Hearing management method based on hearing aid device and related device

Abstract

The invention discloses a hearing management method based on hearing aid equipment and a related device, and the method comprises the steps: carrying out the feature extraction of a time series data set collected by the hearing aid equipment in the audio playing process, and obtaining the dynamic change features of a plurality of hearing state parameters in different time windows; predicting a hearing loss upper limit change curve and a hearing loss lower limit change curve of the plurality of frequency bands according to the dynamic change characteristics to obtain a hearing loss prediction interval of each frequency band; matching the target hearing loss change curve with the hearing loss prediction interval, and determining whether an abnormal time sequence change exists or not; when abnormal time sequence change is detected, the continuity, the restorability and the inducement relevance of the abnormal time sequence change are analyzed, and the hearing loss type is determined; and adjusting audio output parameters of the hearing aid device according to the hearing loss type. According to the application, the personalized and scenarized adaptation capability of the hearing aid device can be improved, so that the hearing aid effect better fits the actual hearing demand of the user, and the accurate hearing management requirement is met.

Description

Hearing management methods and related devices based on hearing aids Technical Field

[0001] This application relates to the field of hearing aid technology, and in particular to a hearing management method and related device based on hearing aids. Background Technology

[0002] Hearing aids are widely used in various scenarios requiring hearing assistance, providing corresponding auditory support to those with hearing needs. However, most hearing aids on the market currently employ fixed sound processing schemes, with audio adjustment and hearing compensation modes largely preset. This makes it difficult to tailor them to the individual user's hearing condition and actual needs, nor can they be optimized based on external environmental sound conditions and scene changes. Such limitations result in the actual hearing aid effect failing to match the user's true hearing needs, limiting the hearing compensation effect, leading to a poor overall user experience and failing to meet the user's hearing assistance requirements. Summary of the Invention

[0003] This application provides a hearing management method and related device based on hearing aids, so as to improve the personalization and scenario-based adaptation capabilities of hearing aids, make the hearing aid effect more in line with the user's actual hearing needs, and meet the requirements of precise hearing management.

[0004] In a first aspect, embodiments of this application provide a hearing management method based on a hearing aid device, comprising: extracting features from a time-series dataset collected by the hearing aid device during audio playback to obtain dynamic change features of multiple hearing status parameters in different time windows, wherein the time-series dataset includes user-side data, device-side data, and environmental-side data, and each side of the data corresponds to at least one hearing status parameter; predicting upper limit and lower limit change curves of hearing loss for multiple frequency bands based on the dynamic change features to obtain a hearing loss prediction interval for each frequency band; matching a target hearing loss change curve with the hearing loss prediction interval to determine whether there is an abnormal time-series change, wherein the target hearing loss change curve is determined based on the dynamic change features of the shortest time window; detecting an abnormal time-series change, analyzing the persistence, recoverability, and causal correlation of the abnormal time-series change to determine the type of hearing loss; and adjusting the audio output parameters of the hearing aid device according to the type of hearing loss.

[0005] The process of detecting abnormal temporal changes and analyzing their persistence, recoverability, and causal correlation to determine the type of hearing loss includes: determining the duration of the abnormal temporal changes in different time windows; obtaining the slope of the sub-hearing loss change curve corresponding to the abnormal temporal changes; analyzing the persistence of the abnormal temporal changes based on the duration, the number of time windows covered by the duration, and the slope to obtain a persistence analysis result; dividing the sub-hearing loss change curve into a baseline hearing loss curve and a recovery hearing loss curve; analyzing the recoverability of the abnormal temporal changes based on the baseline hearing loss curve, the recovery hearing loss curve, and the sub-hearing loss change curve to obtain a recoverability analysis result; analyzing the correlation between the abnormal temporal changes and user behavior, environmental factors, and device status to obtain a causal correlation analysis result; and determining the type of hearing loss based on the persistence analysis result, the causal correlation analysis result, and the recoverability analysis result.

[0006] The step of determining the type of hearing loss based on the results of the persistence analysis, the recoverability analysis, and the cause-related analysis includes: determining whether the hearing loss is true based on the results of the persistence analysis, the recoverability analysis, and the cause-related analysis, to obtain a first type of result; determining whether the hearing loss is temporary based on the results of the persistence analysis and the recoverability analysis, to obtain a second type of result; determining whether the hearing loss is reversible based on the results of the recoverability analysis and the cause-related analysis, to obtain a third type of result; and determining the type of hearing loss based on the first type of result, the second type of result, and the third type of result.

[0007] The step of determining the hearing loss type based on the first type of result, the second type of result, and the third type of result includes: detecting at least two non-negative results among the first type of result, the second type of result, and the third type of result, thus determining the existence of multiple hearing loss types; if the multiple hearing loss types do not include actual hearing loss, then determining at least one hearing impairment factor based on the causal correlation analysis results; determining the matching relationship between each hearing impairment factor and the abnormal temporal change in the time dimension, logical correlation dimension, and impairment characteristic dimension, thus obtaining the contribution of each hearing impairment factor; and determining the hearing loss type based on the contribution.

[0008] The step of analyzing the recoverability of the abnormal temporal change based on the baseline hearing loss curve, the recovery hearing loss curve, and the sub-hearing loss change curve to obtain a recoverability analysis result includes: determining the recovery efficiency of the abnormal temporal change based on the baseline hearing loss curve, the recovery hearing loss curve, and the sub-hearing loss change curve; determining the recovery duration of the abnormal temporal change based on the sub-hearing loss change curve and the recovery hearing loss curve; determining the stability of the abnormal temporal change after recovery based on the recovery hearing loss curve; and analyzing the recoverability of the abnormal temporal change based on the recovery efficiency, the recovery duration, and the stability to obtain the recoverability analysis result.

[0009] The step of extracting features from the time-series dataset collected by the hearing aid device during audio playback to obtain the dynamic change features of multiple hearing status parameters in different time windows includes: segmenting the time-series dataset according to preset different time windows to obtain multiple data sequences for each hearing status parameter; performing a fitting operation on each of the multiple data sequences to obtain a fitting curve for each hearing status parameter in the corresponding time window; obtaining the trend features of each hearing status parameter in different time windows based on the direction of change and the first rate of change of the fitting curve; identifying abrupt change points in the data sequences based on the trend features; obtaining the abrupt change features of each hearing status parameter in different time windows based on the magnitude of change and the second rate of change of the abrupt change points in the corresponding time window; obtaining the fluctuation features of each hearing status parameter in different time windows based on the fluctuation frequency and fluctuation amplitude of the abrupt change points in the corresponding time window; performing correlation analysis on hearing status parameters of different dimensions to obtain the correlation features between the multiple hearing status parameters; and obtaining the dynamic change features of each hearing status parameter in different time windows based on the trend features, the abrupt change features, the fluctuation features, and the correlation features.

[0010] The step of matching the target hearing loss change curve with the hearing loss prediction interval to determine whether there is an abnormal temporal change includes: detecting that the target hearing loss change curve exceeds the hearing loss prediction interval, determining the duration for which the target hearing loss change curve exceeds the hearing loss prediction interval; determining the fluctuation amplitude of the target hearing loss change curve; and determining whether there is an abnormal temporal change based on the duration and / or the fluctuation amplitude.

[0011] Secondly, embodiments of this application provide a hearing management device based on a hearing aid, comprising: a feature extraction unit, used to extract features from a time-series dataset collected by the hearing aid during audio playback, to obtain dynamic change features of multiple hearing status parameters in different time windows, wherein the time-series dataset includes user-side data, device-side data, and environmental-side data, and each side data corresponds to at least one hearing status parameter; a prediction unit, used to predict upper limit and lower limit change curves of hearing loss in multiple frequency bands based on the dynamic change features, to obtain a hearing loss prediction interval for each frequency band; a matching unit, used to match a target hearing loss change curve with the hearing loss prediction interval to determine whether there is an abnormal time-series change, wherein the target hearing loss change curve is determined based on the dynamic change features of the shortest time window; an analysis unit, used to detect the existence of abnormal time-series changes, analyze the persistence, recoverability, and causal correlation of the abnormal time-series changes, and determine the type of hearing loss; and an adjustment unit, used to adjust the audio output parameters of the hearing aid based on the type of hearing loss.

[0012] Thirdly, embodiments of this application provide an electronic device, including a memory, a processor, and executable program code stored in the memory and executable on the processor, wherein the processor executes the executable program code and performs the steps of the method described in the first aspect.

[0013] Fourthly, embodiments of this application provide a computer-readable storage medium storing executable program code, the executable program code including execution instructions for performing the steps of the method as described in the first aspect.

[0014] Fifthly, embodiments of this application provide a computer program product, wherein the computer program product includes a non-transitory computer-readable storage medium storing a computer program operable to cause a computer to perform some or all of the steps described in the first aspect of embodiments of this application. The computer program product may be a software installation package.

[0015] As can be seen, in this embodiment, firstly, feature extraction is performed on the time-series dataset collected by the hearing aid device during audio playback to obtain the dynamic change characteristics of multiple hearing status parameters in different time windows. The time-series dataset includes user-side data, device-side data, and environmental-side data, with each side data corresponding to at least one hearing status parameter. Then, based on the dynamic change characteristics, the upper limit change curve and lower limit change curve of hearing loss for multiple frequency bands are predicted to obtain the hearing loss prediction interval for each frequency band. Next, the target hearing loss change curve is matched with the hearing loss prediction interval to determine whether there is an abnormal time-series change. The target hearing loss change curve is determined based on the dynamic change characteristics of the shortest time window. Then, if an abnormal time-series change is detected, the persistence, recoverability, and causal correlation of the abnormal time-series change are analyzed to determine the type of hearing loss. Finally, the audio output parameters of the hearing aid device are adjusted according to the type of hearing loss.

