of the user's pressure curve in a specific period of time according to the sound signals acquired by the user's hearing aid in multiple historical periods of time and the user's basic information, wherein the specific period of time covers the multiple historical periods of time and the future period of time;

[0008] The existing pressure curve in the specific time period is matched segmentally with the standard pressure curve of the pressure curve type, and the curve after the segment matching the existing pressure curve in the standard pressure curve is used as the user's future pressure curve, wherein the existing pressure curve is monitored by the hearing aid system.

[0009] This method reflects the user's pressure change pattern through a pressure curve over a period of time, and extracts normalized information with little correlation with the time domain position from the dynamic sound signal extracted by the hearing aid through a pressure curve type recognition model, and determines the type of pressure change of the user over a period of time after multimodal fusion with the user's basic information; finally, the pressure trend in the future is predicted through the standard pressure change curve under this type. Compared with short-term pressure prediction, such as pressure prediction at a single moment, this embodiment pays more attention to the change pattern of pressure over a period of time, and through clustering, learning and identification of pressure curves, the prediction time domain is longer, which is more conducive to early intervention or guidance of the user's psychological stress state.

[0010] In a second aspect, an embodiment of the present invention further provides a hearing aid system for real-time monitoring of hormone levels, characterized in that it includes a hearing aid, a flexible sweat sensor patch, and a disease monitoring system:

[0011] The outer side of the flexible sweat sensor patch is used to adhere to the skin behind the ear, and the inner side is attached to the outer surface of the hearing aid, and is electrically connected to the hearing aid chip of the hearing aid through contact electrodes;

[0012] The flexible sweat sensor patch includes a sweat collection hole, an electrochemical electrode and a microchannel; after the hearing aid is worn, the sweat of the skin behind the ear flows into the microchannel through the sweat collection hole, and a corresponding potential is applied to the electrochemical electrode by differential pulse voltammetry, so that the sweat in the microchannel can undergo an electrochemical reaction, and the hormone level is detected by the electrochemical reaction and transmitted to the hearing aid chip for processing; wherein the hormones to be detected include tryptophan and cortisol;

[0013] The disease monitoring system records the detection data, analyzes the user's psychological stress level according to the hormone level, and links the hearing aid to relieve the pressure in the ear when the data is abnormal.

[0014] The hearing aid system combines the advantages of portable hearing aids and non-invasive sweat sensors to achieve the following beneficial effects:

[0015] 1. Traditional hormone testing is mostly done by blood sampling with needles. Long-term invasive testing increases the user's physical pain and easily leads to psychological resistance. This embodiment infers the human hormone level by detecting the hormone level in sweat to assist in stress management.

[0016] 2. As the existing hormone detection devices are inconvenient to carry, hearing aids are the preferred product to solve hearing problems. They have the advantages of miniaturization, portability and unified audience. Combining sweat sensor detection devices with hearing aids can achieve real-time management of psychological stress;

[0017] 3. In view of the problem that conventional self-examination in existing technologies lacks records and analysis, and it is difficult to independently judge the direction of the disease, the real-time monitoring of hormone levels in the hearing aid system with a real-time warning system can be used for health management, real-time recording of hormone levels and disease analysis, auxiliary treatment, and real-time warning. At the same time, with the help of hearing aids, the user's hearing condition is improved to solve the problem of users not being able to hear or not being able to hear clearly. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] In order to more clearly illustrate the specific implementation methods of the present invention or the technical solutions in the prior art, the drawings required for use in the specific implementation methods or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are some implementation methods of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0019] Figure 1 is a functional structure diagram of a hearing aid system for real-time monitoring of hormone levels provided by an embodiment of the present invention;

[0020] Figure 2 This is a main structural diagram of a hearing aid system for real-time monitoring of hormone levels provided by an embodiment of the present invention;

[0021] Figure 3 is a diagram of an installation method of a hearing aid and a sweat sensor provided by an embodiment of the present invention;

[0022] Figure 4 is a schematic structural diagram of a flexible sweat sensor patch provided by an embodiment of the present invention;

[0023] Figure 5 This is a schematic diagram of a method of stimulating sweating by iontophoresis provided by an embodiment of the present invention;

[0024] Figure 6 is a schematic diagram of a differential pulse voltammetry method provided by an embodiment of the present invention;

[0025] Figure 7 is a flow chart of a method for predicting pressure of a hearing aid system based on real-time monitoring of hormone levels provided by an embodiment of the present invention;

[0026] Figure 8 It is a data processing schematic diagram of a pressure curve type identification model provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0027] In order to make the purpose, technical solution and advantages of the present invention clearer, the technical solution of the present invention will be described clearly and completely below. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work belong to the scope of protection of the present invention.

