Multi-mode switching control method of earphone integrated micro-current stimulation device

Abstract

The application discloses a multi-mode switching control method of earphone integrated micro-current stimulation equipment, and relates to the technical field of switching control, and comprises the following steps: acquiring polarity pair detection data sets, and generating polarity asymmetry marks and eccentric direction marks based on the polarity pair detection data sets; generating migration marks based on the polarity asymmetry marks and the eccentric direction marks, and calculating a migration occurrence rate based on the migration marks; calculating an asymmetry proportion based on the polarity asymmetry marks; calculating an eccentric direction dominant degree based on the eccentric direction marks; determining a risk eccentric direction category based on all candidate modes of the earphone integrated micro-current stimulation equipment; determining a target mode based on the asymmetry proportion, the eccentric direction dominant degree, the migration occurrence rate and the risk eccentric direction category, and outputting a micro-current stimulation sequence in the target mode; and the application improves the distinguishing ability of the difference between the change of the conductive path and the real mode effect.

Description

Technical Field

[0001] This invention relates to the field of switching control technology, and in particular to a multi-mode switching control method for an earphone-integrated microcurrent stimulation device. Background Technology

[0002] With the development of wearable health devices, headphone-integrated microcurrent stimulation devices are gradually being applied to various scenarios such as nerve regulation, relaxation, focus assistance, and sleep improvement. These devices typically place electrodes in specific areas of the ear to output weak currents to human tissues in different modes, thereby achieving different physiological regulation purposes. In actual use, the devices are often equipped with multiple stimulation modes, and different regulation targets can be switched through mode switching. In real-world applications, the conditions in the ear area are quite complex. For example, some users have tragus piercings or wear small metal earrings. The contact state between the electrodes and the skin can also be affected by sweat, sebum, or cleaning solution residue. These factors can form a localized moist conductive channel between the electrodes and the skin. At the same time, during head micro-movements, speaking, or swallowing, the connectivity or position of this conductive channel will change, causing the path of the current into human tissue to shift spatially. This results in the same stimulation parameters exhibiting different conductive behaviors and response results at different times.

[0003] Existing technologies typically establish a correspondence between stimulation modes and effects based on feedback signals such as current and voltage, and control mode switching accordingly. However, due to changes in the position of the local moist conductive path, the actual conduction path of the current in the tissue is inconsistent with the set path. Even if the external detection signal is still within an acceptable range, the actual stimulation effect may deviate, leading to misjudgments in the mode determination criteria established based on feedback. This results in inappropriate selections when the device switches between different modes, and may even cause abnormal or uncomfortable stimulation reactions. Existing technologies lack identification and constraint mechanisms for the spatial offset of the conductive path and its dynamic changes, making it difficult to effectively distinguish the relationship between response differences caused by changes in the conductive path and differences in the actual mode effect. In the process of multi-mode switching control, there are still problems of insufficient stability and difficulty in avoiding adverse events. Summary of the Invention

[0004] The purpose of this invention is to address the shortcomings of existing technologies that make it difficult to effectively distinguish the relationship between response differences caused by changes in the conductive path and the differences in the actual mode effect, and to propose a multi-mode switching control method for an integrated microcurrent stimulation device for headphones.

[0005] To address the problems existing in the prior art, the present invention adopts the following technical solution:

[0006] A multi-mode switching control method for an earphone-integrated microcurrent stimulation device includes:

[0007] S1. Obtain the polarity pair detection dataset, and generate polarity asymmetry labels and eccentricity direction labels based on the polarity pair detection dataset;

[0008] S2. Generate migration markers based on polarity asymmetry markers and eccentricity direction markers, and calculate the migration occurrence rate based on the migration markers;

[0009] S3. Calculate the asymmetry ratio based on polarity asymmetry markers;

[0010] S4. Calculate the dominance of the eccentric direction based on the eccentric direction marker;

[0011] S5. Based on all candidate modes of the earphone-integrated microcurrent stimulation device, determine the risk eccentricity direction category;

[0012] S6. Determine the target pattern based on the asymmetry ratio, the dominance of the eccentric direction, the migration rate, and the risk eccentric direction category, and output the microcurrent stimulation sequence under the target pattern.

[0013] Compared with the prior art, the beneficial effects of the present invention are:

[0014] 1. This invention constructs a polarity pair detection dataset and generates polarity asymmetry markers, eccentricity direction markers, and migration markers to characterize the interface state from three aspects: directional differences, spatial offsets, and dynamic changes in the conductive path, thereby improving the ability to distinguish the differences between conductive path changes and the actual mode effect.

[0015] 2. The present invention further calculates the asymmetry ratio, the dominance of the eccentric direction, and the migration occurrence rate, and combines the mode switching event count, adverse event count, and risk eccentric direction category of the candidate mode to constrain and screen different modes, thereby reducing the probability of mode switching misjudgment.

[0016] 3. After determining the target mode, the present invention outputs the corresponding microcurrent stimulation sequence, so that the aforementioned screening results are directly implemented into the actual stimulation process. Even under the state of local wet conductive path deviation, the stimulation output can still be completed with a low risk of adverse events, thus improving the stability of multi-mode switching control. Attached Figure Description

[0017] The accompanying drawings, which are included to provide a further understanding of the invention and form part of this application, illustrate exemplary embodiments of the invention and, together with their description, serve to explain the invention and do not constitute an undue limitation thereof. In the drawings:

[0018] Figure 1 This is a flowchart illustrating a multi-mode switching control method for an integrated microcurrent stimulation device for headphones, provided in an embodiment of the present invention. Detailed Implementation

[0019] The technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments.

[0020] This embodiment provides a multi-mode switching control method for an integrated microcurrent stimulation device for headphones. (See also...) Figure 1 Specifically, including:

[0021] S1. Obtain the polarity pair detection dataset, and generate polarity asymmetry labels and eccentricity direction labels based on the polarity pair detection dataset;

[0022] In an embodiment of the present invention, obtaining a polarity pair detection dataset includes:

[0023] Acquire the raw data stream within one control cycle, collected by the headphone-integrated microcurrent stimulation device under the current stimulation mode;

[0024] The current stimulation mode refers to the waveform and parameter combination of the current applied to the human ear tissue by the earphone-integrated microcurrent stimulation device at a certain moment. This state corresponds to the specific current amplitude, pulse width, and output rhythm. The control cycle refers to a data processing time period divided according to a predetermined time during the continuous operation of the device, which is used to uniformly collect and process the stimulation output and feedback response within this time period. The raw data stream refers to the electrical signal sequence acquired by the device's detection circuit within the control cycle. This electrical signal sequence reflects the current conduction process between the electrode and the human tissue, as well as the changes in the interface electrical characteristics.

