A control method and system for exoskeleton rehabilitation equipment based on electroencephalogram signals

Abstract

The present invention discloses a control method and system for exoskeleton rehabilitation equipment based on EEG signals, which relates to the field of artificial intelligence technology. The method comprises: inputting multiple signal features into an intention classification model for intention classification, receiving multiple signal features through an input layer, extracting multiple signal features through a feature extraction layer to obtain multiple features required for calculation, performing calculations based on the required features through a motion intention index calculation layer to obtain a motion intention index, determining the patient's intention type based on the motion intention index through a decision layer; generating motion parameter adjustment instructions based on the patient's intention type; and adjusting the movement speed and movement amplitude of the leg support based on the motion parameter adjustment instructions. The technical solution of the present invention can solve the problem in the prior art that there is a lack of close real-time feedback and linkage between the exoskeleton rehabilitation equipment and the patient's neurological state, which limits the personalization, accuracy and real-time adaptability of rehabilitation.

Description

Technical Field

[0001] The present invention relates to the field of artificial intelligence technology, and in particular to a method and system for controlling exoskeleton rehabilitation equipment based on electroencephalogram (EEG) signals. Background Art

[0002] Exoskeleton rehabilitation devices currently play an important role in neurological rehabilitation, particularly for conditions like stroke and spinal cord injury. Exoskeleton rehabilitation devices can assist patients in regaining basic functions like walking and standing. However, current control technology for exoskeleton rehabilitation devices has many limitations.

[0003] On the one hand, many exoskeleton rehabilitation devices use control systems based on fixed parameters. These preset, fixed motion parameters, such as gait, speed, and stride length, are often based on the general population or average rehabilitation patient levels. However, individuals with neurological injuries vary significantly, with each patient having a unique degree of nerve damage and physical function. This lack of personalized consideration fails to meet the specific needs of patients, can easily lead to poor rehabilitation outcomes, and can even hinder recovery progress.

[0004] In the process of realizing the present invention, the inventors found that the existing technology lacks close real-time feedback and linkage between the exoskeleton rehabilitation equipment and the patient's neurological state, and does not fully utilize EEG signals to guide exoskeleton rehabilitation, which limits the personalization, accuracy and real-time adaptability of exoskeleton rehabilitation. Summary of the Invention

[0005] In view of this, the present invention proposes a control method and system for exoskeleton rehabilitation equipment based on EEG signals to solve the problem in the prior art that there is a lack of close real-time feedback and linkage between the exoskeleton rehabilitation equipment and the patient's neural state, which limits the personalization, accuracy and real-time adaptability of exoskeleton rehabilitation.

[0006] To achieve the above object, the present invention provides the following technical solutions:

[0007] A method for controlling an exoskeleton rehabilitation device based on electroencephalogram (EEG) signals, comprising:

[0008] The patient's EEG signals are collected in real time by an EEG signal collection device;

[0009] Converting the EEG signal from the time domain to the frequency domain by short-time Fourier transform, calculating the power spectral density corresponding to the EEG signal, calculating the coherence and weighted phase delay index based on the power spectral density, performing multiple comparisons on the functional connectivity values ​​calculated for each frequency band using the coherence and weighted phase delay index, and extracting multiple signal features;

[0010] Inputting multiple signal features into an intention classification model for intention classification, the intention classification model comprising an input layer, a feature extraction layer, a motion intention index calculation layer, a decision layer, and an output layer, wherein the input layer receives multiple signal features, the feature extraction layer extracts multiple signal features to obtain multiple features required for calculation, the motion intention index calculation layer performs calculation based on the required features to obtain a motion intention index, the decision layer determines a patient intention type based on the motion intention index, and the output layer outputs the patient intention type;

[0011] generating an exercise parameter adjustment instruction based on the patient's intention type;

[0012] The movement speed and movement amplitude of the leg support are adjusted based on the movement parameter adjustment instruction.