[0028] In the description of the present invention, it should be noted that the terms "center", "upper", "lower", "left", "right", "vertical", "horizontal", "inner", "outer", etc., indicating the orientation or positional relationship, are based on the orientation or positional relationship shown in the drawings, and are only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore cannot be understood as limiting the present invention. In addition, the terms "first", "second", and "third" are used for descriptive purposes only, and cannot be understood as indicating or implying relative importance.

[0029] In the description of the present invention, it is also necessary to explain that, unless otherwise clearly specified and limited, the terms "installed", "connected", and "connected" should be understood in a broad sense, for example, it can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can be a direct connection, or it can be indirectly connected through an intermediate medium, or it can be the internal communication of two components. For ordinary technicians in this field, the specific meanings of the above terms in the present invention can be understood according to specific circumstances.

[0030] The hormones to be detected in this embodiment include cortisol and tryptophan. The relationship between the detection of these two types of hormones and stress management is as follows:

[0031] Cortisol is a stress hormone (commonly known as "stress hormone"), a chemical substance secreted by the human body in an emergency, which causes physiological reactions such as accelerated heart rate, increased blood pressure, faster breathing, and muscle tension. Under the influence of stress, anxiety, and fear, the secretion of cortisol increases.

[0032] Tryptophan can participate in the renewal of plasma proteins in the human body and promote the action of vitamin B2. It is an essential amino acid for the human body and is also a necessary synthetic substance for serotonin and melatonin. Testing the concentration of tryptophan in sweat can effectively assess physical condition and mental stress. Among them, serotonin is converted by tryptophan in the brain and participates in mood regulation and nerve signal transmission. Its content is closely related to depression and anxiety. At the same time, tryptophan can participate in the synthesis of melatonin in the human body. Melatonin can regulate the human sleep system and improve human diet.

[0033] Reference Figure 1-Figure 5The embodiment of the present invention provides a hearing aid system for real-time monitoring of hormone levels, which can monitor the concentration of tryptophan and cortisol in the human body in real time. Specifically, the system includes: a hearing aid, a flexible sweat sensor patch and a disease monitoring system.

[0034] The hearing aid includes at least one microphone, a hearing aid chip and a receiver. The microphone is a high-precision microphone, which is used to receive external sounds into the hearing aid and transmit them to the hearing aid chip. The hearing aid chip integrates a hearing aid algorithm and a hormone analysis algorithm. The hearing aid algorithm can optimize the sound and output it (for example, amplify, reduce, denoise, etc. the sound and then output it). The hormone analysis algorithm will be described in the flexible sweat sensor patch section. The receiver is used to output the sound into the user's ear to solve the hearing loss problem.

[0035] The flexible sweat sensor patch includes sweat collection holes, electrochemical electrodes (i.e. Figure 4 The outer side of the flexible sweat sensor patch is used to adhere to the skin behind the ear, and the inner side is attached to the outer surface of the hearing aid, and is electrically connected to the hearing aid chip of the hearing aid through the contact electrode. For example, Figure 2 , Figure 3 As shown, the inner side of the flexible sweat sensor patch is connected to the hearing aid host through contact electrodes for power supply and communication, and is fixed to the back of the hearing aid through adhesive. After the hearing aid is worn, the outer side of the flexible sweat sensor patch is attached to the skin behind the ear. Figure 4 , the sweat from the skin behind the ear flows into the microchannel through the sweat collection hole, the sweat in the microchannel contacts the electrochemical electrode, and an electrochemical reaction occurs under the action of the electrochemical electrode. The hormone level can be detected through the electrochemical reaction and transmitted to the hearing aid chip for processing. At the same time, an exhaust port is provided at the end of the microchannel, and waste liquid is converted into gas through electrochemical reaction and heating in the microchannel, which is discharged through the exhaust port. Optionally, the electrochemical electrode can be a nickel-gold alloy electrode. In the hormone analysis algorithm, the hormone level in sweat can be converted into the hormone level in the human body according to the corresponding relationship. The specific process is a prior art and will not be repeated here.