[0025] Specifically, the stimulation output electrode outputs a microcurrent stimulation sequence corresponding to the current stimulation mode to the ear tissue. The control circuit synchronously acquires the loop current signal and loop voltage signal corresponding to the microcurrent stimulation sequence in a continuous sampling manner. The loop current signal and loop voltage signal are written into the storage area in the actual sampling order to form continuous time data covering the start time to the end time of the control cycle. Each sampled data in the continuous time data has a corresponding sampling time record. The sampling time record is directly written by the internal timing unit of the device when the sampling action occurs, which is used to indicate the acquisition time of the corresponding loop current signal and the corresponding loop voltage signal, thereby obtaining the original data stream within the control cycle.

[0026] Based on the acquisition times of the positive and negative detection stimulus response segments in the original data stream, a polarity pair detection dataset is formed.

[0027] A positive detection stimulus response segment refers to the time period signal corresponding to the voltage and current changes between the electrode and the tissue when a detection current with the first polarity is applied to the human body; a negative detection stimulus response segment refers to the corresponding time period signal formed when a detection current with the opposite polarity is applied to the human body; the acquisition time refers to the specific time point or time interval boundary at which the device samples and records the above-mentioned electrical signals; a polarity pair detection dataset refers to a data set formed by combining a set of positive detection stimulus response segments and a set of negative detection stimulus response segments according to the time correspondence, used to characterize the differences in the conductive path and interface state between the electrode and human tissue under different current directions.

[0028] Specifically, in the raw data stream, the detection stimulus period in which the current direction is determined to be positive by the device output direction is searched. The continuously acquired loop current signal and loop voltage signal within this detection stimulus period are extracted as positive detection stimulus response segments. The start and end sampling times of this positive detection stimulus response segment are recorded. Then, in the raw data stream, the detection stimulus period in which the current direction is opposite to the positive detection stimulus response segment is searched. The continuously acquired loop current signal and loop voltage signal within this detection stimulus period are extracted as reverse detection stimulus response segments. The start and end sampling times of this reverse detection stimulus response segment are recorded. A positive detection stimulus response segment and a reverse detection stimulus response segment that are temporally corresponding and whose detection actions belong to the same detection output process are paired. The pairs are written into the same data set in the order in which they are formed, resulting in a polarity paired detection dataset.

[0029] In embodiments of the present invention, generating polarity asymmetry labels and eccentricity direction labels based on a polarity pair detection dataset includes:

[0030] Extract the positive and negative response pattern summaries corresponding to the polarity pair detection dataset;

[0031] The forward response morphology summary refers to the representative set of values ​​obtained by extracting the signal of the loop voltage or loop current changing with time during the forward current action. This set of values ​​is used to describe the response amplitude, rate of change and steady state when the current enters human tissue. The reverse response morphology summary refers to the representative set of values ​​obtained by extracting the corresponding loop signal during the reverse current action. This set of values ​​is used to describe the response when the current enters human tissue from the opposite direction.

[0032] Specifically, each pair of positive detection stimulus response segments and their paired negative detection stimulus response segments are sequentially read from the polarity pair detection dataset. Continuous sampling data is extracted according to their respective start and end sampling times. The sampling data in each response segment is divided into a current establishment phase, a continuous conduction phase, and a current removal phase in chronological order. For each positive detection stimulus response segment, the average value of the loop voltage data, the average value of the loop current data, and the average value of the change between adjacent sampling points are calculated for each of the above phases. The obtained values ​​are written into the same data record in a fixed order to form a positive response pattern summary corresponding to that set of data. The same data reading method, phase division method, numerical calculation method, and arrangement method are used for the negative detection stimulus response segments to form a negative response pattern summary. Each set of positive response pattern summaries and their corresponding negative response pattern summaries are stored in the summary data area in order of the detection pair number.

[0033] Determine the sign consistency of the corresponding component differences between the positive and negative response morphological summaries;

[0034] The corresponding component difference refers to the result obtained by subtracting the corresponding values ​​in the positive response pattern summary and the negative response pattern summary. This result is used to reflect the response difference at the same measurement location under different current directions. The sign consistency refers to the case where the above difference shows the same sign in multiple corresponding components, which is used to indicate that the current introduction path shows a consistent offset trend in both directions.

[0035] Specifically, each pair of positive and negative response pattern summaries is read sequentially according to the detection pair number. The values ​​in the positive and negative response pattern summaries that are in the same arrangement position are taken as corresponding components. A subtraction operation is performed on each pair of corresponding components to obtain the corresponding component difference. The corresponding component difference that is greater than zero is recorded as positive, the corresponding component difference that is less than zero is recorded as negative, and the corresponding component difference that is equal to zero is recorded as zero. The signs of all corresponding component differences in the same detection pair are counted item by item to obtain the number of positive signs, the number of negative signs, and the number of zero signs. The sign consistency result of the corresponding component differences of the detection pair is determined based on whether the number of positive signs and the number of negative signs exist at the same time and whether a single sign is dominant.

[0036] Based on symbol consistency, generate polarity asymmetric tags;

[0037] Specifically, the sign consistency results of each group of test pairs in the consistency record area are read sequentially. Cases where only positive signs or only negative signs exist, or where the number of positive signs is greater than the number of negative signs or vice versa, are identified as cases of polarity asymmetry. Cases where the number of positive signs and negative signs are the same and the number of zero signs covers the other corresponding components are identified as cases of non-polarity asymmetry. Polarity asymmetry flags are written for test pairs identified as having polarity asymmetry, and polarity symmetry flags are written for test pairs identified as not having polarity asymmetry.

[0038] Determine the sign relationship between the corresponding component differences in the positive and negative response morphology summaries;

[0039] Polarity asymmetry marking refers to the category identification made based on the consistency of signs of the current electrode and human tissue interface, which is used to indicate the existence of response differences under different current directions; the sign quantity relationship refers to the comparison result between the number of positive values ​​and the number of negative values ​​in the corresponding component difference, which is used to reflect the degree of bias of the conductive path in spatial distribution.

[0040] Specifically, each positive response pattern summary and its corresponding negative response pattern summary are read sequentially according to the detection pair sequence number. A one-to-one correspondence is established according to the arrangement of each component in the two. The value at each position in the positive response pattern summary is subtracted from the value at the same position in the negative response pattern summary to obtain the corresponding component difference. The corresponding component difference greater than zero is recorded as positive, the corresponding component difference less than zero is recorded as negative, and the corresponding component difference equal to zero is recorded as zero. The number of positive, negative, and zero signs is counted for all corresponding component differences of the same detection pair. The counted number of positive, negative, and zero signs is written together with the corresponding detection pair sequence number into the quantity relationship record area, thereby obtaining the positive and negative quantity relationship of the corresponding component differences of the detection pair.

[0041] Generate eccentric direction markers based on the relationship between the number of positive and negative signs.

