Cerebral stroke lower limb rehabilitation system integrating functional electrical stimulation and virtual reality scene interaction

Abstract

The invention relates to the technical field of cerebral apoplexy lower limb rehabilitation, and discloses a cerebral apoplexy lower limb rehabilitation system integrating functional electrical stimulation and virtual reality scene interaction. The concentration of sodium ions and chloride ions in sweat is continuously monitored through a data collecting and processing module, and the ion strength change is calculated according to the electrolyte theory; extracting a recruitment inflection point and solving a Debye collapse slope based on a Debye length theory; and generating an interface stability weight as a regulation and control parameter through a mapping module. The stimulation amplitude and the pulse width are dynamically adjusted according to the interface stability weight, the virtual reality module adaptively adjusts the training difficulty, the transcranial magnetic stimulation module achieves accurate phase synchronization, and the assistance intensity is adjusted according to the interface stability weight. According to the invention, multi-mode cooperative closed-loop control is realized, and the safety, stability and intelligent level of rehabilitation training are enhanced.

Description

Technical Field

[0001] This invention relates to the field of lower limb rehabilitation technology for stroke, and more specifically, to a lower limb rehabilitation system for stroke that integrates functional electrical stimulation with virtual reality scenario interaction. Background Technology

[0002] Lower limb rehabilitation for stroke patients typically involves functional electrical stimulation (fEP) to induce muscle contraction and restore motor function. To improve patient engagement and attention, virtual reality (VR) scenario-based tasks have been increasingly incorporated into rehabilitation training in recent years, enabling patients to achieve goal-oriented movements through visual and auditory interaction. However, the high level of immersion and complex tasks in virtual environments often trigger significant sympathetic nerve activation, accompanied by increased skin temperature, sweating, and local ion concentration. Sodium, chloride, and water ions from sweat seep beneath the electrodes, altering the electrolyte environment and double-layer thickness of the skin surface, causing a rapid decrease in the impedance at the electrode-skin interface. Since electrical stimulation systems typically assume stable interface impedance and cannot detect these microscopic environmental abrupt changes in real time, when impedance suddenly decreases, the actual output current exceeds the set value, resulting in a momentary increase in stimulation intensity; conversely, as impedance gradually recovers, the stimulation intensity weakens. This cycle repeats, causing irregular fluctuations in stimulation energy.

[0003] In a closed-loop training system for stroke rehabilitation, functional electrical stimulation signals must maintain strict timing matching with the patient's gait phase, motor intentions, and virtual reality feedback. When the electrode interface impedance changes rapidly, the actual waveform, amplitude, and effective duration of the stimulation pulse deviate from the control command, leading to a misalignment between muscle contraction timing and gait events. Simultaneously, a time difference arises between the patient's action feedback in the virtual environment and the actual mechanical output, disrupting the nervous system's integration process of external stimuli. Over long-term training, these small but frequent mismatches weaken the central nervous system's response to peripheral signals, causing neural remodeling to plateau. Therefore, this problem is not simply hardware noise or mechanical error, but stems from a dynamic imbalance among physiological responses, the electrochemical interface, and control signals. Summary of the Invention

[0004] This invention provides a lower limb rehabilitation system for stroke that integrates functional electrical stimulation and virtual reality scenario interaction, solving the technical problem mentioned in the background art: during immersive rehabilitation training, the rapid changes in the electrical state of the electrode-skin interface lead to unstable functional electrical stimulation output, thereby causing closed-loop control deviation and a decline in training effect.

[0005] This invention provides a lower limb rehabilitation system for stroke patients that integrates functional electrical stimulation and virtual reality scenario interaction, comprising: The data acquisition and processing module continuously acquires the sodium ion concentration and chloride ion concentration in sweat; based on the sodium ion concentration and chloride ion concentration, it calculates the change in ion intensity over time according to electrolyte theory; it determines the recruitment inflection point in the process of ion intensity change, and calculates the Debye collapse slope at the recruitment inflection point based on Debye length correlation theory. The mapping module inputs the Debye collapse slope into a preset single-parameter mapping function to generate interface-stable weights. The functional electrical stimulation modulation module scales the preset base amplitude and preset base pulse width based on interface stability weights at a preset fixed frequency, and ensures that the rate of change of the scaled amplitude and pulse width does not exceed the preset rate of change upper limit. The virtual reality control module performs a callback on the preset virtual reality difficulty scale based on the interface stability weight, and the difficulty scale after the callback is between the preset lower limit and the preset upper limit. The transcranial magnetic stimulation phase modulation module applies a delay to the original phase based on a preset original phase and interface stability weights to obtain the real-time phase, and the delay does not exceed the preset maximum allowable delay; wherein the original phase is generated based on the EEG intention signal and the lower limb gait phase signal. The exoskeleton assist control module incrementally adjusts the preset basic assist coefficient based on the interface stability weight, and the increment of the adjusted assist coefficient does not exceed the preset assist increment limit.

