A synergistic neuromodulation system for treatment-resistant depression

Abstract

The application discloses a kind of synergic neuromodulation systems for refractory depression, it is related to neuromodulation technical field, including: obtaining individual historical conversation data to establish reference quantity, collect cortical physiology and salivary cortisol data to calculate cortical inhibition proxy quantity and axis drive proxy quantity;The above proxy quantity is standardized based on individual reference quantity and generates continuous flip point intensity;According to flip point intensity, the total stimulation load benchmark is continuously distributed as cortical stimulation component load and vagus-related regulation component load, and each component load is mapped into synergic control instruction to drive equipment to execute stimulation.The application quantifies physiological state and adaptively distributes load, avoids the reversal of stimulation effect under specific state, and improves treatment stability.

Description

Technical Field

[0001] This invention relates to the field of neuromodulation technology, and more particularly to a synergistic neuromodulation system for treatment-resistant depression. Background Technology

[0002] Patients with treatment-resistant depression often experience severe sleep disturbances and circadian rhythm disruptions. Clinically, cortical stimulation methods such as transcranial magnetic stimulation (TMS) or vagus nerve-related modulation are frequently used for intervention, or a combination of both may be employed to achieve complementary therapeutic effects. These patients often exhibit abnormalities in the hypothalamic-pituitary-adrenal axis, leading to significant fluctuations in the balance between cortical excitability and inhibition due to circadian rhythms and stress states. During treatment, how to rationally apply stimulation based on the patient's current physiological state to improve depressive symptoms while avoiding excessive arousal or increased sleep burden due to inappropriate stimulation is a pressing technical problem that needs to be solved in the field of neuromodulation and physical therapy for mental illnesses.

[0003] Current synergistic neuromodulation methods generally employ treatment strategies with fixed schedules and fixed stimulation parameters, or adjust parameters solely based on long-term clinical scale scores. This approach lacks continuous quantitative characterization of the patient's immediate physiological chassis during each session, making it difficult to accurately identify whether the patient is in a special physiological state characterized by significantly enhanced stress axis drive and severely diminished cortical inhibitory function. Due to the lack of a targeted dynamic allocation mechanism for stimulation load, if patients in this special state are still subjected to conventional high-intensity cortical stimulation, it is highly likely to produce the opposite of the expected net clinical effect, manifesting as increased irritability, heightened anxiety, or greater difficulty falling asleep, thus seriously affecting the stability and safety of the synergistic treatment process. Summary of the Invention

[0004] The purpose of this invention is to address the problem in the prior art that there is a lack of continuous quantitative representation and adaptive load allocation of an individual's real-time physiological state, which leads to a reversal of the direction of the net stimulus effect under specific physiological states, thereby causing sleep deterioration or emotional arousal surge, and undermining the stability and safety of synergistic therapy. Therefore, this invention proposes a synergistic neuromodulation system for treatment-resistant depression.

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

[0006] A synergistic neuromodulation system for treatment-resistant depression, comprising:

[0007] The baseline construction module is used to acquire historical session data of the regulated individuals and establish individual reference values ​​based on the historical session data;

[0008] The state quantification module is used to collect cortical physiological data to calculate the cortical inhibitory proxy level for this session, and to collect salivary cortisol data to calculate the axis-driven proxy level for this session.

[0009] The flip point module is used to standardize the cortical inhibition proxy and axis-driven proxy of the current session based on individual reference values, and generate continuous flip point intensities.

[0010] The load allocation module is used to obtain the total stimulus load benchmark for this session, continuously allocate the total stimulus load benchmark into cortical stimulus component load and vagal-related modulation component load based on the inversion point intensity, and map each component load into a co-control command.

[0011] The collaborative execution module is used to drive the device to execute collaborative stimulation according to the collaborative control instructions.

[0012] Preferably, individual reference values ​​are established based on historical session data, including:

[0013] Obtain the cortical inhibitory surrogate sequence, axis-driven surrogate sequence, and cortical quiescent period sequence from historical session data;

[0014] For the cortical suppression surrogate sequence and axis-driven surrogate sequence in the historical session data, the median is calculated as the cortical suppression reference and axis-driven reference, respectively, and the absolute deviation of the median is calculated as the cortical suppression scale and axis-driven scale, respectively.

[0015] The median of the cortical silent period sequence in historical session data is used as a reference value for the silent period.

[0016] The individual reference value is obtained by combining the cortical inhibition reference value, the axis driving reference value, the cortical inhibition scale value, the axis driving scale value, and the silent period reference value.

[0017] Preferably, the cortical physiological data collected are used to calculate the cortical inhibitory proxy level for this session, including:

[0018] Cortical physiological data include unconditioned evoked amplitude, conditioned evoked amplitude, and cortical quiescence period for the regulated individuals;

[0019] Calculate the ratio of the conditionally induced amplitude to the unconditionally induced amplitude, and subtract the ratio from the result to obtain the first inhibition feature;

[0020] The ratio of the cortical silent period to the reference cortical silent period is calculated to obtain the second inhibition characteristic;

[0021] The median of the first and second inhibitory features is used to obtain the cortical inhibitory proxy for this session.

[0022] Preferably, calculating the axis drive proxy quantity for this session includes:

[0023] Salivary cortisol concentrations were collected at three time points within the same 24-hour period: immediately upon waking, at a preset time after waking, and before bedtime.

[0024] Calculate the ratio of salivary cortisol concentration at a preset time after waking up to salivary cortisol concentration immediately upon waking up, and take the natural logarithm of the ratio as the wake-up response component.