[0042] Eccentric direction marking refers to the direction category determined by the positive and negative sign relationship, used to indicate the spatial direction of current deviation when entering human tissue; the positive and negative sign relationship is also used to describe the relative relationship of the proportion of response in different directions, thus characterizing the offset direction of the conductive path; eccentric direction marking is further used to distinguish the dominant conductive path of current in different directions; eccentric direction marking obtained by the positive and negative sign relationship is used to indicate the directional distribution state when current is introduced into human tissue.

[0043] Specifically, the number of positive signs, negative signs, and zero signs corresponding to each group of test pairs in the quantity relationship record area are read sequentially. The relationship between the number of positive signs and the number of negative signs is compared. When the number of positive signs is greater than the number of negative signs, the group of test pairs is marked as positive eccentricity. When the number of negative signs is greater than the number of positive signs, the group of test pairs is marked as negative eccentricity. When the number of positive signs and the number of negative signs are equal, the group of test pairs is marked as directionally unstable. The generated eccentricity direction marks are written into the direction mark record area in a one-to-one correspondence with the corresponding test pair sequence number, forming an eccentricity direction mark sequence arranged according to the test pair sequence number.

[0044] S2. Generate migration markers based on polarity asymmetry markers and eccentricity direction markers, and calculate the migration occurrence rate based on the migration markers;

[0045] In embodiments of the present invention, migration markers are generated based on polarity asymmetry markers and eccentricity direction markers, including:

[0046] A micro-motion trigger dataset is formed based on the acquisition time of micro-motion trigger response segments in the original data stream;

[0047] A micro-motion triggered response segment refers to the range of electrical signal changes caused by changes in the contact state between the electrode and the skin during speech, swallowing, or slight head movements. The current or voltage waveform corresponding to this range exhibits fluctuations or abrupt changes relative to a stable state. The acquisition time refers to the time position or time interval boundary at which the above electrical signals are sampled and recorded, used to determine the positional relationship of the signal segment on the time axis. The micro-motion triggered dataset refers to a data set formed by combining multiple micro-motion triggered response segments in chronological order, used to reflect the electrical signal performance when the contact state changes dynamically.

[0048] Specifically, the sampling time records and loop response data corresponding to each sampled data are read sequentially along the time axis of the original data stream. The voltage and current changes between adjacent sampling points in the loop response data are calculated item by item. The sampling intervals in which the change direction changes continuously and the change amplitude increases continuously are determined as response fluctuation intervals. The response fluctuation intervals are then aligned with the stimulus output records in the same time period. The sampling intervals corresponding to the stimulus initiation and stimulus withdrawal segments are filtered out, and the sampling intervals in the continuous stimulation period and accompanied by changes in contact state are retained as micro-motion trigger response segments. The start sampling time and end sampling time of each micro-motion trigger response segment are recorded, and the continuous loop voltage data and continuous loop current data within the corresponding time range are extracted. Each micro-motion trigger response segment is written into the same data storage area in the order of its start sampling time to form a micro-motion trigger dataset.

[0049] A stable snap-in dataset is formed based on the acquisition time of the stable snap-in response segment in the original data stream;

[0050] A stable contact dataset refers to a data set composed of multiple stable contact response segments, used to characterize the electrical signal reference situation when the contact state is stable.

[0051] Specifically, the sampling time records and loop response data corresponding to each sampled data are read sequentially along the time axis of the original data stream. The voltage and current changes between consecutive sampling points are compared item by item. The continuous sampling intervals where the voltage and current changes are consistent are determined as stable response intervals. These stable response intervals are aligned with the stimulus output records. The sampling intervals corresponding to the stimulus initiation and withdrawal segments are filtered out. The sampling intervals where the loop voltage and loop current data change continuously during the continuous stimulus period are retained as stable contact response segments. The start and end sampling times of each stable contact response segment are recorded. The continuous loop voltage and continuous loop current data within the corresponding time range are extracted. The stable contact response segments are written sequentially into the same data storage area according to the order of their start sampling times to form a stable contact dataset.

[0052] Based on the polarity asymmetry label and eccentricity direction label corresponding to the micro-motion triggered dataset and the polarity asymmetry label and eccentricity direction label corresponding to the stable attachment dataset, a migration label is generated.

[0053] Migration labels are category identifiers derived from the changes in electrical signal response between micro-motion triggered datasets and stable contact datasets. They are used to represent the changes in the conductive path at the electrode-human tissue contact interface between different states.

[0054] Specifically, polarity asymmetry markers and eccentricity direction markers corresponding to each micro-motion trigger response segment are read from the marker recording area and arranged sequentially according to the starting sampling time of the micro-motion trigger response segments to form a micro-motion marker sequence. Then, polarity asymmetry markers and eccentricity direction markers corresponding to each stable contact response segment are read from the marker recording area and arranged sequentially according to the starting sampling time of the stable contact response segments to form a stable contact marker sequence. A correspondence is established between each micro-motion trigger response segment and the stable contact response segments that are before and after it in time. The correspondence is determined by comparing the starting sampling time and the ending sampling time of each response segment, so that each micro-motion trigger response segment corresponds to at least one stable contact response segment that is closest to it in time.

[0055] Next, for each group of data with established correspondence, the polarity asymmetry markers in the micro-motion marker sequence are compared with those in the stable contact marker sequence to see if they are the same. The eccentricity direction markers in the micro-motion marker sequence are also compared with those in the stable contact marker sequence to see if they are the same. The corresponding groups where the polarity asymmetry markers or eccentricity direction markers change are identified as having a conductive path migration. The corresponding groups where neither the polarity asymmetry markers nor the eccentricity direction markers change are identified as having no conductive path migration. A migration category marker is written for the corresponding groups identified as having a conductive path migration, and a non-migration category marker is written for the corresponding groups identified as having no conductive path migration. The written category markers are then associated with and saved one by one with the corresponding micro-motion trigger response segment number and the corresponding stable contact response segment number to form a migration marker sequence arranged in chronological order.

[0056] In an embodiment of the present invention, calculating the migration occurrence rate based on migration markers includes:

[0057] Generate a migration tag sequence based on the migration tags;

[0058] A migration marker sequence refers to a data set formed by arranging multiple migration markers in chronological order, which reflects the continuous state of the conductive path as it changes over time.

[0059] Specifically, the migration markers corresponding to each micro-motion trigger response segment, as well as the corresponding micro-motion trigger response segment number, stable contact response segment number, start sampling time, and end sampling time, are sequentially read from the migration marker storage area. The read migration markers are arranged in order from earliest to latest according to the start sampling time of the corresponding micro-motion trigger response segment. For migration markers with the same start sampling time, they are arranged in order from earliest to latest according to the start sampling time of the corresponding stable contact response segment. The arranged migration markers, along with their corresponding time information and segment number, are sequentially written into the same sequence storage area to form a migration marker sequence that is continuously recorded in chronological order.