[0016] This application achieves real-time capture and trend analysis of hearing status by extracting features from multi-dimensional time-series data across multiple time windows; it predicts multi-frequency hearing loss intervals based on dynamic features and identifies abnormal temporal changes, providing timely warnings of hearing loss risks; it accurately identifies the type of hearing loss by analyzing the persistence, recoverability, and correlation of causes of abnormalities; and it adaptively adjusts audio output parameters according to the type of hearing loss, effectively improving the personalization and scenario-based adaptation capabilities of hearing aids, making the hearing aid effect more in line with the user's actual hearing needs, and meeting the requirements of precise hearing management. Attached Figure Description

[0017] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0018] Figure 1 is a system architecture diagram of a hearing management system provided in an embodiment of this application; Figure 2 is a flowchart of a hearing management method based on hearing aids provided in an embodiment of this application; Figure 3 is a schematic diagram of a visual interactive interface provided in an embodiment of this application; Figure 4 is a schematic diagram of a detailed hearing health trend interface provided in an embodiment of this application; Figure 5 is a schematic diagram of an abnormal hearing reminder interactive interface provided in an embodiment of this application; Figure 6 is a schematic diagram of a gain adjustment interactive interface provided in an embodiment of this application; Figure 7 is a functional unit block diagram of a hearing management device based on hearing aids provided in an embodiment of this application; Figure 8 is a functional unit block diagram of another hearing management device based on hearing aids provided in an embodiment of this application; Figure 9 is a structural schematic diagram of an electronic device proposed in an embodiment of this application. Detailed Implementation

[0019] To enable those skilled in the art to better understand the present application, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present application, and not all embodiments. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the scope of protection of the present application.

[0020] The terms "first," "second," etc., in the specification, claims, and accompanying drawings of this application are used to distinguish different objects, not to describe a specific order. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion. For example, a process, method, system, product, or apparatus that includes a series of steps or units is not limited to the listed steps or units, but may optionally include steps or units not listed, or may optionally include other steps or units inherent to these processes, methods, products, or apparatuses.

[0021] In this document, the term "embodiment" means that a particular feature, structure, or characteristic described in connection with an embodiment may be included in at least one embodiment of this application. The appearance of this phrase in various places throughout the specification does not necessarily refer to the same embodiment, nor is it a separate or alternative embodiment mutually exclusive with other embodiments. It will be explicitly and implicitly understood by those skilled in the art that the embodiments described herein can be combined with other embodiments.

[0022] Hearing aids have been widely used in various scenarios requiring hearing assistance. However, existing hearing aids have insufficient adaptive adjustment capabilities, failing to adapt to individual user differences, fitting needs, and changes in the external environment. This results in a low degree of matching between the hearing aid effect and the user's actual hearing needs, making it difficult to meet the requirements of precise hearing management.

[0023] To address the aforementioned issues, this application provides a hearing management method and related apparatus based on hearing aids. The embodiments of this application will be described in detail below with reference to the accompanying drawings.

[0024] The hearing aid device described in this application uses a hearing aid speaker as its product carrier. Based on the hardware architecture of general audio playback devices, it adds an audio signal processing module adapted to people with hearing loss, thus forming a corresponding audio output device. This type of device uses a speaker as its appearance and sound-producing carrier. While retaining the speaker's ability to play in all scenarios, it can perform personalized audio signal adjustment based on the hearing impairment of people with hearing loss.

[0025] Specifically, the architecture of hearing aids is designed around hearing adaptation needs, building a multi-level signal processing chain, including a frequency band separation module, a dynamic range compression unit, and a loudness compensation path. This is used to improve the problem of insufficient audio perception caused by hearing loss in individuals with hearing impairment, and to finely adjust for different frequency hearing threshold losses. It also features an intelligent amplitude limiting protection mechanism to prevent secondary damage to residual hearing from excessively high sound pressure levels, achieving personalized audio signal adjustment. This architecture adopts a modular integration approach, integrating professional hearing aid functions into a general-purpose speaker system, forming a dual-track working mode of basic playback and hearing adaptation. The device can perform basic playback operations based on mature general-purpose audio hardware, such as regular audio playback, volume adjustment, and wireless signal connection. It can also achieve precise audio signal adaptation and optimization by adapting to the specific auditory needs of individuals with hearing loss through architectural design.

[0026] For example, in a home setting, a hearing aid speaker can be placed in the center of a coffee table or on one side of a sofa in the living room, close to the user's daily activity area. When working, it adjusts the frequency response gain of the played audio according to the type of hearing loss identified in the user, ensuring that the user can clearly and completely perceive and receive sound, without having to wear any devices, thus maintaining a comfortable user experience.

[0027] The following section details the hearing management method and related devices based on hearing aids provided in this application.

[0028] Please refer to Figure 1, which is a system architecture diagram of a hearing management system provided in an embodiment of this application. As shown in Figure 1, the hearing management system 100 includes a data acquisition module 101, a preprocessing module 102, a feature extraction module 103, a predictive analysis module 104, an anomaly detection module 105, a hearing recognition module 106, an adjustment module 107, and a feedback module 108. All of these modules are interconnected and communicate with each other.

[0029] The data acquisition module 101 is used to collect user-side data, device-side data, and environmental data. The user-side data includes user interaction data and user vital sign data. The user interaction data is used to indicate the user's active operation behavior and operation parameters on the hearing aid device, including operation information such as volume adjustment, frequency band adjustment, and mode switching. The user's volume adjustment amount, frequency band adjustment parameters, and mode switching commands can be collected through the operation panel of the hearing aid device or the associated APP, and the trigger time, duration, and adjustment range of each operation can be recorded.

[0030] Among them, user vital signs data are used to indicate the user's activity and physiological status, including the user's activity level, sleep duration, whether they have a cold, whether they are fatigued, and other physiological and behavioral states; the user's activity level, sleep duration, and other data can be obtained through the associated health APP, and the physiological status information input by the user can be received through the device's interactive interface to form a time-series record of the user's status.

[0031] The environmental data includes playback environment data, which indicates the acoustic environment and scene characteristics of the hearing aid device, including ambient noise level, noise spectrum, room reverberation time, scene type, etc. The scene type includes quiet environment, noisy environment, TV scene, music scene, etc. It can collect ambient noise signals through the device's built-in microphone and analyze the noise level and noise spectrum distribution in real time; it can collect room reverberation time through acoustic sensors and combine it with scene recognition algorithms to determine the current scene type.

[0032] The device-side data includes the deployment status data and audio content data of the hearing aid. The deployment status data is used to indicate the spatial position, posture and operating status of the hearing aid, including the distance between the user and the device, the device placement angle, and the device working status. The relative distance between the user and the hearing aid can be collected by a distance sensor, the device placement angle and the user's head orientation can be collected by a posture sensor, and operating parameters such as the device speaker output power and microphone sensitivity can be collected to determine whether the device is in normal working condition.

[0033] The preprocessing module 102 aligns the raw time-series data from the user side, device side, and environment side with a unified timestamp, filters outliers, format errors, duplicate invalid data, completes short-term missing data, normalizes parameters of different dimensions and maps them to a unified interval, and finally outputs a regularized time-series dataset as input to the feature extraction module 103.

[0034] Among them, the feature extraction module 103 performs feature extraction based on the preprocessed time series data and divides different time windows; it extracts hearing state parameters that are directly related to the user's hearing adaptation from the preprocessed time series data; for the hearing state parameters in each time window, it extracts trend, mutation, fluctuation and multi-dimensional correlation features respectively to obtain dynamic change features under multi-scale time windows.

[0035] Among them, the predictive analysis module 104 constructs a multi-band temporal hearing loss trend model with a "temporal prediction network and multi-band branch" structure based on the user's historical hearing baseline data and dynamic change characteristics.

[0036] In the model training phase, the model is trained based on the user's historical time-series dataset. The mean square error between the predicted hearing loss value and the actual hearing loss value is used as the loss function. The model parameters are optimized by minimizing the loss value. Finally, the normal prediction range and trend of hearing loss in each frequency band in the next 1-3 months are output.

[0037] The anomaly detection module 105 is used to compare the dynamic change features extracted in real time under the short window with the corresponding prediction interval output by the model to determine whether there are abnormal time series changes.

[0038] Among them, the hearing recognition module 106 is used to perform multi-dimensional correlation analysis on abnormal temporal changes. It distinguishes between true hearing loss, temporary threshold shift, pseudo-hearing loss, and mixed hearing loss by combining the persistence, recoverability, and causal correlation of the changes. By analyzing the existence of abnormalities in different time windows and their correlation with external factors, the specific type of hearing loss is determined.

[0039] The adjustment module 107 is used to generate a hearing loss type determination report, which includes information such as hearing status, abnormality type, and cause, and supports visual display on the APP. It adapts differentiated strategies according to different types of hearing loss, evaluates the effectiveness of the strategies daily, and makes fine adjustments.

[0040] Among them, the feedback module 108 visualizes the trend of hearing status changes and hearing assessment reports through the associated APP. At the same time, it transmits the strategy evaluation effect and newly collected user feedback data to the predictive analysis and feature extraction modules to form a complete closed-loop management, ensuring that the system continuously adapts to changes in the user's hearing status.

[0041] Based on this, this application provides a hearing management method based on hearing aids, which will be described in detail below with reference to the accompanying drawings.

[0042] Please refer to Figure 2. Figure 2 is a flowchart of a hearing management method based on a hearing aid device provided in an embodiment of this application. As shown in Figure 2, the method includes the following steps: S210, extracting features from the time-series dataset collected by the hearing aid device during audio playback to obtain the dynamic change features of multiple hearing state parameters in different time windows.

[0043] The time-series dataset includes user-side data, device-side data, and environmental-side data, with each side of the data corresponding to at least one hearing status parameter.