[0025] Calculate the ratio of salivary cortisol concentration immediately upon waking to salivary cortisol concentration before sleep, and take the natural logarithm of the ratio as the diurnal attenuation component;

[0026] The sum of the wake-up response component and the daytime decay component is used as the axis drive agent for this session.

[0027] Preferably, the cortical inhibition proxy and axis drive proxy of this session are standardized, including:

[0028] Calculate the difference between the axis drive proxy quantity and the axis drive reference quantity for this session, and obtain the standardized axis drive quantity based on the ratio of the difference to the axis drive dimensional quantity.

[0029] The difference between the cortical inhibition proxy quantity and the cortical inhibition reference quantity in this session is calculated. Based on the ratio of the difference to the cortical inhibition scale quantity, the standardized cortical inhibition quantity is obtained.

[0030] Preferably, generating continuous inversion point intensity includes:

[0031] The difference between the standardized axial drive quantity and the standardized cortical inhibition quantity is calculated, and the difference is mapped to obtain the inversion point intensity.

[0032] Preferably, the total stimulus load is continuously allocated from the inversion point intensity into a cortical stimulus component and a vagal-related regulatory component, including:

[0033] The vagal-related regulatory component load is obtained by multiplying the inversion point intensity with the baseline total stimulus load for this session.

[0034] Calculate the difference between the intensity of the flip point and the intensity of the flip point, and then multiply the difference by the total stimulus load baseline for this session to obtain the cortical stimulus component load.

[0035] Preferably, mapping each component load to a coordinated control command includes:

[0036] Set baseline levels for cortical stimulation and vagal-related regulation;

[0037] Calculate the ratio of the cortical stimulus component load to the baseline total stimulus load, take the square root of this ratio and multiply it by the baseline cortical stimulus quantity to obtain the cortical stimulus execution quantity;

[0038] Calculate the ratio of the vagal-related modulatory component load to the baseline total stimulus load, and multiply this ratio by the baseline vagal-related modulatory quantity to obtain the vagal-related modulatory execution quantity;

[0039] Generate coordinated control instructions that include cortical stimulation execution and vagal-related regulatory execution.

[0040] Compared with the prior art, the beneficial effects of the present invention are:

[0041] 1. This invention constructs an individual reference quantity based on historical conversation data and uses robust statistical methods to continuously quantify and characterize cortical inhibitory and axonal drive surrogate quantities, overcoming the limitations of single-point-of-time measurement or subjective assessment. By collecting cortical physiological data and salivary cortisol data in real time, it accurately calculates the inflection point intensity reflecting the relative state of an individual's current stress axonal drive strength and cortical inhibitory function. This enables precise identification and continuous tracking of the specific physiological chassis of patients with treatment-resistant depression accompanied by insomnia or circadian rhythm disorders, providing an objective and interpretable quantitative basis for subsequent differentiated regulation.

[0042] 2. This invention adaptively and continuously allocates the total stimulation load benchmark based on the inversion point intensity. It can dynamically adjust the ratio of cortical stimulation components and vagal-related regulatory components according to the patient's real-time physiological state within the same treatment course. When the patient is detected to be in the state of the inversion point of the inhibition trough where axillary drive is enhanced and cortical inhibition is declining, the cortical stimulation component load is automatically reduced to avoid the risk of emotional arousal surge or sleep deterioration caused by the reversal of the net effect direction. At the same time, the vagal-related regulatory component load is increased accordingly to maintain the total amount of regulation and focus on homeostatic regulation. This significantly improves the stability, safety and reproducibility of clinical efficacy of the synergistic neuromodulation process under complex pathophysiological fluctuations. Attached Figure Description

[0043] 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:

[0044] Figure 1 This is a functional block diagram of a synergistic neuromodulation system for treatment-resistant depression, provided as an embodiment of the present invention. Detailed Implementation

[0045] 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.

[0046] Example: This example provides a synergistic neuromodulation system for treatment-resistant depression. See [link to example]. Figure 1 Specifically, including:

[0047] The baseline construction module is used to acquire historical session data of the regulated individuals and establish individual reference values ​​based on the historical session data;

[0048] In embodiments of this method, historical conversation data of the regulated individual is acquired, and an individual reference value is established based on the historical conversation data, including:

[0049] Obtain the cortical inhibitory surrogate sequence, axis-driven surrogate sequence, and cortical quiescent period sequence from historical session data;

[0050] For the cortical suppression surrogate sequence and axis-driven surrogate sequence in the historical session data, the median is calculated as the cortical suppression reference and axis-driven reference, respectively, and the absolute deviation of the median is calculated as the cortical suppression scale and axis-driven scale, respectively.

[0051] Specifically, the cortical inhibitory surrogate sequence refers to the set of representations of the strength of inhibitory gating arranged in the order of the conversation. These representations are derived from the influence of inhibitory regulation on the amplitude and duration of induced responses after cortical stimulation, and are used to characterize the trend of changes in the ability of the intracortical inhibitory circuit to limit excitation propagation with the conversation. The axial drive surrogate sequence refers to the set of representations of axial drive strength arranged in the order of the conversation. These representations are formed by the amplitude of changes in stress-related endocrine rhythms and intraday gradients, and are used to characterize the trend of changes in the degree to which the adrenocorticotropic hormone-releasing hormone-related axis boosts arousal drive and stress load with the conversation. The cortical quiescent period sequence refers to the set of cortical quiescent period durations arranged in the order of the conversation. The duration of the cortical quiescent period reflects the duration of temporary inhibition of motor output after stimulation, and reflects the inhibitory strength and recovery process of the descending motor pathway mediated by inhibitory neurotransmitters.

