Lower limb passive rehabilitation training system, method and device

Abstract

The invention discloses a lower limb passive rehabilitation training system, method and device. Relates to the technical field of rehabilitation medical instruments. The system is characterized in that a resting signal acquisition module acquires a first electroencephalogram signal of a prefrontal lobe area of a patient in a resting state within a first preset duration; the activation signal acquisition module acquires a second electroencephalogram signal in a second preset duration of a prefrontal lobe area in an activated state of the patient; a feature extraction module performs concentration feature extraction on the first electroencephalogram signal and the second electroencephalogram signal, and calculates a resting state concentration index and an activated state concentration index; the change rate calculation module calculates the activation resting relative change rate of the patient; the parameter adjusting module adjusts training parameters of the motion execution equipment according to the activation resting relative change rate and a preset change rate threshold value; the rehabilitation control module sends a control instruction to the motion execution device according to the adjusted training parameters, and passive rehabilitation training of the lower limbs is achieved. Active rehabilitation of early and middle-stage patients can be achieved.

Description

Technical Field

[0001] This application relates to the field of rehabilitation medical device technology, and in particular to a passive rehabilitation training system, method and device for the lower limbs. Background Technology

[0002] Early rehabilitation for paralyzed patients (such as those with lower limb paralysis due to stroke) is crucial for preventing complications such as muscle atrophy, joint contractures, and venous thrombosis. Its core objective is to stimulate nerve remodeling and maintain joint range of motion through continuous passive lower limb movement. However, early-stage paralyzed patients often lack voluntary limb movement. Traditional rehabilitation relies on manual assistance to complete lower limb flexion and extension, making it difficult to standardize training angles and frequencies. Furthermore, limited medical resources prevent meeting patients' needs for multiple daily training sessions. Patients are often in a passive state, lacking conscious participation, resulting in low efficiency in restoring nerve-muscle coordination and impacting the rehabilitation cycle.

[0003] In the prior art, taking the published patent CN117503551A (a brain-controlled lower limb rehabilitation robot) as an example, automated training of the lower limbs of paralyzed patients is achieved through a motor-driven exoskeleton structure. This solution includes an EEG cap and a lower limb robot. The EEG cap collects the patient's movement intention signals, and the lower limb robot simulates normal gait through hip, knee, and ankle drive mechanisms, thus replacing manual assistance to a certain extent and solving the problems of training standardization and efficiency. Simultaneously, it achieves basic linkage between intention and action through EEG signals.

[0004] However, the existing technology mainly focuses on recognizing motor intentions in EEG signals. But early paralysis patients often have blurred consciousness and impaired cognitive function, making it difficult for them to generate clear subjective motor consciousness. As a result, the existing technology cannot capture effective motor intention signals, nor can it drive training devices through these signals. This renders brain control function ineffective in early rehabilitation scenarios, and passive training still needs to be completed by relying on preset programs. Summary of the Invention

[0005] This application provides a passive rehabilitation training system, method, and device for the lower limbs. By collecting data on the patient's attentional characteristics, this application achieves safe and precise passive rehabilitation for early-stage paralyzed patients who are unconscious and unable to generate motor intentions.

[0006] To achieve the above objectives, the technical solutions adopted in the embodiments of this application are as follows: In a first aspect, embodiments of this application provide a lower limb passive rehabilitation training system, comprising: The resting signal acquisition module is used to acquire the first EEG signal in the prefrontal cortex region of the patient in a resting state within a first preset duration; the resting state is the first physiological state in which the patient's whole body muscles are relaxed, and the duration of maintaining the first physiological state in a preset quiet environment is greater than the first preset duration. The activation signal acquisition module is used to acquire the second EEG signal in the prefrontal cortex region within a second preset duration in the patient's activated state; wherein, the second preset duration is less than the first preset duration; the activated state is the second physiological state in which the patient is in a preset quiet environment and is in the process of passive rehabilitation training of the lower limbs, driven by the motor execution module to perform passive flexion and extension movements of the lower limbs. The feature extraction module is used to extract focus features from the first and second EEG signals respectively, and calculate the resting state focus index and the active state focus index based on the extracted focus features; the focus features include: alpha brain waves and beta brain waves; The rate of change calculation module is used to calculate the relative rate of change between active and resting states of a patient's attention based on the resting state attention index and the active state attention index; the relative rate of change between active and resting states is used to characterize the proportion of change of the active state attention index relative to the resting state attention index. The parameter adjustment module is used to adjust the training parameters of the motion execution device based on the activation-resting relative rate of change and a preset rate of change threshold; the training parameters include: flexion-extension speed and flexion-extension angle range; The rehabilitation control module is used to send control commands to the motion execution device based on the adjusted training parameters, thereby enabling the patient to complete passive rehabilitation training of the lower limbs.