[0044] The dynamic change features are used to characterize the temporal distribution of each hearing state parameter and the correlation between multiple hearing state parameters. The dynamic change features include trend features, abrupt change features, fluctuation features, and correlation features. Among them, trend features are derived from short-term, medium-term, and long-term time window analysis; abrupt change features are used to identify sudden changes; fluctuation features are used to determine fluctuating hearing loss; and correlation features are used to analyze the relationship between accommodation and factors such as environment, equipment, and content.

[0045] The user-side data includes user interaction data and user vital sign data; the environmental-side data includes playback environment data; and the device-side data includes the deployment status data of the hearing aid and audio content data. User interaction data includes volume adjustment, frequency band adjustment, and mode switching. User vital sign data includes activity level and physiological state. Playback environment data includes noise level, noise spectrum, reverberation, and scene type. Deployment status data includes the distance between the user and the device and the device's posture. Audio content data includes content type and spectral characteristics.

[0046] This involves aligning the raw time-series data from the user side, device side, and environment side according to timestamps, preprocessing the time-series data, including filtering invalid data, completing missing data, and standardizing transformation, to form a well-organized time-series dataset.

[0047] Among them, hearing status parameters are measurable variables used to characterize the user's current hearing status and device compatibility status. Each side of the data corresponds to at least one hearing status parameter, such as the user's volume adjustment, activity level, and mode switching frequency; the user's distance from the device and device posture; and the environmental noise level and scene type.

[0048] In one possible embodiment, the step of extracting features from the time-series dataset collected by the hearing aid device during audio playback to obtain the dynamic change features of multiple hearing status parameters in different time windows includes: segmenting the time-series dataset according to preset different time windows to obtain multiple data sequences for each hearing status parameter; performing a fitting operation on each of the multiple data sequences to obtain a fitting curve for each hearing status parameter in the corresponding time window; obtaining the trend features of each hearing status parameter in different time windows based on the direction of change and the first rate of change of the fitting curve; identifying abrupt change points in the data sequences based on the trend features; obtaining the abrupt change features of each hearing status parameter in different time windows based on the magnitude of change and the second rate of change of the abrupt change points in the corresponding time window; obtaining the fluctuation features of each hearing status parameter in different time windows based on the fluctuation frequency and fluctuation amplitude of the abrupt change points in the corresponding time window; performing correlation analysis on hearing status parameters of different dimensions to obtain the correlation features between the multiple hearing status parameters; and obtaining the dynamic change features of each hearing status parameter in different time windows based on the trend features, the abrupt change features, the fluctuation features, and the correlation features.

[0049] First, three time windows are preset: short-term, medium-term, and long-term. For example, short-term is 24 hours, medium-term is 3 days, and long-term is 30 days. Based on a unified timestamp, the time-series dataset containing user-side, environment-side, and device-side data is divided into continuous and non-overlapping subsets according to the duration of different time windows. For each hearing status parameter, all values ​​of the parameter within the corresponding time window are extracted from each subset and arranged in chronological order to form a data sequence of each hearing status parameter under different time windows.

[0050] Specifically, for each time window, a coordinate system is constructed for a single hearing state parameter data sequence, such as 24 data points of volume adjustment within a short window. "Time" is used as the horizontal axis and "parameter value" as the vertical axis. The corresponding fitting method is selected based on the parameter variation pattern; for example, linear fitting is used for volume adjustment, and polynomial fitting is used for ambient noise level. The least squares method is used to calculate the fitting coefficients of the fitted curves, such as the slope and intercept of the linear fit, eliminating random noise in the original data and generating smooth fitted curves. Ultimately, each data sequence corresponds to one fitted curve.

[0051] When extracting trend features of hearing status parameters, the slope of the fitted curve for each parameter, i.e., the first rate of change, is first calculated. For different types of hearing status parameters, such as ambient noise level and user volume adjustment, a grading rule for the first rate of change is preset, and the rate level is determined accordingly. Then, the direction of parameter change is determined based on the sign of the slope: a positive slope indicates an upward trend, a negative slope indicates a downward trend, and a slope close to 0 indicates a stable trend. Finally, the trend feature of the hearing status parameter within the corresponding time window is output based on the direction of change and the first rate of change. For example, for user volume adjustment within a short time window, the slope of its fitted curve is calculated to be 0.3. According to the preset rule, the rate level is determined to be medium speed. Combined with the positive slope, the trend feature is that within the short window, the user volume adjustment increases, exhibiting a medium-speed change, and the corresponding bound value is output.

[0052] The normal fluctuation range of parameters is determined based on trend characteristics. For example, when the short-term trend is stable, the normal fluctuation range is less than or equal to 0.3 dB; when the trend is upward, the normal fluctuation range is less than or equal to 0.5 dB. Each data point in the data sequence is traversed, and the deviation between the actual value at that point and the predicted value at the corresponding time point of the fitted curve is calculated. If the deviation exceeds the normal fluctuation range, and the instantaneous rate of change of that point compared to the previous data point is much greater than the first rate of change, then that point is marked as a breakpoint.

[0053] For each labeled mutation point, the change amplitude is first calculated, which is the absolute value of the difference between the parameter value of the mutation point and the previous normal data point. Then, the second change rate is calculated, which is obtained by dividing the change amplitude by the time interval between the mutation point and the previous normal data point, and is used to characterize the instantaneous speed of change at the mutation point. Using the change amplitude and the second change rate, mutation intensity, mutation level, etc., are calculated, and the mutation feature vector for each time window is output. The mutation intensity is the product of the change amplitude and the second change rate, and the mutation level is determined based on the amplitude and rate.

[0054] Among them, the fluctuation frequency is the number of effective fluctuations caused by the mutation point within the time window. Effective fluctuations are used to characterize the opposite direction of change of adjacent mutation points and the amplitude is less than or equal to the minimum fluctuation threshold. The fluctuation amplitude is used to characterize the average or maximum amplitude of all effective fluctuations within the time window.

[0055] The process involves determining the fluctuation frequency of each window based on the number of effective fluctuations and the window duration. Simultaneously, the difference between the maximum and minimum values ​​within the window is calculated, representing the overall fluctuation range. The mean of the amplitude changes at all abrupt change points within the window is also calculated, reflecting the degree of fluctuation driven by the abrupt change, i.e., the fluctuation amplitude. Based on the fluctuation frequency and amplitude, the fluctuation level and stability are determined, and a fluctuation feature vector for each time window is output.

[0056] This study utilizes correlation analysis of multi-dimensional hearing state parameters to uncover the linkage patterns between different parameters. Specifically, all hearing parameters are first aligned by timestamps to form a structured multi-dimensional time-series matrix. Then, correlation indices are calculated. For any two parameters within the same time window, appropriate quantitative indicators are selected based on the correlation attributes. For example, Pearson correlation coefficient is used to calculate linear correlations, mutual information values ​​are used for non-linear correlations, and time-series linkage relationships are analyzed using time-series lag correlation analysis, while significance testing is performed. Finally, correlation features are extracted, including correlation type, correlation strength, and lag duration.

[0057] This involves fusing the trend, abrupt change, and fluctuation characteristics of the same hearing state parameter within the same time window, and combining this parameter with the correlation characteristics of other dimensions to form the complete dynamic change characteristics of this parameter within the corresponding time window.

[0058] As can be seen, in this embodiment, by segmenting the time-series dataset into different time windows and extracting trend features from the fitted curves, the changing patterns of various hearing state parameters at different times can be captured; sudden hearing abnormalities can be captured through mutation features, and the oscillation patterns of hearing state can be identified through fluctuation features; through multi-dimensional parameter correlation analysis, the linkage relationship between parameters can be explored, achieving a comprehensive assessment of hearing state. The above-mentioned trend, mutation, fluctuation, and correlation features are integrated to form dynamic change features, constructing a complete hearing state profile, dynamically capturing hearing change trends, and providing data support for personalized adaptation, real-time monitoring, and abnormal early warning of hearing aids.

[0059] S220, based on the dynamic change characteristics, predict the upper limit change curve and lower limit change curve of hearing loss for multiple frequency bands to obtain the hearing loss prediction range for each frequency band.

[0060] This process involves collecting users' initial pure-tone audiometry reports and previous hearing assessment data. The initial pure-tone audiometry reports include hearing threshold test results for each frequency band, while previous hearing assessment data includes audiometry records at different time points and records of changes in hearing status. Baseline hearing threshold values ​​for each frequency band are extracted to determine the hearing threshold data for each band at different historical time points. For example, three months ago, a user's hearing threshold in the 250Hz band was 20dB and in the 500Hz band was 18dB; one month ago, the hearing threshold in the 250Hz band was 22dB and in the 500Hz band was 20dB. This data is then categorized and archived by frequency band as the foundational data for model construction.

[0061] The system employs a "temporal prediction network + multi-band branch" structure to build a multi-band temporal hearing loss trend model. The input data for the input layer consists of dynamically changing features. The hidden layer uses a Long Short-Term Memory (LSTM) network to capture temporal dependencies, with multiple LSTM units. The number of units in each layer is adjusted according to the data scale; for example, three LSTM units are used, with each layer containing 64 neurons. This structure is used to mine the correlations between hearing data at different time points and reconstruct the temporal logic of hearing changes. The output layer has separate output branches for each frequency band from 250Hz to 8000Hz, with each branch corresponding to one frequency band, outputting the hearing loss trend prediction results for that band, thus achieving simultaneous prediction across multiple frequency bands.

[0062] The model is trained using a historical time-series dataset of users, divided into a training set and a validation set. The training set is used for learning model parameters, while the validation set is used to verify the model's predictive performance. The mean squared error between the predicted and actual hearing loss values ​​is used as the loss function. Gradient descent is employed to optimize the model parameters, iteratively adjusting the weights and biases of the LSTM network. For example, if the loss function value exceeds a preset threshold in a particular iteration, the network learning rate is adjusted, and training continues iteratively until the loss function value stabilizes and the model's prediction accuracy reaches the preset standard. After training, the model can output the normal prediction range for hearing loss in each frequency band for the next 1-3 months.