[0052] In detail, at the end of each co-modulation session, the control system writes the session number, acquisition timestamp, unconditioned evoked amplitude, conditioned evoked amplitude, cortical quiescence duration, and salivary cortisol sampling value used to calculate the axis-driven surrogate quantity into the historical session data table. Using the session number as the sorting key, a time-incrementing sequence of records is formed in the local database or hospital information system database. Subsequently, when constructing reference and scale quantities, at least thirty recent session records are read from the historical session data table in ascending order of session number as a statistical window. If fewer than thirty records are read, all existing session records are read, and the first thirty records after the current session are used as the preferred stable window. After reading, the cortical inhibitory surrogate quantity for each session is calculated based on the unconditioned and conditioned evoked amplitudes of each record, and a cortical inhibitory surrogate quantity sequence is formed according to the reading order. The cortical inhibitory surrogate quantity is calculated as one minus the ratio of the conditioned evoked amplitude to the unconditioned evoked amplitude to characterize the inhibitory gating. The system detects the inhibition ratio of induced responses and discards records when the amplitude of uninduced responses is less than the lower limit of signal-noise ratio to avoid ratio distortion. The lower limit of signal-noise ratio is three times the root mean square value of the idle noise of the acquisition channel and is automatically estimated through device power-on self-test. The system also directly reads the cortical quiescent period duration from each session record and forms a cortical quiescent period sequence according to the reading order for use as a normalized benchmark for calculating the suppression proxy in subsequent sessions. The system then calculates the axis-driven proxy in each session based on the salivary cortisol sampling value and forms an axis-driven proxy sequence according to the reading order. The axis-driven proxy is the sum of the logarithmic ratio of cortisol shortly after waking up to that immediately upon waking up and the logarithmic ratio of cortisol immediately upon waking up to that before sleep, to reflect the arousal drive and the diurnal decay gradient. When any sampling value is less than the detection limit, the detection limit is used instead to avoid logarithmic divergence. The detection limit is the nominal detection limit value of the immunoassay kit used and is obtained through the kit instructions.

[0053] After constructing the three types of sequences, the system sorts the cortical suppression surrogate quantity sequences and takes the value in the middle of the sequence as the cortical suppression reference quantity. When the sequence length is even, the arithmetic mean of the two middle numbers is taken as the median to maintain continuity. The system calculates the axis-driven reference quantity using the same rule for the axis-driven surrogate quantity sequences. Then, the system calculates the absolute value of the difference between each cortical suppression surrogate quantity and the cortical suppression reference quantity to form an absolute deviation sequence. The absolute deviation sequence is then used as the cortical suppression scale quantity according to the median rule to obtain a typical fluctuation amplitude characterization that is insensitive to abnormal sessions. The system also calculates the absolute deviation sequence relative to the axis-driven reference quantity for the axis-driven surrogate quantity sequences and takes the median as the axis-driven scale quantity. Finally, the cortical suppression reference quantity, axis-driven reference quantity, cortical suppression scale quantity, and axis-driven scale quantity are written into the session parameter table and their corresponding statistical window session number ranges are marked.

[0054] The median of the cortical silent period sequence in historical session data is used as a reference value for the silent period.

[0055] The individual reference value is obtained by combining the cortical inhibition reference value, the axis driving reference value, the cortical inhibition scale value, the axis driving scale value, and the silent period reference value.

[0056] In detail, after completing the reading of historical session data and constructing the cortical suppression surrogate quantity sequence, axis-driven surrogate quantity sequence, and cortical quiescent period sequence, the control system first performs validity screening on the cortical quiescent period sequence to remove records with measurement failures or zero duration. The remaining quiescent period durations are then sorted in ascending order by session number to form an effective quiescent period sequence. Subsequently, the median of the effective quiescent period sequence is calculated to obtain the quiescent period reference value. Specifically, the effective quiescent period sequences are sorted by numerical value, and the quiescent period duration in the middle of the sequence is taken as the quiescent period reference value. When the sequence length is even, the arithmetic mean of the two middle quiescent period durations is taken as the quiescent period reference value to maintain the continuity of the reference value with sample variation. If the effective quiescent period sequence length is less than three, the statistical window is extended to earlier session records until the number of effective samples reaches three to ensure the representativeness of the reference value. Furthermore, this extension rule is based on covering at least one complete cycle to reduce the impact of short-term state fluctuations. After obtaining the quiescent period reference value, the system reads the cortical inhibition reference value, axis-driven reference value, cortical inhibition scale value, and axis-driven scale value from the session parameter table, and combines the above four values ​​with the quiescent period reference value to form an individual reference value. The combination method is to write the data into the same data structure in a fixed field order for subsequent calculation and calling. The individual reference value can be stored in a vector or key-value pair structure and include the start and end session numbers of the statistical window as traceability information. Finally, the individual reference value is written into the individual reference value table and indexed with the individual identifier and generation timestamp, so that subsequent sessions can directly call the corresponding fields in the individual reference value to complete the normalization and scaling processing when calculating the robust normalized values ​​of the cortical inhibition surrogate value and the axis-driven surrogate value.

[0057] The state quantification module is used to collect cortical physiological data to calculate the cortical inhibitory proxy level for this session, and to collect salivary cortisol data to calculate the axis-driven proxy level for this session.

