Time domain interference electrical stimulation closed-loop adjustment method and system for tremor rehabilitation

Abstract

The invention discloses a time domain interference electrical stimulation closed-loop adjustment method and system for tremor rehabilitation, and belongs to the technical field of biomedical engineering. According to the method, the electromyographic signals of the patient are monitored in real time, the neuromuscular activity characteristics related to tremor are extracted, and the severity of tremor is evaluated based on the characteristics. According to the method, a closed-loop feedback mechanism is utilized, parameters of time domain interference electrical stimulation are dynamically adjusted, the parameters comprise the position of a stimulation electrode, current intensity, carrier frequency and difference frequency, and an optimal stimulation scheme for minimizing the tremor degree is found by gradually optimizing target spot selection and parameter combination. In addition, the invention provides a complete closed-loop regulation and control system, full-process automation can be realized, human intervention is reduced, and the treatment stability is improved.

Description

Technical Field

[0001] The present invention belongs to the technical field of biomedical engineering, and particularly relates to a closed-loop regulation method and system for time-domain interference electrical stimulation for tremor rehabilitation. Background Art

[0002] Motor tremor is an involuntary, rhythmic shaking that occurs during limb movement. It is common in neurological diseases such as Parkinson's disease or cerebellar lesions, and may also be caused by factors such as drugs and metabolic abnormalities. This tremor usually worsens during movement execution, especially when approaching the target, severely affecting the patient's fine motor ability and quality of life. Currently, the treatment methods for motor tremor are limited. Although drugs and surgery have certain effects, they have limitations. Therefore, developing new treatment options has important clinical significance. Currently, deep brain stimulation (DBS) is one of the effective methods for treating motor tremor symptoms. By stimulating targets such as the subthalamic nucleus (STN) or the globus pallidus interna (GPi), the patient's motor function can be significantly improved. However, as an invasive surgery, DBS has certain disadvantages, such as surgical risks, device infections, electrode displacements, and complications that may be caused by long-term implantation. Therefore, its clinical use is limited.

[0003] Recently, temporal interference stimulation (TI) as an emerging non-invasive neuromodulation technology has gradually attracted attention. Temporal interference stimulation applies sinusoidal currents with slightly different frequencies (in the kilohertz range) outside the skull to generate low-frequency envelope waveforms (from a few hertz to dozens of hertz) in the deep brain, thereby achieving deep stimulation of specific brain regions. Temporal interference stimulation utilizes the interference effect of frequency-difference currents to focus the electric field in the deep brain regions, avoiding the disadvantages of traditional transcranial electrical stimulation techniques, such as shallow regulation depth and difficulty in accurately stimulating targets. Preliminary clinical studies have shown that temporal interference stimulation can effectively stimulate deep brain region targets and has shown good application prospects in the treatment of motor tremor. However, currently, the rehabilitation interventions for motor tremor using temporal interference stimulation all adopt fixed parameter schemes (stimulation targets and stimulation parameters). Therefore, how to efficiently and accurately adjust individual stimulation targets and parameters based on individual differences to achieve individual precise intervention remains an urgent problem to be solved. Summary of the Invention

[0004] In view of this, the object of the present invention is to provide a time-domain interference electrical stimulation closed-loop regulation method and system for tremor rehabilitation. Taking the individual tremor degree of the patient as a feedback parameter, by real-time monitoring the tremor signal, the stimulation parameters (electrical stimulation target, electrode position, current intensity and frequency) are dynamically adjusted, so as to achieve precise individual neurorehabilitation regulation. The system aims to overcome the deficiencies of the prior art and provide a closed-loop personalized, efficient and safe non-invasive rehabilitation regulation scheme for the treatment of patients with motor tremors.

[0005] The specific technical solutions adopted by the present invention are as follows:

[0006] In the first aspect, the present invention provides a time-domain interference electrical stimulation closed-loop regulation method for tremor rehabilitation, which includes:

[0007] S1. Select candidate targets in the target selection area of the brain;

[0008] S2. For each currently selected candidate target, use the personalized head model constructed for the stimulation object to simulate different stimulation electrode positions and stimulation current intensities. With the maximization of the peak electric field and focusing degree at the candidate target as the optimization goal, determine the optimal stimulation electrode position and the optimal stimulation current intensity of each stimulation electrode through an optimization algorithm;

[0009] S3: For each currently selected candidate target, after arranging the stimulation electrodes according to the optimal stimulation electrode position, sample the carrier frequency and difference frequency of the stimulation current of each stimulation electrode to form a series of candidate stimulation frequency parameter combinations. When the stimulation electrodes are fixed to output the optimal stimulation current intensity, continuously adjust the candidate stimulation frequency parameter combinations, sequentially generate the stimulation current control signals corresponding to different candidate stimulation frequency parameter combinations and output them to the stimulation electrodes, and simultaneously collect the myoelectric signals of the stimulation object after being stimulated by the stimulation current, extract the tremor frequency characteristics and tremor intensity characteristics therefrom and calculate the tremor degree score, and select the candidate stimulation frequency parameter combination with the smallest tremor degree score as the optimal stimulation frequency parameter combination;

[0010] S4. Execute S2 and S3 respectively for the candidate targets in the target selection area. With the minimization of the tremor degree score as the goal, search for the optimal target, and save the optimal stimulation electrode position, the optimal stimulation current intensity and the optimal stimulation frequency parameter combination corresponding to the optimal target.