[0006] The beneficial effects of this invention include: by introducing sweat ion monitoring and Debye collapse theory, real-time quantitative assessment of electrode-skin interface stability is achieved, and a multimodal closed-loop control system integrating functional electrical stimulation, virtual reality interaction, transcranial magnetic stimulation, and exoskeleton assistance is established. This system can automatically adjust electrical stimulation parameters, virtual training difficulty, brain stimulation phase, and assistance intensity according to the dynamic changes in the patient's physiological state, thereby significantly reducing stimulation deviations caused by interface impedance mutations, maintaining temporal consistency between the neuromuscular and virtual environment, improving the safety, comfort, and neural remodeling efficiency of rehabilitation training, and ultimately achieving a significant enhancement in lower limb rehabilitation outcomes for stroke patients. Attached Figure Description

[0007] Figure 1 This is a block diagram of the present invention; Figure 2 This is a flowchart of the present invention. Detailed Implementation

[0008] The subject matter described herein will now be discussed with reference to exemplary embodiments. It should be understood that these embodiments are discussed only to enable those skilled in the art to better understand and implement the subject matter described herein, and changes may be made to the function and arrangement of the elements discussed without departing from the scope of this specification. Various processes or components may be omitted, substituted, or added as needed in the examples. Furthermore, features described in some examples may be combined in other examples.

[0009] like Figures 1 to 2 As shown, a lower limb rehabilitation system for stroke patients that integrates functional electrical stimulation and virtual reality scenario interaction includes: The data acquisition and processing module continuously acquires the sodium ion concentration and chloride ion concentration in sweat; based on the sodium ion concentration and chloride ion concentration, it calculates the change in ion intensity over time according to electrolyte theory; it determines the recruitment inflection point in the process of ion intensity change, and calculates the Debye collapse slope at the recruitment inflection point based on Debye length correlation theory. The mapping module inputs the Debye collapse slope into a preset single-parameter mapping function to generate interface-stable weights. The functional electrical stimulation modulation module scales the preset base amplitude and preset base pulse width based on interface stability weights at a preset fixed frequency, and ensures that the rate of change of the scaled amplitude and pulse width does not exceed the preset rate of change upper limit. The virtual reality control module performs a callback on the preset virtual reality difficulty scale based on the interface stability weight, and the difficulty scale after the callback is between the preset lower limit and the preset upper limit. The transcranial magnetic stimulation phase modulation module applies a delay to the original phase based on a preset original phase and interface stability weights to obtain the real-time phase, and the delay does not exceed the preset maximum allowable delay; wherein the original phase is generated based on the EEG intention signal and the lower limb gait phase signal. The exoskeleton assist control module incrementally adjusts the preset basic assist coefficient based on the interface stability weight, and the increment of the adjusted assist coefficient does not exceed the preset assist increment limit.

[0010] In one embodiment of the present invention, the sodium ion concentration and chloride ion concentration of sweat are continuously acquired; based on the sodium ion concentration and chloride ion concentration, the change of ion strength over time is calculated according to electrolyte theory, including: Continuously obtain the sodium ion concentration and chloride ion concentration of sweat; Based on the sodium and chloride ion concentrations in sweat, the change in ionic strength over time was calculated according to electrolyte theory, including: Establish a bounded time set consisting of consecutive time points; On the bounded time set, define a sweat sodium ion concentration function to represent the change of sweat sodium ion concentration with time and a sweat chloride ion concentration function to represent the change of sweat chloride ion concentration with time. For each time point in the bounded time set, the value of the sweat sodium ion concentration function and the value of the sweat chloride ion concentration function corresponding to that time point are added together and then halved to obtain the ion intensity value corresponding to that time point, thus forming the result of the change of ion intensity over time.

[0011] Continuous acquisition of sodium and chloride ion concentrations in sweat refers to the real-time and uninterrupted collection of sodium and chloride ion concentration data in the patient's sweat using wearable sweat sensors and other devices during rehabilitation training, ensuring that the dynamic changes in the concentrations of these two ions over training time can be captured.