[0058] In embodiments of the present invention, collecting cortical physiological data to calculate the cortical inhibitory proxy level for the current session, and collecting salivary cortisol data to calculate the axis-driven proxy level for the current session, includes:

[0059] The unconditioned evoked amplitude, conditioned evoked amplitude, and cortical quiescent period were collected from the regulated individuals.

[0060] Calculate the ratio of the conditionally induced amplitude to the unconditionally induced amplitude, and subtract the ratio from the result to obtain the first inhibition feature;

[0061] The ratio of the cortical silent period to the reference cortical silent period is calculated to obtain the second inhibition characteristic;

[0062] The median of the first and second inhibitory features is used to obtain the cortical inhibitory proxy for this session.

[0063] Specifically, the unconditioned evoked amplitude refers to the amplitude of the evoked potential output generated by the target cortex through the descending pathway in the peripheral muscles when only the test stimulus is applied without any preceding modulated stimulus. It is extracted by the electromyography acquisition channel within a specified sampling window after baseline correction of the evoked waveform, based on the peak value or equivalent amplitude. It is used to characterize the individual's baseline cortical excitability and efferent pathway conduction capacity at the current session moment. The conditioned evoked amplitude refers to the amplitude of the evoked potential output obtained under the same acquisition and extraction rules after applying a conditioned stimulus to modulate the intracortical inhibitory circuit before or simultaneously with the test stimulus. Its change relative to the unconditioned evoked amplitude reflects the effect of the conditioned stimulus on the intracortical inhibitory gating. A relatively decreased conditioned evoked amplitude indicates a weakening or strengthening of the inhibitory circuit's effect on the evoked output; a relatively increased conditioned evoked amplitude indicates a weakening of the inhibitory gating or excitability dominance. The first inhibitory feature refers to... The inhibition ratio, constructed from the relative relationship between conditioned and unconditioned evoked amplitudes, reflects the degree to which the inhibitory circuit in the cortex weakens the amplitude of the evoked response under a given stimulus. A larger value indicates that the evoked output is more significantly suppressed after the introduction of the conditioned stimulus, thus corresponding to a stronger inhibitory gating effect. The second inhibitory feature is a normalized representation of the cortical quiescent period relative to the quiescent period reference value. It reflects the degree of deviation of the ability to temporarily suppress motor output after stimulation from the individual's typical level. A value greater than one indicates a longer quiescent period corresponding to stronger inhibition, while a value less than one indicates a shorter quiescent period corresponding to weaker inhibition. The cortical inhibitory surrogate is a conversational inhibition gating strength representation obtained by robustly fusing the first and second inhibitory features. It is used to comprehensively reflect the combined results of the inhibitory circuit's suppression of evoked amplitude and maintenance of inhibition duration without relying on a single observation.

[0064] In detail, before the start of this session, unconditioned and conditioned stimuli are applied to the target cortical area of ​​the individual being regulated using a transcranial stimulation device, and evoked potential waveforms are simultaneously recorded by the surface electromyography (EMG) acquisition channel. After each stimulation, the waveform is baseline-corrected within a preset sampling window, and the evoked amplitude is extracted in a peak-to-peak manner. The unconditioned evoked amplitude is recorded as the unconditioned evoked amplitude, and the conditioned evoked amplitude is recorded as the conditioned evoked amplitude. Simultaneously, the cortical quiescent period is calculated from the inhibition period of the EMG output after stimulation and recorded as the cortical quiescent period. To avoid distortion of the amplitude ratio under low signal-to-noise conditions, the system first calculates the root mean square value of the EMG idle noise and uses three times it as the lower signal-to-noise limit. When the unconditioned evoked amplitude is not greater than the lower signal-to-noise limit, the recording is discarded, and the acquisition is repeated until at least three valid recordings are obtained. Then, for each valid recording, the ratio of the conditioned evoked amplitude to the unconditioned evoked amplitude is calculated, and the ratio is subtracted from the first value to obtain the first inhibition feature. A larger value for the first inhibition feature indicates a stronger conditioned stimulus response. The greater the proportion of weakened output after input, the stronger the inhibition gating. Then, the cortical quiescent period reference value is read from the individual reference value, and the ratio of the cortical quiescent period to the cortical quiescent period reference value is calculated as the second inhibition feature. The second inhibition feature is greater than one, which means that the quiescent period is longer than the individual typical level and the inhibition is stronger. Subsequently, the system forms a binary set with the first inhibition feature and the second inhibition feature corresponding to the same valid record, and takes the median of the binary set to obtain the cortical inhibition surrogate value of the valid record. The median is equivalent to the arithmetic mean of the two in the binary set, so as to avoid the introduction of artificial weights. After obtaining the cortical inhibition surrogate value for no less than three valid records, the median of these cortical inhibition surrogate values ​​is taken again as the cortical inhibition surrogate value of the current session and written into the session parameter table. At the same time, the unconditional evoked amplitude, conditional evoked amplitude, cortical quiescent period, first inhibition feature and second inhibition feature are archived as traceable fields so that they can be directly called when constructing historical sequences and individual reference values ​​in subsequent sessions.

[0065] Salivary cortisol concentrations were collected at three time points within the same 24-hour period: immediately upon waking, at a preset time after waking, and before bedtime.

[0066] Calculate the ratio of salivary cortisol concentration at a preset time after waking up to salivary cortisol concentration immediately upon waking up, and take the natural logarithm of the ratio as the wake-up response component.