[0013] On the basis of the above technical solution, the present invention can also be improved as follows:

[0014] Optionally, the calculating a coherence and weighted phase delay index based on the power spectral density includes:

[0015] The correlation is calculated using formula (1);

[0016] Formula (1);

[0017] Where, is the cross power spectral density, For electrodes The autopower spectral density of For electrodes The autopower spectral density of and Indicates different electrodes;

[0018] The weighted phase delay index is calculated by formula (2);

[0019] Formula (2);

[0020] Where, is a discrete time point, N is the number of data points, k is the frequency, is a symbolic function, is the phase difference. When the phase difference is greater than 0, , when the phase difference is less than 0, , when the phase difference is equal to 0, .

[0021] Optionally, the calculating by the movement intention index calculation layer based on the features required for the calculation to obtain the movement intention index includes:

[0022] The patient's movement intention index was calculated using formula (3);

[0023] Formula (3);

[0024] Where, is the movement intention index, EEG The frequency band set, EEG The power spectral density, EEG The weight coefficient of is the set of electrode pairs, For electrode pairs The coherence value of For electrode pairs The coherence weight of is the set of electrode pairs, For electrode pairs The weighted phase delay index of For electrode pairs The weighted phase delay index weight, is the adjustment coefficient for balancing the power spectrum density, is the coherent adjustment coefficient, is the adjustment coefficient of the weighted phase delay index.

[0025] Optionally, determining the patient's intention type based on the movement intention index by the decision layer includes:

[0026] determining whether the movement intention index is greater than a first preset threshold, and if so, determining the patient's intention type as a first type;

[0027] determining whether the exercise intention index is within a second preset threshold range, and if so, determining the patient's intention to exercise as a second type;

[0028] Determine whether the movement intention index is less than a third preset threshold; if so, determine it as a third patient intention type.

[0029] Optionally, the exoskeleton rehabilitation device control method based on EEG signals further includes:

[0030] Recording the EEG signals of a normal person in different motion states, and storing the EEG signals of the normal person in different motion states in a storage unit;

[0031] Determine whether there is a difference between the patient's EEG signal and that of a normal person under the same motion state. If so, determine difference data and adjust electrical stimulation parameters based on the difference data.

[0032] Optionally, the determining whether there is a difference between the patient's EEG signal and that of a normal person under the same motion state, and if so, determining difference data, and adjusting electrical stimulation parameters based on the difference data, includes:

[0033] Calculate the difference data using formula (4);

[0034] Formula (4);

[0035] Where, For differential data, is the weight of the time domain difference data, is the time domain difference data, is the weight of the frequency domain difference data, is the frequency domain difference data;

[0036] The adjusted electrical stimulation intensity was calculated by formula (5);

[0037] Formula (5);

[0038] Where, is the adjusted electrical stimulation intensity, is the initial electrical stimulation intensity, is the electrical stimulation intensity adjustment coefficient, For differential data;

[0039] The adjusted electrical stimulation frequency is calculated by formula (6);

[0040] Formula (6);

[0041] Where, is the adjusted electrical stimulation frequency, is the initial electrical stimulation frequency, is the electrical stimulation frequency adjustment coefficient, For differential data;

[0042] The adjusted stimulation duration is calculated by formula (7);

[0043] Formula (7);

[0044] Where, is the adjusted stimulation duration, is the initial stimulation duration, is the stimulus duration adjustment coefficient, For differential data.

[0045] An exoskeleton rehabilitation device control system based on EEG signals, comprising:

[0046] An electroencephalogram (EEG) signal acquisition device, used to acquire the patient's EEG signals in real time;

[0047] a multiple signal feature extraction unit, configured to convert the EEG signal from the time domain to the frequency domain by short-time Fourier transform, calculate the power spectral density corresponding to the EEG signal, calculate the coherence and weighted phase delay index based on the power spectral density, perform multiple comparisons on the functional connectivity values ​​calculated for each frequency band using the coherence and weighted phase delay index, and extract multiple signal features;

[0048] an intention type determination unit, configured to input a plurality of signal features into an intention classification model for intention classification, the intention classification model comprising an input layer, a feature extraction layer, a motion intention index calculation layer, a decision layer, and an output layer, wherein the input layer receives a plurality of signal features, the feature extraction layer extracts the plurality of signal features to obtain a plurality of features required for calculation, the motion intention index calculation layer performs calculation based on the required features to obtain a motion intention index, the decision layer determines a patient's intention type based on the motion intention index, and the output layer outputs the patient's intention type;

[0049] an adjustment instruction generating unit, configured to generate an exercise parameter adjustment instruction based on the patient's intention type;

[0050] The power unit is used to adjust the movement speed and movement amplitude of the leg support based on the movement parameter adjustment instruction.