[0007] Optionally, the feature extraction module includes: a preprocessing submodule and a feature extraction submodule; The preprocessing submodule is used to preprocess the first EEG signal and the second EEG signal respectively, and output the preprocessed resting signal and the preprocessed activation signal. The feature extraction submodule is used to perform focused feature extraction on the preprocessed resting signal and the preprocessed activated signal, respectively.

[0008] Optionally, the preprocessing submodule includes: a baseline drift removal unit, an artifact removal unit, and a filtering unit; The baseline drift removal unit is used to perform baseline drift removal processing on the first EEG signal and the second EEG signal respectively, and output the first resting signal and the first activation signal. The artifact removal unit is used to perform artifact removal processing on the first resting signal and the first activation signal respectively, and output the second resting signal and the second activation signal. The filtering unit is used to filter the second resting signal and the second activation signal respectively, and output the preprocessed resting signal and the preprocessed activation signal.

[0009] Optionally, the feature extraction submodule includes: an alpha brainwave extraction unit and a beta brainwave extraction unit; The alpha brainwave extraction unit is used to extract features from the preprocessed resting signal and the preprocessed activation signal using the first preset frequency band, and output the resting alpha brainwave and the activation alpha brainwave. The beta brainwave extraction unit is used to extract features from the preprocessed resting signal and the preprocessed activation signal using a second preset frequency band, and output resting beta brainwaves and activation beta brainwaves; wherein, the first preset frequency band is smaller than the second preset frequency band.

[0010] Optionally, the alpha brainwave extraction unit includes: a discrete processing unit, a weighted processing unit, a Fourier transform unit, a power spectral density calculation unit, and a brainwave determination unit; The discrete processing unit is used to perform discrete processing on the preprocessed resting signal and the preprocessed activation signal according to a preset unit time window, and output the first discrete resting signal and the first discrete activation signal. The weighting processing unit is used to multiply the first discrete resting signal and the first discrete activation signal by a preset Hanning window function, and output the first weighted resting signal and the first weighted activation signal, respectively. The Fourier transform unit is used to perform a preset Fourier transform on the first weighted resting signal and the first weighted activated signal respectively, and output the first resting frequency domain sequence and the first activated frequency domain sequence. The power spectral density calculation unit is used to calculate the first resting frequency domain sequence and the first active frequency domain sequence respectively, and output the first resting power spectral density and the first active power spectral density. The brainwave determination unit is used to determine the resting alpha brainwave and the activated alpha brainwave based on the first resting power spectral density and the first activated power spectral density.

[0011] Optionally, the preset rate of change thresholds include: a positive threshold, a stable threshold, and a negative threshold; The parameter adjustment module includes: an adjustment module; the adjustment module is used to receive the active resting relative change rate transmitted by the change rate calculation module; when the active resting relative change rate is greater than or equal to the positive threshold, it outputs the speed increase and angle expansion parameter adjustment instructions to increase the lower limb flexion and extension speed of the motion execution device by a preset speed increase value on the current flexion and extension speed, and at the same time expand the flexion and extension angle range by a preset angle expansion value. When the activation-resting relative rate of change is at a stable threshold, the output command is to maintain the current parameters so as to keep the current training parameters unchanged. When the activation resting relative change rate is less than or equal to the negative threshold, a deceleration and angle parameter adjustment command is output to reduce the lower limb flexion and extension speed of the motion execution device by a preset deceleration value from the current flexion and extension speed, while keeping the flexion and extension angle range unchanged.

