A closed-loop reproductive nerve stimulation system based on multimodal biofeedback

Abstract

This invention relates to the field of surface stimulation technology and proposes a closed-loop reproductive nerve stimulation system based on multimodal biofeedback. The system includes: a remote server for analyzing the prior relative effect and prior precision of various stimulation parameter combinations in control trial data; a data processing module for collecting multimodal biofeedback data reflecting reproductive nerve function and calculating the current effect index of the current reproductive nerve; a microcontroller for performing Bayesian updates on the prior matrix, calculating the SUCRA probability corresponding to each stimulation parameter using a network meta-sorting method, and using the stimulation parameter corresponding to the highest SUCRA probability as the output parameter for the next beat; and a skin conduction nerve stimulator controlled by the microcontroller to perform electrical stimulation treatment on the current reproductive nerve based on the output parameter for the next beat. This invention solves the problems of insufficient regulation of reproductive nerve stimulation parameters and inadequate adaptability to user physiological changes.

Description

Technical Field

[0001] This invention relates to a closed-loop reproductive nerve stimulation system based on multimodal biofeedback, belonging to the field of surface stimulation technology. Background Technology

[0002] Electrical stimulation modulation of reproductive nerve function has significant clinical implications in areas such as reproductive health intervention and adjuvant treatment of reproductive system diseases. Specifically, electrical stimulation modulation of reproductive nerve function refers to regulating the activity of neural pathways through appropriate electrical stimulation parameters, thereby improving the physiological state of the reproductive system. Due to the significant heterogeneity in the physiological basis and responsiveness of individual reproductive nerves, stimulation parameters need to be adaptively adjusted according to the patient's real-time physiological signals to ensure the efficacy and safety of electrical stimulation modulation. With technological advancements, the demand for electrical stimulation modulation of reproductive nerve function is gradually shifting towards automation, that is, automatically adjusting electrical stimulation parameters without human intervention. This automated electrical stimulation modulation can not only provide stimulation programs adapted to changes in physiological signals, but also has the potential to reduce reliance on physician experience, thereby improving treatment efficiency.

[0003] Traditional electrical stimulation modulation techniques are divided into open-loop electrical stimulation control, manual parameter titration, and single-factor optimization. First, open-loop electrical stimulation control refers to continuously releasing electrical stimulation signals according to preset parameters without monitoring the patient's physiological feedback after stimulation. Therefore, there is no issue of matching stimulation parameters with the patient's physiological changes. Second, manual parameter titration involves the doctor subjectively selecting stimulation parameters based on the patient's physiological signal detection results when setting initial stimulation parameters. Subsequent adjustments also require the doctor to combine the patient's physiological indicators with subjective feedback. However, doctors cannot be with patients 24 hours a day, which means that when the doctor is not present, the stimulation parameters cannot adapt to the patient's physiological changes. The signal changes adaptively. Finally, the single-factor optimization method refers to adjusting only one parameter at a time when adaptively adjusting stimulus parameters for the patient. After determining the range of this one parameter, other parameters are gradually determined. The principle of this method is to assume that each type of parameter is independent of each other. Only one parameter is adjusted at a time, and the other parameters are set to default values. In addition, this method is based on the current state of the patient, trying parameters one by one from zero to find the local optimum. Therefore, this method cannot reflect the mutual coupling relationship between different types of parameters, nor does it refer to historical parameter information and physiological information. These two defects lead to insufficient accuracy of parameter regulation in adapting to the patient's physiological signals.

[0004] Therefore, existing technologies suffer from insufficient adaptability to changes in the physiological state of users and the regulation of reproductive nerve stimulation parameters. Summary of the Invention

[0005] This invention provides a closed-loop reproductive nerve stimulation system based on multimodal biofeedback, the main purpose of which is to solve the problem of insufficient adaptation of reproductive nerve stimulation parameters to user physiological changes.

[0006] To achieve the above objectives, the present invention provides a closed-loop reproductive nerve stimulation system based on multimodal biofeedback, the system comprising a remote server, a data processing module, a microcontroller, and a skin conduction nerve stimulator; A remote server is used to collect controlled trial data on reproductive nerve samples under various stimulation parameters. The effect index sample reflecting reproductive nerve function is used as the outcome indicator. The prior relative effect and prior precision of each combination of stimulation parameters in the controlled trial data are analyzed to form a prior matrix. The data processing module is used to collect multimodal biofeedback data reflecting the function of the reproductive nerves, and to calculate the current effect index of the current reproductive nerves based on the multimodal biofeedback data. A microcontroller is used to connect to the data processing module and the remote server respectively. It performs Bayesian update on the prior matrix through the current effect index to obtain the posterior relative effect and posterior precision. Based on the posterior relative effect and posterior precision, it uses the network meta-ranking method to calculate the SUCRA probability corresponding to each stimulus parameter and uses the stimulus parameter corresponding to the highest SUCRA probability as the output parameter of the next beat. The skin electrostimulator is controlled by the microcontroller to perform electrical stimulation of the current reproductive nerve based on the output parameters of the next beat.

[0007] Optionally, using an effect index sample reflecting reproductive nerve function as the outcome indicator, the prior relative effect and prior precision of each stimulus parameter combination in the controlled trial data are analyzed, including: The aforementioned outcome index is used as a prior relative effect; The effective-false variance of each type of physiological signal is calculated using the standard deviation and sample size of each physiological signal in the second multimodal biofeedback data and the standard deviation and sample size of the corresponding type of physiological signal in the first multimodal biofeedback data. The specific formula is as follows:

[0008] in, Indicates effective - pseudo variance. This represents the standard deviation of the i-th type of physiological signal in the second multimodal biofeedback data. This represents the sample size of the i-th type of physiological signal in the second multimodal biofeedback data. The 1 in the equation represents the sequence number of the first prior precision calculation relative to the 0 subscript of the sham stimulus group. This represents the standard deviation of the physiological signal of the i-th type in the first multimodal biofeedback data. This represents the sample size of the i-th type of physiological signal in the first multimodal biofeedback data. and The subscript 0 in the figure represents the sequence number of the 0th baseline data in the sham stimulation group; By using weighting coefficients to sum the effective-false variances of various types of physiological signals, the effective-false variances corresponding to each combination of stimulus parameters are obtained. The inverse of the effective-false variance corresponding to each stimulus parameter combination is taken as the prior accuracy.