[0051] Optionally, the exoskeleton rehabilitation device control system based on EEG signals further includes a storage unit and a control unit;

[0052] The storage unit is used to store EEG signals of normal people in different motion states;

[0053] The control unit is used to determine whether there is a difference between the patient's EEG signal and the EEG signal of a normal person under the same movement state. If so, determine the difference data and adjust the electrical stimulation parameters based on the difference data.

[0054] An electronic device comprises a memory, a processor and a computer program stored in the memory and running on the processor, wherein the steps of the method are implemented when the processor executes the computer program.

[0055] A non-transitory computer-readable storage medium stores a computer program, which implements the steps of the method when executed by a processor.

[0056] The present invention has the following advantages:

[0057] The EEG-based exoskeleton rehabilitation device control method proposed in this paper, leveraging the collaborative operation of an EEG signal acquisition device and an intention classification model, can closely track changes in a patient's neural activity, analyze movement intentions in real time, and dynamically optimize the leg support's motion parameters accordingly. This method breaks the limitations of traditional fixed models, providing patients with a personalized rehabilitation path tailored to their current condition, significantly improving the accuracy and effectiveness of rehabilitation training and significantly enhancing rehabilitation outcomes.

[0058] The exoskeleton rehabilitation equipment control method based on EEG signals in the present invention uses non-invasive EEG signals as the control source, avoiding the risks and discomfort that may be caused by traditional invasive stimulation methods, providing patients with a safe and non-invasive rehabilitation experience, and improving patients' acceptance of rehabilitation treatment.

[0059] The exoskeleton rehabilitation equipment control method based on EEG signals in the present invention can customize an exclusive rehabilitation plan for each patient according to the EEG signal characteristics of each patient, enhance the targetedness and effectiveness of treatment, and further improve the patient's compliance and satisfaction during the rehabilitation process. BRIEF DESCRIPTION OF THE DRAWINGS

[0060] For purposes of illustration and not limitation, the present invention will now be described with reference to embodiments thereof and the accompanying drawings, in which:

[0061] Figure 1 Schematic diagram of a first process of a method for controlling an exoskeleton rehabilitation device based on EEG signals in an embodiment of the present invention;

[0062] Figure 2 Schematic diagram of a second process of the method for controlling an exoskeleton rehabilitation device based on EEG signals in an embodiment of the present invention;

[0063] Figure 3 Schematic diagram of the main components of the control system of the exoskeleton rehabilitation device based on EEG signals in an embodiment of the present invention;

[0064] Figure 4 This is a schematic diagram of the physical structure of the electronic device provided by the present invention. DETAILED DESCRIPTION

[0065] In order to enable those skilled in the art to better understand the solutions of the present invention, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative work should fall within the scope of protection of the present invention.

[0066] It should be noted that the terms "first," "second," and the like in the description of the present invention and the accompanying drawings are used to distinguish similar objects and are not necessarily used to describe a particular order or precedence. It should be understood that the terms used in this manner are interchangeable where appropriate for the embodiments of the present invention described herein. In addition, the terms "including," "having," and any variations thereof are intended to cover non-exclusive inclusions. For example, a process, method, system, product, or apparatus comprising a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units that are not explicitly listed or that are inherent to these processes, methods, products, or apparatuses.

[0067] It should be noted that, in the absence of conflict, the embodiments of the present invention and the features thereof can be combined with each other. The embodiments of the present invention will be described in detail below with reference to the accompanying drawings.

[0068] Figure 1 FIG. 1 is a schematic diagram of a first process of a method for controlling an exoskeleton rehabilitation device based on EEG signals in an embodiment of the present invention. Figure 1 As shown, the first exoskeleton rehabilitation device control method based on EEG signals provided by the embodiment of the present invention includes the following steps S101 to S105.