[0012] Establish a bounded time set consisting of consecutive time points; where bounded refers to setting a clear time range (such as the duration of a single rehabilitation training session, e.g., 30 minutes), and consecutive time points refer to selecting time nodes at fixed intervals (e.g., once per second) within this time range to form a coherent and uninterrupted time sequence.

[0013] Based on the bounded time set, each time point is associated with the corresponding sodium ion concentration value and chloride ion concentration value, forming two functions (i.e., the correspondence between time and concentration). Through these two correspondences, the changing trends of sodium ion and chloride ion concentrations over time can be presented.

[0014] According to electrolyte theory, sodium and chloride ions in sweat are both monovalent ions (with a charge of ±1), and are the main ions affecting the electrolyte activity of sweat. Therefore, for each time point, the sodium ion concentration value and chloride ion concentration value at that time point are added together and then divided by 2 to obtain the ion intensity value at that time point. This calculation is performed for all time points to obtain a series of ion intensity values ​​that change over time, forming complete data on the change of ion intensity with training time.

[0015] In one embodiment of the present invention, determining the recruitment inflection point during the ion intensity change process, and calculating the Debye collapse slope at the recruitment inflection point based on the Debye length correlation theory, includes: Based on the results of the change of ion intensity over time, the rate of change of ion intensity over time and the time change of the rate of change of ion intensity are calculated. The time point when the rate of change of ion intensity is zero over time and the rate of change of ion intensity over time is greater than zero is defined as the inflection point time of recruitment. Based on electrolyte theory, the change of Debye length over time is defined based on the change of ionic strength over time, and the rate of change of Debye length over time is calculated. By taking the negative of the rate of change of the Debye length with time corresponding to the inflection point of fundraising, we obtain the Debye collapse slope at the inflection point of fundraising.

[0016] Based on the ionic intensity function over the time set. The first derivative of ionic strength is: ;in, This represents the result of how ionic strength changes over time. Represents discrete time within a time set. This indicates the rate of change of ionic strength over time.

[0017] Based on the ionic intensity function over the aforementioned time set The second derivative of ionic strength is defined as follows: ;in, The time-varying quantity that characterizes the rate of change of ion intensity.

[0018] The inflection point in fundraising is For the time set to satisfy and Element.

[0019] Based on the Debye length correlation theory, the Debye length function is defined as follows for the aforementioned time set: ;in, is a constant parameter used to convert the ionic strength function to the Debye length function.

[0020] On the aforementioned time set, based on the Debye length function Define the first derivative of the Debye length as: ;in, Characterizes the rate of change of the Debye length over time.

[0021] The slope of the Debye collapse is defined as: .

[0022] The rate of change of ionic strength over time (i.e., the first derivative) is used to characterize how fast the ionic strength increases or decreases at different times. The time-dependent change in the rate of change of ion intensity (i.e., the second derivative function) is used to characterize the trend of the rate of increase or decrease of ion intensity itself (such as the rate of increase becoming faster, slower, or unchanged).

[0023] Within the entire timeframe, time points that simultaneously meet two conditions are selected as recruitment inflection points. The first condition is that the rate of change of ion intensity over time is zero (i.e., the second derivative is zero), signifying a turning point in the rate of increase or decrease of ion intensity (from accelerating to decelerating, or vice versa). The second condition is that the rate of change of ion intensity over time is greater than zero (i.e., the first derivative is positive), ensuring that this turning point occurs during a phase of overall increase in ion intensity. This time point corresponds to the transition from sporadic activation of sweat glands to coordinated recruitment of a group of sweat glands, characterizing a change in the interfacial environment.

[0024] According to electrolyte theory, the Debye length is a parameter reflecting the charge shielding effect in an electrolyte solution (the higher the ionic strength, the stronger the charge shielding, and the shorter the Debye length). Therefore, by introducing a fixed conversion constant, a function of the Debye length changing with time is constructed, which is determined by the ionic strength at each moment (ionic strength is the independent variable, and Debye length is the dependent variable). Based on this function, the rate of change of the Debye length with time (i.e., the first derivative of the Debye length) is calculated to characterize the rate of increase or decrease of the Debye length at different times.