[0009] Optionally, the determination of the effect index sample includes: The difference between each physiological signal in the second multimodal biofeedback data and the corresponding physiological signal in the first multimodal biofeedback data is calculated to obtain the effective-false difference value of each type of physiological signal. By using weighting coefficients to sum the effective-false differences of various types of physiological signals, the effect index sample corresponding to each combination of stimulus parameters is obtained.

[0010] Optional, the determination of weighting coefficients includes: The effect index values ​​corresponding to physiological signals with different numerical combinations were analyzed using the expert scoring method. Establish a regression model between physiological signals with different numerical combinations and their corresponding effect index values; The weight coefficients in the regression model are solved using the least squares method.

[0011] Optionally, control trial data for reproductive nerve samples under multiple stimulation parameters can be collected, including: first multimodal biofeedback data of sham stimulation of reproductive nerve samples under multiple combinations of stimulation parameters and second multimodal biofeedback data of effective stimulation of reproductive nerve samples. The control data were determined by combining the first multimodal biofeedback data with the second multimodal biofeedback data.

[0012] Optionally, a priori matrix is ​​formed, including: A prior matrix is ​​constructed using the prior relative effect and prior precision corresponding to each combination of stimulus parameters as row data.

[0013] Optionally, multimodal biofeedback data reflecting the reproductive neural function may be collected, including: Collect initial multimodal biological feedback data at each moment continuously within a fixed time period; The mean of the initial multimodal biofeedback data at each consecutive time point is used as the multimodal biofeedback data over a continuous fixed time period.

[0014] Optionally, based on the multimodal biofeedback data, the current effect index of the current reproductive nerve is calculated, including: The current effect index is obtained by weighting and summing the multimodal biological feedback data of different types within the same continuous fixed time period using weighting coefficients.

[0015] Optionally, the prior matrix is ​​updated using the current effect index via Bayesian method to obtain the posterior relative effect and posterior accuracy, including: Based on the current effect index and the prior matrix, the posterior relative effect is calculated using the following formula:

[0016] in, This represents the posterior relative effect. This represents the relative prior effects in the prior matrix. This indicates the prior precision in the prior matrix. This indicates the real-time signal accuracy of the current effect index. Indicates the current effect index; Based on the real-time signal accuracy corresponding to the current effect index and the prior accuracy in the prior matrix, the posterior accuracy is calculated using the following formula:

[0017] in, Indicates posterior precision. Indicates prior precision. Indicates the real-time signal accuracy.

[0018] Optionally, based on the posterior relative effect and the posterior precision, a network meta-ranking method is used to calculate the SUCRA probability corresponding to each stimulus parameter, including: The network meta-sorting method includes: The effect value is randomly drawn from the effect interval determined by the posterior relative effect and the posterior precision; Compare the effect values ​​corresponding to different combinations of stimulus parameters to select the maximum effect value in each round; The number of times the maximum effect value occurs for each combination of stimulus parameters is counted, and the SUCRA probability corresponding to each combination of stimulus parameters is calculated based on the total number of random sampling rounds and the number of times the maximum effect value occurs for each combination of stimulus parameters.

[0019] Compared to the problems described in the background art, the embodiments of the present invention collect control trial data to calculate subsequent effect index samples and to provide historical parameter and physiological information for reference in subsequent steps. Furthermore, the embodiments of the present invention use prior relative effects to reflect the expected effects of various stimulation parameter combinations relative to the sham stimulation group before the actual collection of real-time feedback data from individual patients. These prior relative effects also provide a starting point for subsequent Bayesian updates, allowing new data to be adjusted and refined based on this. Simultaneously, the provided prior precision, together with the subsequent posterior precision, determines the shape of the posterior distribution. Furthermore, the embodiments of the present invention collect current real-time multimodal biofeedback data, distinct from the aforementioned historical sample data, to prepare for subsequent adaptive customization of stimulation parameters that change with physiological variations for each individual patient. Furthermore, the embodiments of the present invention integrate the real-time collected multimodal biofeedback data into the current effect... The current effect index is used to match the numerical type of the aforementioned effect index samples, facilitating subsequent updates of the effect index samples. Furthermore, in this embodiment, the current effect index and prior matrix are transmitted to a microcontroller carried by the patient, enabling the system to automatically calculate control commands and directly stimulate the patient without doctor intervention. Further, this embodiment updates the prior matrix using Bayesian methods, updating the prior data with the patient's personalized data, resulting in posterior data that is more adapted to the patient's physiological changes. Further, this embodiment quantifies the relative advantage of each parameter combination by calculating the number of times each combination of stimulation parameters achieves its maximum effect value. Furthermore, this embodiment uses Bayesian and network meta-sorting methods to determine stimulation parameters based on biofeedback data, addressing the problem of insufficient adaptation of reproductive nerve stimulation parameters to user physiological changes. Therefore, this invention can solve the problem of insufficient adaptation of reproductive nerve stimulation parameters to user physiological changes. Attached Figure Description