[0011] As a preference of the above first aspect, for the candidate targets in the target selection area, taking the STN-GPi loop as the target selection area, search for the optimal target through the dichotomy method. The process of the dichotomy method is as follows:

[0012] First, take the STN area and the GPi area as two candidate targets, and respectively execute the above S2 and S3 to obtain the tremor degree scores corresponding to the two candidate targets

[0013] Then, judge the relative magnitudes of the tremor degree scores of the two candidate targets, retain the candidate target with the smaller tremor degree score, and at the same time select the middle position between the two targets as a new candidate target to replace the other candidate target with a larger tremor degree score, and re-execute the above S2 and S3 for the updated two candidate targets respectively to obtain the tremor degree scores corresponding to the two candidate targets

[0014] Approximate the optimal solution by continuously iterating and updating the candidate targets until the termination condition is reached, and obtain the optimal target with the smallest tremor degree score

[0015] As a preference of the above first aspect, the optimization algorithm adopts a genetic algorithm. Each set of feasible solutions of the genetic algorithm is at least four electrode positions in the EEG 10 - 10 standard lead system and the current intensity of each electrode. The fitness function of the genetic algorithm is defined as the ratio of the peak electric field to the focusing degree at the candidate target, and the focusing degree is defined as the brain volume greater than a preset percentage of the peak electric field

[0016] As a preference of the above first aspect, when calculating the tremor degree score, use the root mean square value of the electromyogram signal as the tremor intensity feature and the zero crossing rate of the electromyogram signal as the tremor frequency feature, and sum the two feature parameters after weighting to obtain the tremor degree score

[0017] As a preference of the above first aspect, when calculating the tremor degree score, it is necessary to first call the weighted weight value pre - fitted with clinical data for the tremor severity determined by clinical evaluation to weight and sum the tremor frequency feature and the tremor intensity feature

[0018] As a preference of the above first aspect, the electromyogram signal is collected in real - time by arranging surface electromyogram sensors at the limb muscle parts of the stimulation object

[0019] As a preference of the above first aspect, the personalized head model is obtained by brain tissue segmentation and three - dimensional reconstruction from the head MRI and / or CT scan data of the stimulation object. During the execution of the optimization algorithm, for each set of candidate stimulation electrode positions and stimulation current intensities, load the stimulation electrode model at the corresponding position of the personalized head model, then perform finite - element mesh division and assign dielectric parameters to the brain tissue and the stimulation electrode model in the personalized head model respectively, and finally apply the stimulation current according to the corresponding stimulation current intensity to simulate and obtain the brain electric field distribution, and further calculate the peak electric field and the focusing degree at the candidate target

[0020] Second aspect, the present invention provides a time-domain interference electrical stimulation closed-loop regulation system for tremor rehabilitation, which is used to implement the time-domain interference electrical stimulation closed-loop regulation method for tremor rehabilitation described in any one of the above first aspect solutions, and it includes:

[0021] A signal acquisition unit, which is used to collect the electromyogram signals of the limb muscle parts of the stimulation object in real time;

[0022] A stimulation execution unit, which is used to generate a stimulation current control signal that meets the intensity and frequency requirements through a signal generator according to the received stimulation current intensity and stimulation frequency parameters, and output it to the stimulation electrode to generate a corresponding stimulation current;

[0023] A closed-loop regulation unit, which is used to continuously regulate the stimulation current output by the stimulation execution module according to the electromyogram signals collected in real time by the signal acquisition module in the process of iteratively executing the above S2 and S3 according to the time-domain interference electrical stimulation closed-loop regulation method for tremor rehabilitation described in any one of the above first aspect solutions, until the optimal target point with the smallest tremor degree score, and the best stimulation electrode position, the best stimulation current intensity, and the best stimulation frequency parameter combination corresponding to the optimal target point are obtained.

[0024] Third aspect, the present invention provides a computer program product, including computer programs / instructions, which when executed by a processor, can implement the time-domain interference electrical stimulation closed-loop regulation method for tremor rehabilitation described in any one of the above first aspect solutions.

[0025] Fourth aspect, the present invention provides a computer electronic device, which includes a memory and a processor;

[0026] The memory is used to store computer programs;

[0027] The processor is used to, when executing the computer program, be able to implement the time-domain interference electrical stimulation closed-loop regulation method for tremor rehabilitation described in any one of the above first aspect solutions.

[0028] Compared with the prior art, the present invention has the following beneficial effects:

[0029] (1) The time-domain interference electrical stimulation closed-loop regulation method for tremor designed by the present invention, through the closed-loop feedback method, the present invention can dynamically adjust the stimulation parameters according to the real-time tremor degree of the patient, and realize the generation of personalized stimulation parameters. This dynamic regulation method based on the individual characteristics of the patient can better adapt to the changes in the patient's condition and improve the individual accuracy of the stimulation parameters.

[0030] (2) The present invention provides a complete closed-loop regulation system, including a signal acquisition unit, a closed-loop regulation unit, and a stimulation unit. This systematic solution realizes the full-process automation from signal acquisition to stimulation implementation, reduces human intervention, improves the stability and repeatability of treatment, can quickly determine the intervention plan, and improves the treatment efficiency. Description of the Drawings

[0031] The drawings are used to better understand the present solution and do not constitute a limitation to this application. Among them:

[0032] Figure 1 It is a schematic diagram of the steps of the time-domain interference electrical stimulation closed-loop regulation method for tremor rehabilitation;

[0033] Figure 2 It is a schematic diagram of the control feedback process in a single-round parameter optimization cycle;

[0034] Figure 3 It is a schematic diagram of an exemplary process of time-domain interference electrical stimulation closed-loop regulation;

[0035] Figure 4 It is a schematic diagram of the module composition of the time-domain interference electrical stimulation closed-loop regulation system for movement tremor;

[0036] Figure 5 It is a schematic diagram of the structure of a computer electronic device. Detailed Embodiments

[0037] To make the above objects, features, and advantages of the present invention more obvious and understandable, the following describes the detailed embodiments of the present invention in conjunction with the drawings. Many specific details are set forth in the following description to fully understand the present invention. However, the present invention can be implemented in many other ways different from those described herein, and those skilled in the art can make similar improvements without departing from the connotation of the present invention. Therefore, the present invention is not limited by the specific embodiments disclosed below. The technical features in each embodiment of the present invention can be combined correspondingly without conflict.

