Closed-loop stimulation control method and system

Abstract

The invention relates to a closed-loop stimulation control method and system. The method comprises the following steps: acquiring an initial stimulation strategy corresponding to a user; acquiring a feature change matrix formed by the latest N-time continuous blood volume wave images and the corresponding image features thereof; updating a stimulation strategy based on the feature change matrix; according to the method, quantitative stimulation control description is carried out from multiple image feature dimensions, subsequent closed-loop automatic adjustment is supported, a stimulation strategy dynamically adapting to the current condition of a user is dynamically adapted, fine-grained adjustment is carried out, and therefore the stimulation control efficiency is greatly improved.

Description

Technical Field

[0001] The present invention belongs to the technical field of closed-loop stimulation control, and particularly relates to a closed-loop stimulation control method and system.

Background Art

[0002] Blood pressure is one of the vital signs and an important reference index for clinical decision-making and intervention. During the perioperative period, methods such as manual drug administration, intermittent or continuous infusion of vasoactive drugs with a microinfusion pump are often used for blood pressure management. Traditional blood pressure management methods mainly rely on manual drug administration or intermittent infusion of vasoactive drugs, and these methods have significant limitations. For example, when titrating drugs manually, about 42% of blood pressure values may be below the recommended target range. In addition, the "overshoot" phenomenon of the target blood pressure (i.e., the blood pressure fluctuation exceeds the target value by more than 20 mmHg) and the significant fluctuations during blood pressure management may lead to organ damage and dysfunction. The method of closed-loop stimulation for blood pressure control is a new medical technology. To overcome the deficiencies of traditional methods, a closed-loop feedback control system has emerged. This system monitors blood pressure in real time and automatically adjusts drug infusion or stimulation parameters according to a preset target range, so as to quickly respond and adjust blood pressure in real time. The closed-loop system shows significant advantages in perioperative blood pressure management and can control blood pressure within the target range to the greatest extent. With the progress of wearable devices and continuous monitoring technologies, blood pressure monitoring is developing towards wearable, continuous measurement, and multi-parameter intelligent analysis. These technologies provide a more accurate and real-time monitoring basis for closed-loop blood pressure control. In some new closed-loop systems, piezoelectric thin film sensors and photothermal nanomaterials are combined to accurately locate and ablate renal sympathetic nerves, thereby achieving the treatment of refractory hypertension. This system dynamically monitors blood pressure signals and locates the target, guiding the precise ablation of renal sympathetic nerves, and has the characteristics of safety, effectiveness, minimally invasive and non-invasive.

[0003] Currently, on the one hand, blood pressure monitoring and blood pressure management are set separately. Blood pressure monitoring devices do not have the function of blood pressure management, and blood pressure management methods cannot obtain real-time blood pressure monitoring results, and of course cannot adjust blood pressure intervention strategies in real time according to the monitoring results. Combining blood pressure monitoring and blood pressure management to form a blood pressure closed-loop management is beneficial to improving the real-time performance and effectiveness of blood pressure control. On the other hand, the stimulation parameters in the prior art need to be set manually. Once the stimulation parameters are fixed, the stimulation electrodes will apply stimulation in a fixed stimulation manner, and the stimulation parameters cannot change in real time and adaptively with the changes of the user's dynamic blood pressure, heart rate, respiratory rate, etc. The degree of human dependence on the stimulation strategy is very high, and it cannot be adjusted in real time according to the user's dynamic situation; the dynamic adaptability of the stimulation strategy is not carried out dynamically for the real-time state of each user.

[0004] Based on the above problems, the present invention converts real-time monitoring data into image features of blood volume wave images, so as to be able to quantitatively describe stimulus control from multiple image feature dimensions, support subsequent closed-loop automatic regulation, dynamically adapt to the current situation of the user's stimulus strategy, and make fine-grained adjustments, thereby greatly improving the efficiency of stimulus control.