[0036] More specifically, the main body of the flexible sweat sensor patch is a microfluidic chip system. Microfluidic chip, also known as Lab-on-chip, is a science and technology that is mainly characterized by the manipulation of fluids in micrometer-scale space. It has good biocompatibility, electrical insulation, heat dissipation, optical properties, etc., and its surface has good modifiability. The sensor made based on this principle has the characteristics of high sensitivity, fast response speed and small size. The sensor of this embodiment uses polydimethylsiloxane as the base material, depicts microchannels to collect sweat, and detects hormone levels through electrical and chemical reactions.

[0037] Furthermore, in a specific implementation, the sweat sensor can be divided into a sweat monitoring system, a sweat removal system, a sweat collection system, and a sweat analysis system from a functional perspective.

[0038] Sweat monitoring system: The microchannel substrate is made of polydimethylsiloxane material with a rough surface. Different refractive indexes of visible light will cause the microchannel to change color in the empty state and the water-filled state, so as to detect the sweating of the human body in real time and analyze and judge whether the detection state is met. When there is a detection demand but the sweat flow of the human body is not enough to support the detection, active sweating will be started. Optional, such as Figure 5 As shown, iontophoresis can be used, where the electrochemical electrodes act as iontophoresis electrodes, and a voltage is applied between the iontophoresis electrodes to generate an electric current under the skin surface, thereby stimulating sweat secretion.

[0039] Sweat discharge system: The liquid in the microchannel undergoes electrochemical reaction and heating at the electrochemical electrode to generate gas, which is then discharged from the sensor through the discharge hole.

[0040] Sweat collection system: Sweat behind the ear enters the sensor through the unidirectional sweat collection hole and is absorbed by the microchannel.

[0041] Sweat analysis system: After the sweat is absorbed into the microchannel, an electrochemical reaction occurs under the action of the electrochemical electrodes. The levels of tryptophan and cortisol can be measured in the microchannel through the electrochemical reaction and transmitted to the internal chip of the hearing aid host for processing. Specifically, the hormone detection principle of this embodiment is as follows: tryptophan is an electroactive substance and will be oxidized under a specific voltage. Cortisol is a non-electroactive substance, so Prussian blue needs to be added as a redox probe. The content can be measured using differential pulse voltammetry. In order to illustrate this method, cyclic voltammetry is first introduced. Cyclic voltammetry is an electrochemical analysis method in which a triangular wave potential is applied to the working electrode, such as Figure 6 As shown in the figure, the current potential curve obtained includes two parts, one half-wave potential scans toward the cathode, so that the active substance is reduced on the electrode to obtain electrons, generating a reduction waveform, and the other half-wave potential scans toward the anode, and the reduction product on the electrode loses electrons and oxidizes, generating an oxidation waveform. One triangle wave scan completes a cycle of reduction and oxidation. Differential pulse voltammetry is an improvement on cyclic voltammetry. It can reduce the impact of current generated by non-analytes and increase detection sensitivity. A continuous potential function is applied to the electrode system in the microchannel, and the electron movement of the oxidation reaction will generate a current of a specific waveform. The concentration of tryptophan and cortisol is analyzed based on the change in the generated current.

[0042] The disease monitoring system can record and analyze the measured data, and when the data is abnormal, it can link the hearing aid to relieve the pressure in the ear or warn. Among them, the disease monitoring system can be deployed on a mobile phone or other electronic devices. In a specific embodiment, the disease monitoring system can be divided into a data processing system, a data monitoring system and a disease warning system. Among them, the data processing system can record and analyze hormone data, such as converting the measured hormone levels into quantitative data of the user's psychological stress level; the data monitoring system can analyze the health status according to the processed data, form a health report, and provide a reference for psychological stress management; and the disease warning system will link the hearing aid to relieve the pressure in the ear or remind when the data is abnormal, and warn the bound relatives or doctors through the mobile phone APP, so as to carry out stress relief or examination and treatment in time.