[0038] In the description of the present invention, it should be understood that when an element is considered to be "connected" to another element, it can be directly connected to the other element or indirectly connected, that is, there is an intermediate element. On the contrary, when an element is referred to as being "directly" connected to another element, there is no intermediate element.

[0039] The present invention extracts tremor-related neuromuscular activity characteristics by real-time monitoring of the patient's electromyographic signals, and assesses the severity of tremor based on these characteristics. This allows a closed-loop feedback mechanism to be constructed. By dynamically adjusting the parameters of time-domain interferometric electrical stimulation (including the position of the stimulation electrode, current intensity, carrier frequency, and difference frequency), and gradually optimizing target selection and electrical stimulation parameter combinations, the optimal stimulation scheme that minimizes the degree of tremor is found.

[0040] Reference Figure 1 As shown, in one embodiment of the present invention, a time-domain interferometric electrical stimulation closed-loop regulation method for tremor rehabilitation is provided, which includes four steps S1 to S4. Each step is described in detail below.

[0041] S1. Select candidate targets in the target selection area of the brain.

[0042] It should be noted that the target selection region can be selected based on actual stimulation needs and can be determined by clinician diagnosis or expert experience. Since the STN-GPi circuit is a core pathway for basal ganglia motor regulation, and research has shown that stimulating targets such as the subthalamic nucleus (STN) or globus pallidus interna (GPi) can significantly improve patients' motor function, in embodiments of the present invention, the STN-GPi circuit is recommended as the target selection region from which candidate targets are selected. The number of candidate targets selected from the STN-GPi circuit can be determined based on actual needs, and the candidate targets can be selected in batches during the iterative optimization process or all at once. That is, all candidate targets can be selected in the target selection region at once, and then subsequent optimization and screening processes can be performed one by one to select the optimal target. Alternatively, a subset of targets can be selected in the target selection region, and based on the optimization results for these subset of targets, the target selection region can be narrowed down and candidate targets updated, with continued screening and optimization. The specific method of selecting candidate targets needs to be determined according to the selected optimization algorithm and is not limited to this.

[0043] S2. For each candidate target currently selected, simulate different stimulation electrode positions and stimulation current intensities using a personalized head model built for the stimulation subject. Maximizing the peak electric field and focus at the candidate target is the optimization goal. The optimal stimulation electrode position and the optimal stimulation current intensity for each stimulation electrode are determined through an optimization algorithm.

[0044] It should be noted that when performing time-domain interference electrical stimulation on candidate target points in the present invention, it is necessary to synchronously apply stimulation currents through four stimulation electrodes. Therefore, a set of feasible solutions for the positions of the stimulation electrodes and the stimulation current intensities in the optimization algorithm includes the spatial positions of the four stimulation electrodes relative to the head and the current intensities output by each of the four stimulation electrodes. Each set of feasible solutions during simulation also needs to correspond to setting stimulation electrode models at four positions and applying simulation currents to each stimulation electrode according to the corresponding current intensities. It should be noted that four electrodes are the minimum configuration for time-domain interference electrical stimulation, and if necessary, the number of stimulation electrodes can be more than four.

[0045] In the present invention, it is recommended to apply the stimulation current using the EEG 10-10 standard lead system. Since the positions of the stimulation electrodes in the EEG 10-10 standard lead system are relatively fixed, the optimization algorithm can search for the optimal electrode positions from all the electrode positions of the EEG 10-10 standard lead system. When optimizing the stimulation current intensity, certain limiting conditions should be met, that is: the current intensity of a single pair is less than 2.0 mA, and the total current intensity is 4 mA, so as to avoid adverse reactions that may be caused by too high current intensity.

[0046] In addition, the above optimization algorithm can be implemented using different algorithms. Since the optimization objectives are to maximize the peak electric field and the focusing degree at the candidate target point, it is necessary to jointly optimize these two optimization objectives. In practical applications, the spatial positions of the four stimulation electrodes relative to the head and the current intensities output by each of the four stimulation electrodes can be optimized through a multi-objective optimization algorithm with the peak electric field and the focusing degree at the candidate target point as two optimization objectives respectively, or the two optimization objectives can be fused into a single optimization objective, and a single-objective optimization algorithm can be used to optimize the spatial positions of the four stimulation electrodes relative to the head and the current intensities output by each of the four stimulation electrodes.

[0047] In the embodiments of the present invention, the above optimization algorithm can adopt a genetic algorithm. Each set of feasible solutions of the genetic algorithm is the positions of four electrodes in the EEG 10-10 standard lead system and the current intensities of each electrode. The fitness function of the genetic algorithm is defined as the ratio of the peak electric field to the focusing degree at the candidate target point, and the focusing degree is defined as the brain volume greater than a preset percentage of the peak electric field. Thus, there is no need to unify the two optimization objectives of the peak electric field and the focusing degree into a single optimization objective in the form of a ratio, reducing the complexity of the optimization algorithm.

[0048] It should be noted that the preset percentage of the peak electric field used to calculate the focusing degree can be optimized and adjusted according to the actual situation. In the embodiments of the present invention, it is recommended to adopt 85% of the peak electric field, that is, after extracting the peak electric field E max from the brain electric field distribution, further determine whether the electric field intensity of each brain voxel is greater than 85%Emax The voxel corresponding to the electric field strength greater than 85%E max is used as the focusing degree of the brain volume size.