[0020] Figure 1 A schematic diagram of a traditional nerve stimulation system for a closed-loop reproductive nerve stimulation system based on multimodal biofeedback provided in an embodiment of the present invention; Figure 2 A schematic diagram of the traditional stimulation process of a closed-loop reproductive nerve stimulation system based on multimodal biofeedback provided in an embodiment of the present invention; Figure 3 This is a schematic diagram of the modules for implementing the closed-loop reproductive nerve stimulation system based on multimodal biofeedback, according to an embodiment of the present invention. Figure 4 A data transmission flow diagram between modules of a closed-loop reproductive nerve stimulation system based on multimodal biofeedback provided in an embodiment of the present invention; Figure 5 A star-shaped evidence network for a closed-loop reproductive nerve stimulation system based on multimodal biofeedback, provided in an embodiment of the present invention; Figure 6 This is the prior matrix of a closed-loop reproductive nerve stimulation system based on multimodal biofeedback provided in an embodiment of the present invention; Figure 7 A flowchart illustrating the determination of posterior data for a closed-loop reproductive nerve stimulation system based on multimodal biofeedback, provided in an embodiment of the present invention. Figure 8 The normal distribution diagram of a closed-loop reproductive nerve stimulation system based on multimodal biofeedback provided in an embodiment of the present invention; Figure 9 SUCRA probability ranking diagram of a closed-loop reproductive nerve stimulation system based on multimodal biofeedback provided in an embodiment of the present invention; Figure 10 A diagram of a skin electroreceptor stimulator for a closed-loop reproductive nerve stimulation system based on multimodal biofeedback, provided in an embodiment of the present invention; Figure 11 The stimulation flowchart of a closed-loop reproductive nerve stimulation system based on multimodal biofeedback provided in an embodiment of the present invention is shown below. Figure 12 This is a flowchart illustrating a closed-loop reproductive nerve stimulation method based on multimodal biofeedback provided in an embodiment of the present invention. Figure 13 A schematic diagram of a computer device for a closed-loop reproductive nerve stimulation system based on multimodal biofeedback provided in an embodiment of the present invention; The objectives, features, and advantages of this invention will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0021] It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.

[0022] This application provides a closed-loop reproductive nerve stimulation system based on multimodal biofeedback. The executing entity of the closed-loop reproductive nerve stimulation system based on multimodal biofeedback includes, but is not limited to, at least one of the following electronic devices that can be configured to execute the method provided in this application: a server, a terminal, etc. In other words, the closed-loop reproductive nerve stimulation system based on multimodal biofeedback can be executed by software or hardware installed on a terminal device or a server device. The server includes, but is not limited to, a single server, a server cluster, a cloud server, or a cloud server cluster.

[0023] See Figure 1The diagram shown is a schematic of a traditional neural stimulation system for a closed-loop reproductive neural stimulation system based on multimodal biofeedback, provided in an embodiment of the present invention. Figure 1 The image shows a traditional reproductive nerve stimulation system. 1 represents an indicator light, 2 represents a control switch, and 3 represents a ring. The electrical stimulator is housed within the ring. Stimulation of subcutaneous nerves is achieved by bringing the ring into contact with the skin surface of the genital area. Figure 1 It can be seen that this traditional reproductive nerve stimulation system requires the user to manually control the switch to adjust the stimulation frequency.

[0024] See Figure 2 The diagram shown illustrates the traditional stimulation process of a closed-loop reproductive nerve stimulation system based on multimodal biofeedback, according to an embodiment of the present invention. Figure 2 The image shows the traditional process of reproductive nerve stimulation, with the ring representing the stimulation. Figure 1 Traditional nerve stimulation systems are fitted onto the genitals; the skin patch indicates that... Figure 1 Traditional nerve stimulation systems are applied close to the skin. Users select parameters such as the stimulation frequency. After selecting the parameters, they manually press a switch to turn the electrical stimulation of the reproductive nerves on or off.

[0025] Based on the above scenarios, it can be seen that traditional reproductive nerve stimulation processes rely on doctors or users to select stimulation parameters and then manually stimulate the reproductive nerves by controlling a switch. This traditional method cannot adaptively adjust the stimulation parameters based on the physiological changes in reproductive nerve function, which may lead to overstimulation or ineffective stimulation.

[0026] In view of this, the closed-loop reproductive nerve stimulation system based on multimodal biofeedback provided in this application collects multimodal biofeedback data on reproductive nerve function and uses the multimodal biofeedback data to adaptively adjust stimulation parameters, thereby reducing the mismatch between stimulation parameters and physiological changes in reproductive nerves.

[0027] The technical solution of this application and how the technical solution of this application solves the above-mentioned technical problems are described in detail below with specific embodiments.

[0028] like Figure 3 The diagram shown is a schematic diagram of the module of the closed-loop reproductive nerve stimulation system based on multimodal biofeedback of the present invention.

[0029] The closed-loop reproductive nerve stimulation system 300 based on multimodal biofeedback described in this invention can be installed in an electronic device. Depending on the functions implemented, the closed-loop reproductive nerve stimulation system includes a remote server 301, a data processing module 302, a microcontroller 303, and a skin conduction nerve stimulator 304. The module described in this invention can also be called a unit, referring to a series of computer program segments that can be executed by the processor of an electronic device and perform a fixed function, stored in the memory of the electronic device.

[0030] In this embodiment of the invention, the functions of each module / unit are as follows: See Figure 4 The diagram shown is a data transmission flow diagram between modules of a closed-loop reproductive nerve stimulation system based on multimodal biofeedback provided in an embodiment of the present invention. Figure 4 In this process, data such as literature and experimental records collected in reality are transmitted to a remote server, which analyzes the prior matrix. The data processing module collects real-time biofeedback data of the reproductive nerves and calculates a current effect index. The prior matrix and the current effect index are then transmitted to the microcontroller, which outputs the next step parameters to the electrodermal nerve stimulator. The electrodermal nerve stimulator stimulates the reproductive nerves, thus forming a closed-loop reproductive nerve stimulation system.

[0031] The remote server 301 is used to collect control trial data for reproductive nerve samples under multiple stimulation parameters, using the effect index sample reflecting reproductive nerve function as the outcome indicator, and to analyze the prior relative effect and prior precision of each combination of stimulation parameters in the control trial data to form a prior matrix.

[0032] In this embodiment of the invention, by collecting data from control experiments, the purpose is twofold: firstly, to calculate subsequent effect index samples, and secondly, to provide historical parameter and physiological information for reference in subsequent steps.