Summary of the Invention

[0005] To solve the above problems in the prior art, the present invention proposes a closed-loop stimulus control method and system, and the method includes:

[0006] Step S1: Obtain an initial stimulus strategy corresponding to the user; where: the stimulus strategy st u includes three stimulus parameters: stimulus pulse pattern md v , stimulus frequency fr, and stimulus intensity ac; represent the u-th stimulus strategy as st u =f(md v , ac, fr)=ac×md v (fr);

[0007] Step S2: Obtain a feature change matrix composed of the most recent N consecutive blood volume wave images and their corresponding image features; specifically: for the n-th blood volume wave image among the most recent N consecutive blood volume wave images, extract K image features from it, and arrange the K image features in order to form the n-th row in the feature change matrix MT=[mt n，k ; the size of the feature change matrix is N×K; where: n∈1~N;

[0008] Step S3: Update the stimulus strategy based on the feature change matrix; specifically: normalize the feature change matrix; update the stimulus strategy based on the normalized feature change matrix; apply the stimulus based on the updated stimulus strategy.

[0009] Further, the obtaining of the initial stimulus strategy is to select a stimulus strategy suitable for the current user from the stimulus strategy set {st u}, u = 1~U, as the initial stimulus strategy; U is the number of stimulus strategies in the stimulus strategy set.

[0010] Further, the stimulus intensity is related to the stimulus voltage and / or stimulus current.

[0011] Further, the updating of the stimulus strategy based on the normalized feature change matrix is specifically: obtain the key image features of the normalized feature change matrix; determine whether it is necessary to adjust the stimulus pattern, if so, keep the current stimulus pulse pattern unchanged, adjust the stimulus frequency fr and / or stimulus intensity ac to update the stimulus strategy; otherwise, adjust the stimulus pattern to update the stimulus strategy.

[0012] Further, updating the stimulation strategy based on the normalized feature change matrix specifically includes the following steps:

[0013] Step S3A1: Obtain the key image features of the normalized feature change matrix; specifically: calculate the entropy of each image feature k in the feature change matrix based on the following formula (1)-(3) or (2)(3); determine kk∈1~K based on the entropy, such that kk satisfies w kk = max(w k ); Take the image feature kk as the key image feature;

[0014]

[0015] Step S3A2: Keep the current stimulation pulse pattern unchanged, and adjust the stimulation frequency fr and / or the stimulation intensity ac to update the stimulation strategy; specifically: keep the current stimulation pulse pattern unchanged, determine which stimulation parameter of the key image feature is more sensitive to the stimulation frequency fr and the stimulation intensity ac, and adjust the more sensitive stimulation parameter to update the stimulation strategy; when determining the adjustment unit stimulation frequency or stimulation intensity, the one with a larger change amplitude of the key image feature is the more sensitive stimulation parameter; when the larger one is the stimulation frequency, update the stimulation mode to st u = f(md v , ac, fr’), where: fr’ is the adjusted stimulation frequency; when the larger one is the stimulation intensity, update the stimulation mode to st u = f(md v , ac’, fr), where: ac’ is the adjusted stimulation intensity.

[0016] Further, the closed-loop stimulation is for blood pressure.

[0017] A closed-loop stimulation control system for implementing the above-mentioned closed-loop stimulation control method.

[0018] A closed-loop stimulation control device for implementing the above-mentioned closed-loop stimulation control method.

[0019] A closed-loop stimulation control terminal for implementing the above-mentioned closed-loop stimulation control method.

[0020] A closed-loop stimulation control circuit, characterized in that the closed-loop stimulation control module is used to implement the above-mentioned closed-loop stimulation control method.

[0021] The beneficial effects of the present invention include:

[0022] (1) Convert the real-time monitoring data into image features of blood volume wave images, so as to enable a quantitative description of stimulation control from multiple image feature dimensions. Represent it with a feature change matrix substituting the continuous stimulation results, and dynamically adjust the stimulation strategy by comprehensively considering the user's recent continuous stimulation performance. Thus, it supports the subsequent closed-loop control automatic adjustment customized by the user, a stimulation strategy that dynamically adapts to the user's current situation, and makes fine-grained adjustments, thereby greatly improving the stimulation control efficiency;

[0023] (2) Represent the stimulation strategy functionally, so that the adjustment of the stimulation strategy is combined with the image features. Find the important information in the local range by discovering the key image features, and realize the adjustment of the stimulation strategy with different levels of coarse and fine granularity, achieving flexible customization of the stimulation strategy. Further, through the entropy quantity and transformation quantity, discover the consistency of the user's performance of the stimulation strategy. When the performance is consistent, it indicates the adjustment direction of local optimization, and when the performance is found to be inconsistent, it can perform global adjustment of the stimulation strategy.