[0049] In addition, after obtaining the optimal stimulation electrode position and the optimal stimulation current intensity for each currently selected candidate target, the two stimulation frequency parameter combinations of the carrier frequency and the difference frequency of the stimulation current can be further optimized to obtain the optimal stimulation frequency parameter combination for the candidate target. It should be noted that when optimizing the stimulation frequency parameter combination (i.e., the carrier frequency and the difference frequency) of the stimulation current, the corresponding optimization algorithm can also be selected according to actual needs. In theory, more complex optimization algorithms such as genetic algorithms can also be used. However, since the solution space for optimizing the stimulation frequency parameter combination is small, in step S3 of the present invention, the entire solution space can be directly traversed. The computational resources and time consumed are also relatively low. The carrier frequency and the difference frequency can be directly uniformly sampled within their respective value ranges, and then combined to form all candidate solutions. Each group of candidate solutions is traversed to select the optimal solution.

[0050] In addition, it should be noted that during the execution of the above optimization algorithm, for each group of candidate stimulation electrode positions and stimulation current intensities, finite element simulation needs to be performed. This simulation technology can be implemented with reference to the prior art. In the embodiment of the present invention, the personalized head model required for the simulation can be obtained by segmenting the brain tissue (which can be implemented using U-Net) and three-dimensional reconstruction of the head MRI and / or CT scan data of the stimulation object (it is recommended to use MRI data). During the execution of the optimization algorithm, for each group of candidate stimulation electrode positions and stimulation current intensities, the stimulation electrode model is loaded at the corresponding position of the personalized head model, and then finite element mesh division is performed, and dielectric parameters are assigned to the brain tissue and the stimulation electrode model in the personalized head model respectively. Finally, the stimulation current is applied according to the corresponding stimulation current intensity, and the brain electric field distribution is simulated to calculate the peak electric field and the focusing degree at the candidate target.

[0051] S3: For each currently selected candidate target, after arranging the stimulation electrodes according to the optimal stimulation electrode position, sample the carrier frequency and the difference frequency of the stimulation current of each stimulation electrode to form a series of candidate stimulation frequency parameter combinations. When the stimulation electrodes are fixed to output the optimal stimulation current intensity, continuously adjust the candidate stimulation frequency parameter combinations, sequentially generate the stimulation current control signals corresponding to different candidate stimulation frequency parameter combinations and output them to the stimulation electrodes, and simultaneously collect the myoelectric signals of the stimulation object after being stimulated by the stimulation current, extract the tremor frequency characteristics and the tremor intensity characteristics therefrom and calculate the tremor degree score, and select the candidate stimulation frequency parameter combination with the smallest tremor degree score as the optimal stimulation frequency parameter combination.

[0052] It should be noted that when sampling candidate stimulus frequency parameter combinations, the carrier frequency and the difference frequency, these two stimulus frequency parameters, each have their corresponding value spaces. A series of carrier frequency values and difference frequency values can be sampled from them respectively in a uniform sampling manner, and then cross-combined pairwise to form a series of candidate stimulus frequency parameter combinations. Each pair of candidate stimulus frequency parameter combinations includes a specific carrier frequency value and a specific difference frequency value. Subsequently, under the optimal stimulus electrode position and the optimal stimulus current intensity, a stimulus current control signal can be generated according to this pair of candidate stimulus frequency parameter combinations, thereby controlling the stimulus electrode to generate a stimulus current that meets the required current intensity, carrier frequency value, and difference frequency value. Since there is a series of candidate stimulus frequency parameter combinations, in this step, it is necessary to sequentially traverse each candidate stimulus frequency parameter combination, then generate the corresponding stimulus current control signal and apply it to the stimulation object, and then collect the electromyographic signal of the stimulation object after being stimulated by the stimulus current for tremor degree scoring. The control feedback process is shown in Figure 2 as shown. After all candidate stimulus frequency parameter combinations have been traversed, the corresponding candidate stimulus frequency parameter combination can be selected as the optimal stimulus frequency parameter combination based on the minimum tremor degree score.

[0053] In addition, in the present invention, when calculating the tremor degree score, it is necessary to extract the tremor frequency feature and the tremor intensity feature from the electromyographic signal. The specific forms of these two features can be optimized according to the actual signal. In the embodiments of the present invention, it is recommended to use the root mean square value of the electromyographic signal as the tremor intensity feature and the zero-crossing rate of the electromyographic signal as the tremor frequency feature, and the two feature parameters are weighted and summed to obtain the tremor degree score. The electromyographic signal can be collected in real time by arranging surface electromyographic sensors at the limb muscle parts of the stimulation object.

[0054] S4. Perform the above S2 and S3 on the candidate targets in the target selection area respectively, search for the optimal target with the goal of minimizing the tremor degree score, and save the optimal stimulus electrode position, the optimal stimulus current intensity, and the optimal stimulus frequency parameter combination corresponding to this optimal target.

[0055] It should be noted that the above S2 and S3 steps need to be performed separately for each candidate target. Each time S2 and S3 are performed, the optimal stimulus electrode position, the optimal stimulus current intensity, and the optimal stimulus frequency parameter combination for the current candidate target can be selected. After all candidate targets have been traversed, the optimal target can be selected. In theory, by traversing all candidate targets, the optimal target can surely be found, but the search efficiency of this traversal method is relatively low. Therefore, in the embodiments of the present invention, it is recommended to use the binary search method to search for the optimal target. For the candidate targets in the target selection area, taking the STN-GPi loop as an exemplary target selection area, the optimal target is searched by the binary search method. The process of the binary search is as follows:

[0056] First, taking the STN region and the GPi region as two candidate target points, respectively execute the above S2 and S3 to obtain the tremor degree scores corresponding to the two candidate target points respectively;

[0057] Then, judge the relative magnitudes of the tremor degree scores of the two candidate target points, retain the candidate target point with the smaller tremor degree score, and at the same time select the middle position between the two target points as a new candidate target point to replace the other candidate target point with the larger tremor degree score, and re-execute the above S2 and S3 for the updated two candidate target points respectively to obtain the tremor degree scores corresponding to the two candidate target points respectively;

[0058] Approximate the optimal solution by continuously iteratively updating the candidate target points until the termination condition is reached, and obtain the optimal target point with the smallest tremor degree score. Among them, the termination condition can be that the number of iterations reaches the preset maximum number or the change trend of the tremor degree score converges.