[0069] S101, collecting the patient's EEG signals in real time through an EEG signal collection device.

[0070] The specific operation can be to closely attach the electrode array of the EEG signal acquisition device to the patient's scalp, so as to accurately monitor the patient's brain wave activity in real time. Regardless of whether the patient is in a resting state or in the process of exercise, their neural state information can be obtained comprehensively and detailedly.

[0071] S102, extracting multiple signal features from the EEG signal.

[0072] Specifically, the EEG signals are preprocessed to obtain preprocessed data, where the data preprocessing includes filtering, segmentation, bad segment removal, artifact removal, and re-referencing. Marking allows for precise segmentation and in-depth analysis of the EEG signals, enabling the control unit to accurately understand the patient's motor intentions corresponding to EEG signal changes during different task phases.

[0073] The pre-processed data is converted from the time domain to the frequency domain by short-time Fourier transform, and the power spectrum density corresponding to the pre-processed data is calculated.

[0074] Coherence and weighted phase delay indices are calculated based on the power spectral density.

[0075] The correlation is calculated using formula (1);

[0076] Formula (1);

[0077] Where, is the cross power spectral density, For electrodes The autopower spectral density of For electrodes The autopower spectral density of and Indicates different electrodes;

[0078] The weighted phase delay index is calculated by formula (2);

[0079] Formula (2);

[0080] Where, is a discrete time point, N is the number of data points, k is the frequency, is a symbolic function, is the phase difference. When the phase difference is greater than 0, , when the phase difference is less than 0, , when the phase difference is equal to 0, .

[0081] The functional connectivity values ​​calculated for each frequency band using the coherence and weighted phase delay indices were subjected to multiple comparisons using the Network Solution Toolbox. After correction for multiple comparisons, the connectivity of the brain functional network was visualized using BrainNet Viewer, allowing for in-depth exploration of brain functional network characteristics, discovery of potential brain functional differences and associations, and extraction of various signal features.

[0082] Multiple signal features are input into the intention classification model to perform intention classification and obtain the patient's movement intention.

[0083] S103: Input multiple signal features into the intent classification model to perform intent classification.

[0084] Specifically, the intention classification model includes an input layer, a feature extraction layer, a motion intention index calculation layer, a decision layer and an output layer. The input layer receives multiple signal features, the feature extraction layer extracts multiple signal features to obtain multiple features required for calculation, the motion intention index calculation layer performs calculation based on the features required for calculation to obtain the motion intention index, the decision layer determines the patient's intention type based on the motion intention index, and the output layer outputs the patient's intention type.

[0085] The patient's movement intention index was calculated using formula (3);

[0086] Formula (3);

[0087] Where, is the movement intention index, EEG The frequency band set, EEG The power spectral density, EEG The weight coefficient of is the set of electrode pairs, For electrode pairs The coherence value of For electrode pairs The coherence weight of is the set of electrode pairs, For electrode pairs The weighted phase delay index of For electrode pairs The weighted phase delay index weight, is the adjustment coefficient for balancing the power spectrum density, is the coherent adjustment coefficient, is the adjustment coefficient of the weighted phase delay index.

[0088] determining whether the movement intention index is greater than a first preset threshold, and if so, determining the patient's intention type as a first type;

[0089] determining whether the exercise intention index is within a second preset threshold range, and if so, determining the patient's intention to exercise as a second type;

[0090] Determine whether the movement intention index is less than a third preset threshold; if so, determine it as a third patient intention type.

[0091] One embodiment is:

[0092] After a large number of clinical experiments and analysis of a large number of patient data, the first preset threshold was set at 0.6, the second preset threshold range was 0.3 - 0.6 (including 0.3, excluding 0.6), and the third preset threshold was 0.3.

[0093] When the movement intention index is greater than 0.6, it is determined to be the first patient intention type, defined as "strong active movement intention", which means that the patient has a high willingness and relatively sufficient physical strength to perform autonomous movements of relatively large amplitude and speed.