[0012] Optionally, the parameter adjustment module also includes a pause / restart submodule; The pause / restart submodule is used to continuously receive the relative change rate of the active and resting states. When the parameter adjustment module outputs deceleration and hold angle parameter adjustment commands, if the relative change rate of activation and resting time corresponding to a consecutive preset number of preset unit time windows is less than or equal to the negative threshold, the pause restart submodule sends a pause and restart control command to the rehabilitation control module.

[0013] Secondly, embodiments of this application provide a passive rehabilitation training method for the lower limbs, applied to any of the passive rehabilitation training systems for the lower limbs described in the first aspect, comprising: Acquire the first EEG signal in the prefrontal cortex region of the patient in a resting state within a first preset duration; the resting state is the first physiological state in which the patient's whole body muscles are relaxed, and the duration of maintaining the first physiological state in a preset quiet environment is greater than the first preset duration. Acquire the second EEG signal in the prefrontal cortex region of the patient in the activated state within a second preset duration, wherein the second preset duration is less than the first preset duration; the activated state is the second physiological state in which the patient is in a preset quiet environment and is in the process of passive rehabilitation training of the lower limbs, and is driven by the motor execution module to perform passive flexion and extension movements of the lower limbs. Attention features were extracted from the first and second EEG signals, respectively, and the resting state attention index and the active state attention index were calculated based on the extracted attention features. Attention features included alpha brain waves and beta brain waves. The relative change rate between active and resting states of attention was calculated based on the resting state attention index and the active state attention index. The training parameters of the motion execution device are adjusted based on the activation-resting relative rate of change and the preset rate of change threshold. The training parameters include: flexion-extension speed and flexion-extension angle range. Based on the adjusted training parameters, control commands are sent to the motion execution device to guide the patient through passive rehabilitation training of the lower limbs.

[0014] Thirdly, this application provides a lower limb passive rehabilitation training device, which integrates any of the lower limb passive rehabilitation training systems described in the first aspect. The lower limb passive rehabilitation training device includes: a headband acquisition device, a control box, and a motion execution device; the motion execution device includes: a fixed drive rod, a drive assembly, and a passive moving rod. The headband data acquisition device communicates with the control box. The control box is electrically connected to the drive assembly; One end of the drive assembly is hinged to the fixed drive rod, and the other end of the drive assembly is hinged to the passive movable rod to form a three-bar linkage structure. The fixed drive rod is used to fix the device to the preset nursing equipment. The passive moving rod has a medical strap at the end away from the drive assembly to fix the patient's lower leg. The fixed drive rod has a medical strap at the other end away from the drive assembly to fix the patient's thigh.

[0015] Optionally, the control box includes: Memory containing executable program code; and memory-coupled processors; The processor calls the executable program code stored in memory to execute the passive rehabilitation training method for the lower limbs, as described in the second aspect.