[0059] Therefore, in a preferred embodiment of the present invention, an exemplary process of the entire time-domain interference electrical stimulation closed-loop regulation can be seen as shown in Figure 3 shown. In this process, taking the STN-GPi loop as the recommended target selection area, and combining the dichotomy method to perform efficient target search, the specific implementation steps are as follows:

[0060] Step 1), First determine the initial positions of the stimulation target points: Take the STN-GPi loop as the target selection area and select two initial candidate target points from it (denoted as a and b respectively). The initial candidate target point a is set in the STN region, and the initial candidate target point b is set in the GPi region.

[0061] Step 2), According to the currently selected target points a and b, obtain the optimal stimulation electrode position and current intensity through the target optimization algorithm.

[0062] In this embodiment, the number of stimulating electrodes is 4, and the optimization method for the respective optimal electrode positions and current intensities is the genetic algorithm. Specifically, the EEG 10-10 standard lead system can be used as the search space for electrode positions, and all the electrodes therein are numbered. The chromosome consists of the numbers of the four electrodes and their respective current intensities. The current intensity of a single pair is less than 2.0 mA, and the total current intensity is 4 mA to avoid adverse reactions that may be caused by too high current intensity and ensure sufficient stimulation effect at the same time; the electric field distribution is simulated and calculated according to the finite element method; the fitness function can be defined as the ratio of the peak electric field at the target point to the focusing degree, where the focusing degree is defined as the brain volume greater than 85% of the peak electric field; the initial population starts from 1000 randomly generated individuals. For each individual, its fitness value is calculated. According to the fitness value, individuals with higher fitness are selected to enter the next generation, and 100 elites are retained in each generation; 500 parent individuals are randomly selected and crossed in pairs to generate new offspring individuals; the newly generated offspring individuals are randomly mutated with a mutation frequency of 20% to increase the population diversity; the operations of selection, crossover, and mutation are repeated until the maximum number of iterations is reached or the fitness no longer increases significantly; the optimal electrode positions (determined according to the numbers), current intensities, and fitness values of the 4 stimulating electrodes are output, and thus the optimal stimulating electrode positions of the 4 stimulating electrodes and the optimal stimulating current intensity of each stimulating electrode are obtained.

[0063] Step 3): Based on the determined optimal stimulating electrode positions and optimal stimulating current intensities, optimize and adjust the carrier frequency and difference frequency parameter combinations to find the optimal stimulating frequency parameter combination for this target point. The parameter optimization method for the carrier frequency and difference frequency is as follows:

[0064] On the premise of ensuring safety, according to the existing research and clinical experience of movement tremor treatment, set the safety ranges of the carrier frequency and difference frequency. Specifically: the carrier frequency is selected in the high-frequency range (such as 1000~5000 Hz) to meet the high-frequency requirements of time-domain interference electrical stimulation; the difference frequency is set within the tremor-related frequency range of movement tremor patients (such as 1~200 Hz) to accurately regulate tremor symptoms. Within the above safety ranges, sample and freely combine the carrier frequency and difference frequency according to the preset step size to form a series of candidate stimulating frequency parameter combinations. Then, when the 4 stimulating electrodes fixedly output their respective optimal stimulating current intensities, the current frequency parameters output by the electrodes are adjusted one by one in a traversal manner to each candidate stimulating frequency parameter combination, and the stimulating current control signals corresponding to different candidate stimulating frequency parameter combinations are generated and output to the stimulating electrodes in turn, and the electromyographic signals of the stimulated object after being stimulated by the current are collected in real time, the tremor frequency characteristics and tremor intensity characteristics are extracted therefrom, and the tremor degree score is calculated. The candidate stimulating frequency parameter combination with the smallest tremor degree score is selected as the optimal stimulating frequency parameter combination.

[0065] Among them, the process of calculating the tremor severity score based on the electromyogram signal can objectively, quantitatively, and continuously evaluate the tremor state by analyzing the electrophysiological characteristics of the tremor-related muscle groups. In the embodiments of the present invention, surface electromyogram sensors disposed on muscle parts such as the patient's forearm, leg, or wrist can be used to obtain muscle activity signals in real time, and the original electromyogram signals are preprocessed by denoising, filtering, and normalization to extract key parameters closely related to the tremor intensity and frequency. In this embodiment, it is recommended to select RMS (root mean square value) and ZCR (zero crossing rate) as the core indicators for tremor state analysis. Among them, RMS is used as the tremor intensity feature to reflect the overall activation level and contraction intensity of the muscle. The higher the value, the greater the degree of continuous contraction or abnormal activation of the muscle; while ZCR is used as the tremor frequency feature to characterize the frequency change of the periodic components in the electromyogram signal, and its value is positively correlated with the tremor frequency. Based on these two physiological indicators, a tremor severity score TremorSeverity Score (TSS) model is constructed, and the model can be expressed as:

[0066]

[0067] where α and β are weight coefficients, which can be optimized by fitting clinical data. RMS_norm and ZCR_norm are the RMS value and ZCR value after normalization processing respectively. This scoring model quantitatively maps the tremor state to a continuous numerical interval.

[0068] Through research, it is found that the weight coefficients in the above TSS model are related to the severity of the patient's tremor. Therefore, in the embodiments of the present invention, it is necessary to first determine the severity of the tremor of the stimulation object according to the clinical evaluation. The clinical evaluation can be achieved by using the clinical scale evaluation method. Then, the weighted weight values pre-fitted with clinical data for this tremor severity are called to perform weighted summation on the tremor frequency feature and the tremor intensity feature to obtain the above TSS score. The results of the clinical scale evaluation can be divided into four levels: "no tremor, mild tremor, moderate tremor, severe tremor" to describe the severity of the patient's tremor. Suppose the total score score obtained from the clinical evaluation is normalized to 0-1, and the level division rules are as follows: no tremor: score < 0.1; mild tremor: 0.1 ≤ score < 0.3; moderate tremor: 0.3 ≤ score < 0.6; severe tremor: score ≥ 0.6. The fitting values of the relevant weight coefficients corresponding to the four tremor levels are:

[0069]

[0070] In addition to this embodiment, the tremor degree score can also be obtained by fusing and solving indicators such as the waveform length of electromyogram, the average absolute value, the tremor band power ratio, the power spectral density, the center frequency, the bandwidth, and the instantaneous tremor activity enhancement area.