[0063] This can be achieved by setting a weekly update cycle, continuously collecting time-series data during audio playback from the user's hearing aid device, and integrating the newly collected time-series data into the existing training set to form an updated training set. Based on the updated training set, the model parameters are fine-tuned, adjusting only weights and biases that are less adapted to the new data, without changing the overall model structure. For example, if newly collected hearing loss data in the 2000Hz band shows a slight upward trend within a week, the relevant parameters of the output branch for that band are fine-tuned to ensure the model can adapt to long-term changes in the user's hearing status and guarantee prediction accuracy.

[0064] Among them, based on the constructed multi-band temporal hearing loss trend model, the dynamic change characteristics of each hearing status parameter are input. The model captures the temporal dependency through the LSTM network, and combines the historical baseline data and temporal change patterns of each frequency band to predict the upper limit change curve and lower limit change curve of hearing loss in the future for each frequency band, so as to obtain the hearing loss prediction range of each frequency band.

[0065] As can be seen, in the embodiments of this application, based on the multi-band temporal hearing loss trend model, the short-term fluctuations and long-term changes in hearing status of each frequency band can be monitored simultaneously. When the hearing status approaches or exceeds the prediction range and shows signs of abnormal decline, an early warning can be issued to promptly detect potential hearing damage risks, provide a reliable basis for adjusting hearing aid parameters, effectively improve the timeliness and pertinence of intervention, and protect the user's hearing health.

[0066] S230, Match the target hearing loss change curve with the hearing loss prediction interval to determine whether there are abnormal temporal changes.

[0067] The target hearing loss change curve is determined based on the dynamic change characteristics of the shortest time window.

[0068] Specifically, the dynamic change features extracted from the shortest time window are transformed into a numerical feature vector. With time on the horizontal axis and hearing loss value on the vertical axis, a continuous curve is fitted to the discrete feature data within the shortest time window. During the fitting process, the true trajectory of the hearing loss value within the window is preserved, resulting in the target hearing loss change curve. This curve is used to characterize subtle changes in the user's hearing status over a short time period.

[0069] In one possible embodiment, matching the target hearing loss change curve with the hearing loss prediction interval to determine whether there is an abnormal temporal change includes: detecting that the target hearing loss change curve exceeds the hearing loss prediction interval; determining the duration for which the target hearing loss change curve exceeds the hearing loss prediction interval; determining the fluctuation amplitude of the target hearing loss change curve; and determining whether there is an abnormal temporal change based on the duration and / or the fluctuation amplitude.

[0070] First, the target hearing loss change curve is matched with the upper and lower limit curves output by the model at each time node to detect whether the target curve value at each time node falls within the prediction interval. If it is detected that some values ​​of the target hearing loss change curve exceed the hearing loss prediction interval, the start time of exceeding the interval is recorded, and the duration of exceeding the interval is continuously monitored and calculated.

[0071] In this process, the fluctuation amplitude of the target hearing loss change curve is calculated simultaneously. The difference between the maximum and minimum values ​​of the target curve within the window is used to calculate the degree of oscillation of the target hearing loss change curve. Combined with the duration and fluctuation amplitude, the presence of abnormal temporal changes is determined according to preset judgment rules.

[0072] Specifically, if the target hearing loss change curve does not exceed the hearing loss prediction range and the fluctuation amplitude is less than or equal to the first threshold, it is determined to be a normal temporal fluctuation of the user's hearing status, and the system returns to the data collection stage to continue collecting real-time data; if the target hearing loss change curve exceeds the hearing loss prediction range and the duration of the excess exceeds the second threshold, it is marked as a hearing-related abnormal temporal change; or if the fluctuation amplitude exceeds the first threshold, regardless of whether it exceeds the prediction range, it is marked as a hearing-related abnormal temporal change.

[0073] If an abnormal time series change is identified, the abnormal start time, abnormal feature type, and corresponding multi-dimensional data are recorded.

[0074] As can be seen, in this embodiment, the target hearing loss change curve is constructed by the dynamic change characteristics of the shortest time window. Combined with the hearing loss prediction interval, abnormal temporal changes are identified by duration and fluctuation amplitude. This allows for real-time tracking of short-term changes in the user's hearing, accurately distinguishing between normal fluctuations and abnormal deviations. It avoids misjudging small fluctuations as abnormalities and can also promptly capture real abnormal situations such as continuous over-limits and large fluctuations, thus achieving sensitive and accurate monitoring of hearing status.

[0075] S240, an abnormal temporal change is detected, and the persistence, recoverability, and causal correlation of the abnormal temporal change are analyzed to determine the type of hearing loss.

[0076] In one possible embodiment, the detection of abnormal temporal changes and the analysis of the persistence, recoverability, and causal correlation of the abnormal temporal changes to determine the type of hearing loss include: determining the duration of the abnormal temporal changes in different time windows; obtaining the slope of the sub-hearing loss change curve corresponding to the abnormal temporal changes; analyzing the persistence of the abnormal temporal changes based on the duration, the number of time windows covered by the duration, and the slope of the change, to obtain a persistence analysis result; dividing the sub-hearing loss change curve into a baseline hearing loss curve and a recovery hearing loss curve; analyzing the recoverability of the abnormal temporal changes based on the baseline hearing loss curve, the recovery hearing loss curve, and the sub-hearing loss change curve, to obtain a recoverability analysis result; analyzing the correlation between the abnormal temporal changes and user behavior, environmental factors, and device status, to obtain a causal correlation analysis result; and determining the type of hearing loss based on the persistence analysis result, the causal correlation analysis result, and the recoverability analysis result.

[0077] The process involves obtaining the start and end times of the abnormal temporal changes, determining the duration of the abnormality within each time window, and identifying the number of time windows covered by the abnormal duration. Sub-hearing loss change curves are extracted for the abnormal period. These sub-curves only contain the curves showing the change in hearing loss value over time during the abnormal period. With time on the horizontal axis and hearing loss value on the vertical axis, the curves are linearly fitted using the least squares method, and the slope of the fitted curve is calculated. Finally, the persistence of the abnormality is analyzed by combining the above information to determine whether it is a short-term, temporary occurrence or a long-term, continuous process, thus obtaining the persistence analysis results.

[0078] For example, if the anomaly lasts only 2 hours in a short window, covers one window, and the absolute value of the slope is less than the third threshold, it is judged as "short-term temporary anomaly, without persistence"; if the anomaly lasts for more than 72 hours, exists continuously in multiple time windows, and the absolute value of the slope is greater than or equal to the third threshold, it is judged as "medium-term persistent anomaly, with hearing loss value showing a continuous downward trend".

[0079] The study uses the sub-hearing loss change curve as a benchmark to divide the baseline hearing loss curve and the recovery hearing loss curve. The baseline hearing loss curve is used as a reference standard for normal hearing status, and the recovery hearing loss curve is used as a reference for hearing recovery after the occurrence of abnormality. By comparing the numerical changes, regression trends and overlap of the baseline hearing loss curve, the recovery hearing loss curve and the sub-hearing loss change curve, the study determines whether the abnormality can be recovered, the recovery speed and the final degree of recovery, and obtains the recoverability analysis results.

[0080] In one possible embodiment, the step of analyzing the recoverability of the abnormal temporal change based on the baseline hearing loss curve, the recovery hearing loss curve, and the sub-hearing loss change curve to obtain a recoverability analysis result includes: determining the recovery efficiency of the abnormal temporal change based on the baseline hearing loss curve, the recovery hearing loss curve, and the sub-hearing loss change curve; determining the recovery duration of the abnormal temporal change based on the sub-hearing loss change curve and the recovery hearing loss curve; determining the stability of the abnormal temporal change after recovery based on the recovery hearing loss curve; and analyzing the recoverability of the abnormal temporal change based on the recovery efficiency, the recovery duration, and the stability to obtain the recoverability analysis result.

[0081] The baseline hearing loss curve, the recovery hearing loss curve, and the sub-hearing loss change curve all correspond to the same time window. The baseline hearing loss curve serves as a reference standard for normal hearing loss before the abnormality occurs. The abnormal extreme point of the sub-hearing loss change curve is used as the recovery starting point. The magnitude and rate of hearing loss value change as the sub-hearing loss change curve approaches the recovery hearing loss curve, and the recovery hearing loss curve regresses to the baseline hearing loss curve, are calculated. The rate of reduction of the numerical difference between the curves characterizes the recovery efficiency; the faster the difference converges, the higher the recovery efficiency.

[0082] The process involves first determining the first moment when the sub-hearing loss change curve enters an abnormal state and reaches its extreme value, and then determining the second moment when the recovered hearing loss curve tends to coincide with the baseline hearing loss curve without significant deviation. The time span between the first and second moments is the recovery time of the abnormal temporal change.

[0083] This process involves extracting data from the period when the recovered hearing loss curve regresses to near the baseline hearing loss curve, calculating the fluctuation amplitude and frequency of the curve values ​​within that period, and noting that smaller and more stable fluctuations indicate a higher degree of stability after recovery. Finally, based on the recovery efficiency, recovery duration, and level of stability, it is determined whether the abnormal temporal changes are fully recoverable, partially recoverable, or unrecoverable, ultimately forming the corresponding recoverability analysis results.

[0084] As can be seen, in the embodiments of this application, recoverability analysis determines the reversibility of abnormal temporal changes through three dimensions: recovery efficiency, recovery duration, and stability. It can accurately distinguish whether the abnormality is a temporary and reversible state fluctuation or an irreversible progression of hearing loss, providing a data basis for subsequent hearing loss type determination and personalized intervention strategy formulation. This helps to adjust hearing aid device parameters or take intervention measures in a targeted manner, improve the accuracy and effectiveness of hearing management, and protect users' hearing health and hearing aid experience.