[0016] Compared with the prior art, this application has the following beneficial effects: This application relates to a lower limb passive rehabilitation training system, method, and device. It falls within the field of rehabilitation medical device technology. The system comprises a resting signal acquisition module, an activation signal acquisition module, a feature extraction module, a rate of change calculation module, a parameter adjustment module, and a rehabilitation control module. First, the resting signal acquisition module acquires the first EEG signal from the prefrontal cortex region within a first preset duration in the patient's resting state. The activation signal acquisition module acquires the second EEG signal from the prefrontal cortex region within a second preset duration in the patient's activated state. By acquiring EEG signals corresponding to two physiological states in the same preset quiet environment, effective signals can be generated even if the patient's consciousness is impaired, without requiring active motor intention. Simultaneously, the identical environment avoids EEG signal distortion, fundamentally solving the problem of failed motor intention signal capture and improving the adaptability of the lower limb passive rehabilitation training system to different patients. Second, the feature extraction module extracts focus features from the first and second EEG signals respectively, ensuring that the extracted focus features accurately reflect the alpha and beta brainwaves of the patient's physiological state, rather than relying on brainwave features dependent on motor intention. Based on the extracted focus features, the resting state focus index and the activated state focus index are calculated respectively, transforming the abstract EEG signal into a quantifiable indicator. This system addresses the analytical challenge of extracting movement intention-related features. Furthermore, the rate of change calculation module calculates the relative rate of change between the patient's active and resting states based on the resting state attention index and the active state attention index. This quantifies the fluctuation of the patient's current state relative to their resting state baseline, allowing for objective assessment without patient feedback and filling the gap in existing technologies that cannot quantify the state of unconscious patients. Simultaneously, the parameter adjustment module adjusts the training parameters of the motion execution device based on the relative rate of change between the active and resting states and a preset rate of change threshold. This automatically outputs control commands based on the relative rate of change between the active and resting states and the preset rate of change threshold, dynamically adapting to the patient's real-time physiological state. This eliminates reliance on preset programs, avoids discomfort caused by muscle strength fluctuations and tolerance differences in the early stages, reduces the risk of joint injury and muscle soreness, and improves the safety and comfort of rehabilitation training. Finally, the rehabilitation control module sends control commands to the motion execution device based on the adjusted training parameters, guiding the patient to complete passive lower limb rehabilitation training. This ensures smooth and controllable passive flexion and extension movements, meeting the needs of early and mid-stage rehabilitation patients with limited joint mobility and requiring precise training.Therefore, the lower limb passive rehabilitation training system of this application is realized by relying on the hardware of the lower limb passive rehabilitation training device. The data transmission between each module adopts a standardized interface, with low transmission latency and high stability. Moreover, the lower limb passive rehabilitation training system operates automatically throughout the entire process. Data collection, analysis, parameter adjustment, and execution do not require active operation by the patient, which also reduces the real-time intervention cost of medical staff. It can also be adapted to various scenarios such as bedside and wheelchair, making it convenient for patients to use for rehabilitation at home or in the hospital. It completely fills the application gap of existing brain control technology in early rehabilitation scenarios and truly realizes safe and precise passive rehabilitation for early paralyzed patients with blurred consciousness and inability to generate motor intentions. Attached Figure Description

[0017] To more clearly illustrate the embodiments of this application, the accompanying drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0018] Figure 1 A schematic diagram of the structure of a lower limb passive rehabilitation training device provided in this application embodiment. Figure 1 ; Figure 2 A schematic diagram of the structure of a lower limb passive rehabilitation training device provided in this application embodiment. Figure 2 ; Figure 3 A schematic diagram of the structure of a lower limb passive rehabilitation training system provided in this application embodiment. Figure 1 ; Figure 4 A schematic diagram of the structure of a lower limb passive rehabilitation training system provided in this application embodiment. Figure 2 ; Figure 5 A schematic diagram of the structure of a lower limb passive rehabilitation training system provided in this application embodiment. Figure 3 ; Figure 6 This is a schematic diagram of the structure of a control box provided in an embodiment of this application; Figure 7 This is a flowchart illustrating a passive rehabilitation training method for the lower limbs provided in an embodiment of this application. Detailed Implementation

[0019] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those of ordinary skill in the art without creative effort are within the protection scope of this application.

[0020] It should be noted that, in the description of this application, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or apparatus. The terms "first," "second," etc., in this application are used to distinguish similar objects and are not used to describe a specific order or sequence.

[0021] To enable those skilled in the art to better understand the present application, the present application will be further described in detail below with reference to the accompanying drawings and specific embodiments.

[0022] The passive rehabilitation training system, method, and apparatus for the lower limbs according to this application will be described in detail below with reference to the accompanying drawings and through multiple embodiments.

[0023] Figure 1 A schematic diagram of the structure of a lower limb passive rehabilitation training device provided in this application embodiment. Figure 1 .like Figure 1 As shown, the lower limb passive rehabilitation training device 100 includes: a headband data acquisition device 110, a control box 120, and a motion execution device 130.

[0024] The motion execution device 130 includes: a fixed drive rod 131, a drive assembly 132, and a passive moving rod 133.