[0033] The reproductive nerve samples refer to reproductive nerve information collected from various literature and previous experimental records. The purpose of identifying these reproductive nerve samples is to study the biological responses of the reproductive nerves under different stimulation parameters. It should be noted that this study focuses on only one type of reproductive nerve, such as the dorsal penile nerve or the pudendal nerve. Since the types of biofeedback data exhibited by each reproductive nerve are different, subsequent electrical stimulation adjustments based on biofeedback data can only be performed on the same type of reproductive nerve. For example, either only the dorsal penile nerve or only the pudendal nerve will be studied. The stimulation parameters refer to the parameters preset when the electrodermal nerve stimulator performs surface stimulation.

[0034] In one embodiment of the present invention, the collection of control test data for reproductive nerve samples under multiple stimulation parameters includes: collecting first multimodal biofeedback data when sham stimulation is applied to reproductive nerve samples under a combination of multiple types of stimulation parameters, and second multimodal biofeedback data when effective stimulation is applied to reproductive nerve samples; and determining control test data through the first multimodal biofeedback data and the second multimodal biofeedback data.

[0035] The control experiment data includes the names of different stimulation parameters under sham and effective stimuli, the specific values ​​of different combinations of stimulation parameters, and the multimodal biofeedback data corresponding to different combinations of stimulation parameters.

[0036] For example, the first multimodal biofeedback data when sham stimulation is applied to a reproductive nerve sample, and the second multimodal biofeedback data when effectively stimulating a reproductive nerve sample, are collected under a combination of various types of stimulation parameters. Examples of these include various types of stimulation parameters such as frequency, amplitude, and pulse width. Sham stimulation refers to parameters that are not set or are set to extremely low values ​​that have no stimulating effect, such as setting frequency, amplitude, and pulse width to 0. Effective stimulation is the opposite of sham stimulation; it consists of stimulation parameters that can affect the patient's reproductive nerve. The first multimodal biofeedback data is the patient's physiological signal data during sham stimulation. For example, when the reproductive nerve is the dorsal penile nerve, the types of multimodal biofeedback data include skin sympathetic potential, temperature, and photocapacitance. The first multimodal biofeedback data, such as pulse waves, is collected under sham stimulation. Therefore, the values ​​of the first multimodal biofeedback data are generally the baseline values ​​when there is no stimulation. The second multimodal biofeedback data, on the other hand, is collected under effective stimulation. It should be noted that the second multimodal biofeedback data consists of data under different combinations of stimulation parameters. For example, if there are stimulation parameter combinations A, B, and C, where A is a frequency of 5Hz, an amplitude of 10mA, and a pulse width of 200µs, B is a frequency of 10Hz, an amplitude of 15mA, and a pulse width of 250µs, and C is a frequency of 20Hz, an amplitude of 8mA, and a pulse width of 300µs, then the second multimodal biofeedback data includes the data collected under combination A, combination B, and combination C.

[0037] See Figure 5 The image shown illustrates a star-shaped evidence network for a closed-loop reproductive neural stimulation system based on multimodal biofeedback, provided in an embodiment of the present invention. Figure 5 In the diagram, A, B, and C represent effective stimulus groups, and 4 represents sham stimulus group. A, B, and C are connected to 4 respectively, indicating that each effective stimulus group is compared with the sham stimulus group.

[0038] It should be noted that when each term appears later, it indicates the type of data, such as frequency, amplitude, and pulse width. When each term appears, it indicates the specific value under a fixed type of parameter, such as frequency 10Hz, amplitude 15mA, and pulse width 250µs. It needs to be interpreted adaptively based on the actual descriptions that follow.

[0039] Furthermore, in this embodiment of the invention, the prior relative effect is used to reflect the expected effect of each combination of stimulation parameters relative to the sham stimulation group before the actual collection of real-time feedback data from individual patients. The prior relative effect can also provide a starting point for subsequent Bayesian updates, so that new data can be adjusted and refined on this basis. At the same time, the provided prior accuracy, together with the subsequent posterior accuracy, determines the shape of the posterior distribution.

[0040] The prior relative effect represents the estimate of the expected effect size of different combinations of stimulation parameters relative to the sham stimulation group. The prior precision describes the reliability of the prior relative effect estimate, i.e. the uncertainty of the prior effect estimate. High precision means that the prior effect estimate is more reliable and has less impact on the posterior update. Low precision means that the estimate is less certain and the posterior data has a greater impact on the posterior effect. The outcome index refers to the numerical value for evaluating the efficacy of electrical stimulation. Here, the effect index sample is directly used as the outcome index.

[0041] In one embodiment of the present invention, the step of using an effect index sample reflecting reproductive nerve function as the outcome indicator to analyze the prior relative effect and prior precision of each stimulus parameter combination in the control trial data includes: using the outcome indicator as the prior relative effect; calculating the effective-false variance of each type of physiological signal using the standard deviation and sample size of each physiological signal in the second multimodal biofeedback data and the standard deviation and sample size of the corresponding type of physiological signal in the first multimodal biofeedback data; weighting and summing the effective-false variance of each type of physiological signal using weighting coefficients to obtain the effective-false variance corresponding to each stimulus parameter combination; and taking the reciprocal of the effective-false variance corresponding to each stimulus parameter combination as the prior precision.

[0042] In another embodiment of the present invention, the step of calculating the effective-false variance of each type of physiological signal using the standard deviation and sample size of each physiological signal in the second multimodal biofeedback data and the standard deviation and sample size of the corresponding type of physiological signal in the first multimodal biofeedback data includes: calculating the effective-false variance using the following formula:

[0043] in, Indicates effective - pseudo variance. This represents the standard deviation of the i-th type of physiological signal in the second multimodal biofeedback data. This represents the sample size of the i-th type of physiological signal in the second multimodal biofeedback data. The 1 in the equation represents the sequence number of the first prior precision calculation relative to the 0 subscript of the sham stimulus group. This represents the standard deviation of the physiological signal of the i-th type in the first multimodal biofeedback data. This represents the sample size of the i-th type of physiological signal in the first multimodal biofeedback data. The 0 subscript in the figure indicates the sequence number of the 0th baseline data in the sham stimulation group.