[0043] In summary, this embodiment provides a hearing aid system for real-time monitoring of hormone levels, which combines the advantages of the portability of hearing aids and the non-invasiveness of sweat sensors, and can achieve the following beneficial effects:

[0044] 1. Traditional hormone testing is mostly done by blood sampling with needles. Long-term invasive testing increases the user's physical pain and easily leads to psychological resistance. This embodiment infers the human hormone level by detecting the hormone level in sweat to assist in stress management.

[0045] 2. As the existing hormone detection devices are inconvenient to carry, hearing aids are the preferred product to solve hearing problems. They have the advantages of miniaturization, portability and unified audience. Combining sweat sensor detection devices with hearing aids can achieve real-time management of psychological stress;

[0046] 3. In view of the problem that conventional self-examination in existing technologies lacks records and analysis, and it is difficult to independently judge the direction of the disease, the real-time monitoring of hormone levels in the hearing aid system with a real-time warning system can be used for health management, real-time recording of hormone levels and disease analysis, auxiliary treatment, and real-time warning. At the same time, with the help of hearing aids, the user's hearing condition is improved to solve the problem of users not being able to hear or not being able to hear clearly.

[0047] Based on the above hearing aid system that monitors hormone levels in real time, Figure 7 This is a flowchart of a method for predicting pressure in a hearing aid system based on real-time monitoring of hormone levels provided by an embodiment of the present invention. The method is applicable to predicting the pressure change pattern of a user in the future. The method can be executed by the disease monitoring system in the hearing aid system or other electronic devices. Figure 7 As shown, the method specifically includes:

[0048] S110. Identify the type of pressure curve of the user in a specific time period according to the sound signals acquired by the hearing aid of the user in multiple historical time periods and the basic information of the user, wherein the specific time period covers the multiple historical time periods and the future time period.

[0049] Among them, the stress curve refers to the curve of the user's stress level changing over time. Specifically, the above-mentioned disease monitoring system can monitor the user's stress level in real time and generate the user's stress curve. For a certain user, a larger fixed time length can be selected, and multiple curves (the time range of each curve can be repeated) can be intercepted from the user's existing historical stress curve through a moving window of the time length as multiple stress curve samples. Performing the above operation on multiple users can obtain stress curve samples from a large number of users, and the duration of each sample is the fixed time length. Then, these pressure curve samples are classified according to the change law of the curve, and each type of pressure curve corresponds to a pressure curve type, and each pressure curve type represents a pressure change law with universal significance. Optionally, all pressure curve samples can be clustered, each cluster cluster corresponds to a pressure curve type, and the curve corresponding to each cluster center is used as the standard pressure curve of this type. There are many specific curve clustering methods. For example, each curve sample can be discretized into a pressure data sequence at the same time interval, and each sequence can be used as a vector to execute a vector clustering algorithm to obtain multiple clusters. The vectors corresponding to each cluster center are interpolated into a continuous curve, which is the standard pressure curve of each cluster center.

[0050] This step divides the specific time period to be processed into two parts, one is the historical time period that has occurred, and the other is the future time period that has not occurred. The pressure curve type shown by the user in the entire time period of the specific time period is identified through the information related to the historical time period and the user's basic information, which is used to predict the pressure trend of the future time period.

[0051] In a specific implementation, a pressure curve type recognition model can be constructed by a deep learning method, and the model is used to process the user's basic information and the sound signal obtained by the hearing aid to obtain the user's pressure curve type in a specific period of time. Optionally, the user's basic information includes one or more of occupation, age, gender, height, weight, and average daily sleep duration. Optionally, the sound signal from the hearing aid includes semantic information and audio information, wherein relatively stable information (hereinafter referred to as normalized information) that can affect the law of pressure change of the user is information that is conducive to the recognition of the pressure curve type, while those unstable information that is greatly affected by the user's immediate state are not very helpful for the recognition of the pressure curve type. Exemplarily, the semantic information may include the user's conversation content, wherein the normalized information may come from the user's high-frequency activities, focal points, or relatively stable activities, focal points, etc. in a period of time; the audio information includes the user's sound spectrum, the noise spectrum of the environment, etc., wherein the normalized information may come from the user's overall sound state in a period of time, or the basic environmental state of some high-frequency places, etc. In order to extract these normalized information, this embodiment selects multiple historical periods from the specific period to be predicted to avoid a single historical period from effectively covering the normalized information. Of course, the duration of the specific time period is less than or equal to the fixed time period, so the time spanned by the multiple historical time periods will not be too long, thereby ensuring that the pressure change pattern of the user in the specific time period is learnable.