[0085] Specifically, the synchronicity and correspondence between abnormal temporal changes and user behavior, environmental factors, and equipment status are analyzed to determine whether the abnormality occurs or disappears simultaneously with a certain type of factor, identify the inducing factors of the abnormality, and obtain the results of the causal correlation analysis.

[0086] The study integrates the results of persistence analysis, recoverability analysis, and cause correlation analysis, and combines them with the diagnostic characteristics of different types of hearing loss to ultimately determine the corresponding type of hearing loss.

[0087] In one possible embodiment, determining the type of hearing loss based on the persistence analysis results, the recoverability analysis results, and the causation correlation analysis results includes: determining whether the hearing loss is true based on the persistence analysis results, the recoverability analysis results, and the causation correlation analysis results, obtaining a first type result; determining whether the hearing loss is temporary based on the persistence analysis results and the recoverability analysis results, obtaining a second type result; determining whether the hearing loss is reversible based on the recoverability analysis results and the causation correlation analysis results, obtaining a third type result; and determining the type of hearing loss based on the first type result, the second type result, and the third type result.

[0088] The analysis process involves retrieving three sets of results to determine whether the abnormal temporal changes persist across multiple time windows, cover multiple time windows, and have a stable slope. It also determines whether the correlation between the abnormality and external factors such as environment, equipment, and content is less than a preset threshold, and whether the recoverability analysis shows it is unrecoverable. If all of these conditions are met, the first type of result is determined to be true hearing loss; if any one condition is not met, the first type of result is determined to be non-true hearing loss.

[0089] In this study, the duration of the abnormality in the persistence analysis is considered to determine whether the abnormality only appears within a short window. At the same time, the results of the recoverability analysis are considered to determine whether the abnormality has the ability to recover. If both conditions are met, the second type of result is determined to be temporary hearing loss, which is a type of pseudo-hearing loss.

[0090] The process involves several steps. First, a recoverability analysis determines whether the abnormality is recoverable. Second, a correlation analysis of the underlying causes determines if there are clear and reversible causes, such as environmental noise, equipment location, or physiological state. The correlation between the cause and the abnormal change must be greater than or equal to a preset threshold, and the abnormality must recover synchronously after the cause is eliminated. If both conditions are met, the third type of result is classified as reversible hearing loss, which is considered pseudo-hearing loss.

[0091] For example, during continuous monitoring, a user discovered that their equivalent hearing loss value consistently exceeded the predicted range in both the short-term and medium-term windows, with the abnormal duration covering both time windows. Through correlation analysis, the correlation between this abnormality and environmental noise, device deployment status, and playback content was all less than the fourth threshold, ruling out external causes. At the same time, multiple frequency bands showed specific declines; for example, a continuous decline in the high-frequency band was determined to be a true hearing loss. Further, by combining the fitting slope, the user was able to distinguish between gradual, sudden, and fluctuating true hearing loss.

[0092] For example, if the abnormal timing change only occurs within a short window, and there is strong noise exposure before the abnormality occurs, such as a noise level greater than or equal to 85 dB(A) and a duration greater than or equal to 30 minutes, and it can subsequently recover to the predicted range, then it is determined to be a temporary threshold shift and ear fatigue.

[0093] For example, if the abnormal change is synchronized with an increase in ambient noise, and its correlation is greater than or equal to the fifth threshold, and the situation returns to normal after changing the scene, it is determined to be a false decrease in environmental factors; if the abnormal change is synchronized with the movement of the device or changes in its placement angle, and its correlation is greater than or equal to the fifth threshold, and the situation returns to normal after adjusting the device's layout, it is determined to be a false decrease in device factors; if the abnormal change only occurs during the playback of a specific type of content, and its correlation is greater than or equal to the fifth threshold, and the situation returns to normal after switching content, it is determined to be a false decrease in content factors; if the abnormal change is synchronized with the user's physiological state such as a cold or fatigue, and its correlation is greater than or equal to the fifth threshold, and the abnormality disappears after the user's state recovers, it is determined to be a false decrease in physiological factors.

[0094] In one possible embodiment, determining the hearing loss type based on the first type of result, the second type of result, and the third type of result includes: detecting at least two non-negative results among the first type of result, the second type of result, and the third type of result, thus determining the existence of multiple hearing loss types; if the multiple hearing loss types do not include true hearing loss, then determining at least one hearing impairment factor based on the causal correlation analysis results; determining the matching relationship between each hearing impairment factor and the abnormal temporal change in the time dimension, logical correlation dimension, and impairment characteristic dimension, thus obtaining the contribution of each hearing impairment factor; and determining the hearing loss type based on the contribution.

[0095] If, based on the above results, multiple types of hearing loss are identified, and it is determined whether there is true hearing loss among these multiple types, and if there is true hearing loss, and the correlation degree corresponding to each cause is greater than or equal to the sixth threshold, then it is identified as mixed hearing loss. A weighted allocation algorithm is used to calculate the contribution weight based on the correlation degree of each cause, identify the dominant cause, and then generate a hearing protection strategy based on the dominant cause.

[0096] If the identified multiple types of hearing loss do not include actual hearing loss, then at least one hearing impairment factor is extracted and identified based on the causal correlation analysis results. Specifically, all factors associated with abnormal temporal changes in the causal correlation analysis are extracted, such as environmental noise, device movement, user physiological state, and playback content. Factors with a correlation degree greater than or equal to a preset threshold are selected as hearing impairment factors, while irrelevant factors with low correlation degrees are excluded. This ensures that the extracted impairment factors have a clear correlation with the abnormal changes. For example, "increased environmental noise" and "distance between user and device" are selected as impairment factors with correlation degrees of 0.8 and 0.6, respectively.

[0097] This process involves determining the matching relationship between each hearing impairment factor and abnormal temporal changes across the time, logical correlation, and impairment characteristic dimensions, and then calculating the contribution of each impairment factor. It also involves determining whether the appearance / disappearance time of the impairment factor is synchronized with the start / end time of the abnormal temporal change; verifying whether the abnormal temporal change is synchronously alleviated or delayed after the impairment factor is eliminated; and determining the degree of feature overlap between the characteristics corresponding to the impairment factor and the impairment characteristics of the abnormal temporal change.

[0098] The contribution is calculated based on the matching results of the three dimensions mentioned above. If all three dimensions match completely, a high score is assigned, and the contribution is in the high range. If two dimensions match completely and one dimension matches partially, a medium score is assigned, and the contribution is in the medium range. If only one dimension matches completely or multiple dimensions match partially, a low score is assigned, and the contribution is in the low range. If none of the three dimensions match, the contribution is 0. The contribution of each hearing impairment factor is calculated by assigning values ​​based on multi-dimensional matching to ensure that the contribution calculation is consistent with the actual correlation.

[0099] The primary type of hearing loss is determined by prioritizing the factor with the highest contribution. If two or more factors have similar contribution levels (e.g., a difference of less than or equal to 0.1), it is classified as a mixed type. For example, if environmental noise contributes 0.9 and the distance between the user and the device contributes 0.3, it is classified as reversible temporary hearing loss caused by environmental factors. If both environmental noise and user fatigue contribute 0.8, it is classified as reversible temporary hearing loss caused by a combination of environmental and physiological factors.

[0100] As can be seen, in this embodiment, various types of hearing loss can be accurately distinguished. By detecting multiple types of results, various types of hearing loss can be identified, avoiding omissions and improving the comprehensiveness of hearing status assessment. After excluding true hearing loss, the factors causing hearing damage are located based on the correlation analysis of causes. By combining multi-dimensional matching and contribution calculation, the influence degree of different causes is distinguished, the primary and secondary causes are clarified, the probability of misjudgment is reduced, and the judgment results are more consistent with the actual causes. At the same time, through multi-dimensional time-series data fusion and correlation analysis, various types of hearing loss are accurately identified, improving the accuracy of type identification.

[0101] S250, adjust the audio output parameters of the hearing aid device according to the type of hearing loss.

[0102] Among them, the corresponding gain adjustment strategy, reminder strategy and hearing protection strategy are dynamically adapted according to different hearing loss types, and the effectiveness of the strategy is evaluated and dynamically optimized regularly.

[0103] Specifically, hearing loss type assessment reports are generated in chronological order based on timestamps. The reports include hearing status, abnormality type, abnormal cause, duration, and trend for each time period. The assessment results can be visualized and displayed via an app associated with the hearing aid, generating trend curves.

[0104] Among them, a differentiated fitting strategy is implemented for different types of hearing loss to ensure the effectiveness of hearing aids while avoiding secondary damage.

[0105] For example, in the case of gradual, genuine hearing loss, the personalized frequency response gain curve is gradually adjusted according to the rate of hearing loss change in each frequency band, with the adjustment range strictly controlled to be less than or equal to 2dB / week. This avoids discomfort caused by sudden gain changes and achieves long-term stable adaptation. For example, in the case of sudden, genuine hearing loss, a medical testing alert is immediately triggered, and a safety gain mode is simultaneously activated, limiting the maximum output volume to within 85dB(A). This ensures basic listening needs are met while preventing further hearing damage. For example, in the case of fluctuating, genuine hearing loss, the compression ratio of the gain curve is dynamically adjusted based on the amplitude and frequency of hearing loss fluctuations, adapting to the dynamic changes in hearing in real time to ensure a good listening experience under different fluctuation conditions.

[0106] For example, if it is a temporary threshold shift, the compensation gain of the corresponding frequency band is increased in a short period of time, and the gain adjustment is less than or equal to 3dB. At the same time, an ear rest reminder is issued to guide the user to reduce auditory load. When it is detected that the hearing loss prediction range has been restored, the gain is automatically reduced to the normal level, and the long-term hearing loss trend model is not updated to avoid temporary fluctuations affecting the long-term adaptation strategy.