[0067] Calculate the ratio of salivary cortisol concentration immediately upon waking to salivary cortisol concentration before sleep, and take the natural logarithm of the ratio as the diurnal attenuation component;

[0068] The sum of the wake-up response component and the daytime decay component is used as the axis drive agent for this session.

[0069] Specifically, salivary cortisol concentration refers to the cortisol level detected through saliva samples. It reflects the endocrine output intensity of the hypothalamic-pituitary-adrenal axis at the time of sampling, and can reflect the axial drive state of an individual during wakefulness initiation, daytime stress load, and nighttime decline. The wake-up response component is a measure of the relative increase in salivary cortisol concentration at a preset time after waking up compared to the immediate salivary cortisol concentration upon waking up. It is obtained by taking the natural logarithm of the ratio between the two and is used to characterize the magnitude of the rapid increase in cortisol during the wakefulness initiation phase, thereby characterizing the boosting strength of axial drive on wakefulness and mobilization. The diurnal decay component refers to the immediate salivary cortisol concentration upon waking up. The relative decline gradient of salivary cortisol concentration relative to bedtime salivary cortisol concentration is obtained by taking the natural logarithm of the ratio between the two. It is used to characterize the rhythmic decline from morning to night, thus characterizing the overall gradient strength of axial drive in the diurnal rhythm. The axial drive surrogate quantity is a session-level axial drive strength characterization quantity obtained by summing the wake-up response component and the diurnal decay component. It comprehensively reflects the combined result of the arousal initiation increase and the diurnal rhythm gradient. It is used as an input quantity for subsequent flip point strength calculation and co-stimulus load allocation to characterize the individual's stress and arousal drive level in the diurnal corresponding to the current session.

[0070] In detail, to obtain a representation of the driving strength of the corticotropin-releasing hormone-related axis within a 24-hour period, saliva samples were taken from the regulated individuals at three points within the same 24-hour period corresponding to this session, and the salivary cortisol concentration was measured. These three points were: the immediate time upon waking, the preset time after waking, and the bedtime. The immediate time upon waking was defined as within five minutes of the individual opening their eyes and getting out of bed to minimize short-term disturbances caused by changes in position from lying down to standing. The bedtime was defined as within thirty minutes before the planned time to fall asleep to avoid rapid fluctuations in the sleep-onset process. The preset time after waking was preferably set at thirty minutes after waking, based on the fact that the salivary cortisol wake-up response typically peaks between thirty and forty-five minutes after waking, thus providing a stable representation of the arousal drive amplitude. The system recorded the salivary cortisol concentration at the immediate time upon waking as follows: The concentration of salivary cortisol at a preset time after waking up is recorded as follows: Record the concentration of salivary cortisol before bedtime as To ensure the numerical stability of logarithmic calculations, if any sample concentration is lower than the detection limit, the detection limit is used instead. This detection limit is taken as the nominal detection limit of the test kit used, and the value is directly read from the kit's instruction manual. After concentration quantification, the ratio of salivary cortisol concentration at a preset time after waking to salivary cortisol concentration immediately upon waking is calculated. The natural logarithm of this ratio is then taken to obtain the wake-up response component, denoted as [missing information]. To characterize the axial drive increase during the wakefulness initiation phase, the ratio of salivary cortisol concentration immediately upon waking to that before sleep is calculated, and the natural logarithm of this ratio is taken to obtain the diurnal attenuation component, denoted as [missing information]. The axial drive strength is used to characterize the circadian rhythm gradient. Finally, the wake-up response component and the intra-day decay component are added together to obtain the axial drive proxy quantity for this session. The axial drive proxy quantity is denoted as C=a+d. The wake-up response component, the intra-day decay component, and the axial drive proxy quantity are written into the session parameter table.

[0071] The flip point module is used to standardize the cortical inhibition proxy and axis-driven proxy of the current session based on individual reference values, and generate continuous flip point intensities.

[0072] In an embodiment of the present invention, the cortical inhibition proxy and axis drive proxy of the current session are standardized, and continuous flip point intensities are generated, including:

[0073] Calculate the difference between the axis drive proxy quantity and the axis drive reference quantity for this session, and obtain the standardized axis drive quantity based on the ratio of the difference to the axis drive dimensional quantity.

[0074] The difference between the cortical inhibition proxy quantity and the cortical inhibition reference quantity in this session is calculated, and the standardized cortical inhibition quantity is obtained based on the ratio of the difference to the cortical inhibition scale quantity.

[0075] Calculate the difference between the standardized axial drive and the standardized cortical inhibition, and map the difference to obtain the inversion point intensity;