[0071] Step 4: Based on the currently selected two candidate targets a and b, through the above steps, the best stimulation electrode position, the best stimulation current intensity, and the best stimulation frequency parameter combination corresponding to each optimal target can be obtained, and at the same time, the corresponding tremor degree score TSS can be obtained. Then, the targets are gradually selected according to the distance iteration. The best stimulation electrode position is optimized according to the selected target. The target update selection method is the dichotomy method, and the selection method is as follows:

[0072] For the target a initially set in the STN region and the target b initially set in the GPi region, the tremor degree scores under the best stimulation electrode position, the best stimulation current intensity, and the best stimulation frequency parameter combination can be compared. The smaller the tremor degree score, the better the stimulation effect. Therefore, the relative advantages and disadvantages of the stimulation effects of the two targets can be compared, and then the candidate target with the smaller tremor degree score is retained. At the same time, the middle position of the two targets is selected as the new candidate target to replace the other candidate target with a larger tremor degree score, and the above-mentioned optimization steps are respectively executed for the updated two candidate targets. In the present invention, taking the target a and the target b as examples, the update process is as follows:

[0073] Compare the stimulation effects produced by the stimulation parameter combinations corresponding to the target a and the target b. If the stimulation effect produced by the stimulation parameter (current intensity and current frequency parameter) combination corresponding to the target a is better than that of the target b, the next target selection area is reduced to the loop between the target a and the midpoint c of a and b, that is, the position of the target b is changed to c, and the stimulation parameters are gradually optimized again, and its stimulation effect is compared with that of the target a; if the stimulation effect produced by the stimulation parameter combination corresponding to the target a is inferior to that of the target b, the next target selection area is reduced to the loop between the target b and the midpoint c of a and b, that is, the position of the target a is changed to c, and the stimulation parameters are gradually optimized again, and its stimulation effect is compared with that of the target b.

[0074] Thus, through continuously narrowing the target selection range, the above dichotomy method can quickly and efficiently find the target that minimizes the patient's tremor degree and avoid global traversal. When the set number of iterations is completed or the tremor degree does not improve significantly after multiple optimizations of the stimulation, the algorithm execution ends, and the target and the stimulation parameter combination corresponding to the best stimulation effect are selected.

[0075] It can be seen that the above-mentioned time-domain interference electrical stimulation closed-loop regulation method for tremor rehabilitation provided by the present invention can dynamically adjust the stimulation parameters according to the real-time tremor degree of the patient through the closed-loop feedback method, and generate a personalized stimulation plan. This dynamic regulation method based on the individual characteristics of the patient can better adapt to the changes in the patient's condition.

[0076] It should be specifically noted that the above-mentioned closed-loop regulation process of time-domain interference electrical stimulation in steps S1 to S4 only describes the regulation output of the stimulation signal and the process of feedback optimization based on the collected myoelectric signals, but does not involve the specific electrical stimulation treatment process. That is to say, the above-mentioned S1 to S4 of the present invention correspond to the signal feedback control and optimization at the instrument and equipment level, but the output of the electrical stimulation current and its action on the target position of the actual patient belong to the process outside S1 to S4 of the present invention. Therefore, this technical solution only involves the signal processing process, but does not involve the disease treatment process.

[0077] In another embodiment of the present invention, a time-domain interference electrical stimulation closed-loop regulation system for tremor rehabilitation can be further provided. This system is used to implement the time-domain interference electrical stimulation closed-loop regulation method shown in the above-mentioned S1 to S4 steps. In this system, its core includes a signal acquisition unit, a stimulation execution unit, and a closed-loop regulation unit. The specific forms and mutual cooperation relationships of each unit will be described in detail below.

[0078] The above-mentioned signal acquisition unit is used to collect the myoelectric signals of the limb muscle parts of the stimulation object in real time;

[0079] The above-mentioned stimulation execution unit is used to generate a stimulation current control signal that meets the intensity and frequency requirements through a signal generator according to the received stimulation current intensity and stimulation frequency parameters, and output it to the stimulation electrode to generate a corresponding stimulation current;

[0080] The above-mentioned closed-loop regulation unit is used to continuously regulate the stimulation current output by the stimulation execution module according to the myoelectric signals collected in real time by the signal acquisition module in the process of iteratively executing S2 and S3 according to the time-domain interference electrical stimulation closed-loop regulation method for tremor rehabilitation in the foregoing embodiment, until the optimal target point with the smallest tremor degree score and the combination of the best stimulation electrode position, the best stimulation current intensity, and the best stimulation frequency parameters corresponding to the optimal target point are obtained.

[0081] It should be noted that the above-mentioned signal acquisition unit can be a sensor capable of accurately detecting myoelectric signals and the corresponding signal processing circuit. For example, a mature combination scheme of a myoelectric amplifier and a myoelectric electrode can be used to implement it. In addition,

[0082] It should be noted that the above signal acquisition unit and stimulation execution unit are hardware modules, and the main function of the closed-loop regulation unit is implemented through software modules, but it can be implemented by being mounted on an electronic device with data processing and signal transceiver capabilities. The electronic device cooperates with the signal acquisition unit and the stimulation execution unit to integrally implement the time-domain interference electrical stimulation closed-loop regulation method for tremor rehabilitation in the foregoing embodiments, and finally outputs the stimulation current under the combination limit of the optimal stimulation electrode position, the optimal stimulation current intensity, and the optimal stimulation frequency parameter through the stimulation execution unit, and then acts on the optimal target point of the stimulation object, that is, the patient, to obtain the best treatment effect.