[0060] Count the migration marker sequences to obtain non-migration counts and migration counts;

[0061] Non-migration count refers to the number of items in the migration marker sequence that are marked as having no change in conductive path. This number is used to characterize the degree to which the interface between the electrode and human tissue maintains a stable conductive state. Migration count refers to the number of items in the migration marker sequence that are marked as having a change in conductive path. This number is used to characterize the frequency of changes in the conductive path at the interface between the electrode and human tissue.

[0062] Specifically, the category content of each migration mark is read sequentially along the migration mark sequence. Migration marks whose category content indicates that no migration has occurred in the conductive path are recorded as non-migration marks, and migration marks whose category content indicates that the conductive path has occurred are recorded as migration marks. For each non-migration mark read, the non-migration count record is incremented by one, and for each migration mark read, the migration count record is incremented by one, until all migration marks in the migration mark sequence have been read and classified and accumulated, resulting in the non-migration count and migration count corresponding to the migration mark sequence.

[0063] The migration occurrence rate is calculated based on the non-migration count and the migration count.

[0064] Migration occurrence rate refers to the proportion of migration counts in the total number of migration marker sequences. This proportion is used to describe the relative extent to which conductive paths change within the observation time range.

[0065] Specifically, the non-migration count and migration count corresponding to the same control cycle are read from the counting result storage area. The migration count and non-migration count are added together to obtain the total number of migration marks. Then, the migration count is divided by the total number of migration marks to obtain the migration occurrence rate.

[0066] It should be noted that the current conduction process between the electrode and the human ear tissue can be regarded as a continuous evolution of the conductive path in time. In this evolution process, each time segment corresponds to a conductive state. When the contact interface is stable, the conductive path remains continuous and its spatial distribution remains basically unchanged. However, when micro-motion or interface disturbance occurs, the conductive path will shift or redistribute, thus forming discrete state change events. The migration marker is regarded as the event record of the conductive path redistribution. The non-migration count corresponds to the number of events in which the conductive path remains stable, and the migration count corresponds to the number of events in which the conductive path changes. Then, the relative degree of change of the conductive path within a given time range can be characterized by the ratio between the number of changing events and the total number of events. This ratio is consistent with the statistical expression of the frequency of random events and can reflect the probability and frequency of the conductive path reconstruction per unit time. Therefore, the value obtained by dividing the migration count by the sum of the migration count and the non-migration count can directly characterize the relative probability of the conductive path migrating within this time range, and is thus defined as the migration occurrence rate.

[0067] S3. Calculate the asymmetry ratio based on polarity asymmetry markers;

[0068] In embodiments of the present invention, calculating the asymmetry ratio based on polarity asymmetry markers includes:

[0069] Arrange all polarity asymmetry markers to obtain a polarity asymmetry marker sequence;

[0070] A polarity asymmetry marker sequence refers to a data set formed by arranging multiple polarity asymmetry markers in chronological order. This data set reflects the changes in the conductive path over a continuous period of time.

[0071] Specifically, the polar asymmetry markers corresponding to each detection pair are read one by one from the polar asymmetry marker storage area, along with the detection pair number, the start sampling time of the positive detection stimulus response segment, the start sampling time of the negative detection stimulus response segment, and the corresponding control period identifier. All polar asymmetry markers belonging to the same control period are arranged sequentially from earliest to latest according to the start sampling time of the positive detection stimulus response segment corresponding to the detection pair. If the start sampling times of the positive detection stimulus response segments are the same, they are arranged again from earliest to latest according to the start sampling time of the corresponding negative detection stimulus response segment. The arranged polar asymmetry markers, along with their corresponding detection pair numbers and time information, are sequentially written into the same sequence storage area to form a polar asymmetry marker sequence recorded continuously in chronological order.

[0072] Counting the polarity-asymmetric labeled sequences yields symmetric and asymmetric counts.

[0073] Symmetrical counting refers to the number of markers in a polarity asymmetric marker sequence that are determined to have small differences in response and even distribution of conductive paths under positive and negative currents. This number is used to characterize the degree to which conductive paths remain relatively consistent in both directions. Asymmetric counting refers to the number of markers in a polarity asymmetric marker sequence that are determined to have significant differences in response and offset distribution of conductive paths under positive and negative currents. This number is used to characterize the degree to which conductive paths offset under different current directions.

[0074] Specifically, the category content of each polar asymmetric marker is read sequentially along the arrangement order of the polar asymmetric marker sequence. The polar asymmetric marker whose category content represents the symmetrical response of the conductive path under the action of positive and negative detection stimuli is recorded as a symmetrical marker, and the polar asymmetric marker whose category content represents the asymmetrical response of the conductive path under the action of positive and negative detection stimuli is recorded as an asymmetric marker. For each symmetrical marker read, the count is increased by one, and for each asymmetric marker read, the count is increased by one, until all polar asymmetric markers in the polar asymmetric marker sequence are read and classified and accumulated, and the symmetrical count and asymmetric count corresponding to the polar asymmetric marker sequence are obtained.

[0075] Calculate the asymmetric proportion based on symmetric and asymmetric counting.

[0076] The asymmetry proportion refers to the ratio of asymmetry counts to the total number of polarity asymmetry labeled sequences. This proportion is used to describe the relative degree to which conductive paths exhibit directional dependence differences over the observation time range.

[0077] Specifically, the symmetrical and asymmetrical counts corresponding to the same control cycle are read from the counting result storage area. The asymmetrical counts are added to the symmetrical counts to obtain the total number of polar asymmetrical markers. Then, the asymmetrical counts are divided by the total number of polar asymmetrical markers to obtain the asymmetrical proportion.

[0078] It should be noted that the conductivity between the electrode and the human ear tissue is influenced by the interface contact state, local wetting degree, and current flow direction. When the conductivity path maintains the same or approximately the same spatial distribution under forward and reverse current, the corresponding loop response exhibits a symmetrical state. However, when the conductivity path becomes dependent on the current direction due to non-uniform conductivity at the interface, local bias current conduction, or changes in the contact boundary, the loop response under forward and reverse current is asymmetrical. Therefore, each polarity asymmetry marker can be regarded as a discrete determination result of whether a single conductivity state has a direction dependence. Within the observation time range, the symmetric count corresponds to the number of events in which the conductive path remains in an unbiased state, and the asymmetric count corresponds to the number of events in which the conductive path exhibits a biased state. The asymmetric count and the symmetric count are added together to obtain the total number of all observed events. The asymmetric count is then divided by the total number of all observed events. The resulting value is the proportion of direction-dependent conductive states in all observed states. This proportion conforms to the basic method in statistical physics of characterizing the degree of state occupancy by the frequency of event occurrence. It can directly reflect the relative degree to which the conductive path exhibits an asymmetric state within the observation interval. Therefore, this calculation result is defined as the asymmetric proportion.