[0025] The Debye collapse slope quantifies the rate of collapse of the Debye length at the inflection point. Since the ionic strength at the inflection point is in an increasing phase (satisfying a rate of change greater than zero), the Debye length typically shows a shortening trend (i.e., the rate of change of the Debye length over time is negative, indicating that the Debye length is collapsing). Therefore, taking the negative of the rate of change of the Debye length at the inflection point yields the Debye collapse slope. This slope is positive, quantifying the instantaneous rate of collapse of the Debye length at the inflection point. This instantaneous collapse rate is calculated using a conversion constant, the ionic strength at the inflection point, and the rate of change of the ionic strength, ensuring its correlation with the dynamic changes in the interfacial electrochemical environment.

[0026] In one embodiment of the present invention, the Debye collapse slope is input into a preset single-parameter mapping function to generate interface stability weights, including: Determine a preset single-parameter mapping function, which includes a mapping intensity constant that is used to adjust the mapping attenuation intensity and does not change with time; The Debye collapse slope is used as the independent variable of the single-parameter mapping function; The interface stability weights are calculated using this single-parameter mapping function.

[0027] Set the single-parameter mapping function as ;in, It is a single-parameter mapping function. The independent variable of the single-parameter mapping function, is the mapping intensity constant.

[0028] The pre-defined single-parameter mapping function is a fixed function preset according to the rehabilitation training scenario and the characteristics of the electrode-skin interface. A single parameter means the function only requires one input variable to calculate the output. The mapping intensity constant included in the function is a fixed value that does not change with dynamic factors such as training time and ion concentration; its function is to adjust the decay rate of the mapping process. For example, the larger the constant value, the more significant the decay of the output result will be when the input Debye collapse slope changes, thus adapting to the sensitivity requirements of different patients and different training stages for interface stability assessment.

[0029] The Debye collapse slope is an indicator of the rate of double layer collapse at the electrode-skin interface (the larger the slope, the faster the double layer collapses and the more unstable the interface). Substituting it into the function as an input variable ensures that the mapping process is directly related to the microscopic electrochemical state of the interface.

[0030] Based on the characteristics of the pre-defined exponential single-parameter mapping function, its output (interface stability weight) always remains between 0 and 1. When the input Debye collapse slope is small (slow interface double-layer collapse, stable interface), the weight value output by the function is close to 1; when the slope is large (fast interface double-layer collapse, unstable interface), the weight value is close to 0. This transforms the abstract Debye collapse slope into an intuitive and directly controllable stability weight.

[0031] In one embodiment of the present invention, scaling a preset base amplitude and a preset base pulse width based on interface stability weights at a preset fixed frequency includes: A preset fixed frequency constant is set for the output frequency of functional electrical stimulation, and this preset fixed frequency constant remains unchanged throughout the entire modulation process; Set the preset baseline amplitude constant and preset baseline pulse width constant for functional electrical stimulation; Multiply the preset base amplitude constant by the interface stability weight to obtain the real-time amplitude that changes over time. The real-time pulse width, which varies over time, is obtained by multiplying the preset base pulse width constant by the interface stability weight.

[0032] The output frequency of functional electrical stimulation determines the muscle contraction pattern; for example, low frequency induces single twitch, while medium frequency induces tetanic contraction. A fixed frequency constant needs to be set in advance according to the rehabilitation goals for the lower limbs after stroke (such as activating specific muscle groups and restoring gait rhythm). This constant remains unchanged throughout the entire modulation process to avoid frequency changes causing disorder in the muscle contraction mechanism and to ensure the stability of the contraction pattern required for rehabilitation training.

[0033] The preset baseline amplitude and baseline pulse width constant are reference electrical stimulation parameters determined in advance based on the patient's individual condition (such as muscle strength level, skin tolerance, and rehabilitation stage). The baseline amplitude determines the intensity benchmark of the stimulation signal, and the baseline pulse width determines the duration benchmark of the stimulation signal. Together, the preset baseline amplitude and baseline pulse width constant ensure that the initial stimulation energy can effectively activate the target muscle group without causing discomfort to the patient.

[0034] The interface stability weight reflects the stability of the electrode-skin interface (the closer the weight is to 1, the more stable the interface; the closer it is to 0, the less stable the interface). The real-time amplitude is obtained by multiplying the base amplitude constant by the real-time interface stability weight. When the interface is stable, the weight is high, and the real-time amplitude is close to the base value, ensuring that the stimulation intensity meets the standard. When the interface is unstable (e.g., a sudden drop in impedance), the weight is low, and the real-time amplitude decreases accordingly, avoiding excessively high actual stimulation current due to impedance drop, which could cause stinging or muscle overactivation in the patient.