[0044] It should be noted that since the changes in biofeedback caused by stimulation at a certain moment are not obvious, the biofeedback data considered here and in the following discussion are actually data within a fixed duration. For example, if the above-mentioned stimulation parameter combination A is used on a patient, and the patient is stimulated 10 times, with each stimulation lasting 50 seconds, then the sample size is 10. The value of each sample is the mean within 50 seconds of each stimulation, resulting in a total of 10 sample sizes = 10 means. The standard deviation is also the standard deviation among these 10 samples.

[0045] In another embodiment of the present invention, the determination of the effect index sample includes: calculating the difference between each physiological signal in the second multimodal biofeedback data and the corresponding type of physiological signal in the first multimodal biofeedback data to obtain the effective-false difference value of each type of physiological signal; and using weighting coefficients to perform weighted summation of the effective-false difference values ​​of each type of physiological signal to obtain the effect index sample corresponding to each stimulus parameter combination.

[0046] Similarly, when calculating the difference here, if the fixed duration is 50 seconds, then the calculation is the difference between the mean of the i-th physiological signal within 50 seconds and the mean of the i-th physiological signal in the sham stimulation group. Here, and for each subsequent physiological signal, the value does not represent the value at a certain moment, but the mean over a certain period of time. If the same combination of stimulation parameters is used for multiple stimulations, the mean of the multiple stimulations needs to be taken before calculating the difference. This will not be elaborated on in detail here.

[0047] In another embodiment of the present invention, the determination of the weight coefficients includes: analyzing the effect index values ​​corresponding to physiological signals with different numerical combinations using an expert scoring method; establishing a regression model between physiological signals with different numerical combinations and their corresponding effect index values; and solving for the weight coefficients in the regression model using the least squares method.

[0048] The specific steps of the expert scoring method include: in a controlled environment, applying different combinations of stimulus parameters to a group of patients, having experts score the effect of each combination of stimulus parameters based on the patients' multimodal biofeedback data, and taking the average of the scores from multiple experts as the effect index value corresponding to this group of multimodal biofeedback data. The least squares method refers to the existing conventional technique for solving unknown weight coefficients in a regression model that minimizes the sum of squares of the differences between the model's predicted values ​​and the actual observed values.

[0049] In one embodiment of the present invention, forming the prior matrix includes: establishing a prior matrix with the prior relative effect and prior precision corresponding to each stimulus parameter combination as row data.

[0050] The prior matrix, for example, has the following structure: the first row and first column represent the prior relative effect corresponding to the above-mentioned combination of A stimulus parameters, and the first row and second column represent the prior precision corresponding to the combination of A stimulus parameters.

[0051] See Figure 6 The figure shown is the prior matrix of a closed-loop reproductive nerve stimulation system based on multimodal biofeedback provided in an embodiment of the present invention. Figure 6 In the middle, the three rows of data represent the prior relative effect and prior precision corresponding to the combination of stimulus parameters A, B, and C, respectively.

[0052] The data processing module 302 is used to collect multimodal biofeedback data reflecting the function of the reproductive nerves, and to calculate the current effect index of the current reproductive nerves based on the multimodal biofeedback data.

[0053] This invention collects real-time multimodal biofeedback data, which differs from the aforementioned historical sample data, to prepare for the subsequent adaptive customization of stimulation parameters that change with physiological variations for each individual patient.

[0054] In one embodiment of the present invention, the acquisition of multimodal biofeedback data reflecting the reproductive nerve function includes: acquiring initial multimodal biofeedback data at each continuous moment within a continuous fixed time period; and using the average of the initial multimodal biofeedback data at each continuous moment as the multimodal biofeedback data within the continuous fixed time period.

[0055] It should be noted that the fixed time period selected for the appearance of physiological signals in the aforementioned control experiment data should be consistent with the continuous fixed time period mentioned here.

[0056] Furthermore, in this embodiment of the invention, multimodal biological feedback data collected in real time is integrated into a current effect index to maintain numerical type matching with the aforementioned effect index samples, facilitating subsequent updates of the effect index samples using the current effect index.

[0057] In one embodiment of the present invention, the step of calculating the current effect index of the current reproductive nerve based on the multimodal biofeedback data includes: using weighting coefficients to perform weighted summation of different types of multimodal biofeedback data within the same continuous fixed time period to obtain the current effect index.

[0058] The microcontroller 303 is used to connect to the data processing module and the remote server respectively. It performs Bayesian update on the prior matrix through the current effect index to obtain the posterior relative effect and posterior precision. Based on the posterior relative effect and posterior precision, it uses the network meta-sorting method to calculate the SUCRA probability corresponding to each stimulus parameter and uses the stimulus parameter corresponding to the highest SUCRA probability as the output parameter for the next beat.

[0059] In this embodiment of the invention, the current effect index and prior matrix are transmitted to a microcontroller carried by the patient, so that the control instructions can be calculated automatically without the participation of a doctor, and the stimulation control can be directly applied to the patient.

[0060] In one embodiment of the present invention, after the microcontroller is connected to the data processing module and the remote server respectively, the method further includes: the data processing module transmitting the current effect index to the microcontroller, and the remote server transmitting the prior matrix to the microcontroller.

[0061] Furthermore, in this embodiment of the invention, the prior matrix is ​​updated using Bayesian methods to update the prior data with the current patient's personalized data, resulting in posterior data that is more adapted to the patient's physiological changes than the prior data.

[0062] In one embodiment of the present invention, the step of performing a Bayesian update on the prior matrix using the current effect index to obtain the posterior relative effect and posterior accuracy includes: calculating the posterior relative effect based on the current effect index and the prior matrix; and calculating the posterior accuracy based on the real-time signal accuracy corresponding to the current effect index and the prior accuracy in the prior matrix.