BRIEF DESCRIPTION OF THE DRAWINGS

[0024] The drawings described herein are used to provide a further understanding of the present invention, form a part of this application, but do not constitute an improper limitation of the present invention. In the drawings:

[0025] Figure 1 It is a schematic diagram of the closed-loop stimulation control method provided by the present invention.

DETAILED DESCRIPTION OF THE EMBODIMENTS

[0026] The present invention will be described in detail below in conjunction with the drawings and specific embodiments, and the illustrative embodiments and descriptions are only used to explain the present invention, but do not limit the present invention.

[0027] The present invention provides a closed-loop stimulation control method and system. As shown in the attached Figure 1 figure, the method includes the following steps:

[0028] Step S1: Obtain the initial stimulation strategy corresponding to the user; where: the stimulation strategy st u includes three stimulation parameters: the stimulation pulse pattern md v , the stimulation frequency fr, and the stimulation intensity ac; obtaining the initial stimulation strategy is to select a stimulation strategy suitable for the current user from all stimulation strategy sets {st u}, u = 1 to U, as the initial stimulation strategy; for example: select the single-pulse mode, and use 77.5 Hz, 15 mA transcranial alternating current stimulation (tACS) for the forehead and mastoid regions as the initial stimulation strategy; U is the number of stimulation strategies in the stimulation strategy set;

[0029] Preferably: Represent the u-th stimulation strategy as st u= f(md v , ac, fr) = ac × md v (fr); where: the stimulation intensity is the main adjustment factor, and fr is the auxiliary adjustment factor; in the case of the v-th stimulation mode md v being determined, the adjustable values of the stimulation frequency are relatively determined and in small quantity, while ac can be adjusted more continuously within a larger range, and of course this adjustment can be continuous or discontinuous; it can be understood that for the stimulation strategy of the user starting from the initial stimulation strategy and through closed-loop control and dynamic adjustment, from the start of applying stimulation to the user, the blood volume wave image corresponding to the monitoring data changes continuously and moves towards the target blood volume wave image; and the quantitative representation of the image features in the blood volume wave image can more sensitively detect the changes or effects of stimulation control;

[0030] Preferably: the stimulation intensity is related to the stimulation voltage or stimulation current; the stimulation voltage and stimulation current are high voltage or current; the stimulation intensity is the amplification or reduction degree of the stimulation mode, or defines the magnitude of the pulsed current in the stimulation mode;

[0031] Preferably: one or more stimulation pulses are organized in a certain manner to form a stimulation cycle ω corresponding to the stimulation mode in this organization manner. Each stimulation mode contains one or more stimulation pulses; for example: single pulse (width 100 μS, current 5 mA); the stimulation strategy is represented as st u = f(md v , ac, fr) = ac × md v (fr); for each stimulation mode (for example: for the above stimulation mode (width 100 μS, current 5 mA), the stimulation frequency is 1 Hz, and the stimulation intensity is 2; when the stimulation intensity is 2, the current in the single pulse mode is amplified to 10 mA, and the pulse width may / may not increase with the increase of the stimulation intensity), under the stimulation mode limitation conditions, by adjusting the stimulation frequency fr and / or the stimulation intensity ac, the change direction of z image features can be made consistent with the adjustment direction of this stimulation frequency fr and / or stimulation intensity ac; the set of these z image features is called the single sensitive set of the stimulation frequency fr and / or the stimulation intensity ac in this stimulation mode; where: z = 1 to K; the limitation conditions define the effective conditions of the stimulation mode in the stimulation strategy; the stimulation conditions are the limitation conditions related to user attributes, the stimulation frequency fr and / or the stimulation intensity ac; for example: defining the user attribute as female, when the stimulation frequency is within the range of 70 - 72 Hz, the stimulation mode md vEffective; at this time, the adjustment of the stimulation intensity ac will make the change directions of the z image features consistent with the adjustment direction of the stimulation intensity ac; that is to say, this stimulation condition defines the limiting conditions of the stimulation mode; the change directions of the image features are towards or in the opposite direction of the change directions of the target image features; and the adjustment directions of the stimulation frequency fr and / or the stimulation intensity ac are to increase or decrease;