[0035] The real-time pulse width is obtained by multiplying the base pulse width constant by the real-time interface stability weight. The pulse width affects the energy of a single stimulus (the longer the pulse width, the higher the energy). By scaling the pulse width using weights, the total energy of a single stimulus can be controlled in conjunction with the amplitude. When the interface is unstable, the real-time pulse width decreases as the weight decreases, further controlling the stimulus energy and avoiding energy fluctuations caused by changes in interface impedance, thus ensuring stable and safe electrical stimulation output.

[0036] In one embodiment of the present invention, the callback of a preset virtual reality difficulty scale based on interface stability weights includes: Set the batch flags used to mark callback batches and the sampling time points corresponding to each batch flag; Obtain the interface stability weight at each sampling time point; Set a callback gain constant for calculating the difficulty callback magnitude, subtract the interface stability weight corresponding to the sampling time point from 1, and then multiply the result by the callback gain constant to obtain the difficulty callback amount for this batch. Set the current value of the virtual reality difficulty scale for the current batch, and subtract the difficulty callback amount from the current value of the virtual reality difficulty scale to obtain unconstrained virtual reality difficulty scale candidate values; Set a lower limit constant and an upper limit constant for virtual reality difficulty to limit the range of difficulty; If an unconstrained candidate value for the virtual reality difficulty scale is less than the virtual reality difficulty lower limit constant, then the virtual reality difficulty scale for the next batch will be the virtual reality difficulty lower limit constant. If an unconstrained candidate value for the virtual reality difficulty scale is greater than the virtual reality difficulty upper limit constant, then the virtual reality difficulty scale for the next batch will be the virtual reality difficulty upper limit constant. If an unconstrained virtual reality difficulty scale candidate value is greater than or equal to the lower limit constant of virtual reality difficulty and less than or equal to the upper limit constant of virtual reality difficulty, then the virtual reality difficulty scale for the next batch will be taken from the unconstrained virtual reality difficulty scale candidate value.

[0037] To avoid frequent and irregular adjustments to the difficulty of virtual reality, the difficulty adjustment is divided into multiple batches, with batch markers distinguishing different adjustment stages. At the same time, each batch marker corresponds to a sampling time point, which is used to extract the current interface stability weight, ensuring that each batch adjustment has a clear physiological state basis, and achieving orderly and controllable difficulty adjustment.

[0038] At each sampling time point corresponding to each batch, the interface stability weight (which reflects the current stability of the electrode-skin interface) is extracted and used as the basis for difficulty adjustment. The more unstable the interface (the smaller the weight), the greater the subsequent difficulty adjustment needs to be, in order to alleviate the interface fluctuations caused by sympathetic excitation from the source.

[0039] The callback gain constant is a fixed value set in advance to control the sensitivity of the difficulty callback (the larger the gain, the greater the difficulty adjustment under the same weight change); when calculating, first subtract the interface stability weight at the sampling time point from 1 (the smaller the weight, the larger this difference, and the greater the difficulty reduction required), and then multiply the difference by the callback gain constant. The result is the difficulty reduction amount (difficulty callback amount) required for this batch.

[0040] Based on the current batch's virtual reality difficulty scale, the calculated difficulty adjustment amount is subtracted to obtain preliminary candidate values ​​for the next batch's difficulty. These preliminary candidate values ​​are based on theoretical adjustments for interface stability, but do not yet consider constraints related to rehabilitation effectiveness.

[0041] To ensure the effectiveness of rehabilitation training, a lower limit (to avoid the difficulty being too low to achieve the goal of neural remodeling) and an upper limit (to avoid the difficulty being too high to further activate the sympathetic nervous system, aggravate sweating and interface instability) are set to form an effective range for difficulty control.

[0042] The final difficulty is determined based on the relationship between the unconstrained candidate values ​​and the upper and lower limits of difficulty. If the candidate value is lower than the lower limit, the lower limit is used (to ensure that the training is challenging enough); if the candidate value is higher than the upper limit, the upper limit is used (to avoid further deterioration of the interface); if the candidate value is between the upper and lower limits, the candidate value is used directly (balancing interface stability and training effectiveness), ultimately achieving difficulty adjustment based on physiological state and in line with rehabilitation goals.