[0107] For example, if the hearing loss is pseudo-hearing loss, the adaptation operation is performed according to the cause. For environmental causes, the gain curve is automatically adjusted to counteract noise masking, or suggestions to change the listening position are pushed; for device causes, the user is promptly prompted to adjust the device position or angle, or the device's operating status is checked to troubleshoot the problem; for content causes, the gain fine-tuning parameters of the current content type are optimized to adapt to the acoustic characteristics of different content; for physiological causes, the user is prompted to rest or seek symptomatic treatment, and the long-term gain strategy is not adjusted for the time being to avoid ineffective intervention.

[0108] For example, in the case of mixed hearing loss, the adaptation strategy is first implemented for the dominant cause, while distinguishing between true hearing loss and pseudo-hearing loss. Long-term gain adjustment is performed on the true hearing loss, and short-term compensation is implemented on the pseudo-hearing loss, so as to achieve an adaptation effect that takes into account both long-term improvement and short-term relief. The strategy weights corresponding to various causes are updated weekly based on the latest dynamic changes to ensure that the adaptation strategy is accurately matched with the changes in hearing status.

[0109] The strategy evaluation cycle is set daily, integrating user operation data and dynamic change characteristics to assess the effectiveness of the current adaptation strategy. If the user's volume adjustment continues to be abnormal, the strategy parameters are fine-tuned in a timely manner to form a complete closed loop and continuously improve the accuracy and comfort of hearing aid fitting.

[0110] As can be seen, in this embodiment, multi-dimensional time-series data is used to extract features over multiple time windows, enabling real-time capture and trend analysis of hearing status; based on dynamic features, multi-frequency hearing loss intervals are predicted and abnormal temporal changes are identified, providing timely warnings of hearing loss risks; through analysis of the persistence, recoverability, and causal correlation of abnormalities, the type of hearing loss is accurately identified; finally, audio output parameters are adaptively adjusted according to the type of hearing loss, effectively improving the personalization and scenario-based adaptation capabilities of hearing aids, making the hearing aid effect more in line with the user's actual hearing needs, and meeting the requirements of precise hearing management.

[0111] It also realizes closed-loop hearing health management, constructing a complete closed loop of data collection, type determination, strategy adaptation, feedback optimization and medical guidance management. It can output precise gain adjustment strategies for different types of hearing loss, and provide professional medical intervention guidance for abnormal hearing states, thereby enhancing the medical value of the plan and realizing scientific hearing health management.

[0112] In addition, the embodiments of this application can adapt to the usage needs of multiple scenarios, dynamically adjust the adaptation strategy according to different environmental scenarios and audio content types, and ensure that users can obtain a stable and clear listening experience in various scenarios such as quiet environment, noisy environment, listening to music and using navigation, so as to meet diverse daily usage needs.

[0113] In one possible implementation, users cannot intuitively understand the reasons for hearing changes and the system's judgment logic, lack trust in operations such as gain adjustment, and lack convenient auditory feedback channels, making it difficult to achieve bidirectional adaptation optimization. Therefore, this embodiment builds a visual interactive interface to intuitively present the hearing status and system judgment logic, while providing convenient auditory feedback entry points to enhance user trust and facilitate bidirectional adaptation optimization.

[0114] Specifically, please refer to Figure 3, which is a schematic diagram of a visual interactive interface provided in an embodiment of this application. As shown in Figure 3, the visual interactive interface includes a hearing health scoring module, a time-series trend module, a status label interpretation module, and a function interaction entry module. The hearing health scoring module is used to present the user's current hearing health level and data timeliness. Specifically, it includes a real-time hearing health score (87 points) on a scale of 0 to 100, the corresponding health level label (good), the data update time (today 14:32), and the module name identifier (hearing health scoring card).

[0115] The time-series trend module is used to display the long-term changes in a user's hearing status and mark abnormal nodes. Specifically, it includes a trend mini-chart of the hearing health score or equivalent hearing loss value over 30 days, as well as time points marked with red dots indicating abnormal fluctuations in hearing status.

[0116] The status label interpretation module is used to display the judgment result and adaptation status of the current hearing status, specifically including labels representing the nature of hearing fluctuations (normal fluctuations), labels representing the system adaptation optimization status (optimized), and labels explaining the judgment result of abnormal causes (none).

[0117] The functional interaction entry module provides an operational channel for viewing in-depth data and providing two-way feedback. Specifically, it includes two function buttons: "View Detailed Trends" and "Hearing Feedback," which respectively provide users with an in-depth tracing path for hearing data and an entry point for submitting listening experience feedback to the system.

[0118] The hearing health scoring module and the function interaction entry module are related, meaning that the hearing health score can be further viewed or feedback can be submitted through the function interaction entry; the time-series trend module and the status label interpretation module are related, with changes in hearing trends serving as the basis for interpreting status labels.

[0119] After clicking the "View Detailed Trends" button, the user enters the detailed hearing health trend interface, as shown in Figure 4. Figure 4 is a schematic diagram of a detailed hearing health trend interface provided in this embodiment of the application. As shown in Figure 4, the detailed hearing health trend interface is a secondary page that presents in-depth hearing status data and system judgment results, including a multi-band hearing loss trend display module, a correlation factor display module, and a hearing loss type judgment display module. The multi-band hearing loss trend display module is used to present the changes in the user's multi-band hearing loss values ​​over time. The horizontal axis within the module covers a 30-day monitoring period in days, representing the temporal sequence of changes in hearing status over a long period. The vertical axis is marked with decibels to represent the hearing loss values ​​at different time points, reflecting the changes in hearing loss values. The black curves within the module are multi-band hearing loss trend lines, with each line corresponding to an independent hearing frequency band, showing the trend of hearing loss values ​​in that frequency band over time. Through the distribution of multiple black lines, the synchronous changes and differences in hearing status across different frequency bands can be visually observed.

[0120] The multi-band hearing loss trend display module also shows a reference range for normal hearing loss, presented as a light green curve. This range represents the model's predicted range and serves as a reference for determining whether hearing loss is abnormal. When the actual hearing loss trend exceeds the normal range, it is displayed with a bold red line, helping users intuitively identify the frequency band and time of abnormal hearing loss fluctuations. The light green rectangles within the module highlight key periods of abnormal hearing loss fluctuations. The area within the rectangle corresponds to the time range where the hearing loss trend significantly exceeds the normal reference range, allowing users to quickly locate key periods of significant changes in hearing status and focus on analyzing the related influencing factors within those periods.

[0121] The correlation factor display module is used to present the temporal changes of various influencing factors related to hearing loss. For example, on the 6th day of this month, the influencing factor for hearing loss is the change in accommodation, showing the curve of accommodation changing over time; on the 10th day of this month, the influencing factor for hearing loss is the ambient noise level, showing the curve of noise level changing over time; and on the 17th day of this month, the influencing factor for hearing loss is the change in device distance, showing the curve of device distance changing over time, so as to show the synchronous relationship between these factors and hearing loss changes.

[0122] The hearing loss type determination and display module is used to present the system's determination results and basis for the current hearing status. For example, if hearing loss is detected on the 17th day of this month, the determination type is temporary threshold shift. At the same time, it displays no less than 3 determination criteria and confidence information to help users understand the logic and results of the system's determination.

[0123] In one possible embodiment, when the system detects an abnormal hearing decline, a pop-up alert appears, leading to an abnormal hearing alert interactive interface. Please refer to Figure 5, which is a schematic diagram of an abnormal hearing alert interactive interface provided in an embodiment of this application. As shown in Figure 5, the abnormal hearing alert interactive interface includes a detection result prompt module, a cause analysis display module, and a function operation button module. The detection result prompt module is used to inform the user of the specific situation of the abnormal hearing change and the system's preliminary judgment. For example, it includes a title text indicating that an abnormal hearing change has been detected, and a prompt text such as "Your high-frequency hearing has declined significantly in the past three days, exceeding the normal prediction range, and is suspected to be a real hearing loss. It is recommended to conduct a professional test as soon as possible," clearly presenting the specific manifestations of the hearing abnormality and the preliminary conclusion given by the system.

[0124] The cause analysis module lists the possible causes of hearing abnormalities identified by the system, i.e., the cause analysis. Examples include "increased recent noise exposure," "continuous increase in high-frequency accommodation," and "no obvious equipment or environmental abnormalities." Each cause is accompanied by a corresponding number, visually presenting the direction of the abnormality's cause deduced by the system. Clicking on the number leads to a detailed analysis interface for that cause. This interface displays the time-series change curve of the corresponding cause, its synchronous correlation with changes in hearing abnormalities, and the specific monitoring data supporting the cause analysis, helping users gain a deeper understanding of the relationship between the cause and the hearing abnormality.

[0125] The function operation button module provides users with subsequent operation guidance and function access, including function buttons such as viewing detailed analysis, scheduling hearing tests, and providing reminders later. These provide users with in-depth viewing paths for the causes of abnormalities, channels for scheduling professional tests, and operation options for delayed reminders.

[0126] In one possible embodiment, after the system automatically executes a personalized gain adjustment strategy based on the hearing loss type determination result, a gain adjustment interactive interface will pop up. Please refer to Figure 6. Figure 6 is a schematic diagram of a gain adjustment interactive interface provided in an embodiment of this application. As shown in Figure 6, the gain adjustment interactive interface displays the adjustment scheme to the user and collects listening feedback so as to optimize the adaptation strategy in the future.

[0127] The gain adjustment interface includes a gain curve comparison module, an adjustment instructions module, a scene adaptation module, and a listening feedback module. The gain curve comparison module displays the changes in the gain curve before and after the adjustment, including two curves before and after the adjustment. Furthermore, it can also indicate the frequency band and amplitude of the adjustment, such as high frequency +2dB, to intuitively present the specific content of the gain adjustment.