[0079] S4. Calculate the dominance of the eccentric direction based on the eccentric direction marker;

[0080] In an embodiment of the present invention, calculating the dominance of the eccentric direction based on the eccentric direction marker includes:

[0081] Arrange all the eccentric direction markers to obtain the eccentric direction marker sequence;

[0082] An eccentric direction marker sequence refers to a data set formed by arranging multiple eccentric direction markers in chronological order. This data set reflects the directional changes of the conductive path over a continuous period of time.

[0083] Specifically, the eccentric direction markers corresponding to each detection pair are read one by one from the eccentric direction marker storage area, along with the detection pair number, the start sampling time of the positive detection stimulus response segment, the start sampling time of the negative detection stimulus response segment, and the corresponding control period identifier. All eccentric direction markers belonging to the same control period are arranged in order from earliest to latest according to the start sampling time of the positive detection stimulus response segment corresponding to the detection pair. If the start sampling time of the positive detection stimulus response segment is the same, it is arranged in order from earliest to latest according to the start sampling time of the corresponding negative detection stimulus response segment. The arranged eccentric direction markers, together with their corresponding detection pair number and time information, are written into the same sequence storage area to form an eccentric direction marker sequence recorded continuously in chronological order.

[0084] Count the eccentric direction marker sequence to obtain the positive side count and the negative side count;

[0085] The positive side count refers to the number of markers in the eccentric direction marker sequence that are determined to be conductive paths biased to the positive side. This number is used to characterize the frequency of conductive paths shifting to the positive side. The negative side count refers to the number of markers in the eccentric direction marker sequence that are determined to be conductive paths biased to the negative side. This number is used to characterize the frequency of conductive paths shifting to the negative side.

[0086] Specifically, the category content of each eccentric direction mark is read sequentially along the arrangement order of the eccentric direction mark sequence. Eccentric direction marks whose category content represents the conductive path biased to the positive side are recorded as positive side marks, and eccentric direction marks whose category content represents the conductive path biased to the negative side are recorded as negative side marks. For each positive side mark read, the count is increased by one in the positive side count record, and for each negative side mark read, the count is increased by one in the negative side count record, until all eccentric direction marks in the eccentric direction mark sequence have been read and classified and accumulated, thus obtaining the positive side count and negative side count corresponding to the eccentric direction mark sequence.

[0087] Obtain the total number of detection pairs in the eccentric orientation marker sequence;

[0088] Specifically, all eccentric direction markers corresponding to the current control cycle and their corresponding detection pair numbers are read from the eccentric direction marker sequence storage area. The detection pair numbers are read one by one in the order of arrangement and the count is accumulated one by one. Each time a detection pair number is read, the count is increased by one to the total number of detection pairs. This process continues until all the detection pair numbers corresponding to all eccentric direction markers in the current control cycle have been read and accumulated, thus obtaining the total number of detection pairs corresponding to the eccentric direction marker sequence.

[0089] The dominance of the eccentric direction is calculated based on the positive side count, the negative side count, and the total number of detection pairs.

[0090] The total number of detection pairs refers to the total number of detection pairs corresponding to the eccentric direction marking sequence. This number reflects the total number of measurements involved in direction determination within the observation time range. The eccentric direction dominance refers to the larger of the proportions of the positive side count and the negative side count in the total number of detection pairs. This value is used to describe the degree to which the conductive path occupies the main offset direction within the observation time range.

[0091] Specifically, the positive and negative counts corresponding to the current control cycle are read from the counting result storage area, and the total number of detection pairs corresponding to the current control cycle is read from the total count result storage area. The positive count is divided by the total number of detection pairs to obtain the positive side percentage, and the negative count is divided by the total number of detection pairs to obtain the negative side percentage. The positive side percentage and the negative side percentage are compared and the larger one is selected as the eccentricity direction dominance.

[0092] It should be noted that the current conduction path between the electrode and the human ear tissue is not completely uniformly distributed in space. When there is geometric bias or local conductivity difference at the contact interface, the probability of current passing in different directions will be unevenly distributed, resulting in the phenomenon that the conductive path gathers to one side. In the continuous detection process, each eccentric direction mark can be regarded as the spatial pointing result of the conductive path at that moment. The positive side count and the negative side count correspond to the number of times the conductive path points in different directions, respectively. The total number of detection pairs corresponds to the total number of observations. The proportion obtained by dividing the number of occurrences in each direction by the total number of observations is the probability that the direction is occupied. When the larger proportion is selected as the result, this value represents the probability of occurrence corresponding to the dominant conductive path direction in the entire observation process. This conforms to the basic law of statistical physics that the frequency of passage reflects the degree of state occupancy. Therefore, this larger proportion can directly characterize the dominance of the conductive path biased in a certain direction, and is thus defined as the eccentric direction dominance.

[0093] S5. Based on all candidate modes of the earphone-integrated microcurrent stimulation device, determine the risk eccentricity direction category;

[0094] In embodiments of the present invention, based on all candidate modes of the earphone-integrated microcurrent stimulation device, the risk eccentricity direction category is determined, including:

[0095] Obtain all candidate modes of the headphone-integrated microcurrent stimulation device;

[0096] Specifically, all mode configuration records stored in the mode configuration storage area are read. For each mode configuration record, the mode identifier, corresponding current output amplitude data, pulse duration data, pulse interval data, polarity output rule data, and mode activation status data are extracted. Mode configuration records that are represented as executable by the mode activation status data are retained, while those that are represented as non-executable by the mode activation status data are removed. The retained mode configuration records are arranged according to the encoding order of the mode identifiers, and the arranged mode configuration records are written into the candidate mode storage area in sequence to form a set of all candidate modes that are consistent with the current operating configuration of the earphone integrated microcurrent stimulation device.

[0097] Determine the count of mode switching events and the count of adverse events for candidate modes;

[0098] Candidate modes refer to the different microcurrent output methods that the device can select and execute. Each candidate mode corresponds to a set of defined current amplitude, pulse duration, pulse interval, and polarity output rules. Mode switching event count refers to the cumulative number of times the device switches from the current stimulation mode to a certain candidate mode. This number reflects the frequency with which the corresponding candidate mode is called in actual operation. Adverse event count refers to the cumulative number of events such as stimulation abnormality, interface conductivity abnormality, user discomfort feedback, or output interruption that occur after switching to a certain candidate mode. This number is used to characterize the frequency with which the candidate mode causes abnormal states under actual conductivity conditions.

[0099] Divide the number of adverse events by the number of mode switching events to obtain the correlation strength;

[0100] Association strength refers to the ratio of adverse event count to mode switching event count. This ratio is used to represent the relative probability that a candidate mode will be accompanied by adverse results during its invocation.