[0043] In one embodiment of the present invention, based on a preset original phase and interface stability weight, a delay is applied to the original phase to obtain a real-time phase, wherein the delay does not exceed a preset maximum allowable delay, including: Acquire a preset raw phase generated from EEG intention signals and lower limb gait phase signals; Set the phase delay gain constant, which is used to convert the interface stability weight, as a delay amount; Subtract the interface stability weight from 1, and then multiply the result by the phase delay gain constant to obtain the basic value of the delay. Set a preset maximum allowable delay constant to limit the range of delay amounts; If the base value of the delay is greater than the preset maximum allowable delay constant, the actual delay shall be the preset maximum allowable delay constant. If the base value of the delay is less than or equal to the preset maximum allowable delay constant, then the actual delay is taken as the base value of the delay. The real-time phase is obtained by adding the preset original phase to the actual delay.

[0044] The primitive phase is the initial timing reference for transcranial magnetic stimulation, generated by two types of signals: EEG signals that reflect the patient’s motor intentions (ensuring that the stimulation matches the patient’s intention to move actively).

[0045] Gait phase signals characterizing the rhythm of lower limb movement (ensuring that stimuli match the gait cycle).

[0046] The combination of EEG signals and gait phase signals enables the original phase to accurately correspond to the neuromotor needs of rehabilitation training.

[0047] Setting the phase delay gain constant is a fixed parameter that is set in advance. Its function is to establish the conversion relationship between the interface stability weight and the specific delay amount, and to control the sensitivity of the delay amount to the weight change (e.g., the larger the constant, the greater the adjustment range of the delay amount under the same weight change).

[0048] The smaller the interface stability weight (the less unstable the interface), the larger the difference between 1 and the weight, indicating that the stimulus needs to be delayed more significantly to avoid the unstable window; multiplying this difference by the phase delay gain constant, the resulting base value of the delay is the theoretical delay magnitude calculated based on the current interface stability.

[0049] To avoid excessive delay causing a complete disconnect between the transcranial magnetic stimulation and the patient's movement intentions and gait phase (disrupting the temporal coordination of rehabilitation training), an upper limit for the delay needs to be set to ensure that the delay is always within a reasonable range that can adapt to the interface stably and does not deviate from the training requirements.

[0050] If the base value of the delay exceeds the maximum allowable delay, the maximum value is taken (to prevent excessive delay from affecting training); if it does not exceed the maximum value, the base value is used directly (to accurately match the interface stability requirements). This judgment balances the interface stability control and the effectiveness of training timing.

[0051] The preset original phase is added to the actual delay to obtain the adjusted real-time phase. This allows the triggering time of transcranial magnetic stimulation to dynamically shift with the stability of the interface, avoiding periods of interface instability and preventing closed-loop oscillations caused by the superposition of central stimulation and peripheral interface abrupt changes, thus ensuring that the stimulation sequence is adapted to the interface state.

[0052] In one embodiment of the present invention, incremental adjustment of a preset basic assist coefficient based on interface stability weight includes: Set the preset base assist coefficient for the exoskeleton, the assist gain constant for converting interface stability weights into assist increments, and the preset assist increment upper limit for limiting the range of assist increments; Subtract the interface stability weight from 1, and then multiply the result by the boost gain constant to obtain the boost coefficient increment. If the increase in the assist coefficient is greater than the preset upper limit of the assist increment, then the preset upper limit of the assist increment is taken as the assist coefficient increase after the limit is applied. If the increase in the assist coefficient is less than or equal to the preset upper limit of the assist coefficient, then the increase in the assist coefficient is taken as the increase in the assist coefficient after the limit is applied. The real-time assist coefficient of the exoskeleton is obtained by adding the preset basic assist coefficient to the assist coefficient increment after the amplitude limit.

[0053] A preset baseline assistance coefficient is set as the benchmark value for exoskeleton assistance. It is set in advance based on the patient's muscle strength level and rehabilitation stage to ensure that the initial assistance can match the needs of routine training. The assist gain constant is a fixed conversion parameter used to control the sensitivity of the conversion from interface stability weight to assist increment (the larger the constant, the more obvious the adjustment of assist increment under the same weight change). The preset upper limit for assist increment is the maximum limit for assist increment, which avoids excessive assistance causing patients to become dependent on the exoskeleton or affecting their active movement ability.

[0054] The smaller the interface stability weight (the less unstable the electrode-skin interface), the larger the difference between 1 and this weight, which means that the patient's spontaneous muscle metabolic burden needs to be reduced by increasing the exoskeleton assistance (indirectly reducing sweating and helping to stabilize the interface); multiplying this difference by the assistance gain constant, the result is the additional assistance that the exoskeleton needs to provide (assistance coefficient increment).