[0094] When the movement intention index is between 0.3 and 0.6, it is determined to be the second patient intention type, defined as "moderate movement intention", which means that the patient has a certain willingness to exercise but may need partial auxiliary support from the exoskeleton device.

[0095] When the movement intention index is less than 0.3, it is determined to be the third patient intention type, which is defined as "weak movement intention or no obvious movement intention". At this time, the patient may be in a resting state or have only very weak movement intentions.

[0096] Based on this judgment result, the control unit of the exoskeleton rehabilitation device generates corresponding motion parameter adjustment instructions. For "moderate motion intention", the exoskeleton rehabilitation device may appropriately assist the patient's movement, adjust the movement speed of the leg support to a medium speed, and set the movement amplitude to a medium amplitude to match the patient's current motion intention and help the patient conduct more effective rehabilitation training.

[0097] In this way, the exoskeleton rehabilitation device can accurately determine the patient's intention type based on the patient's real-time movement intention index and make corresponding adjustments to achieve personalized and precise rehabilitation training assistance.

[0098] S104: Generate motion parameter adjustment instructions based on the patient's intention type.

[0099] S105, adjusting the movement speed and movement amplitude of the leg support based on the movement parameter adjustment instruction.

[0100] Specifically, in terms of adjusting the movement speed, the control unit will adjust the speed parameter value in the instruction based on the movement parameters to accurately control the speed of the drive motor. By sending a specific pulse signal or analog voltage signal to the motor driver, the motor speed can be finely adjusted. For example, if the instruction requires speeding up the movement of the leg support, the control unit will increase the pulse frequency sent to the motor driver or increase the analog voltage amplitude, thereby increasing the motor speed and driving the leg support to run at a faster speed to match the movement intention conveyed by the patient's brain, such as being able to respond in time when the patient expects to walk quickly or perform rapid leg movements.

[0101] As for the adjustment of the range of motion, the control unit will accurately control the range of motion of the leg support joints according to the amplitude information in the instructions. Advanced sensor feedback is combined with precise mechanical structure design to ensure that the amplitude of the leg support during flexion, extension, rotation and other movements is accurate. For example, when the instruction requires an increase in the range of motion, the control unit will adjust the limit device or control signal of the relevant joint to enable the leg support to move within a larger angle range, which may be manifested as an increase in the leg lifting amplitude, a wider stride, etc., so that patients can perform more natural and coordinated leg movement training according to their own exercise intentions during the rehabilitation training process, further improving the effectiveness and adaptability of rehabilitation training, and better promoting the recovery and reconstruction of patients' nerve and muscle functions.

[0102] In this EEG-based exoskeleton rehabilitation device control method, the leg support's movement speed and amplitude are instantly adjusted based on the patient's real-time EEG signals. During rehabilitation training, the patient's physical condition and neural responses may change at any time. For example, a patient may experience fatigue at a certain moment, and their exercise intentions may change accordingly. By collecting and analyzing EEG signals in real time, the control unit can quickly detect such changes and promptly adjust the exoskeleton's movement parameters to ensure that rehabilitation training is always carried out in optimal conditions.

[0103] This exoskeleton rehabilitation device control method based on EEG signals enables patients to train according to parameters that are consistent with their own movement intentions, can better mobilize the coordination of their own nerves and muscles, and accelerate the rehabilitation process. For example, during the rehabilitation process of a stroke patient, the personalized movement parameters generated according to their EEG signals can allow the patient to perform leg movements within a comfortable and effective range, avoiding overtraining or undertraining, and thus more efficiently recovering the ability to walk. Real-time adaptability helps prevent patients from being injured due to inappropriate movement parameters during rehabilitation training. For example, if a patient suddenly experiences an abnormal condition such as muscle spasm, the EEG signal will change. After detecting it, the control unit will immediately adjust the movement amplitude and speed of the leg support to avoid secondary injury to the patient due to the forced movement of the exoskeleton, thereby ensuring the safety of rehabilitation training.