[0052] Furthermore, in order to make full use of the above information, we can construct Figure 8 The multi-modal pressure curve type identification model shown in Figure 8 As shown in Figure 1, the model includes semantic branch, audio branch, basic information branch, multimodal fusion layer and classifier. The following is an explanation of the structure and function of each part in combination with the specific data processing flow.

[0053] The sound signals acquired by the user's hearing aid in multiple historical periods are obtained. First, semantic content is extracted from the sound signal by a sound-to-text method, and the semantic content is input into a semantic branch for encoding to obtain an embedding vector of the semantic content. Optionally, the semantic branch can adopt a Glove structure or other context encoders for text vectorization.

[0054] At the same time, the sound signal is converted into the frequency domain, and the spectrum information is input into the audio branch for encoding to obtain an embedding vector of the audio information. Optionally, the original audio waveform can be converted into the frequency domain to generate a log-Mel spectrum graph, and the audio branch uses a VGGish structure to convert the log-Mel spectrum graph into an embedding vector. The sound signals of each historical period can be processed separately to obtain an embedding vector respectively.

[0055] At the same time, one-hot encoding or hash encoding is used to vectorize each user information and input it into the basic information branch for encoding. Optionally, the basic information branch can adopt a multi-layer perceptron structure.

[0056] After encoding of each branch is completed, the encoding vectors output by each branch are input into the multimodal fusion layer for feature fusion to obtain multimodal fusion features. Optionally, the multimodal fusion layer can adopt the encoder structure of Transformer, and the encoding vectors output by each branch are spliced ​​and input into the Transformer encoder to obtain multimodal fusion features.

[0057] Finally, the multimodal fusion features are input into a classifier. The classifier may adopt a multi-layer perceptron structure to output a probability vector of the specific time period belonging to various pressure curve types. The one with the highest probability is the final classification result.

[0058] Furthermore, in order to ensure that normalized information that is helpful for identifying the law of user pressure changes is extracted from the sound signal, the following methods can be used for training in the training stage:

[0059] For a certain pressure curve sample of any user, perform the following operations respectively:

[0060] Step 1: Construct multiple combinations including multiple historical periods, where each historical period is a period covered by the pressure curve sample, and the historical period included in any combination is not completely the same as the historical period included in another combination. For example, if combination A includes historical periods a, b, and c, then combination B includes historical periods e, f, and g, or combination B includes historical periods c, f, and g, which are not completely the same as the three historical time periods of combination A, thereby ensuring rich sampling of the time domain in each combination.

[0061] Step 2: For each combination of the user in the sample, perform the following operations respectively:

[0062] S1-1. Extract semantic content from the sound signal obtained by the user's hearing aid in the current combination, and input the semantic content into the semantic branch for encoding; at the same time, perform frequency domain conversion on the sound signal corresponding to the current combination, and input the spectrum information into the audio branch for encoding; at the same time, input the basic information of the user into the basic information branch for encoding.

[0063] S1-2. The encoding results of each branch are input into the multimodal fusion layer to obtain multimodal fusion features.

[0064] S1-3. Input the multimodal fusion feature into the classifier to identify the pressure curve type under the current combination.

[0065] Step 3: After performing the above operations for each combination, the parameters of the pressure curve type recognition model are updated by constraining the multimodal fusion features under different combinations to be consistent, and the recognition results of the pressure curve types under different combinations to be consistent with the pressure curve type labels of the user in the sample. Optionally, the following loss function can be constructed:

[0066]

[0067] in, represents the loss function value, and They represent the multimodal fusion features corresponding to the multi-period combinations indexed as i and j under the same pressure curve sample; The probability vector corresponding to the pressure curve type label of the same pressure curve sample is represented. The label can come from the above clustering process or from manual annotation based on the clustering result. The probability of the vector element where the label is located is 1, and the probability of other elements is 0; represents the pressure curve type recognition result corresponding to the combination with index n under the same pressure curve sample, that is, the probability vector output by the classifier; and are the weights of the two loss terms, and their values ​​are between [0,1]. Finally, the model parameters are updated by minimizing the Loss.