[0128] The adjustment explanation module explains the strategy basis for this gain adjustment, including a textual explanation of the reasons for the strategy, explaining the logic of the system's adjustment, and allowing users to understand the rationality of the adjustment. For example, the reason for the adjustment is based on your recent high-frequency hearing loss trend, and the system has automatically optimized the high-frequency gain.

[0129] The scene adaptation module displays the adaptation status in different scenes, including adaptation tags for quiet environments, noisy environments, and music scenes. It shows whether each scene has been optimized or provides the default option to help users understand and adjust the adaptation in different scenes.

[0130] The listening feedback module is used to collect users' subjective feelings about the current gain adjustment, including a rating option of 1 to 5 stars to provide feedback on the user's evaluation of the current sound, as well as submit and cancel buttons. Users can provide feedback on their listening experience through rating, which provides a basis for system optimization.

[0131] As can be seen, the embodiments of this application can improve the user interaction experience and trust. The visual interface intuitively displays hearing trends, judgment criteria and adjustment logic, reducing the user's understanding threshold. The addition of hearing feedback channels enables two-way adaptation between the system and the user, improving the user's acceptance and trust in the solution.

[0132] Similar to the above embodiments, please refer to Figure 7. Figure 7 is a functional unit block diagram of a hearing management device based on a hearing aid provided in this application embodiment. As shown in Figure 7, the hearing management device 70 based on a hearing aid includes: a feature extraction unit 71, used to extract features from a time-series dataset collected by the hearing aid during audio playback to obtain dynamic change features of multiple hearing state parameters in different time windows. The time-series dataset includes user-side data, device-side data, and environmental-side data, with each side data corresponding to at least one hearing state parameter; and a prediction unit 72, used to predict the dynamic change features based on the dynamic change features. The system measures the upper and lower limits of hearing loss across multiple frequency bands to obtain a predicted hearing loss range for each band. A matching unit 73 matches the target hearing loss curve with the predicted range to determine if any abnormal temporal changes exist. The target hearing loss curve is determined based on the dynamic characteristics of the shortest time window. An analysis unit 74 detects abnormal temporal changes and analyzes their persistence, recoverability, and causal correlation to determine the type of hearing loss. An adjustment unit 75 adjusts the audio output parameters of the hearing aid device according to the type of hearing loss.

[0133] In one possible embodiment, when an abnormal temporal change is detected, the analysis unit 74 is specifically used to: determine the duration of the abnormal temporal change in different time windows; obtain the slope of the sub-hearing loss change curve corresponding to the abnormal temporal change; analyze the duration of the abnormal temporal change based on the duration, the number of time windows covered by the duration, and the slope of the change, to obtain a duration analysis result; divide the baseline hearing loss curve and the recovery hearing loss curve based on the sub-hearing loss change curve; analyze the recoverability of the abnormal temporal change based on the baseline hearing loss curve, the recovery hearing loss curve, and the sub-hearing loss change curve, to obtain a recoverability analysis result; analyze the correlation between the abnormal temporal change and user behavior, environmental factors, and device status, to obtain a causation correlation analysis result; and determine the type of hearing loss based on the duration analysis result, the causation correlation analysis result, and the recoverability analysis result.

[0134] In one possible embodiment, in determining the type of hearing loss based on the persistence analysis results, the recoverability analysis results, and the cause-related analysis results, the analysis unit 74 is further configured to: determine whether the hearing loss is genuine based on the persistence analysis results, the recoverability analysis results, and the cause-related analysis results, obtaining a first type of result; determine whether the hearing loss is temporary based on the persistence analysis results and the recoverability analysis results, obtaining a second type of result; determine whether the hearing loss is reversible based on the recoverability analysis results and the cause-related analysis results, obtaining a third type of result; and determine the type of hearing loss based on the first type of result, the second type of result, and the third type of result.

[0135] In one possible embodiment, in determining the hearing loss type based on the first type of result, the second type of result, and the third type of result, the analysis unit 74 is further configured to: detect at least two non-negative results among the first type of result, the second type of result, and the third type of result, and determine that there are multiple hearing loss types; if the multiple hearing loss types do not include actual hearing loss, then determine at least one hearing impairment factor based on the causal correlation analysis results; determine the matching relationship between each hearing impairment factor and the abnormal temporal change in the time dimension, logical correlation dimension, and impairment feature dimension, and obtain the contribution of each hearing impairment factor; and determine the hearing loss type based on the contribution.

[0136] In one possible embodiment, in analyzing the recoverability of the abnormal temporal change based on the baseline hearing loss curve, the recovery hearing loss curve, and the sub-hearing loss change curve to obtain a recoverability analysis result, the analysis unit 74 is specifically further configured to: determine the recovery efficiency of the abnormal temporal change based on the baseline hearing loss curve, the recovery hearing loss curve, and the sub-hearing loss change curve; determine the recovery duration of the abnormal temporal change based on the sub-hearing loss change curve and the recovery hearing loss curve; determine the stability of the abnormal temporal change after recovery based on the recovery hearing loss curve; and analyze the recoverability of the abnormal temporal change based on the recovery efficiency, the recovery duration, and the stability to obtain the recoverability analysis result.

[0137] In one possible embodiment, in extracting features from a time-series dataset collected by the hearing aid device during audio playback to obtain the dynamic change features of multiple hearing state parameters in different time windows, the feature extraction unit 71 is further configured to: segment the time-series dataset according to preset different time windows to obtain multiple data sequences for each hearing state parameter; perform a fitting operation on each data sequence in the multiple data sequences to obtain a fitting curve for each hearing state parameter under the corresponding time window; and obtain the trend features of each hearing state parameter under different time windows based on the changing direction and first changing rate of the fitting curve. The trend features are used to identify abrupt changes in the data sequence; based on the magnitude and rate of change of the abrupt changes in the corresponding time window, the abrupt change features of each hearing state parameter in different time windows are obtained; based on the fluctuation frequency and amplitude of the abrupt changes in the corresponding time window, the fluctuation features of each hearing state parameter in different time windows are obtained; correlation analysis is performed on hearing state parameters of different dimensions to obtain correlation features between the multiple hearing state parameters; based on the trend features, the abrupt change features, the fluctuation features, and the correlation features, the dynamic change features of each hearing state parameter in different time windows are obtained.

[0138] In one possible embodiment, in matching the target hearing loss change curve with the hearing loss prediction interval to determine whether there is an abnormal temporal change, the prediction unit 72 is further configured to: detect that the target hearing loss change curve exceeds the hearing loss prediction interval, determine the duration for which the target hearing loss change curve exceeds the hearing loss prediction interval; determine the fluctuation amplitude of the target hearing loss change curve; and determine whether there is an abnormal temporal change based on the duration and / or the fluctuation amplitude.

[0139] It is understood that since the method embodiments and the device embodiments are different presentations of the same technical concept, the content of the method embodiment section in this application should be adapted to the device embodiment section, and will not be repeated here.

[0140] In the case of using integrated units, please refer to Figure 8, which is a functional unit block diagram of another hearing management device based on a hearing aid according to an embodiment of this application. As shown in Figure 8, the hearing management device 70 based on a hearing aid includes a processing module 702 and a communication module 701. The processing module 702 is used to control and manage the actions of the hearing management device 70 based on the hearing aid, for example, executing the steps of the feature extraction unit 71, the prediction unit 72, the matching unit 73, the analysis unit 74, and the adjustment unit 75, and / or to execute other processes of the technology described herein. The communication module 701 is used for interaction between the hearing management device 70 based on the hearing aid and other devices.

[0141] As shown in Figure 8, the hearing management device 70 based on the hearing aid device may further include a storage module 703, which is used to store the program code and data of the hearing management device 70 based on the hearing aid device.

[0142] The processing module 702 can be a processor or controller, such as a central processing unit (CPU), a general-purpose processor, a digital signal processor (DSP), an application-specific integrated circuit (ASIC), a field-programmable gate array (FPGA), or other programmable logic devices, transistor logic devices, hardware components, or any combination thereof. It can implement or execute the various exemplary logic blocks, modules, and circuits described in conjunction with the disclosure of this application. The processor can also be a combination that implements computational functions, such as a combination of one or more microprocessors, a combination of a DSP and a microprocessor, etc.

[0143] The communication module 701 can be a transceiver, RF circuit, or communication interface, etc. The storage module 703 can be a memory.

[0144] All relevant content for each scenario involved in the above method embodiments can be referenced from the functional descriptions of the corresponding functional modules, and will not be repeated here. The hearing management device 70 based on the hearing aid device described above can execute the hearing management method based on the hearing aid device shown in Figure 2.

[0145] Please refer to Figure 9, which is a schematic diagram of the structure of an electronic device proposed in an embodiment of this application. As shown in Figure 9, the electronic device 900 includes a processor 910, a memory 920, a communication interface 930, and one or more programs 921. The one or more programs 921 are stored in the memory and configured to be executed by the processor. When the program is executed, it includes some or all of the steps of any hearing management method based on hearing aid device described in the above method embodiments. The processor, memory, and communication interface are interconnected and complete communication between them.

[0146] The memory can be volatile memory such as Dynamic Random Access Memory (DRAM) or non-volatile memory such as a hard disk drive. The memory stores a set of executable program code, and the processor calls the executable program code stored in the memory to execute some or all of the steps of any of the hearing management methods based on hearing aids described in the above embodiments.

[0147] As can be seen, the electronic device 900 described in this application embodiment first extracts features from the time-series dataset collected by the hearing aid device during audio playback to obtain the dynamic change features of multiple hearing status parameters in different time windows. The time-series dataset includes user-side data, device-side data, and environmental-side data, with each side data corresponding to at least one hearing status parameter. Then, based on the dynamic change features, it predicts the upper limit change curve and lower limit change curve of hearing loss for multiple frequency bands to obtain the hearing loss prediction interval for each frequency band. Next, it matches the target hearing loss change curve with the hearing loss prediction interval to determine whether there is an abnormal time-series change. The target hearing loss change curve is determined based on the dynamic change features of the shortest time window. Then, it detects the existence of abnormal time-series changes and analyzes the persistence, recoverability, and causal correlation of the abnormal time-series changes to determine the type of hearing loss. Finally, it adjusts the audio output parameters of the hearing aid device according to the type of hearing loss.