[0032] Preferably: the consistency is positive correlation or negative correlation; under the limiting conditions for a stimulation mode, the adjustment of the stimulation frequency or the stimulation intensity makes the change directions of the z image features towards the image features corresponding to a better blood volume wave image; the change directions include getting larger, getting smaller or remaining unchanged;

[0033] Obtain the initial stimulation strategy corresponding to the user; specifically: perform 24H ambulatory blood pressure monitoring on the user, and set the initial stimulation strategy corresponding to the user based on the 24H ambulatory blood pressure monitoring results;

[0034] Preferably: perform ambulatory blood pressure monitoring every 0.5 - 1H; after obtaining new monitoring data each time, execute step S2 and step S3 to determine whether the stimulation strategy needs to be updated;

[0035] Step S2: Obtain the feature change matrix composed of the most recent N consecutive blood volume wave images and their corresponding image features; specifically: for the nth blood volume wave image among the most recent N consecutive blood volume wave images, extract K image features from it, and arrange the K image features in sequence to form the feature change matrix MT =

[0036] [mt n ,k] as the nth row; obviously, the size of the feature change matrix is N×K; where: n ∈ 1 - N;

[0037] Preferably: N is greater than or equal to 2, for example: 12; the K image features are some typical image features among all the image features; these typical image features show consistent or stage - by - stage consistent changes as the blood volume wave image changes towards the target blood volume wave image; based on these stage - by - stage consistent changes, the limiting conditions can be set so that under the limitation of the limiting conditions, the changes of the image features are consistent changes under the stimulation mode;

[0038] Preferably, the image features of the blood volume wave image include waveform features, spectral features, time-domain features, statistical features, non-linear features, morphological features, physiological features, spatial features, etc.; among them: the waveform features include the amplitude reflecting the degree of change in blood volume in the blood vessel, the period representing the time interval of the blood volume wave, the rise time indicating the time required for the waveform to rise from the lowest point to the highest point, the fall time indicating the time required for the waveform to return from the highest point to the lowest point, indicating single peaks, double peaks, etc., and the waveform shape reflecting blood vessel elasticity and hemodynamic status, etc.; the spectral features include the frequency components obtained by analyzing the frequency distribution in the waveform through Fourier transform, the main frequency indicating the frequency component with the highest energy in the spectrum, and the harmonics reflecting the non-linear characteristics of the blood vessel; the time-domain features include: the mean indicating the average value of the waveform in time, the variance indicating the degree of fluctuation of the waveform change, and the peak value indicating the maximum and minimum values of the waveform; the statistical features include the skewness indicating the symmetry of the waveform distribution and the kurtosis indicating the sharpness of the waveform distribution; the non-linear features include the fractal dimension describing the complexity and self-similarity of the waveform, etc.

[0039] Step S3: Update the stimulation strategy based on the feature change matrix; specifically: normalize the feature change matrix; update the stimulation strategy based on the normalized feature change matrix; apply stimulation based on the updated stimulation strategy.

[0040] The normalization of the feature change matrix is specifically: perform relative normalization on the feature change matrix to obtain the normalized feature change matrix MTU = [mtu n ,k]ω Obtain the target blood volume wave image ω corresponding to the user attribute Extract K image features corresponding to the target blood volume wave image to form the target K-element vector VD = (vdk), k = 1~Kω Set MTU = [mtu n,k = [|mt n,k -vd k | / span k , where: span k is the numerical range change area of image feature k

[0041] interval; Since the target and its blood volume wave image that each user ultimately needs to achieve are different, only through relative normalization can the effectiveness of the stimulation strategy update be ensured, and at the same time, it can also support phased stimulation plans.

[0042] The obtaining of the target blood volume wave image corresponding to the user attribute; specifically: obtain the stimulation plan or phased stimulation plan of the user, and obtain the target blood volume wave image corresponding to the user stimulation plan.

[0043] Updating the stimulation strategy based on the normalized feature change matrix specifically includes: obtaining the key image features of the normalized feature change matrix; determining whether it is necessary to adjust the stimulation pattern. If so, keep the current stimulation pulse pattern unchanged and adjust the stimulation frequency fr and / or the stimulation intensity ac to update the stimulation strategy; otherwise, adjust the stimulation pattern to update the stimulation strategy.