[0101] Specifically, all mode switching records and all adverse event records corresponding to the current statistical period are read from the running event record area. The target mode identifier in the mode switching record is matched one by one with the mode identifier in the candidate mode storage area. For each candidate mode, the number of records that switch from any other mode to the candidate mode is accumulated to obtain the mode switching event count corresponding to the candidate mode. Then, the occurrence time in the adverse event record is matched with the switching completion time in the mode switching record to extract all adverse events recorded in each candidate mode within the time range from the completion of the switching to the occurrence of the next mode switching. The number of extracted adverse events is accumulated to obtain the adverse event count corresponding to the candidate mode.

[0102] Based on the eccentricity direction marker, the adverse event count is filtered to obtain the number of adverse events under positive conditions, the number of adverse events under negative conditions, and the number of adverse events under unstable conditions;

[0103] The number of adverse events on the positive side refers to the number of events with the corresponding eccentric direction marked as positive when the adverse event occurs. The number indicates the frequency of adverse conditions when the conductive path is biased towards the positive side.

[0104] The number of adverse events on the negative side refers to the number of events with the corresponding eccentric direction marked as negative when the adverse event occurs. The number indicates the frequency of the adverse state when the conductive path is biased to the negative side.

[0105] The number of unstable condition adverse events refers to the number of events whose corresponding eccentric direction is marked as unstable when the adverse event occurs. The number indicates the frequency of adverse state occurrence when the conductive path has not formed a stable unilateral offset.

[0106] Specifically, all adverse event records within the current statistical period are read from the adverse event record area. The occurrence time, candidate mode identifier, and event sequence number of each adverse event record are extracted. Then, the eccentric direction mark corresponding to the occurrence time of the adverse event is read from the eccentric direction mark record area. The correspondence is determined by comparing the occurrence time of the adverse event with the start and end sampling times of each detection pair. The eccentric direction mark whose occurrence time falls within the time range of a certain detection pair is determined as the eccentric direction mark corresponding to that adverse event. Adverse events with positive eccentric direction marks are filtered one by one and the number of positive condition adverse events is accumulated. Adverse events with negative eccentric direction marks are filtered one by one and the number of negative condition adverse events is accumulated. Adverse events with unstable eccentric direction marks are filtered one by one and the number of unstable condition adverse events is accumulated.

[0107] Based on the number of adverse events on the positive side, the number of adverse events on the negative side, and the number of adverse events on the unstable side, the risk eccentricity direction category is determined.

[0108] The risk eccentricity direction category refers to the direction category with the largest number of adverse events on the positive side, adverse events on the negative side, and adverse events on the unstable side. This category is used to indicate the direction of conductive path deviation that is more likely to be accompanied by adverse results during actual operation.

[0109] Specifically, the number of positive-side adverse events, negative-side adverse events, and unstable adverse events corresponding to the current statistical period are read from the directional condition count result storage area. The number of positive-side adverse events, negative-side adverse events, and unstable adverse events are compared. When the number of positive-side adverse events is greater than the number of negative-side adverse events and the number of unstable adverse events, the risk bias direction category is determined as the positive-side risk category. When the number of negative-side adverse events is greater than the number of positive-side adverse events and the number of unstable adverse events, the risk bias direction category is determined as the negative-side risk category. When the number of unstable adverse events is greater than the number of positive-side adverse events and the number of unstable adverse events is greater than the number of negative-side adverse events, the risk bias direction category is determined as the unstable risk category. The determined risk bias direction category and the corresponding statistical period identifier are written into the risk direction result storage area.

[0110] S6. Determine the target pattern based on the asymmetry ratio, the dominance of the eccentric direction, the migration rate, and the risk eccentric direction category, and output the microcurrent stimulation sequence under the target pattern.

[0111] In embodiments of the present invention, the target pattern is determined based on the asymmetry proportion, the dominance of the eccentric direction, the migration incidence rate, and the risk eccentric direction category, including:

[0112] Based on the proportion of asymmetry, the dominance of the eccentric direction, and the migration rate, the dominant statistical item is determined;

[0113] The dominant statistic refers to the statistic with the largest value among the above multiple statistics, which is used to represent the type of change that dominates the current change in conductivity state;

[0114] Specifically, the asymmetric proportion, eccentricity dominance, and migration rate corresponding to the current statistical period are read from the statistical results storage area. The asymmetric proportion, eccentricity dominance, and migration rate are organized according to the same statistical period, and the values ​​of the three are compared one by one. When the asymmetric proportion is greater than the eccentricity dominance and the asymmetric proportion is greater than the migration rate, the dominant statistical item is determined to be the asymmetric proportion. When the eccentricity dominance is greater than the asymmetric proportion and the eccentricity dominance is greater than the migration rate, the dominant statistical item is determined to be the eccentricity dominance. When the migration rate is greater than the asymmetric proportion and the migration rate is greater than the eccentricity dominance, the dominant statistical item is determined to be the migration rate. The determined dominant statistical item, along with the corresponding statistical period identifier and the asymmetric proportion, eccentricity dominance, and migration rate used for comparison, are written into the dominant item results storage area.

[0115] Based on the number of adverse events on the positive side, the number of adverse events on the negative side, and the number of adverse events on the unstable side, determine the number of adverse events on the directional side of the risk eccentricity category;

[0116] Specifically, the system reads the number of positive-side adverse events, negative-side adverse events, and unstable adverse events corresponding to the current statistical period from the directional condition count result storage area. Then, it reads the risk eccentricity direction category corresponding to the current statistical period from the risk directional result storage area. The risk eccentricity direction category is then matched with the number of positive-side adverse events, negative-side adverse events, and unstable adverse events. When the risk eccentricity direction category is a positive-side risk category, the number of positive-side adverse events is selected as the number of directional adverse events. When the risk eccentricity direction category is a negative-side risk category, the number of negative-side adverse events is selected as the number of directional adverse events. When the risk eccentricity direction category is an unstable risk category, the number of unstable adverse events is selected as the number of directional adverse events. Finally, the determined number of directional adverse events, along with the corresponding statistical period identifier, risk eccentricity direction category, number of positive-side adverse events, number of negative-side adverse events, and number of unstable adverse events, are written into the directional condition result storage area.

[0117] The number of adverse events in directional conditions refers to the cumulative number of adverse events recorded when the conductive path is in a certain offset direction. This number is obtained by matching the time of each adverse event with the eccentric direction marker within the corresponding time range. When the conductive path is determined to be in a specific directional offset state within the time range, the corresponding adverse event is counted in the number of adverse events in directional conditions for that direction. This is used to characterize the frequency of abnormal conductivity or abnormal stimulation under the offset direction of the conductive path.

[0118] The target condition is to set the number of adverse events in the directional condition of the risk bias direction category to 0.

[0119] A zero number of adverse events under directional conditions indicates that no abnormal conductivity events were observed in the corresponding offset direction of the conductive path; the target condition refers to the constraint condition used to screen candidate patterns, which limits the conductive path to not exhibiting abnormal conductivity behavior in a specific directional state.