[0104] This EEG-based exoskeleton rehabilitation device control method combines EEG signals with exoskeleton motion control. By adjusting exoskeleton motion based on movement intent analyzed from EEG signals, the patient's brain, nervous system, and exoskeleton form an organic whole, better promoting the recovery of neurological function. For example, when the patient's brain generates the intention to walk, the exoskeleton can promptly and accurately assist the patient in completing the corresponding movement. This synergistic effect helps strengthen the neural connection between the brain and limbs, accelerating neurological recovery.

[0105] Figure 2 FIG. 1 is a schematic diagram of the second process of an embodiment of the method for controlling an exoskeleton rehabilitation device based on EEG signals of the present invention. Figure 2 As shown, in the second process of the method for controlling an exoskeleton rehabilitation device based on EEG signals in an embodiment of the present invention, steps 1 to 5 are the same as those of the first process described above, except that, after step S105, the second process further includes:

[0106] S106, determining whether there is a difference between the patient's EEG signal and that of a normal person under the same motion state; if so, determining difference data, and adjusting electrical stimulation parameters based on the difference data.

[0107] Specifically, the system records the EEG signals of a healthy individual under different motion states and stores them in a storage unit. After receiving the patient's EEG signal, it first conducts a comprehensive and in-depth comparison and analysis with the patient's EEG signals pre-stored in the storage unit under the same motion state. This process involves precise consideration of multiple characteristic dimensions of the EEG signal, including but not limited to the signal's frequency distribution, amplitude, phase characteristics, and the energy ratio between different frequency bands (such as delta, theta, alpha, beta, and gamma).

[0108] Advanced signal processing algorithms and data analysis techniques, such as spectrum analysis, time-frequency analysis, and classification algorithms in machine learning, are used to meticulously extract and quantitatively evaluate these features. By calculating statistical indicators such as the mean, variance, and standard deviation of various characteristic parameters and rigorously comparing them with the corresponding indicator ranges of normal EEG signals, we can accurately determine whether the patient's EEG signal differs from normal EEG signals.

[0109] Assume that for a certain motion state, the collected patient EEG signal sequence is ,in , is the number of patient EEG signal data points; the normal person EEG signal sequence is , and through data preprocessing (such as interpolation, truncation, etc.) (Ensure that the number of data points is the same for both to facilitate subsequent calculations and comparisons).

[0110] First, calculate the mean of the two, which are respectively and ;

[0111] ;

[0112] ;

[0113] Then calculate the absolute difference of the means as the mean difference part ;

[0114] ;

[0115] Calculate the variance of the patient's EEG signal and the normal person's EEG signal respectively, and record it as and ;

[0116] ;

[0117] ;

[0118] Variance Difference Defined as:

[0119] ;

[0120] Assign weights to mean differences and variance differences separately and , , the weights can be determined based on actual experience or experimental results to obtain comprehensive time domain difference data:

[0121] ;

[0122] First, use a suitable transformation method (such as fast Fourier transform) to transform the time domain EEG signal and Convert to the frequency domain and get the corresponding power spectrum density function and ,in Represents frequency, assuming the frequency range is .

[0123] Calculate the absolute difference in power spectral density at each frequency point:

[0124] ;

[0125] Perform weighted average of the absolute differences at all frequency points (assuming the weight is , which can be set according to factors such as the importance of different frequencies), and obtain frequency domain difference data:

[0126] ;

[0127] Taking into account the above time domain difference data and frequency domain difference data , and then assign them weights and ( ), and obtain the final comprehensive difference data :

[0128] Calculate the difference data using formula (4);

[0129] Formula (4);

[0130] Where, For differential data, is the weight of the time domain difference data, is the time domain difference data, is the weight of the frequency domain difference data, is the frequency domain difference data;

[0131] Through this layer-by-layer constructed formula, the differences between the EEG signals of patients and normal people are quantified from different dimensions (time domain, frequency domain, and a combination of the two), which can more comprehensively and accurately reflect the differences between the two. Then, based on this difference data, the electrical stimulation parameters and other operations can be reasonably adjusted to help exoskeleton rehabilitation equipment better serve the patient's rehabilitation process.