[0044] Preferably: Updating the stimulation strategy based on the normalized feature change matrix specifically includes the following steps:

[0045] Step S3A1: Obtaining the key image features of the normalized feature change matrix; specifically: calculating the entropy of each image feature k in the feature change matrix based on the following formula (1)-(3) or (2)(3); determining kk∈1~K based on the entropy such that kk satisfies w kk = max(w k ); taking the image feature kk as the key image feature.

[0046]

[0047] Step S3A2: Keeping the current stimulation pulse pattern unchanged and adjusting the stimulation frequency fr and / or the stimulation intensity ac to update the stimulation strategy; specifically: keeping the current stimulation pulse pattern unchanged, determining which stimulation parameter among the stimulation frequency fr and the stimulation intensity ac the key image feature is more sensitive to, and adjusting the more sensitive stimulation parameter to update the stimulation strategy; when determining the unit adjustment of the stimulation frequency or the stimulation intensity, the one with a larger change amplitude of the key image feature is the more sensitive stimulation parameter; when the larger one is the stimulation frequency, updating the stimulation pattern to st u = f(md v , ac, fr'), where: fr' is the adjusted stimulation frequency; when the larger one is the stimulation intensity, updating the stimulation pattern to st u = f(md v , ac', fr), where: ac' is the adjusted stimulation intensity; that is to say, in general, the stimulation pattern can remain unchanged.

[0048] Preferably: In the above steps, when the key image feature only falls into the single sensitive set of the stimulation frequency under this stimulation pattern, updating the stimulation pattern to st u = f(md v , ac, fr'), where: fr' is to increase (or decrease) the stimulation frequency; and when the key image feature only falls into the single sensitive set of the stimulation intensity under this stimulation pattern, updating the stimulation pattern to st u = f(md v, ac', fr), where: ac' is to increase (or decrease) the stimulation intensity; when simultaneously falling into a single sensitive set of the stimulation frequency and the stimulation intensity in the stimulation mode, it is only determined which stimulation parameter of the stimulation frequency fr and the stimulation intensity ac the key image feature is more sensitive to, and subsequent operations are performed;

[0049] Preferably: The adjustment is in the way of increasing or decreasing, for example: increasing or decreasing the unit stimulation frequency or stimulation intensity;

[0050] Preferably: Determine whether it is necessary to adjust the stimulation mode before updating the stimulation strategy; specifically: Determine the consistency of each image feature change in the feature change matrix; when the consistency is low, update the stimulation mode; otherwise, keep the stimulation mode unchanged; when updating the stimulation mode, select a candidate applicable stimulation mode that has not been used this time based on user attributes for updating; of course, when updating the stimulation mode, set the stimulation frequency and stimulation intensity matching the stimulation mode correspondingly, so as to form an updated stimulation strategy; after updating the stimulation strategy, return to step S2 to wait for new monitoring data, and repeat steps S2 - S3 to achieve closed-loop control; The steps specifically include the following steps:

[0051] Step S3B1: Sequentially obtain each element of the nth row of the normalized feature change matrix; calculate the change entropy of each element in this row of elements; specifically: Calculate the change entropy cgs of the element mtu n，k of n，k ;

[0052] cgs n，k = mtu n，k × w k (4);

[0053] Step S3B2: Obtain the sorting value sq_cgs of the cgs n，k numerical value within the change entropy cgs of this row of elements n，1 ~ cgs n，K ; Set the value av of the kth element in the consistency vector n，k = sq_cgs n,k ; Construct the consistency vector AV n，k =(av n ) corresponding to this nth row of elements; The sorting value only needs to be set in a unified manner, for example: the largest is 1, the second largest is 2, and so on; n，k )

[0054] Step S3B3: Calculate the degree of change between N consistency vectors; specifically: Calculate the consistency vectors (AV n and AV n+1) the distance between; the average value of the distances calculated based on N consistency vectors is used as the degree of change;

[0055] Preferably: calculating the distance between the consistency vectors corresponding to each two adjacent rows specifically includes: calculating the Mahalanobis or Euclidean distance as the distance between the consistency vectors corresponding to each two adjacent rows;

[0056] Alternatively: the step S3B3 specifically includes: calculating the degree of change between every two of the N consistency vectors; specifically: calculating the distance between every two consistency vectors; the average value of all the distances is used as the degree of change; the distance is the Euclidean distance;

[0057] Step S3B4: When the degree of change is greater than the change degree threshold, adjust the stimulation pattern; otherwise, keep the current stimulation pattern;