[0120] Specifically, the risk bias direction category represents the direction of conduction path offset that is most likely to be accompanied by adverse events in historical observations. When the conduction path is in this direction, the interface conduction state has already shown a higher probability of instability or anomalies. Therefore, using the absence of adverse events in this direction as a screening constraint is equivalent to verifying the stability of the model under the most unfavorable conduction path distribution conditions. This can ensure that the selected model still maintains normal output in the conduction state that is most likely to trigger anomalies, thereby fundamentally suppressing adverse responses caused by local conduction bias and uneven contact, and making the selected model have higher overall safety and robustness.

[0121] When the dominant statistical item is the migration occurrence rate, candidate patterns that meet the target conditions are selected from all candidate patterns to form the first candidate subset;

[0122] Specifically, the dominant statistical item corresponding to the current statistical period is read from the dominant item result storage area; all candidate mode identifiers corresponding to the current headphone integrated microcurrent stimulation device are read from the candidate mode storage area; the number of adverse events of the directional condition corresponding to the current statistical period is read from the directional condition result storage area; and the target condition content corresponding to the current statistical period is read from the target condition record area. The dominant statistical item and the migration occurrence rate are compared by category. When they correspond to the same category, condition filtering is performed on all candidate modes one by one. For each candidate mode, its corresponding mode identifier, mode switching event count, adverse event count, and the candidate mode are read from the mode statistical result storage area. The formula stores the number of adverse events on the positive side, the number of adverse events on the negative side, and the number of adverse events on the unstable side in the storage area of ​​the directional condition count results. Then, it determines the number of adverse events on the directional condition corresponding to the candidate pattern based on the risk eccentricity directional category corresponding to the current statistical period. It then performs a consistency judgment between the determined number of adverse events on the directional condition and the target condition. If the number of adverse events on the directional condition is equal to zero, the candidate pattern is retained. If the number of adverse events on the directional condition is not equal to zero, the candidate pattern is removed. The retained candidate patterns are then written into the first candidate subset storage area in the order of the pattern identifier encoding to form the first candidate subset corresponding to the current statistical period.

[0123] If the first candidate subset is not empty, select the candidate pattern with the lowest correlation strength in the first candidate subset as the target pattern.

[0124] The first candidate subset refers to the set of modes that meet the target conditions selected from all candidate modes; the target mode refers to the microcurrent output mode selected for actual execution under the constraint conditions.

[0125] Specifically, when the first candidate subset is not empty, the candidate mode with the smallest correlation strength is selected as the target mode from the first candidate subset because the candidate modes in the first candidate subset already meet the constraint that no adverse events will occur in the direction of risk eccentricity of the conductive path. On this basis, the correlation strength is further compared. The correlation strength reflects the frequency of adverse events after mode switching. Taking the candidate mode with the smallest correlation strength as the target mode can minimize the probability of adverse events after mode switching under the premise of satisfying the directional safety constraint, thereby achieving comprehensive optimization of the stability and safety of the conductive path.

[0126] If the first candidate subset is empty, select the candidate pattern with the lowest correlation strength from all candidate patterns as the target pattern.

[0127] Specifically, when the first candidate subset is empty, the candidate mode with the lowest correlation strength is selected as the target mode from all candidate modes because there is no candidate mode that simultaneously satisfies the directional constraint. In this case, the occurrence of adverse events cannot be eliminated by the directional constraint. Therefore, by selecting the mode with the lowest correlation strength from all candidate modes, the overall frequency of adverse events is reduced as much as possible. This selection method provides an alternative strategy to minimize the risk of adverse events when the constraint cannot be satisfied, thereby ensuring that a relatively safe microcurrent stimulation mode can be selected under any operating conditions.

[0128] Specifically, when the dominant statistical item is asymmetric proportion, all candidate patterns are screened according to the statistical results related to their corresponding asymmetric proportions. The candidate pattern set with the smallest asymmetric proportion is selected first. When the set is not empty, the correlation strength is further compared and the candidate pattern with the smallest correlation strength is selected as the target pattern. When the set is empty, the candidate pattern with the smallest correlation strength is directly selected from all candidate patterns as the target pattern.

[0129] When the dominant statistical term is the dominance of the eccentric direction, all candidate patterns are screened according to their corresponding dominance of the eccentric direction. The candidate pattern set with the lower dominance of the eccentric direction and the more balanced distribution of the conductive path is selected first. When the set is not empty, the candidate pattern with the smallest correlation strength is selected as the target pattern. When the set is empty, the candidate pattern with the smallest correlation strength is selected as the target pattern from all candidate patterns. When the dominant statistical term cannot be uniquely determined or multiple statistical values ​​are the same, all candidate patterns are directly sorted according to the correlation strength as the candidate set and the candidate pattern with the smallest correlation strength is selected as the target pattern.

[0130] The above processing method is based on the correspondence between the stability of the conductive path and the probability of adverse events. Under the premise that the dominant statistical item reflects the main change type of the current conductive state, prioritizing the screening of the change type can directly suppress the most important source of instability. For example, the asymmetry ratio corresponds to the directional dependence of the conductive path, the dominance of the eccentric direction corresponds to the spatial offset concentration of the conductive path, and the migration rate corresponds to the frequency of change of the conductive path over time. After screening the dominant change factors, the candidate modes are compared by correlation strength. The correlation strength reflects the relative probability of adverse events after mode switching. Therefore, selecting the candidate mode with the lowest correlation strength can further reduce the overall frequency of anomalies on the basis of controlling the main unstable factors. When the screening conditions cannot be met, directly comparing the correlation strength is equivalent to selecting the mode with the lowest probability of adverse events among all available modes, thereby ensuring that the target mode with the lowest risk and the best stability can be obtained regardless of the change of conductive state.

[0131] Specifically, the microcurrent stimulation sequence is output under the target mode. This includes reading the mode identifier corresponding to the target mode from the target mode result storage area, and then reading the current amplitude data, pulse duration data, pulse interval data, polarity output rule data, and sequence period data corresponding to the mode identifier from the mode configuration storage area. The current amplitude data, pulse duration data, pulse interval data, polarity output rule data, and sequence period data are written into the stimulation output control area according to the arrangement order corresponding to the target mode. The stimulation output circuit generates the corresponding current pulses item by item according to the arrangement order and outputs them to the human ear tissue through the ear electrode. During each pulse output process, the loop voltage data and loop current data are collected synchronously and a corresponding record is established with the current target mode identifier. This process continues until a complete stimulation sequence output cycle corresponding to the target mode is completed.