[0076] The standardized axial drive quantity refers to the dimensionless representation of the offset of the axial drive surrogate quantity relative to the axial drive reference quantity in the current session, scaled using an axial drive scale. It reflects the degree of enhancement or weakening of the axial drive intensity during the corresponding day and night of the current session relative to the individual's typical level. A positive value indicates that the axial drive is higher than the individual's normal level, and the larger the positive value, the greater the enhancement exceeds the individual's natural fluctuation. A negative value indicates that the axial drive is lower than the individual's normal level, and the smaller the negative value, the more significant the weakening. The standardized cortical inhibition quantity refers to the dimensionless representation of the offset of the cortical inhibition surrogate quantity relative to the cortical inhibition reference quantity in the current session, scaled using a cortical inhibition scale. It reflects the strength of the inhibitory gating in the current session relative to the individual's typical level. The degree of enhancement or weakening is indicated by a positive value, which means that the inhibitory gating is stronger than the individual's normal state, and the larger the positive value, the higher the enhancement exceeds the individual's natural fluctuation. The negative value means that the inhibitory gating is weaker than the individual's normal state, and the smaller the negative value, the more significant the weakening. The flip point intensity is a characterization quantity in the range of zero to one obtained by continuously and monotonically mapping the difference between the standardized axial driving quantity and the standardized cortical inhibition quantity. It reflects the degree to which the individual is close to the flip point of the inhibition trough in the corresponding state of this session. The closer the flip point intensity is to one, the stronger the axial driving is and the weaker the inhibitory gating is, thus making it easier to have an unfavorable state of reversal of the direction of the net stimulus effect. The closer the flip point intensity is to zero, the weaker the axial driving is or the stronger the inhibitory gating is, thus moving away from the flip point state.

[0077] In detail, after obtaining the axial-driven surrogate quantity and cortical-inhibitory surrogate quantity for the current session, the system reads the axial-driven reference quantity, axial-driven scale quantity, cortical-inhibitory reference quantity, and cortical-inhibitory scale quantity from the individual reference quantity and enters the inversion point intensity calculation process. The system first calculates the difference between the axial-driven surrogate quantity and the axial-driven reference quantity for the current session as the axial-driven offset quantity. The axial-driven offset quantity is recorded as the axial-driven surrogate quantity minus the axial-driven reference quantity. Then, the axial-driven offset quantity is divided by the axial-driven scale quantity to obtain the standardized axial-driven quantity. The axial-driven scale quantity is the absolute deviation of the axial-driven surrogate quantity sequence of historical session data relative to the axial-driven reference quantity, which characterizes the typical amplitude of the axial-driven fluctuation of an individual under natural conditions. Thus, this ratio can convert the current offset quantity into a dimensionless intensity relative to the typical fluctuation scale of the individual. To avoid numerical amplification caused by the axial-driven scale quantity being too small in extremely stable individuals or when there is insufficient historical samples, the axial-driven scale quantity is compared with the most recent... The sum of the smallest resolution quantities is used as the denominator, where the smallest resolution quantity is taken as the order of magnitude of the concentration resolution of the salivary cortisol detection kit after natural logarithmic transformation and obtained from the kit instructions, thus obtaining a stable standardized axis driving quantity. Then, the difference between the cortical inhibition surrogate quantity and the cortical inhibition reference quantity in the current session is calculated using the same rules as the cortical inhibition offset quantity. The cortical inhibition offset quantity is divided by the cortical inhibition scale quantity to obtain the standardized cortical inhibition quantity. The cortical inhibition scale quantity is the median absolute deviation of the cortical inhibition surrogate quantity sequence of historical session data relative to the cortical inhibition reference quantity, which characterizes the typical fluctuation amplitude of inhibition gating within an individual. Similarly, the sum of the cortical inhibition scale quantity and the smallest resolution quantity is used as the denominator to avoid the denominator being too small, where the smallest resolution quantity is taken as the smallest resolvable change in the cortical inhibition surrogate quantity determined by the amplitude resolution of the electromyography acquisition system and the timing resolution of the cortical silent period, and is calculated through the nominal parameters of the device.

[0078] The difference between the standardized axial drive and the standardized cortical suppression is calculated as the flip drive difference, which is equal to the standardized axial drive minus the standardized cortical suppression. A positive flip drive difference indicates that the axial drive is relatively enhanced while the suppression gating is relatively weakened, thus getting closer to the suppression trough flip point state. A negative flip drive difference indicates that the axial drive is relatively weakened or the suppression gating is relatively enhanced, thus moving away from the flip point state. The system maps the flip drive difference to obtain the flip point strength. The mapping uses a continuously monotonic logic function to compress any real difference to between zero and one and avoid threshold jumps. Preferably, the mapping is a negative flip drive difference of one divided by an exponential function, where the base of the exponential function is a natural constant. The selection criterion is that the logic function has an approximately linear response when the difference is close to zero and saturates when the difference is extremely large or small, thus suppressing noise amplification. Finally, the flip point strength is obtained, and the axial drive offset, standardized axial drive, cortical suppression offset, standardized cortical suppression, flip drive difference, and flip point strength are written into the session parameter table.

[0079] The load allocation module is used to obtain the total stimulus load benchmark for this session, continuously allocate the total stimulus load benchmark into cortical stimulus component load and vagal-related modulation component load based on the inversion point intensity, and map each component load into a co-control command.

[0080] In embodiments of the present invention, the total stimulus load benchmark is continuously allocated into cortical stimulus component load and vagal-related modulatory component load based on the inversion point intensity, and each component load is mapped to a co-control command, including:

[0081] Obtain the baseline of the total stimulus load for this session;

[0082] The vagal-related regulatory component load is obtained by multiplying the inversion point intensity with the baseline total stimulus load for this session.

[0083] Calculate the difference between the intensity of the flip point and the intensity of the flip point, and obtain the cortical stimulation component load by multiplying the difference by the total stimulation load baseline for this session.