[0132] The adjusted electrical stimulation intensity was calculated by formula (5);

[0133] Formula (5);

[0134] Where, is the adjusted electrical stimulation intensity, is the initial electrical stimulation intensity, is the electrical stimulation intensity adjustment coefficient, For differential data;

[0135] The adjusted electrical stimulation frequency is calculated by formula (6);

[0136] Formula (6);

[0137] Where, is the adjusted electrical stimulation frequency, is the initial electrical stimulation frequency, is the electrical stimulation frequency adjustment coefficient, For differential data;

[0138] The adjusted stimulation duration is calculated by formula (7);

[0139] Formula (7);

[0140] Where, is the adjusted stimulation duration, is the initial stimulation duration, is the stimulus duration adjustment coefficient, For differential data.

[0141] If a discrepancy is determined, further in-depth analysis is conducted to identify the specific discrepancies. The discrepancy data is analyzed in detail for deviations and trends in frequency, amplitude, and phase. For example, the team can determine which frequency bands have experienced abnormal increases or decreases in energy, whether the amplitude fluctuation range exceeds normal thresholds, and whether phase synchronization has been disrupted.

[0142] Based on these precisely determined differential data, the electrical stimulation parameters are then intelligently adjusted. If it is found that the energy of the patient's EEG signal in a certain frequency band is too low, it may indicate that the neural function corresponding to this frequency band is not active enough. In this case, the frequency of electrical stimulation can be appropriately increased to specifically stimulate the neural pathways related to this frequency band and enhance their activity. Or if the amplitude fluctuates too much, it may mean that the neural excitability is unstable. At this time, the intensity or pulse width of the electrical stimulation can be adjusted to stabilize the excitability of the nerves and promote the recovery and balance of neural function. Through this dynamic electrical stimulation parameter adjustment based on differential data, precise regulation of the patient's neural state and personalized rehabilitation treatment can be achieved.

[0143] Figure 3 Schematic diagram of the main components of the exoskeleton rehabilitation device control system based on EEG signals in an embodiment of the present invention. Figure 3 As shown, the exoskeleton rehabilitation equipment control system 1 based on EEG signals provided by an embodiment of the present invention includes an EEG signal acquisition device 10, a multiple signal feature extraction unit 20, an intention type determination unit 30, an adjustment instruction generation unit 40 and a power unit 50.

[0144] The EEG signal acquisition device 10 is used to acquire the patient's EEG signals in real time.

[0145] The multiple signal feature extraction unit 20 is used to convert the EEG signal from the time domain to the frequency domain through short-time Fourier transform, calculate the power spectrum density corresponding to the EEG signal, calculate the coherence and weighted phase delay index based on the power spectrum density, and perform multiple comparisons on the functional connectivity values ​​calculated for each frequency band using the coherence and weighted phase delay index to extract multiple signal features;

[0146] The intention type determination unit 30 is used to input multiple signal features into an intention classification model for intention classification. The intention classification model includes an input layer, a feature extraction layer, a motion intention index calculation layer, a decision layer, and an output layer. The input layer receives multiple signal features, the feature extraction layer extracts the multiple signal features to obtain multiple features required for calculation, the motion intention index calculation layer performs calculation based on the required features to obtain a motion intention index, the decision layer determines the patient's intention type based on the motion intention index, and the output layer outputs the patient's intention type.

[0147] The adjustment instruction generating unit 40 is used to generate a motion parameter adjustment instruction based on the patient's intention type;

[0148] The power unit 50 is used to adjust the movement speed and movement amplitude of the leg support based on the movement parameter adjustment instruction.

[0149] Figure 4A schematic diagram of the physical structure of an electronic device provided in an embodiment of the present invention, such as Figure 4 As shown, the electronic device 60 includes: a processor 601 (processor), a memory 602 (memory) and a bus 603;

[0150] The processor 601 and the memory 602 communicate with each other via the bus 603.

[0151] The processor 601 is configured to call program instructions in the memory 602 to execute the methods provided by the above method embodiments.

[0152] This embodiment provides a non-transitory computer-readable storage medium, which stores computer instructions. The computer instructions enable a computer to execute the methods provided by the above method embodiments.