[0132] It should be noted that when there is a localized shift in the wet conductive path, the wet film adjacent to the metal ear ornament may form a non-uniform conductive channel and its connectivity may change with micro-movement. This means that even if the nominal stimulation parameters remain unchanged, the spatial boundary of the actual current entering the tissue may still undergo implicit rearrangement. This can cause distortion of the pattern-effect correspondence established based on feedback in multi-mode switching control and lead to inappropriate switching. Therefore, only after completing the screening based on asymmetry ratio, eccentricity, migration rate, and the number and correlation strength of adverse events under directional conditions, and then actually outputting the microcurrent stimulation sequence corresponding to the selected target pattern to the ear tissue, can the suppression of the spatial boundary disturbance caused by the wet conductive shift be realized as a real stimulation process. This allows the current to achieve the intended treatment or regulation goal with a lower risk of adverse events and a more stable conductive path distribution under the current contact interface conditions.

[0133] It should be noted that the localized wet conductive path offset state refers to the conductive state in which, in the contact area between the electrode and the human ear tissue, the localized continuous wet film formed by sweat, sebum, or cleaning fluid residue, together with the adjacent metal components, changes the interface conductive boundary, causing the actual conduction path of the current in the tissue to no longer be distributed in a geometrically symmetrical manner, but to shift to one side. This state can be affected by changes in contact pressure or slight movements that cause changes in the shape of the wet film, resulting in the connection and disconnection of the conductive path, thus causing the spatial distribution of the current entering the tissue to exhibit unilateral bias and dynamic changes.

[0134] The above description is only a preferred embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any equivalent substitutions or modifications made by those skilled in the art within the scope of the technology disclosed in the present invention, based on the technical solution and inventive concept of the present invention, should be covered within the scope of protection of the present invention.

Claims

1. A multi-mode switching control method for an earphone-integrated microcurrent stimulation device, characterized in that, Includes the following steps: S1. Obtain the polarity pair detection dataset, and generate polarity asymmetry labels and eccentricity direction labels based on the polarity pair detection dataset; S2. Generate migration markers based on polarity asymmetry markers and eccentricity direction markers, and calculate the migration occurrence rate based on the migration markers; S3. Calculate the asymmetry ratio based on polarity asymmetry markers; S4. Calculate the dominance of the eccentric direction based on the eccentric direction marker; S5. Based on all candidate modes of the earphone-integrated microcurrent stimulation device, determine the risk eccentricity direction category; S6. Determine the target pattern based on the asymmetry ratio, the dominance of the eccentric direction, the migration rate, and the risk eccentric direction category, and output the microcurrent stimulation sequence under the target pattern.

2. The multi-mode switching control method for an earphone-integrated microcurrent stimulation device according to claim 1, characterized in that, Obtain the polarity pair detection dataset, including: Acquire the raw data stream within one control cycle, collected by the headphone-integrated microcurrent stimulation device under the current stimulation mode; Based on the acquisition times of the positive and negative detection stimulus response segments in the original data stream, a polarity pair detection dataset is formed.

3. The multi-mode switching control method for an integrated microcurrent stimulation device for headphones according to claim 2, characterized in that, Based on the polarity pair detection dataset, polarity asymmetry labels and eccentricity direction labels are generated, including: Extract the positive and negative response pattern summaries corresponding to the polarity pair detection dataset; Determine the sign consistency of the corresponding component differences between the positive and negative response morphological summaries; Based on symbol consistency, generate polarity asymmetric tags; Determine the sign relationship between the corresponding component differences in the positive and negative response morphology summaries; Generate eccentric direction markers based on the relationship between the number of positive and negative signs.

4. The multi-mode switching control method for an earphone-integrated microcurrent stimulation device according to claim 1, characterized in that, Migration markers are generated based on polarity asymmetry markers and eccentricity direction markers, including: A micro-motion trigger dataset is formed based on the acquisition time of micro-motion trigger response segments in the original data stream; A stable snap-in dataset is formed based on the acquisition time of the stable snap-in response segment in the original data stream; Based on the polarity asymmetry label and eccentricity direction label corresponding to the micro-motion triggered dataset and the polarity asymmetry label and eccentricity direction label corresponding to the stable attachment dataset, a migration label is generated.

5. The multi-mode switching control method for an earphone-integrated microcurrent stimulation device according to claim 1, characterized in that, The migration occurrence rate is calculated based on migration markers, including: Generate a migration tag sequence based on the migration tags; Count the migration marker sequences to obtain non-migration counts and migration counts; The migration occurrence rate is calculated based on the non-migration count and the migration count.

6. The multi-mode switching control method for an earphone-integrated microcurrent stimulation device according to claim 1, characterized in that, Calculating the asymmetry proportion based on polarity asymmetry markers includes: Arrange all polarity asymmetry markers to obtain a polarity asymmetry marker sequence; Counting the polarity-asymmetric labeled sequences yields symmetric and asymmetric counts. Calculate the asymmetric proportion based on symmetric and asymmetric counting.

7. The multi-mode switching control method for an earphone-integrated microcurrent stimulation device according to claim 1, characterized in that, The dominance of the eccentric direction is calculated based on the eccentric direction marker, including: Arrange all the eccentric direction markers to obtain the eccentric direction marker sequence; Count the eccentric direction marker sequence to obtain the positive side count and the negative side count; Obtain the total number of detection pairs in the eccentric orientation marker sequence; The dominance of the eccentric direction is calculated based on the positive side count, the negative side count, and the total number of detection pairs.

8. The multi-mode switching control method for an integrated microcurrent stimulation device for headphones according to claim 1, characterized in that, Based on all candidate modes of the earphone-integrated microcurrent stimulation device, risk eccentricity direction categories were determined, including: Obtain all candidate modes of the headphone-integrated microcurrent stimulation device; Determine the count of mode switching events and the count of adverse events for candidate modes; Divide the number of adverse events by the number of mode switching events to obtain the correlation strength; Based on the eccentricity direction marker, the adverse event count is filtered to obtain the number of adverse events under positive conditions, the number of adverse events under negative conditions, and the number of adverse events under unstable conditions; Based on the number of adverse events on the positive side, the number of adverse events on the negative side, and the number of adverse events on the unstable side, the risk eccentricity direction category is determined.

9. The multi-mode switching control method for an integrated microcurrent stimulation device for headphones according to claim 1, characterized in that, The target pattern is determined based on the asymmetric proportion, the dominance of the eccentric direction, the migration incidence rate, and the risk eccentric direction category, including: Based on the proportion of asymmetry, the dominance of the eccentric direction, and the migration rate, the dominant statistical item is determined; Based on the number of adverse events on the positive side, the number of adverse events on the negative side, and the number of adverse events on the unstable side, determine the number of adverse events on the directional side of the risk eccentricity category; The target condition is to set the number of adverse events in the directional condition of the risk bias direction category to 0. When the dominant statistical item is the migration occurrence rate, candidate patterns that meet the target conditions are selected from all candidate patterns to form the first candidate subset; If the first candidate subset is not empty, select the candidate pattern with the lowest correlation strength in the first candidate subset as the target pattern. If the first candidate subset is empty, select the candidate pattern with the lowest correlation strength from all candidate patterns as the target pattern.

Images