[0063] In another embodiment of the present invention, the step of calculating the posterior relative effect based on the current effect index and the prior matrix includes: calculating the posterior relative effect based on the current effect index and the prior matrix using the following formula:

[0064] in, This represents the posterior relative effect. This represents the relative prior effects in the prior matrix. This indicates the prior precision in the prior matrix. This indicates the real-time signal accuracy of the current effect index. This represents the current effect index.

[0065] Here, the real-time signal precision is set to 1, which is equivalent to assuming that the variance of the current effect index is 1, and that the current data has basic reliability by default.

[0066] In another embodiment of the present invention, the step of calculating the posterior accuracy based on the real-time signal accuracy corresponding to the current effect index and the prior accuracy in the prior matrix includes: calculating the posterior accuracy using the following formula based on the real-time signal accuracy corresponding to the current effect index and the prior accuracy in the prior matrix:

[0067] in, Indicates posterior precision. Indicates prior precision. Indicates the real-time signal accuracy.

[0068] See Figure 7 The diagram shown is a flowchart illustrating the determination of posterior data for a closed-loop reproductive nerve stimulation system based on multimodal biofeedback, according to an embodiment of the present invention. Figure 7 In this study, the prior relative effect and prior precision are determined by comparing the biofeedback data of the sham stimulus group and the effective stimulus group. The current effect index is calculated by collecting real-time multimodal biofeedback data. The prior relative effect and prior precision are then updated to posterior relative effect and posterior precision using Bayesian methods based on the preset real-time signal precision and current effect index.

[0069] Furthermore, by calculating the number of times each combination of stimulus parameters exhibits its maximum effect value, the relative advantage of each parameter combination can be quantified.

[0070] In one embodiment of the present invention, the step of calculating the SUCRA probability corresponding to each stimulus parameter using a network meta-ranking method based on the posterior relative effect and the posterior precision includes: the network meta-ranking method includes: randomly sampling effect values ​​from the effect interval determined by the posterior relative effect and the posterior precision; comparing the effect values ​​corresponding to different combinations of stimulus parameters to select the maximum effect value in each round; counting the number of times the maximum effect value occurs for each combination of stimulus parameters, and calculating the SUCRA probability corresponding to each combination of stimulus parameters based on the total number of random sampling rounds and the number of times the maximum effect value occurs for each combination of stimulus parameters.

[0071] Wherein, the posterior relative effect and the posterior precision determine the normal distribution. ~N( , The posterior relative effect determines the center position of the normal distribution, and the posterior precision determines the interval width. The higher the posterior precision, the stronger the reliability of the data, and the narrower the interval, the smaller the uncertainty of the effect. The lower the posterior precision, the higher the data dispersion, and the wider the interval, the greater the uncertainty of the effect. The effect interval refers to the interval in which the normal distribution is centered on the posterior relative effect and has an interval width of posterior precision.

[0072] See Figure 8 The figure shown is a normal distribution diagram of a closed-loop reproductive nerve stimulation system based on multimodal biofeedback provided in an embodiment of the present invention. Figure 8 middle, This indicates the central location of the normal distribution.

[0073] It should be noted that the above-mentioned mesh meta-sorting method only uses a portion of the steps of the existing conventional mesh meta-sorting method. This algorithm is not a custom algorithm defined in this solution.

[0074] For example, the comparison of effect values ​​corresponding to different stimulus parameter combinations to select the maximum effect value in each round, for instance, involves 100 random sampling rounds. The first round involves sampling the random effect values ​​corresponding to stimulus parameter combination A, stimulus parameter combination B, and stimulus parameter combination C. The largest effect value among these three is selected as the maximum effect value in the first sampling. Furthermore, the number of times the maximum effect value appears for each stimulus parameter combination is counted to calculate the SUCRA probability corresponding to each stimulus parameter combination based on the total number of random sampling rounds and the number of times the maximum effect value appears for each stimulus parameter combination. For example, the 100 sampling rounds can yield 100 maximum effect values ​​belonging to different stimulus parameter combinations. To calculate the SUCRA probability corresponding to stimulus parameter combination A, the proportion of the number of times the maximum effect value belongs to stimulus parameter combination A among the 100 maximum effect values ​​to the total number of sampling rounds is calculated.

[0075] See Figure 9 The image shown is a SUCRA probability ranking diagram of a closed-loop reproductive nerve stimulation system based on multimodal biofeedback provided in an embodiment of the present invention. Figure 9 In the diagram, the bars in the first row represent the highest probability of SUCRA, and the bars in the last row represent the lowest probability of SUCRA.

[0076] The skin conduction nerve stimulator 304, controlled by the microcontroller, performs electrical stimulation of the current reproductive nerve based on the output parameters of the next beat.

[0077] The embodiments of the present invention determine stimulation parameters based on biofeedback data using Bayesian and reticular meta-sorting methods, which can solve the problem of insufficient regulation of reproductive nerve stimulation parameters and insufficient adaptation to user physiological changes.

[0078] In one embodiment of the present invention, the step of controlling the microcontroller to perform electrical stimulation of the current reproductive nerve according to the next-beat output parameters includes: after receiving the next-beat output parameters from the microcontroller in the electrodermal nerve stimulator, adjusting each stimulation parameter in the electrodermal nerve stimulator according to the next-beat output parameters to drive the electrode pads of the electrodermal nerve stimulator to perform electrical stimulation of the current reproductive nerve.

[0079] See Figure 10 The image shown is a diagram of the electrodermal stimulator in a closed-loop reproductive nerve stimulation system based on multimodal biofeedback, according to an embodiment of the present invention. Figure 10 In the diagram, 5 indicates that the microcontroller and the electrodermal stimulator are integrated on a single board, 6 indicates the positive electrode of the electrodermal stimulator, and 7 indicates the negative electrode of the electrodermal stimulator. The electrodes stimulate the reproductive nerves by contacting the body surface.