[0068] After the above parameters are updated for each stress curve sample of each user, the finally trained semantic branch and audio branch can extract normalized information that is not affected by the time domain position of the historical period, and the finally trained classifier can output the correct stress curve type label.

[0069] S120: segmentally match the existing pressure curve in the specific time period with the standard pressure curve of the pressure curve type, and use the curve after the segment matching the existing pressure curve in the standard pressure curve as the user's future pressure curve.

[0070] The existing pressure curve in the specific period refers to the pressure curve formed by the pressure data that have been monitored in the specific period. Since the line segment of the standard pressure curve is long, the existing pressure curve can only be similar to a segment under the standard pressure curve. By comparing and locating the similar segment segment by segment, the standard pressure curve after the segment can represent the pressure trend of the future period.

[0071] Optionally, a sliding window may be used to divide the standard pressure curve into multiple segments of equal length to the existing pressure curve, and the segments may overlap; then, the similarity between each segment and the existing pressure curve is calculated, and the segment with the highest similarity is used as the matching segment. The method for calculating the curve similarity is prior art and will not be described in detail here.

[0072] In summary, this embodiment provides a pressure prediction method for a hearing aid system based on real-time monitoring of hormone levels, which reflects the user's pressure change pattern through a pressure curve within a period of time, and clusters these pressure change patterns to obtain a variety of universal pressure curve types; then, through the pressure curve type recognition model, extract normalized information that is not highly correlated with the time domain position from the dynamic sound signal extracted by the hearing aid, and after multimodal fusion with the user's basic information, determine the user's pressure change type within a period of time; finally, predict the pressure trend in the future through the standard pressure change curve of this type. Compared with short-term pressure prediction, such as pressure prediction at only a single moment, this embodiment pays more attention to the change pattern of pressure within a period of time, and through clustering, learning and identification of pressure curves, the prediction time domain is longer, which is more conducive to early intervention or guidance of the user's psychological stress state.

[0073] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or replace some or all of the technical features therein by equivalents. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the technical solutions of the embodiments of the present invention.

Claims

1. A pressure prediction method for a hearing aid system based on real-time monitoring of hormone levels, characterized in that: A hearing aid system for real-time monitoring of hormone levels, the hearing aid system comprising a hearing aid, a flexible sweat sensor patch and a disease monitoring system, each of which can cooperate to monitor the hormone level and psychological stress level of the user in real time; The method comprises: According to the sound signals obtained by the user's hearing aid in multiple historical time periods and the user's basic information, the pressure curve type of the user in a specific time period is identified, wherein the specific time period covers the multiple historical time periods and the future time period; specifically, semantic content is extracted from the sound signals obtained by the user's hearing aid in multiple historical time periods, and the semantic content is input into the semantic branch of the pressure curve type recognition model for encoding; the sound signal is converted into the frequency domain, and the spectrum information is input into the audio branch of the pressure curve type recognition model for encoding; the user's basic information is input into the basic information branch of the pressure curve type recognition model for encoding; the encoding results of each branch are input into the multimodal fusion layer of the pressure curve type recognition model to obtain the multimodal fusion feature; the multimodal fusion feature is input into the classifier of the pressure curve type recognition model to obtain the user's pressure curve type; The existing pressure curve in the specific time period is matched segmentally with the standard pressure curve of the pressure curve type, and the curve after the segment matching the existing pressure curve in the standard pressure curve is used as the user's future pressure curve, wherein the existing pressure curve is monitored by the hearing aid system.

2. The pressure prediction method according to claim 1, characterized in that: Before identifying the type of the pressure curve of the user in a specific period of time according to the sound signals obtained by the hearing aid of the user in multiple historical periods and the basic information of the user, the method further includes: Extracting multiple stress curves of multiple users under a fixed duration, wherein the fixed duration is greater than or equal to the duration of the specific time period; Each pressure curve is clustered and a pressure curve type is assigned to each cluster.