[0025] The headband acquisition device 110 and the control box 120 are connected wirelessly / wired to collect EEG signals (including resting and active state signals) from the patient's prefrontal cortex in real time and transmit them to the control box 120. The headband acquisition device 110, which collects EEG information, integrates a flexible electrode array placed in the FP1 area of ​​the patient's prefrontal cortex (the outer third of the left frontal pole), a key brain region connecting the prefrontal and temporal lobes. This brain region integrates emotion, memory, and social cognition (such as empathy and interpersonal judgment) and is closely related to emotion regulation and social behavioral decision-making.

[0026] The control box 120 is electrically connected to the drive assembly 132. After processing the EEG signals received from the headband acquisition device 110, the control box 120 generates control commands (such as flexion / extension speed commands, angle adjustment commands, etc.) and transmits them to the drive assembly 132. One end of the drive assembly 132 is hinged to the fixed drive rod 131, and the other end is hinged to the passive movable rod 133, forming a three-bar linkage structure to drive the patient's lower limbs to passively flex and extend in response to the control commands from the control box 120. The drive assembly 132 can consist of a motor, an angle sensor, and a speed sensor. The output of the motor is connected to the angle sensor and the speed sensor, respectively. Both the angle sensor and the speed sensor are communicatively connected to the control box 120 for real-time feedback of the angle and speed information of the lower limb flexion and extension. The motor model can be flexibly selected according to actual rehabilitation training needs; for example, an SG90 servo motor can be selected to meet the requirements of precise drive.

[0027] One end of the fixed drive rod 131 is fixedly connected to the preset nursing device to provide a support reference for the entire motion execution device 130. The other end of the fixed drive rod 131 away from the drive assembly 132 is provided with a medical strap to fix the patient's thigh, limit the displacement of the patient's thigh, and ensure the stability of the lower limb posture during rehabilitation training.

[0028] The passive moving rod 133 has a medical strap at the end away from the drive component 132 to fix the patient's lower leg. Under the drive of the drive component 132, it rotates around the hinge point to achieve passive flexion and extension of the lower leg, thereby promoting the overall rehabilitation training of the lower limb. The medical strap is set at the fixation positions of the thigh and lower leg respectively, taking into account both the reliability of fixation and the comfort of wearing, and avoiding the lower limb from sliding or being injured by pressure during rehabilitation training.

[0029] The preset nursing equipment can be selected according to the actual situation. For example, the preset nursing equipment can be a bed or a wheelchair.

[0030] For example, Figure 2 A schematic diagram of the structure of a lower limb passive rehabilitation training device provided in this application embodiment. Figure 2 .like Figure 2 As shown, the flexible electrode array on the headband acquisition device 110 collects electroencephalogram (EEG) signals from the patient's prefrontal cortex, providing data input for the focus feature extraction of each module of the lower limb passive rehabilitation training system 200. The data collected by the headband acquisition device 110 is transmitted to the control box 120, processed by the control box 120, and then output to the drive assembly 132. The drive assembly 132 drives the fixed drive rod 131 and the passive moving rod 133 to realize the patient's lower limb rehabilitation training movements.

[0031] The fixed drive rod 131 is fixedly connected to the preset bed frame and simultaneously secures the patient's thigh with a medical strap (not shown in the figure but its function is implicit), providing dual stable support for the motion execution device 130 and the lower limb. The drive assembly 132 is electrically connected to the control box 120 and is used to drive the fixed drive rod 131 and the passive moving rod 133 to move after receiving control commands. One end of the passive moving rod 133 is hinged to the drive assembly 132, and the other end is secured to the patient's lower leg with a medical strap. Under the drive of the drive assembly 132, it rotates around the hinge point to achieve passive flexion and extension of the lower leg, completing the lower limb rehabilitation training movements.

[0032] The drive assembly 132, the fixed drive rod 131, and the passive moving rod 133 constitute a three-bar linkage structure, which ensures that the patient's lower limb movement trajectory is stable and controllable, and meets the precise drive requirements of passive rehabilitation.