[0080] See Figure 11 The diagram shown is a stimulation flowchart of a closed-loop reproductive nerve stimulation system based on multimodal biofeedback provided in an embodiment of the present invention. Figure 11 In the process, the microcontroller first outputs the parameters for the next beat, then the electrodermal stimulator adjusts the stimulation value according to the parameters, and then the electrodermal stimulator transmits the adjusted stimulation signal to the electrode pads. After receiving the signal, the electrode pads stimulate the surface of the genitals, and finally transmit this stimulation to the reproductive nerves.

[0081] Reference Figure 12 The diagram shown is a flowchart illustrating a closed-loop reproductive nerve stimulation method based on multimodal biofeedback provided in an embodiment of the present invention. In this embodiment, the closed-loop reproductive nerve stimulation method based on multimodal biofeedback includes: S1. Collect controlled trial data for reproductive nerve samples under multiple stimulation parameters, and use the effect index sample reflecting reproductive nerve function as the outcome indicator to analyze the prior relative effect and prior precision of each combination of stimulation parameters in the controlled trial data to form a prior matrix. S2. Collect multimodal biofeedback data reflecting the function of the reproductive nerve, and calculate the current effect index of the current reproductive nerve based on the multimodal biofeedback data; S3. Connect to the data processing module and the remote server respectively, perform Bayesian update on the prior matrix through the current effect index to obtain the posterior relative effect and posterior precision, and calculate the SUCRA probability corresponding to each stimulus parameter using the network meta-sorting method based on the posterior relative effect and the posterior precision, and take the stimulus parameter corresponding to the highest SUCRA probability as the output parameter of the next beat. S4. The microcontroller controls the execution of electrical stimulation of the current reproductive nerve based on the output parameters of the next beat.

[0082] In detail, each step in the closed-loop reproductive nerve stimulation method based on multimodal biofeedback described in this embodiment of the invention adopts the same method as described above. Figure 3 The technology used is the same as that of the closed-loop reproductive nerve stimulation system based on multimodal biofeedback described in the article, and can produce the same technical effects, so it will not be elaborated here.

[0083] In one embodiment, a computer device is provided, which may be a server or a client, and its internal structure diagram may be as follows: Figure 13 As shown, the computer device includes a processor, memory, network interface, and database connected via a system bus. The processor provides computational and control capabilities. The memory includes non-volatile and / or volatile storage media and internal memory. The non-volatile storage media stores the operating system, computer programs, and database. The internal memory provides an environment for the operation of the operating system and computer programs in the non-volatile storage media. The network interface is used for communication with external clients via a network connection. When executed by the processor, the computer program implements functions or steps on the server or client side of a closed-loop reproductive nerve stimulation system based on multimodal biofeedback.

[0084] In one embodiment, a computer device is provided, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to perform the following steps: We collected controlled trial data on reproductive nerve samples under various stimulation parameters, using the effect index sample reflecting reproductive nerve function as the outcome indicator, and analyzed the prior relative effect and prior precision of each combination of stimulation parameters in the controlled trial data to form a prior matrix. Collect multimodal biofeedback data reflecting the function of the reproductive nerves, and calculate the current effect index of the current reproductive nerves based on the multimodal biofeedback data; The system is connected to the data processing module and the remote server respectively. The prior matrix is ​​updated using the current effect index to obtain the posterior relative effect and posterior precision. Based on the posterior relative effect and posterior precision, the network meta-ranking method is used to calculate the SUCRA probability corresponding to each stimulus parameter. The stimulus parameter corresponding to the highest SUCRA probability is used as the output parameter for the next beat. The microcontroller controls the execution of electrical stimulation of the current reproductive nerve based on the output parameters of the next beat.

[0085] In one embodiment, a computer-readable storage medium is provided having a computer program stored thereon, the computer program performing the following steps when executed by a processor: We collected controlled trial data on reproductive nerve samples under various stimulation parameters, using the effect index sample reflecting reproductive nerve function as the outcome indicator, and analyzed the prior relative effect and prior precision of each combination of stimulation parameters in the controlled trial data to form a prior matrix. Collect multimodal biofeedback data reflecting the function of the reproductive nerves, and calculate the current effect index of the current reproductive nerves based on the multimodal biofeedback data; The system is connected to the data processing module and the remote server respectively. The prior matrix is ​​updated using the current effect index to obtain the posterior relative effect and posterior precision. Based on the posterior relative effect and posterior precision, the network meta-ranking method is used to calculate the SUCRA probability corresponding to each stimulus parameter. The stimulus parameter corresponding to the highest SUCRA probability is used as the output parameter for the next beat. The microcontroller controls the execution of electrical stimulation of the current reproductive nerve based on the output parameters of the next beat.

[0086] It should be noted that the functions or steps that can be implemented by the computer-readable storage medium or computer device described above can be referred to the relevant descriptions on the server side and client side in the foregoing method embodiments. To avoid repetition, they will not be described one by one here.

[0087] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium. When executed, the computer program can include the processes of the embodiments of the above methods. Any references to memory, storage, databases, or other media used in the embodiments provided in this application can include non-volatile and / or volatile memory. Non-volatile memory may include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memory may include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in a variety of forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), dual data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link DRAM (SLDRAM), RAMbus direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM), etc.

[0088] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the above-described division of functional units and modules is used as an example. In practical applications, the above functions can be assigned to different functional units and modules as needed, that is, the internal structure of the system can be divided into different functional units or modules to complete all or part of the functions described above.

[0089] It will be apparent to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above, and that the present invention can be implemented in other specific forms without departing from the spirit or essential characteristics of the present invention.

[0090] Finally, it should be noted that in the above embodiments, each embodiment can be combined with each other or independent. Deleting any one of them will not affect the technical implementation of other embodiments. The above embodiments are only used to illustrate the technical solutions of the present invention and not to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention.