[0083] The above stimulation execution unit can be implemented with reference to the transcranial electrical stimulation system in the prior art. The signal control part inside it is composed of a constant current source, a boost pump, a waveform generator, etc., and can output an electrical stimulation waveform with adjustable current intensity, frequency, and time interval parameters. The above waveform synthesizer, constant current source, and boost pump can be finished products in the form of integrated chips, and the parameters of such chips can be modified through programming. In the embodiment of the present invention, the output channels of the stimulation execution module are at least 4, because 4 stimulation electrodes need to be arranged simultaneously for the time-domain interference electrical stimulation of tremor rehabilitation.

[0084] The above closed-loop regulation unit is the core of the regulation of the present invention, and the S1-S4 steps executed inside it can be further implemented through multiple software function modules. See Figure 4As shown, in the embodiment of the present invention, it can be divided into a feedback processing subunit, a first optimization subunit, and a second optimization subunit. Among them, the feedback processing subunit is used to control the loop of the parameter closed-loop condition, which includes a tremor degree evaluation module and an execution judgment module. The tremor degree evaluation module is used to extract the tremor frequency feature and tremor intensity feature from the electrocardiogram signal and perform weighted processing to obtain the tremor degree score of each round of loop, so as to quantitatively reflect the severity of the tremor and record it. The execution judgment module decides whether to continue to execute the iterative step according to the tremor degree score. The first optimization subunit and the second optimization subunit are both parameter optimization units. The first optimization subunit includes a target selection module and an electrode position and current intensity optimization module, and the second optimization subunit includes a frequency optimization module and a feedback module. The first optimization subunit optimizes the electrode position based on the EEG 10-10 standard lead system (that is, 4 electrodes are selected from the system for activation output). The target selection module uses the STN-GPi loop as the target selection area and gradually narrows the target range through the dichotomy method. The electrode position and current intensity optimization module obtains the optimal stimulation electrode position and the optimal current intensity through multi-objective optimization according to the selected target. Based on the determined electrode position and the optimal current intensity, the second optimization subunit, the frequency optimization module freely combines the carrier frequency and difference frequency of the current according to the preset safety range and step size, and sequentially outputs the parameter combinations that meet the requirements to the stimulation execution unit. The feedback module finds the stimulation parameter combination that minimizes the tremor degree for the determined target according to the tremor degree evaluation module in the feedback processing unit and stores it, that is, saves the finally optimized parameters (the optimal stimulation current intensity and the optimal stimulation frequency parameter combination).

[0085] It should be noted that the method steps shown in S1~S4 above can essentially be implemented in the form of a computer program or a software function module.

[0086] Thus, based on the same inventive concept, as Figure 5 shown, the present invention also provides a computer electronic device corresponding to the time-domain interference electrical stimulation closed-loop regulation method for tremor rehabilitation provided in the above embodiment, which includes a memory and a processor;

[0087] The memory is used to store a computer program;

[0088] The processor is used to implement the time-domain interference electrical stimulation closed-loop regulation method for tremor rehabilitation as described above when executing the computer program;

[0089] In addition, when the logical instructions in the above-mentioned memory are implemented in the form of software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on such an understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or a part of the technical solution, can be embodied in the form of a software product. The computer software product is stored in a storage medium and includes several instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention.

[0090] Therefore, based on the same inventive concept, the present invention provides a computer-readable storage medium corresponding to a time-domain interference electrical stimulation closed-loop regulation method for tremor rehabilitation. A computer program is stored on the storage medium, and when the computer program is executed by a processor, it can implement the time-domain interference electrical stimulation closed-loop regulation method for tremor rehabilitation as described above.

[0091] Therefore, based on the same inventive concept, the present invention provides a computer program product, including a computer program / instructions. When the computer program / instructions are executed by a processor, they can implement the time-domain interference electrical stimulation closed-loop regulation method for tremor rehabilitation as described above.

[0092] Specifically, in the computer-readable storage media of the above three embodiments, the stored computer program is executed by a processor, and the steps of S1 to S4 described above can be executed.

[0093] It can be understood that the above storage medium may include a random access memory (RAM), and may also include a non-volatile memory (NVM), such as at least one disk memory. At the same time, the storage medium may also be various media such as a USB flash drive, a mobile hard disk, a magnetic disk, or an optical disc that can store program codes.

[0094] It can be understood that the above-mentioned processor may be a general-purpose processor, including a central processing unit (CPU), a network processor (NP), etc.; it may also be a digital signal processor (DSP), an application specific integrated circuit (ASIC), a field-programmable gate array (FPGA), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components.

[0095] It should be further noted that those skilled in the art can clearly understand that for the convenience and brevity of description, the specific working process of the system described above can refer to the corresponding process in the foregoing method embodiments, and will not be elaborated herein. In the embodiments provided in the present application, the division of steps or modules in the system and method is only a logical function division, and there may be other division methods in actual implementation. For example, multiple modules or steps can be combined or integrated together, and a module or step can also be split.

[0096] The above-described embodiments are only some preferred implementation solutions of the present invention, but are not intended to limit the present invention. Those of ordinary skill in the relevant art can make various changes and modifications without departing from the spirit and scope of the present invention. Therefore, all technical solutions obtained by means of equivalent replacement or equivalent transformation fall within the protection scope of the present invention.