[0058] Preferably: the change degree threshold is a preset value; by adjusting the change degree threshold, the sensitivity of the stimulation strategy adjustment can be adjusted. When the change degree threshold is large, the sensitivity is low but the occurrence of strategy adjustment oscillation can be avoided; on the contrary, when the change degree threshold is large, the sensitivity is high but the occurrence of stimulation strategy adjustment oscillation may be brought about,

[0059] Alternatively: updating the stimulation strategy based on the normalized feature change matrix specifically includes: looking up the feature strategy correspondence table based on the normalized feature change matrix to obtain the updated stimulation strategy corresponding to the lookup result; specifically: the feature strategy correspondence table stores the correspondence between the feature change matrix and the stimulation strategy; the stimulation strategy corresponding to the record most similar to the normalized feature change matrix in the correspondence table is used as the lookup result;

[0060] Obtaining the updated stimulation strategy corresponding to the lookup result specifically includes: presetting the correspondence ω between the feature change matrix and the preset stimulation strategy. The preset stimulation strategy is set according to experience ω

[0061] Alternatively: updating the stimulation strategy based on the normalized feature change matrix specifically includes: inputting the normalized feature change matrix into a pre-trained intelligent strategy model to obtain a stimulation strategy adapted to the feature change matrix; the intelligent strategy model is an artificial intelligence model and is trained through historical data; when using the intelligent strategy model to obtain the stimulation strategy, a large amount of historical data support is required;

[0062] Preferably: the artificial intelligence model is a neural network model ω

[0063] Preferably: the neural network model is a convolutional neural network model;

[0064] Preferably, the method further includes step S4: If a stimulation strategy adjustment occurs, obtain the blood volume wave image after the stimulation strategy adjustment, and determine whether the adjusted blood volume wave image is better than the blood volume wave image before the adjustment. If so, maintain the adjusted stimulation strategy; otherwise, perform a stimulation strategy fallback and return to the stimulation strategy before the adjustment.

[0065] The returning to the stimulation strategy before the adjustment is specifically: Select one from the stimulation strategies corresponding to the most recent N monitoring cycles as the fallback stimulation strategy.

[0066] Based on the same inventive concept, the present invention also provides a closed-loop stimulation control system, which is used to complete the above-mentioned closed-loop stimulation control method.

[0067] A computer program (also known as a program, software, software application, script, or code) can be written in any form of programming language, including assembly or interpreted languages, declarative or procedural languages, and can be deployed in any form, including as a stand-alone program or as a module, component, subroutine, object, or other unit suitable for use in a computing environment. A computer program may or may not correspond to a file in a file system. The program can be stored as part of a file that holds other programs or data (such as one or more scripts in a markup language document), in a single file dedicated to the program, or in multiple cooperating files (such as files that store one or more modules, subroutines, or code portions). A computer program can be deployed to execute on one computer or on multiple computers located at one site or distributed across multiple sites and interconnected by a communication network.

[0068] Those skilled in the art should understand that the embodiments of the present invention can be provided as a method, a system, or a computer program product. Therefore, the present invention can take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware aspects. Moreover, the present invention can take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0069] The present invention is described with reference to the flowcharts and / or block diagrams of methods, apparatuses (systems), and computer program products according to the embodiments of the present application. It should be understood that each flow and / or block in the flowchart and / or block diagram, and the combination of flows and / or blocks in the flowchart and / or block diagram, can be realized by computer program instructions. These computer program instructions can be provided to the processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing devices to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing devices generate for realizing the process Figure 1one process or multiple processes and / or blocks Figure 1 means for the functions specified in one block or multiple blocks.

[0070] These computer program instructions can also be stored in a computer-readable memory capable of guiding a computer or other programmable data processing device to work in a specific manner, so that the instructions stored in the computer-readable memory produce a manufactured article including an instruction device, and the instruction device implements the processes Figure 1 one process or multiple processes and / or blocks Figure 1 the functions specified in one block or multiple blocks.

[0071] These computer program instructions can also be loaded onto a computer or other programmable data processing device, so that a series of operation steps are executed on the computer or other programmable device to generate a computer-implemented process. Therefore, the instructions executed on the computer or other programmable device provide steps for implementing the functions specified in Figure 1 one process or multiple processes and / or blocks Figure 1 one block or multiple blocks.