3. The pressure prediction method according to claim 1, characterized in that: Before inputting the semantic content into the semantic branch of the pressure curve type recognition model for encoding, the method further includes: For any pressure curve sample of any user, multiple combinations including multiple historical periods are constructed, wherein the historical period included in any combination is not completely the same as the historical period included in another combination; For each combination, perform the following operations: S1-1, extracting semantic content from the sound signal obtained by the hearing aid of any user in the current combination, and inputting the semantic content into the semantic branch of the pressure curve type recognition model for encoding; at the same time, performing frequency domain conversion on the sound signal corresponding to the current combination, and inputting the spectrum information into the audio branch of the pressure curve type recognition model for encoding; at the same time, inputting the basic information of any user into the basic information branch of the pressure curve type recognition model for encoding; S1-2, inputting the encoding results of each branch into the multimodal fusion layer of the pressure curve type recognition model to obtain a multimodal fusion feature; S1-3, inputting the multimodal fusion feature into the classifier of the pressure curve type recognition model to identify the pressure curve type under the current combination; By constraining the multimodal fusion features under different combinations to be consistent, and the recognition results of the pressure curve type under different combinations to be consistent with the pressure curve type label of any pressure curve sample, the parameters of the pressure curve type recognition model are updated, so that the finally trained semantic branch and audio branch can extract normalized information that is not affected by the time domain position of each historical period.

4. The pressure prediction method according to claim 3, characterized in that: The updating of the parameters of the pressure curve type recognition model by constraining the multimodal fusion features under different combinations to be consistent and the recognition results of the pressure curve types under different combinations to be consistent with the pressure curve type labels of any pressure curve samples includes: Construct the following loss function: ； in, represents the loss function value, represents the similarity between two vectors. and They represent the multimodal fusion features corresponding to the combination of index i and j under the same pressure curve sample, a pressure curve type label representing the same pressure curve sample, represents the pressure curve type recognition result corresponding to the combination with index n under the same pressure curve sample, and are the weights of the two loss terms respectively; The model parameters are updated by minimizing the loss.

5. The pressure prediction method according to claim 1, characterized in that: The outer side of the flexible sweat sensor patch is used to adhere to the skin behind the ear, and the inner side is attached to the outer surface of the hearing aid, and is electrically connected to the hearing aid chip of the hearing aid through contact electrodes; The flexible sweat sensor patch includes a sweat collection hole, an electrochemical electrode and a microchannel; after the hearing aid is worn, the sweat of the skin behind the ear flows into the microchannel through the sweat collection hole, and a corresponding potential is applied to the electrochemical electrode by differential pulse voltammetry, so that the sweat in the microchannel can undergo an electrochemical reaction, and the hormone level is detected by the electrochemical reaction and transmitted to the hearing aid chip for processing; wherein the hormones to be detected include tryptophan and cortisol; The disease monitoring system records the detection data, analyzes the user's psychological stress level according to the hormone level, and links the hearing aid to relieve the pressure in the ear when the data is abnormal.

6. The pressure prediction method according to claim 5, characterized in that: The microchannel substrate is made of polydimethylsiloxane material. Different refractive indexes and transmittances of visible light can make the microchannel present different colors in an empty state and a water-filled state, thereby judging whether the amount of sweat meets the detection state.

7. The pressure prediction method according to claim 5, characterized in that: When there is a detection requirement but the amount of sweat is insufficient for detection, a voltage is applied between the electrochemical electrodes to generate a non-inductive current under the skin surface, thereby stimulating sweat secretion.

8. The pressure prediction method according to claim 5, characterized in that: The hearing aid also includes a microphone and a receiver. The microphone is used to acquire external sounds and transmit them to the hearing aid chip. The hearing aid chip is used to optimize the sounds. The receiver is used to output the optimized sounds to the user's ears.

9. The pressure prediction method according to claim 5, characterized in that: The end of the microchannel comprises an exhaust hole, and the liquid in the microchannel generates gas through electrochemical reaction and heating, and is discharged from the exhaust hole.

Images