[0055] If the calculated increase in the assist coefficient exceeds the preset upper limit of the assist increase, the upper limit is taken as the final increase (to prevent excessive assistance and ensure the patient's active participation in movement); if the increase does not exceed the upper limit, the calculated increase is used directly (to accurately match the stability requirements of the interface) to ensure that the assist increase is within a safe and effective range.

[0056] The real-time assist coefficient is obtained by adding the preset base assist coefficient to the increment of the assist coefficient after the amplitude limit. When the interface is unstable, the real-time assist increases with the increment, helping patients reduce exercise burden and reduce sweating, indirectly improving the interface environment; when the interface is stable, the increment is small, and the real-time assist is close to the base value, ensuring that patients can carry out normal active training and achieving the adaptation of assist adjustment to interface status and rehabilitation needs.

[0057] The embodiments of this example have been described above. However, this example is not limited to the specific implementation methods described above. The specific implementation methods described above are merely illustrative and not restrictive. Those skilled in the art can make many other forms based on the guidance of this example, and all of them are within the protection scope of this example.

Claims

1. A lower limb rehabilitation system for stroke patients that integrates functional electrical stimulation and virtual reality scenario interaction, characterized in that, include: The data acquisition and processing module continuously acquires the sodium ion concentration and chloride ion concentration in sweat; based on the sodium ion concentration and chloride ion concentration, it calculates the change of ion strength over time according to electrolyte theory. The recruitment inflection point during the change of ion intensity was determined, and the Debye collapse slope at the recruitment inflection point was calculated based on the Debye length correlation theory. The mapping module inputs the Debye collapse slope into a preset single-parameter mapping function to generate interface-stable weights. The functional electrical stimulation modulation module scales the preset base amplitude and preset base pulse width based on interface stability weights at a preset fixed frequency, and ensures that the rate of change of the scaled amplitude and pulse width does not exceed the preset rate of change upper limit. The virtual reality control module performs a callback on the preset virtual reality difficulty scale based on the interface stability weight, and the difficulty scale after the callback is between the preset lower limit and the preset upper limit. The transcranial magnetic stimulation phase modulation module applies a delay to the original phase based on a preset original phase and interface stability weights to obtain the real-time phase, and the delay does not exceed the preset maximum allowable delay; wherein the original phase is generated based on the EEG intention signal and the lower limb gait phase signal. The exoskeleton assist control module incrementally adjusts the preset basic assist coefficient based on the interface stability weight, and the increment of the adjusted assist coefficient does not exceed the preset assist increment limit.

2. The stroke lower limb rehabilitation system integrating functional electrical stimulation and virtual reality scenario interaction as described in claim 1, characterized in that, Continuously obtain the sodium ion concentration and chloride ion concentration in sweat; based on the sodium ion concentration and chloride ion concentration, calculate the change in ionic strength over time according to electrolyte theory, including: Continuously obtain the sodium ion concentration and chloride ion concentration of sweat; Based on the sodium and chloride ion concentrations in sweat, the change in ionic strength over time was calculated according to electrolyte theory, including: Establish a bounded time set consisting of consecutive time points; On the bounded time set, define a sweat sodium ion concentration function to represent the change of sweat sodium ion concentration with time and a sweat chloride ion concentration function to represent the change of sweat chloride ion concentration with time. For each time point in the bounded time set, the value of the sweat sodium ion concentration function and the value of the sweat chloride ion concentration function corresponding to that time point are added together and then halved to obtain the ion intensity value corresponding to that time point, thus forming the result of the change of ion intensity over time.

3. The stroke lower limb rehabilitation system integrating functional electrical stimulation and virtual reality scenario interaction according to claim 2, characterized in that, The recruitment inflection point during the ion intensity change process was determined, and the Debye collapse slope at the recruitment inflection point was calculated based on the Debye length correlation theory, including: Based on the results of the change of ion intensity over time, the rate of change of ion intensity over time and the time change of the rate of change of ion intensity are calculated. The time point when the rate of change of ion intensity is zero over time and the rate of change of ion intensity over time is greater than zero is defined as the inflection point time of recruitment. Based on electrolyte theory, the change of Debye length over time is defined based on the change of ionic strength over time, and the rate of change of Debye length over time is calculated. By taking the negative of the rate of change of the Debye length with time corresponding to the inflection point of fundraising, we obtain the Debye collapse slope at the inflection point of fundraising.