[0072] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit them. Although the present invention has been described in detail with reference to the above embodiments, those of ordinary skill in the art should understand that: modifications or equivalent replacements can still be made to the specific implementation manners of the present invention, and any modification or equivalent replacement that does not depart from the spirit and scope of the present invention should be covered by the protection scope of the claims of the present invention.

Claims

1. A closed-loop stimulation control method, characterized in that: The method comprises: Step S1: Obtain the initial stimulation strategy corresponding to the user; wherein: the stimulation strategy st u Including stimulation pulse mode md v , stimulation frequency fr and stimulation intensity ac; the u-th stimulation strategy is represented as st u =f(md v , ac, fr) = ac × md v (fr); Step S2: Obtain the feature change matrix composed of the most recent N consecutive blood volume wave images and their corresponding image features; specifically: for the nth blood volume wave image among the most recent N consecutive blood volume wave images, extract K image features from it, and arrange the K image features in order to form a feature change matrix MT=[mt n ,k]; the feature change matrix size is N×K; where: n∈1~N; Step S3: updating the stimulation strategy based on the feature change matrix; specifically: normalizing the feature change matrix; updating the stimulation strategy based on the normalized feature change matrix; applying stimulation based on the updated stimulation strategy.

2. The closed-loop stimulation control method according to claim 1, characterized in that: The initial stimulation strategy is obtained by selecting the stimulation strategy set {st u }, u=1~U, select a stimulation strategy suitable for the current user as the initial stimulation strategy; U is the number of stimulation strategies in the stimulation strategy set.

3. The closed-loop stimulation control method according to claim 2, characterized in that: The stimulation intensity is related to the stimulation voltage and / or the stimulation current.

4. The closed-loop stimulation control method according to claim 3, characterized in that: The method of updating the stimulation strategy based on the normalized feature change matrix is ​​specifically as follows: obtaining key image features of the normalized feature change matrix; determining whether the stimulation mode needs to be adjusted, and if so, keeping the current stimulation pulse mode unchanged, adjusting the stimulation frequency fr and / or the stimulation intensity ac to update the stimulation strategy; otherwise, adjusting the stimulation mode to update the stimulation strategy.

5. The closed-loop stimulation control method according to claim 3, characterized in that: The updating of the stimulation strategy based on the normalized feature change matrix specifically includes the following steps: Step S3A1: Obtain key image features of the normalized feature change matrix; specifically: calculate the entropy of each image feature k in the feature change matrix based on the following formulas (1)-(3) or (2)(3); determine kk∈1~K based on the entropy, so that kk satisfies w kk =max(w k ); taking the image feature kk as the key image feature; Step S3A2: keep the current stimulation pulse mode unchanged, adjust the stimulation frequency fr and / or stimulation intensity ac to update the stimulation strategy; specifically: keep the current stimulation pulse mode unchanged, determine which stimulation parameter of the stimulation frequency fr and the stimulation intensity ac the key image feature is more sensitive to, and adjust the more sensitive stimulation parameter to update the stimulation strategy; determine that when adjusting the unit stimulation frequency or stimulation intensity, the key image feature with a larger change amplitude is the more sensitive stimulation parameter; when the larger one is the stimulation frequency, update the stimulation mode to st u =f(md v , ac, fr'), where: fr' is the adjusted stimulation frequency; when the larger one is the stimulation intensity, the updated stimulation mode is st u =f(md v , ac', fr), where: ac' is the adjusted stimulus intensity.

6. The closed-loop stimulation control method according to claim 5, characterized in that: The closed-loop stimulation is stimulation targeting blood pressure.

7. A closed-loop stimulation control system, characterized in that: The closed-loop stimulation control system is used to implement the closed-loop stimulation control method described in any one of claims 1 to 6.

8. A closed-loop stimulation control device, characterized in that: The closed-loop stimulation control device is used to implement the closed-loop stimulation control method described in any one of claims 1 to 6.

9. A closed-loop stimulus control terminal, characterized in that: The closed-loop stimulation control terminal is used to implement the closed-loop stimulation control method described in any one of claims 1 to 6.

10. A closed-loop stimulation control circuit, characterized in that: The closed-loop stimulation control module is used to implement the closed-loop stimulation control method described in any one of claims 1 to 6.

Images