[0148] This application achieves real-time capture and trend analysis of hearing status by extracting features from multi-dimensional time-series data across multiple time windows; it predicts multi-frequency hearing loss intervals based on dynamic features and identifies abnormal temporal changes, providing timely warnings of hearing loss risks; it accurately identifies the type of hearing loss by analyzing the persistence, recoverability, and correlation of causes of abnormalities; and it adaptively adjusts audio output parameters according to the type of hearing loss, effectively improving the personalization and scenario-based adaptation capabilities of hearing aids, making the hearing aid effect more in line with the user's actual hearing needs, and meeting the requirements of precise hearing management.

[0149] This application also provides a computer storage medium storing a computer program for electronic data interchange, which causes a computer to perform some or all of the steps of any of the methods described in the above method embodiments, wherein the computer includes an electronic device.

[0150] This application also provides a computer program product, which includes a non-transitory computer-readable storage medium storing a computer program operable to cause a computer to perform some or all of the steps of any of the methods described in the above method embodiments. The computer program product may be a software installation package, and the computer may include an electronic device.

[0151] It should be noted that, for the sake of simplicity, the aforementioned methods are described as a series of actions. However, those skilled in the art should understand that this application is not limited to the described order of actions, as some steps may be performed in other orders or simultaneously. Furthermore, those skilled in the art should also understand that the embodiments described in the specification are optional, and the actions and modules involved are not necessarily essential to this application.

[0152] In the above embodiments, the descriptions of each embodiment have their own emphasis. For parts not described in detail in a certain embodiment, please refer to the relevant descriptions of other embodiments.

[0153] In the several embodiments provided in this application, it should be understood that the disclosed apparatus can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection between devices or units may be electrical or other forms.

[0154] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment, depending on actual needs.

[0155] Furthermore, the functional units in the various embodiments of this application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software program module.

[0156] If the integrated unit is implemented as a software program module and sold or used as an independent product, it can be stored in a computer-readable storage device (CMD). Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a memory and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this application. The aforementioned memory includes various media capable of storing program code, such as USB flash drives, read-only memory (ROM), random access memory (RAM), portable hard drives, magnetic disks, or optical disks.

[0157] Those skilled in the art will understand that all or part of the steps in the various methods of the above embodiments can be implemented by a program instructing related hardware. The program can be stored in a computer-readable storage device, which may include: a flash drive, a read-only memory, a random access memory, a magnetic disk, or an optical disk, etc.

[0158] The embodiments of this application have been described in detail above. Specific examples have been used to illustrate the principles and implementation methods of this application. The above description of the embodiments is only for the purpose of helping to understand the method and core ideas of this application. At the same time, for those skilled in the art, there will be changes in the specific implementation methods and application scope based on the ideas of this application. Therefore, the content of this specification should not be construed as a limitation of this application.

Claims

1. A hearing management method based on hearing aids, characterized in that, include: Feature extraction is performed on the time-series dataset collected by the hearing aid device during audio playback to obtain the dynamic change characteristics of multiple hearing status parameters in different time windows. The time-series dataset includes user-side data, device-side data, and environmental-side data, with each side data corresponding to at least one hearing status parameter. Based on the dynamic change characteristics, upper and lower limit change curves of hearing loss for multiple frequency bands are predicted to obtain the hearing loss prediction interval for each frequency band. The target hearing loss change curve is matched with the hearing loss prediction interval to determine whether there are abnormal time-series changes. The target hearing loss change curve is determined based on the dynamic change characteristics of the shortest time window. If abnormal time-series changes are detected, the persistence, recoverability, and causal correlation of the abnormal time-series changes are analyzed to determine the type of hearing loss. The audio output parameters of the hearing aid device are adjusted according to the type of hearing loss.

2. The method according to claim 1, characterized in that, The detection of abnormal temporal changes involves analyzing the persistence, recoverability, and causal correlation of these abnormal temporal changes to determine the type of hearing loss. This includes: determining the duration of the abnormal temporal changes in different time windows; obtaining the slope of the sub-hearing loss change curve corresponding to the abnormal temporal changes; analyzing the persistence of the abnormal temporal changes based on the duration, the number of time windows covered by the duration, and the slope, to obtain a persistence analysis result; dividing the sub-hearing loss change curve into a baseline hearing loss curve and a recovery hearing loss curve; analyzing the recoverability of the abnormal temporal changes based on the baseline hearing loss curve, the recovery hearing loss curve, and the sub-hearing loss change curve, to obtain a recoverability analysis result; analyzing the correlation between the abnormal temporal changes and user behavior, environmental factors, and device status, to obtain a causal correlation analysis result; and determining the type of hearing loss based on the persistence analysis result, the causal correlation analysis result, and the recoverability analysis result.

3. The method according to claim 2, characterized in that, The step of determining the type of hearing loss based on the results of the persistence analysis, the recoverability analysis, and the cause-related analysis includes: determining whether the hearing loss is true based on the results of the persistence analysis, the recoverability analysis, and the cause-related analysis, to obtain a first type of result; determining whether the hearing loss is temporary based on the results of the persistence analysis and the recoverability analysis, to obtain a second type of result; determining whether the hearing loss is reversible based on the results of the recoverability analysis and the cause-related analysis, to obtain a third type of result; and determining the type of hearing loss based on the first type of result, the second type of result, and the third type of result.

4. The method according to claim 3, characterized in that, The step of determining the hearing loss type based on the first type of result, the second type of result, and the third type of result includes: detecting at least two non-negative results among the first type of result, the second type of result, and the third type of result, thus determining that there are multiple hearing loss types; if the multiple hearing loss types do not include actual hearing loss, then determining at least one hearing impairment factor based on the causal correlation analysis results; determining the matching relationship between each hearing impairment factor and the abnormal temporal change in the time dimension, logical correlation dimension, and impairment characteristic dimension, thus obtaining the contribution of each hearing impairment factor; and determining the hearing loss type based on the contribution.

5. The method according to claim 2, characterized in that, The step of analyzing the recoverability of the abnormal temporal change based on the baseline hearing loss curve, the recovery hearing loss curve, and the sub-hearing loss change curve to obtain a recoverability analysis result includes: determining the recovery efficiency of the abnormal temporal change based on the baseline hearing loss curve, the recovery hearing loss curve, and the sub-hearing loss change curve; determining the recovery duration of the abnormal temporal change based on the sub-hearing loss change curve and the recovery hearing loss curve; determining the stability of the abnormal temporal change after recovery based on the recovery hearing loss curve; and analyzing the recoverability of the abnormal temporal change based on the recovery efficiency, the recovery duration, and the stability to obtain the recoverability analysis result.

6. The method according to any one of claims 1-5, characterized in that, The method of extracting features from the time-series dataset collected by the hearing aid device during audio playback to obtain the dynamic change features of multiple hearing status parameters in different time windows includes: segmenting the time-series dataset according to preset different time windows to obtain multiple data sequences for each hearing status parameter; performing a fitting operation on each of the multiple data sequences to obtain a fitting curve for each hearing status parameter in the corresponding time window; obtaining the trend features of each hearing status parameter in different time windows based on the direction of change and the first rate of change of the fitting curve; identifying abrupt change points in the data sequences based on the trend features; obtaining the abrupt change features of each hearing status parameter in different time windows based on the magnitude of change and the second rate of change of the abrupt change points in the corresponding time window; obtaining the fluctuation features of each hearing status parameter in different time windows based on the fluctuation frequency and fluctuation amplitude of the abrupt change points in the corresponding time window; performing correlation analysis on hearing status parameters of different dimensions to obtain the correlation features between the multiple hearing status parameters; and obtaining the dynamic change features of each hearing status parameter in different time windows based on the trend features, the abrupt change features, the fluctuation features, and the correlation features.

7. The method according to any one of claims 1-5, characterized in that, The step of matching the target hearing loss change curve with the hearing loss prediction interval to determine whether there is an abnormal temporal change includes: detecting that the target hearing loss change curve exceeds the hearing loss prediction interval, determining the duration for which the target hearing loss change curve exceeds the hearing loss prediction interval; determining the fluctuation amplitude of the target hearing loss change curve; and determining whether there is an abnormal temporal change based on the duration and / or the fluctuation amplitude.

8. A hearing management device based on a hearing aid, characterized in that, include: The feature extraction unit extracts features from the time-series dataset collected by the hearing aid device during audio playback to obtain the dynamic change features of multiple hearing status parameters in different time windows. The time-series dataset includes user-side data, device-side data, and environmental-side data, with each side data corresponding to at least one hearing status parameter. The prediction unit predicts the upper and lower limit change curves of hearing loss for multiple frequency bands based on the dynamic change features, obtaining the hearing loss prediction interval for each frequency band. The matching unit matches the target hearing loss change curve with the hearing loss prediction interval to determine whether there is an abnormal time-series change. The target hearing loss change curve is determined based on the dynamic change features of the shortest time window. The analysis unit detects the existence of abnormal time-series changes and analyzes the persistence, recoverability, and causal correlation of the abnormal time-series changes to determine the type of hearing loss. An adjustment unit is used to adjust the audio output parameters of the hearing aid device according to the type of hearing loss.

9. An electronic device, characterized in that, The device includes: a memory, a processor, and executable program code stored in the memory and executable on the processor, wherein when the processor executes the executable program code, it performs the steps of the hearing management method based on a hearing aid device as described in any one of claims 1-7.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores executable program code, which includes execution instructions for performing the steps of the hearing management method based on a hearing aid device as described in any one of claims 1-7.

Images