[0153] Those skilled in the art will understand that all or part of the steps of implementing the above-mentioned method embodiment can be completed by hardware related to program instructions, and the aforementioned program can be stored in a computer-readable storage medium. When the program is executed, it executes the steps of the above-mentioned method embodiment; and the aforementioned storage medium includes: ROM, RAM, disk or optical disk, etc. Various storage media that can store program codes.

[0154] The above specific embodiments do not limit the scope of protection of the present invention. Those skilled in the art will appreciate that various modifications, combinations, sub-combinations, and substitutions may occur depending on design requirements and other factors. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention are intended to be included within the scope of protection of the present invention.

Claims

1. An exoskeleton rehabilitation equipment control system based on EEG signals, characterized in that: include: An electroencephalogram (EEG) signal acquisition device, used to acquire the patient's EEG signals in real time; a multiple signal feature extraction unit, configured to convert the EEG signal from the time domain to the frequency domain by short-time Fourier transform, calculate the power spectral density corresponding to the EEG signal, calculate the coherence and weighted phase delay index based on the power spectral density, perform multiple comparisons on the functional connectivity values ​​calculated for each frequency band using the coherence and weighted phase delay index, and extract multiple signal features; an intention type determination unit, configured to input a plurality of signal features into an intention classification model for intention classification, the intention classification model comprising an input layer, a feature extraction layer, a motion intention index calculation layer, a decision layer, and an output layer, wherein the input layer receives a plurality of signal features, the feature extraction layer extracts the plurality of signal features to obtain a plurality of features required for calculation, the motion intention index calculation layer performs calculation based on the required features to obtain a motion intention index, the decision layer determines a patient's intention type based on the motion intention index, and the output layer outputs the patient's intention type; The step of calculating the movement intention index by the movement intention index calculation layer based on the required calculation features to obtain the movement intention index includes: The patient's movement intention index was calculated using formula (3); Formula (3); Where, is the movement intention index, EEG The frequency band set, EEG The power spectral density, EEG The weight coefficient of is the set of electrode pairs used to calculate coherence, For electrode pairs The coherence value of For electrode pairs The coherence weight of is the set of electrode pairs used to calculate the weighted phase delay index, For electrode pairs The weighted phase delay index of For electrode pairs The weighted phase delay index weight, is the adjustment coefficient for balancing the power spectrum density, is the coherent adjustment coefficient, is the adjustment coefficient of the weighted phase delay index; an adjustment instruction generating unit, configured to generate an exercise parameter adjustment instruction based on the patient's intention type; The power unit is used to adjust the movement speed and movement amplitude of the leg support based on the movement parameter adjustment instruction.

2. The exoskeleton rehabilitation equipment control system based on EEG signals according to claim 1, characterized in that: The calculating of the coherence and weighted phase delay index based on the power spectral density comprises: The coherence is calculated by formula (1); Formula (1); Where, is the cross power spectral density, for The autopower spectral density of the electrode, for The autopower spectral density of the electrode, and Indicates different electrodes; The weighted phase delay index is calculated by formula (2); Formula (2); Where, is a discrete time point, N is the number of data points, k is the frequency, is a symbolic function, is the phase difference. When the phase difference is greater than 0, , when the phase difference is less than 0, , when the phase difference is equal to 0, , not included in the calculation.

3. The exoskeleton rehabilitation equipment control system based on EEG signals according to claim 1, characterized in that: The determining the patient intention type based on the movement intention index by the decision layer includes: determining whether the movement intention index is greater than a first preset threshold, and if so, determining the patient's intention type as a first type; determining whether the exercise intention index is within a second preset threshold range, and if so, determining the patient's intention to exercise as a second type; Determine whether the movement intention index is less than a third preset threshold; if so, determine it as a third patient intention type.

4. The exoskeleton rehabilitation equipment control system based on EEG signals according to claim 1, characterized in that: The exoskeleton rehabilitation device control system based on EEG signals also includes a storage unit and a control unit; The storage unit is used to store EEG signals of normal people in different motion states; The control unit is further configured to: Determine whether there is a difference between the patient's EEG signal and that of a normal person under the same motion state. If so, determine difference data and adjust electrical stimulation parameters based on the difference data.

Images