[0084] Specifically, the total stimulation load baseline refers to the scale of the total equivalent stimulation amount allowed to be applied within this session. It is used to uniformly characterize the overall upper limit of the synergistic neuromodulation input and maintain consistent dosage measurement within the same treatment course, thereby ensuring the comparability and traceability of stimulation allocation results between different sessions. This load baseline can be expressed in units of energy, charge, or dosage defined by the device manufacturer and corresponds to the overall effect level of the stimulation input on neural tissue. The vagal-related modulatory component load refers to the equivalent stimulation amount allocated from the total stimulation load baseline to the vagal-related modulatory component according to the inflection point intensity. It reflects the stimulation used to modulate autonomic neural pathways, represented by the vagus nerve, in this session. Input intensity share: A larger value indicates that a greater proportion of the total effect in this session is used to influence the balance of sympathetic and vagal tone and the physiological regulatory pathways related to arousal drive. Cortical stimulation component load refers to the equivalent stimulus amount allocated to the cortical stimulation component from the total stimulation load baseline minus the inversion point intensity. It reflects the input intensity share used to act on the cortical network in this session. A larger value indicates that more of the total effect in this session is used to directly change the balance of cortical excitation and inhibition and cortical network activity. Thus, it undertakes the main therapeutic input when it is far from the inversion point state of inhibition trough, and automatically contracts when it is close to the inversion point state to reduce the risk of adverse effects caused by the reversal of the net effect direction.

[0085] In detail, after calculating the inversion point intensity, the stimulation load allocation process begins. First, the total stimulation load benchmark for the current session is read from the treatment prescription form or the session parameters issued by the doctor's workstation and used as a unified load scale. The total stimulation load benchmark represents the total equivalent stimulation amount allowed to be applied within this session. It can be expressed in terms of energy, charge, or dimensionless dose units provided by the equipment manufacturer, and the measurement caliber should remain consistent within the same treatment course to ensure comparability of load allocation. Then, the system reads the inversion point intensity corresponding to this session and calculates the vagal-related regulatory component load. Specifically, the inversion point intensity is multiplied by the total stimulation load benchmark to obtain the vagal-related regulatory component load. This vagal-related regulatory component load is recorded as the total stimulation load benchmark multiplied by the inversion point intensity, and this load increases with the inversion point intensity. The system continuously increases the load to reflect the tendency to allocate the stimulation load to the vagal-related modulatory component, which is less sensitive to the direction of cortical plasticity, when the system is at the inversion point of inhibition trough. The system then calculates the cortical residual allocation coefficient by subtracting the inversion point intensity and multiplies the cortical residual allocation coefficient by the total stimulation load benchmark to obtain the cortical stimulation component load. The cortical stimulation component load is recorded as the total stimulation load benchmark multiplied by subtracting the inversion point intensity. This load continuously decreases as the inversion point intensity increases to reflect that the cortical stimulation component can bear more load when it is far from the inversion point and automatically reduces the load when it is close to the inversion point to avoid the risk of reversal of the net effect direction. Finally, the system writes the vagal-related modulatory component load and the cortical stimulation component load into the session parameter table and uses them as direct inputs for generating collaborative control commands.

[0086] Set baseline levels for cortical stimulation and vagal-related regulation;

[0087] Calculate the ratio of the cortical stimulus component load to the baseline total stimulus load, take the square root of this ratio and multiply it by the baseline cortical stimulus quantity to obtain the cortical stimulus execution quantity;

[0088] Calculate the ratio of the vagal-related modulatory component load to the baseline total stimulus load, and multiply this ratio by the baseline vagal-related modulatory quantity to obtain the vagal-related modulatory execution quantity;

[0089] Generate coordinated control instructions that include cortical stimulation execution levels and vagal-related regulatory execution levels;

[0090] In detail, after obtaining the cortical stimulation component load and the vagal-related modulation component load, the system enters the execution quantity mapping and instruction generation process. The system first reads the cortical stimulation baseline quantity and the vagal-related modulation baseline quantity from the treatment prescription table and uses them as the calibrated execution quantity for this treatment when the intensity is at a neutral level at the inversion point. The cortical stimulation baseline quantity corresponds to the calibrated amplitude or equivalent output intensity of the cortical stimulation device in a standard session, and the vagal-related modulation baseline quantity corresponds to the calibrated current amplitude or equivalent output intensity of the vagal-related modulation device in a standard session. The value of the baseline quantity is established by the doctor based on the recommended starting dose range given in the device manual and combined with individual tolerance. It is preferred to take the actual execution quantity when no obvious discomfort occurs in the first session and both the cortical inhibition proxy quantity and the axis drive proxy quantity are close to the individual reference quantity as the baseline quantity to ensure that the baseline quantity represents the typical level that the individual can tolerate.

[0091] The ratio of the cortical stimulation component load to the total stimulation load baseline is calculated, and the square root of this ratio is used to obtain the cortical load amplitude coefficient. The basis for taking the square root is that the equivalent load of the cortical stimulation component is approximately quadratically related to the stimulation amplitude under the energy caliber, so the square root can convert the load ratio into an amplitude ratio. The cortical load amplitude coefficient is multiplied by the cortical stimulation baseline quantity to obtain the cortical stimulation execution quantity. The cortical stimulation execution quantity is used to directly drive the amplitude setting of the cortical stimulation device and is trimmed within the rated output range of the device to meet the safety boundary. Then, the ratio of the vagal-related modulation component load to the total stimulation load baseline is calculated as the vagal load amplitude coefficient, and this ratio is multiplied by the vagal-related modulation baseline quantity. The relevant modulation base quantity is used to obtain the vagal-related modulation execution quantity. The linear mapping is based on the fact that the equivalent load and current amplitude of the vagal-related modulation component are approximately linearly related under fixed frequency and pulse width conditions, so the load ratio can be directly converted into the current ratio. Finally, the cortical stimulation execution quantity and the vagal-related modulation execution quantity are written into the same data structure and a collaborative control instruction is generated. The collaborative control instruction includes at least a cortical stimulation execution quantity field and a vagal-related modulation execution quantity field, and is accompanied by the current session number and generation timestamp for traceability. The collaborative control instruction is sent to the cortical stimulation device control interface and the vagal-related modulation device control interface to complete the collaborative execution of this session.