Claims

1. A time-domain interference electrical stimulation closed-loop regulation method for tremor rehabilitation, characterized in that, Including: S1. Select candidate targets in the target selection area of the brain; S2. For each currently selected candidate target, use a personalized head model constructed for the stimulation object to simulate different stimulation electrode positions and stimulation current intensities. With the maximization of the peak electric field and focusing degree at the candidate target as the optimization goal, determine the optimal stimulation electrode position and the optimal stimulation current intensity for each stimulation electrode through an optimization algorithm; S3. For each currently selected candidate target, after arranging the stimulation electrodes according to the optimal stimulation electrode position, sample the stimulation current carrier frequency and difference frequency of each stimulation electrode to form a series of candidate stimulation frequency parameter combinations. Continuously adjust the candidate stimulation frequency parameter combinations while the stimulation electrodes fixedly output the optimal stimulation current intensity, successively generate the stimulation current control signals corresponding to different candidate stimulation frequency parameter combinations and output them to the stimulation electrodes, and simultaneously collect the electromyogram signals of the stimulation object after being stimulated by the stimulation current in real time. Extract the tremor frequency characteristics and tremor intensity characteristics therefrom and calculate the tremor degree score, and select the candidate stimulation frequency parameter combination with the smallest tremor degree score as the optimal stimulation frequency parameter combination; S4. Execute S2 and S3 respectively for the candidate targets in the target selection area, with the minimization of the tremor degree score as the goal, search for the optimal target, and save the optimal stimulation electrode position, the optimal stimulation current intensity, and the optimal stimulation frequency parameter combination corresponding to the optimal target.

2. The time-domain interference electrical stimulation closed-loop regulation method for tremor rehabilitation according to claim 1, wherein For the candidate targets in the target selection area, with the STN-GPi loop as the target selection area, search for the optimal target through the bisection method. The process of the bisection method is as follows: First, take the STN area and the GPi area as two candidate targets, and execute S2 and S3 respectively to obtain the tremor degree scores corresponding to the two candidate targets respectively; Then, judge the relative magnitudes of the tremor degree scores of the two candidate targets, retain the candidate target with the smaller tremor degree score, and at the same time select the middle position between the two targets as a new candidate target to replace the other candidate target with a larger tremor degree score, and re-execute S2 and S3 respectively for the two updated candidate targets to obtain the tremor degree scores corresponding to the two candidate targets respectively; Approximate the optimal solution by continuously iteratively updating the candidate targets until the termination condition is reached, and obtain the optimal target with the smallest tremor degree score.

3. The time-domain interference electrical stimulation closed-loop regulation method for tremor rehabilitation according to claim 1, wherein The optimization algorithm adopts a genetic algorithm. Each set of feasible solutions of the genetic algorithm is at least four electrode positions in the EEG 10-10 standard lead system and the current intensities of each electrode. The fitness function of the genetic algorithm is defined as the ratio of the peak electric field to the focusing degree at the candidate target, and the focusing degree is defined as the brain volume greater than a preset percentage of the peak electric field.

4. The time-domain interference electrical stimulation closed-loop regulation method for tremor rehabilitation according to claim 1, wherein, When calculating the tremor degree score, use the root mean square value of the electromyogram signal as the tremor intensity characteristic, use the zero crossing rate of the electromyogram signal as the tremor frequency characteristic, and obtain the tremor degree score by weighted summing the two characteristic parameters.

5. The time-domain interference electrical stimulation closed-loop regulation method for tremor rehabilitation according to claim 4, wherein, When calculating the tremor degree score, it is necessary to first determine the tremor severity of the stimulation object according to clinical evaluation, and then call the weighted weight value pre-fitted with clinical data for this tremor severity to perform weighted summation on the tremor frequency feature and the tremor intensity feature.

6. The time-domain interference electrical stimulation closed-loop regulation method for tremor rehabilitation according to claim 1, wherein, The electromyogram signal is collected in real time by arranging surface electromyogram sensors at the limb muscle parts of the stimulation object.

7. The time-domain interference electrical stimulation closed-loop regulation method for tremor rehabilitation according to claim 1, characterized in that, The personalized head model is obtained by brain tissue segmentation and three-dimensional reconstruction from the head MRI and / or CT scan data of the stimulation object. During the execution of the optimization algorithm, for each group of candidate stimulation electrode positions and stimulation current intensities, a stimulation electrode model is loaded at the corresponding position of the personalized head model, and then finite element meshing is performed, and dielectric parameters are assigned to the brain tissue and the stimulation electrode model in the personalized head model respectively. Finally, a stimulation current is applied according to the corresponding stimulation current intensity, and the brain electric field distribution is simulated to calculate the peak electric field and the focusing degree at the candidate target point.

8. A time-domain interference electrical stimulation closed-loop regulation system for tremor rehabilitation, characterized in that, Including: A signal acquisition unit for collecting the electromyogram signal of the limb muscle parts of the stimulation object in real time; A stimulation execution unit for generating a stimulation current control signal that meets the intensity and frequency requirements through a signal generator according to the received stimulation current intensity and stimulation frequency parameters, and outputting it to the stimulation electrode to generate a corresponding stimulation current; A closed-loop adjustment unit for continuously regulating the stimulation current output by the stimulation execution module according to the electromyogram signal collected in real time by the signal acquisition module during the iterative execution of S2 and S3 according to the time-domain interference electrical stimulation closed-loop adjustment method for tremor rehabilitation described in any one of claims 1 to 7 until the optimal target point with the minimum tremor degree score, and the best stimulation electrode position, the best stimulation current intensity, and the best stimulation frequency parameter combination corresponding to the optimal target point are obtained.

9. A computer program product, comprising a computer program / instructions, characterized in that, When the computer program / instructions are executed by a processor, the time-domain interference electrical stimulation closed-loop adjustment method for tremor rehabilitation described in any one of claims 1 to 7 can be implemented.

10. A computer electronic device, characterized in that, Including a memory and a processor; The memory is used to store a computer program; The processor is used to implement the time-domain interference electrical stimulation closed-loop adjustment method for tremor rehabilitation described in any one of claims 1 to 7 when executing the computer program.

Images