Claims

1. A closed-loop reproductive nerve stimulation system based on multimodal biofeedback, characterized in that, The system includes a remote server, a data processing module, a microcontroller, and a skin conduction nerve stimulator; A remote server is used to collect controlled trial data on reproductive nerve samples under various stimulation parameters. The effect index sample reflecting reproductive nerve function is used as the outcome indicator. The prior relative effect and prior precision of each combination of stimulation parameters in the controlled trial data are analyzed to form a prior matrix. The data processing module is used to collect multimodal biofeedback data reflecting the function of the reproductive nerves, and to calculate the current effect index of the current reproductive nerves based on the multimodal biofeedback data. A microcontroller is used to connect to the data processing module and the remote server respectively. It performs Bayesian update on the prior matrix through the current effect index to obtain the posterior relative effect and posterior precision. Based on the posterior relative effect and posterior precision, it uses the network meta-ranking method to calculate the SUCRA probability corresponding to each stimulus parameter and uses the stimulus parameter corresponding to the highest SUCRA probability as the output parameter of the next beat. The skin electrostimulator is controlled by the microcontroller to perform electrical stimulation of the current reproductive nerve based on the output parameters of the next beat.

2. The closed-loop reproductive nerve stimulation system based on multimodal biofeedback as described in claim 1, characterized in that, Using the effect index sample reflecting reproductive nerve function as the outcome indicator, the prior relative effect and prior precision of each stimulus parameter combination in the controlled trial data were analyzed, including: The aforementioned outcome index is used as a prior relative effect; The effective-false variance of each type of physiological signal is calculated using the standard deviation and sample size of each physiological signal in the second multimodal biofeedback data and the standard deviation and sample size of the corresponding type of physiological signal in the first multimodal biofeedback data. The specific formula is as follows: in, Indicates effective - pseudo variance. This represents the standard deviation of the i-th type of physiological signal in the second multimodal biofeedback data. This represents the sample size of the i-th type of physiological signal in the second multimodal biofeedback data. The 1 in the equation represents the sequence number of the first prior precision calculation relative to the 0 subscript of the sham stimulus group. This represents the standard deviation of the physiological signal of the i-th type in the first multimodal biofeedback data. This represents the sample size of the i-th type of physiological signal in the first multimodal biofeedback data. and The subscript 0 in the figure represents the sequence number of the 0th baseline data in the sham stimulation group; By using weighting coefficients to sum the effective-false variances of various types of physiological signals, the effective-false variances corresponding to each combination of stimulus parameters are obtained. The inverse of the effective-false variance corresponding to each stimulus parameter combination is taken as the prior accuracy.

3. The closed-loop reproductive nerve stimulation system based on multimodal biofeedback as described in claim 2, characterized in that, The determination of the effect index sample includes: The difference between each physiological signal in the second multimodal biofeedback data and the corresponding physiological signal in the first multimodal biofeedback data is calculated to obtain the effective-false difference value of each type of physiological signal. By using weighting coefficients to sum the effective-false differences of various types of physiological signals, the effect index sample corresponding to each combination of stimulus parameters is obtained.

4. The closed-loop reproductive nerve stimulation system based on multimodal biofeedback as described in claim 3, characterized in that, The determination of weighting coefficients includes: The effect index values ​​corresponding to physiological signals with different numerical combinations were analyzed using the expert scoring method. Establish a regression model between physiological signals with different numerical combinations and their corresponding effect index values; The weight coefficients in the regression model are solved using the least squares method.

5. The closed-loop reproductive nerve stimulation system based on multimodal biofeedback as described in claim 1, characterized in that, Collect controlled trial data on reproductive nerve samples under various stimulation parameters, including: The study collected first-modal biofeedback data on sham stimulation of reproductive nerve samples under various combinations of stimulation parameters, and second-modal biofeedback data on effective stimulation of reproductive nerve samples. The control data were determined by combining the first multimodal biofeedback data with the second multimodal biofeedback data.

6. The closed-loop reproductive nerve stimulation system based on multimodal biofeedback as described in claim 1, characterized in that, The prior matrix is ​​formed, including: A prior matrix is ​​constructed using the prior relative effect and prior precision corresponding to each combination of stimulus parameters as row data.

7. The closed-loop reproductive nerve stimulation system based on multimodal biofeedback as described in claim 1, characterized in that, Collect multimodal biofeedback data reflecting the reproductive neural function, including: Collect initial multimodal biological feedback data at each moment continuously within a fixed time period; The mean of the initial multimodal biofeedback data at each consecutive time point is used as the multimodal biofeedback data over a continuous fixed time period.

8. The closed-loop reproductive nerve stimulation system based on multimodal biofeedback as described in claim 1, characterized in that, Based on the aforementioned multimodal biofeedback data, the current effect index of the reproductive nervous system is calculated, including: The current effect index is obtained by weighting and summing the multimodal biological feedback data of different types within the same continuous fixed time period using weighting coefficients.

9. The closed-loop reproductive nerve stimulation system based on multimodal biofeedback as described in claim 1, characterized in that, By performing a Bayesian update on the prior matrix using the current effect index, the posterior relative effect and posterior accuracy are obtained, including: Based on the current effect index and the prior matrix, the posterior relative effect is calculated using the following formula: in, This represents the posterior relative effect. This represents the relative prior effects in the prior matrix. This indicates the prior precision in the prior matrix. This indicates the real-time signal accuracy of the current effect index. Indicates the current effect index; Based on the real-time signal accuracy corresponding to the current effect index and the prior accuracy in the prior matrix, the posterior accuracy is calculated using the following formula: in, Indicates posterior precision. Indicates prior precision. Indicates the real-time signal accuracy.

10. The closed-loop reproductive nerve stimulation system based on multimodal biofeedback as described in claim 1, characterized in that, Based on the posterior relative effect and the posterior precision, the SUCRA probability corresponding to each stimulus parameter is calculated using the network meta-ranking method, including: The network meta-sorting method includes: The effect value is randomly drawn from the effect interval determined by the posterior relative effect and the posterior precision; Compare the effect values ​​corresponding to different combinations of stimulus parameters to select the maximum effect value in each round; The number of times the maximum effect value occurs for each combination of stimulus parameters is counted, and the SUCRA probability corresponding to each combination of stimulus parameters is calculated based on the total number of random sampling rounds and the number of times the maximum effect value occurs for each combination of stimulus parameters.

Images