[0092] 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 synergistic neuromodulation system for treatment-resistant depression, characterized in that, include: The baseline construction module is used to acquire historical session data of the regulated individuals and establish individual reference values ​​based on the historical session data; The state quantification module is used to collect cortical physiological data to calculate the cortical inhibitory proxy level for this session, and to collect salivary cortisol data to calculate the axis-driven proxy level for this session. The flip point module is used to standardize the cortical inhibition proxy and axis-driven proxy of the current session based on individual reference values, and generate continuous flip point intensities. The load allocation module is used to obtain the total stimulus load benchmark for this session, continuously allocate the total stimulus load benchmark into cortical stimulus component load and vagal-related modulation component load based on the inversion point intensity, and map each component load into a co-control command. The collaborative execution module is used to drive the device to execute collaborative stimulation according to the collaborative control instructions.

2. The synergistic neuromodulation system for treatment-resistant depression according to claim 1, characterized in that, Individual reference values ​​are established based on historical session data, including: Obtain the cortical inhibitory surrogate sequence, axis-driven surrogate sequence, and cortical quiescent period sequence from historical session data; For the cortical suppression surrogate sequence and axis-driven surrogate sequence in the historical session data, the median is calculated as the cortical suppression reference and axis-driven reference, respectively, and the absolute deviation of the median is calculated as the cortical suppression scale and axis-driven scale, respectively. The median of the cortical silent period sequence in historical session data is used as a reference value for the silent period. The individual reference value is obtained by combining the cortical inhibition reference value, the axis driving reference value, the cortical inhibition scale value, the axis driving scale value, and the silent period reference value.

3. A synergistic neuromodulation system for treatment-resistant depression according to claim 2, characterized in that, Collect cortical physiological data to calculate the cortical inhibitory proxy for this session, including: Cortical physiological data include unconditioned evoked amplitude, conditioned evoked amplitude, and cortical quiescence period for the regulated individuals; Calculate the ratio of the conditionally induced amplitude to the unconditionally induced amplitude, and subtract the ratio from the result to obtain the first inhibition feature; The ratio of the cortical silent period to the reference cortical silent period is calculated to obtain the second inhibition characteristic; The median of the first and second inhibitory features is used to obtain the cortical inhibitory proxy for this session.

4. A synergistic neuromodulation system for treatment-resistant depression according to claim 1, characterized in that, Calculate the axis drive proxy quantity for this session, including: Salivary cortisol concentrations were collected at three time points within the same 24-hour period: immediately upon waking, at a preset time after waking, and before bedtime. Calculate the ratio of salivary cortisol concentration at a preset time after waking up to salivary cortisol concentration immediately upon waking up, and take the natural logarithm of the ratio as the wake-up response component. Calculate the ratio of salivary cortisol concentration immediately upon waking to salivary cortisol concentration before sleep, and take the natural logarithm of the ratio as the diurnal attenuation component; The sum of the wake-up response component and the daytime decay component is used as the axis drive agent for this session.

5. A synergistic neuromodulation system for treatment-resistant depression according to claim 2, characterized in that, The cortical inhibition proxy and axis drive proxy of this session are standardized, including: Calculate the difference between the axis drive proxy quantity and the axis drive reference quantity for this session, and obtain the standardized axis drive quantity based on the ratio of the difference to the axis drive dimensional quantity. The difference between the cortical inhibition proxy quantity and the cortical inhibition reference quantity in this session is calculated. Based on the ratio of the difference to the cortical inhibition scale quantity, the standardized cortical inhibition quantity is obtained.

6. A synergistic neuromodulation system for treatment-resistant depression according to claim 5, characterized in that, Generate continuous flip point intensities, including: The difference between the standardized axial drive quantity and the standardized cortical inhibition quantity is calculated, and the difference is mapped to obtain the inversion point intensity.

7. A synergistic neuromodulation system for treatment-resistant depression according to claim 6, characterized in that, Based on the inversion point intensity, the total stimulus load is continuously allocated into cortical stimulus component load and vagal-related regulatory component load, including: The vagal-related regulatory component load is obtained by multiplying the inversion point intensity with the baseline total stimulus load for this session. Calculate the difference between the intensity of the flip point and the intensity of the flip point, and then multiply the difference by the total stimulus load baseline for this session to obtain the cortical stimulus component load.

8. A synergistic neuromodulation system for treatment-resistant depression according to claim 7, characterized in that, Map each component load to a coordinated control command, including: Set baseline levels for cortical stimulation and vagal-related regulation; Calculate the ratio of the cortical stimulus component load to the baseline total stimulus load, take the square root of this ratio and multiply it by the baseline cortical stimulus quantity to obtain the cortical stimulus execution quantity; Calculate the ratio of the vagal-related modulatory component load to the baseline total stimulus load, and multiply this ratio by the baseline vagal-related modulatory quantity to obtain the vagal-related modulatory execution quantity; Generate coordinated control instructions that include cortical stimulation execution and vagal-related regulatory execution.

Images