4. The stroke lower limb rehabilitation system integrating functional electrical stimulation and virtual reality scenario interaction as described in claim 3, characterized in that, Input the Debye collapse slope into a preset single-parameter mapping function to generate interface stability weights, including: Determine a preset single-parameter mapping function, which includes a mapping intensity constant that is used to adjust the mapping attenuation intensity and does not change with time; The Debye collapse slope is used as the independent variable of the single-parameter mapping function; The interface stability weights are calculated using this single-parameter mapping function.

5. A lower limb rehabilitation system for stroke patients integrating functional electrical stimulation and virtual reality scenario interaction as described in claim 4, characterized in that, At a preset fixed frequency, the preset base amplitude and preset base pulse width are scaled based on interface stability weights, including: A preset fixed frequency constant is set for the output frequency of functional electrical stimulation, and this preset fixed frequency constant remains unchanged throughout the entire modulation process; Set the preset baseline amplitude constant and preset baseline pulse width constant for functional electrical stimulation; Multiply the preset base amplitude constant by the interface stability weight to obtain the real-time amplitude that changes over time. The real-time pulse width, which varies over time, is obtained by multiplying the preset base pulse width constant by the interface stability weight.

6. A lower limb rehabilitation system for stroke patients integrating functional electrical stimulation and virtual reality scenario interaction according to claim 5, characterized in that, The preset virtual reality difficulty scale is callback-based based on interface stability weights, including: Set the batch flags used to mark callback batches and the sampling time points corresponding to each batch flag; Obtain the interface stability weight at each sampling time point; Set a callback gain constant for calculating the difficulty callback magnitude, subtract the interface stability weight corresponding to the sampling time point from 1, and then multiply the result by the callback gain constant to obtain the difficulty callback amount for this batch. Set the current value of the virtual reality difficulty scale for the current batch, and subtract the difficulty callback amount from the current value of the virtual reality difficulty scale to obtain unconstrained virtual reality difficulty scale candidate values; Set a lower limit constant and an upper limit constant for virtual reality difficulty to limit the range of difficulty; If an unconstrained candidate value for the virtual reality difficulty scale is less than the virtual reality difficulty lower limit constant, then the virtual reality difficulty scale for the next batch will be the virtual reality difficulty lower limit constant. If an unconstrained candidate value for the virtual reality difficulty scale is greater than the virtual reality difficulty upper limit constant, then the virtual reality difficulty scale for the next batch will be the virtual reality difficulty upper limit constant. If an unconstrained virtual reality difficulty scale candidate value is greater than or equal to the lower limit constant of virtual reality difficulty and less than or equal to the upper limit constant of virtual reality difficulty, then the next batch of virtual reality difficulty scales will use the unconstrained virtual reality difficulty scale candidate value.

7. A lower limb rehabilitation system for stroke patients integrating functional electrical stimulation and virtual reality scenario interaction according to claim 6, characterized in that, Based on a preset original phase and interface stability weights, a delay is applied to the original phase to obtain the real-time phase, and the delay does not exceed a preset maximum allowable delay, including: Acquire a preset raw phase generated from EEG intention signals and lower limb gait phase signals; Set the phase delay gain constant, which is used to convert the interface stability weight, as a delay amount; Subtract the interface stability weight from 1, and then multiply the result by the phase delay gain constant to obtain the basic value of the delay. Set a preset maximum allowable delay constant to limit the range of delay amounts; If the base value of the delay is greater than the preset maximum allowable delay constant, the actual delay shall be the preset maximum allowable delay constant. If the base value of the delay is less than or equal to the preset maximum allowable delay constant, then the actual delay is taken as the base value of the delay. The real-time phase is obtained by adding the preset original phase to the actual delay.

8. A lower limb rehabilitation system for stroke patients integrating functional electrical stimulation and virtual reality scenario interaction according to claim 7, characterized in that, Incremental adjustments are made to the preset base assist coefficient based on interface stability weights, including: Set the preset base assist coefficient for the exoskeleton, the assist gain constant for converting interface stability weights into assist increments, and the preset assist increment upper limit for limiting the range of assist increments; Subtract the interface stability weight from 1, and then multiply the result by the boost gain constant to obtain the boost coefficient increment. If the increase in the assist coefficient is greater than the preset upper limit of the assist increment, then the preset upper limit of the assist increment is taken as the assist coefficient increase after the limit is applied. If the increase in the assist coefficient is less than or equal to the preset upper limit of the assist coefficient, then the increase in the assist coefficient is taken as the increase in the assist coefficient after the limit is applied. The real-time assist coefficient of the exoskeleton is obtained by adding the preset basic assist coefficient to the assist coefficient increment after the amplitude limit.

Images