[0033] It should be noted that the lower limb passive rehabilitation training device provided in this application is suitable for the early and middle stages of rehabilitation for patients, that is, when patients are in the early and middle stages of rehabilitation where muscle strength has not been fully restored and they need to rely on passive training to maintain joint mobility. It also has flexible training modes: it supports single-leg independent training, that is, for patients with unilateral lower limb dysfunction, only one side of the passive movement bar is driven to complete the passive flexion and extension of one leg, realizing individualized unilateral rehabilitation; it can also realize bilateral synchronous training, that is, drive both sides of the passive movement bar at the same time to meet the collaborative rehabilitation needs of both lower limbs, promote the symmetrical recovery of limb function, promote blood circulation, prevent the formation of lower limb deep vein thrombosis, fully adapt to the individualized rehabilitation scenarios of different patients, and effectively improve the applicability and rehabilitation training efficiency of the device.

[0034] In addition, it should be noted that the fixed drive rod 131 is fixedly connected to the preset bed, providing a stable support benchmark for the entire lower limb passive rehabilitation training device 100; at the same time, the fixed drive rod 131 and the preset bed adopt a detachable connection structure (such as quick-release buckles, threaded connections and other adaptable structures), which can be quickly assembled and disassembled according to rehabilitation scenarios (such as bedside rehabilitation, wheelchair transfer rehabilitation, etc.), flexibly adapting to different preset nursing equipment, greatly improving the scenario adaptability and ease of use of lower limb passive rehabilitation training.

[0035] The passive rehabilitation training device for the lower limbs provided in this application consists of a headband acquisition device, a control box, and a motion execution device. The motion execution device comprises a fixed drive rod, a drive assembly, and a passive movable rod. The headband acquisition device is communicatively connected to the control box to stably transmit the patient's electroencephalogram (EEG) signals. The control box is electrically connected to the drive assembly to reduce the transmission delay of control commands, enabling the motion execution device to quickly respond to parameter adjustment commands and achieve real-time linkage between signal acquisition, parameter adjustment, and motion execution. This allows training parameters to dynamically match the patient's state, improving rehabilitation accuracy. One end of the drive assembly is hinged to the fixed drive rod, and the other end is hinged to the passive movable rod, forming a three-bar linkage structure. It utilizes the rotational characteristics of hinges to achieve smooth and controllable transmission, accurately replicating the flexion and extension angles and speed commands output by the control box. This avoids limb swaying or uneven force during passive movement, making it particularly suitable for patients in the early and middle stages of rehabilitation. It effectively helps patients maintain joint range of motion and promotes limb function recovery. The fixed drive rod is used to fix it to the preset nursing equipment. The end of the passive movement rod away from the drive component is equipped with a medical strap to fix the patient's lower leg. The other end of the fixed drive rod away from the drive component is equipped with a medical strap to fix the patient's thigh, forming a double fixation of the movement execution device and the lower limb. This ensures the stability of the patient's lower limb position during rehabilitation training, prevents slippage or displacement, and ensures training safety.

[0036] Optionally, the lower limb passive rehabilitation training system is built upon the aforementioned lower limb passive rehabilitation training device. The signal input of the lower limb passive rehabilitation training system relies on the real-time acquisition of the patient's electroencephalogram (EEG) signals (including resting and active state signals) by the headband acquisition device 110. The action output relies on the motion execution device 130 responding to control commands to complete the passive flexion and extension movements of the lower limbs. All processing functions are integrated into the control box 120 of the lower limb passive rehabilitation training device 100, so as to form a complete closed loop of signal acquisition, module processing and action execution through the control box 120, ensuring deep collaboration and efficient linkage between the core logic of the lower limb passive rehabilitation training system and the hardware of the lower limb passive rehabilitation training device.

[0037] Figure 3 A schematic diagram of the structure of a lower limb passive rehabilitation training system provided in this application embodiment. Figure 1 .like Figure 3 As shown, the lower limb passive rehabilitation training system 200 may include: a resting signal acquisition module 210, an activation signal acquisition module 220, a feature extraction module 230, a rate of change calculation module 240, a parameter adjustment module 250, and a rehabilitation control module 260.

[0038] Specifically, the feature extraction module 230 establishes data transmission connections with the resting signal acquisition module 210 and the activation signal acquisition module 220, respectively; the rate of change calculation module 240 establishes a data transmission connection with the feature extraction module 230, and the parameter adjustment module 250 establishes a data transmission connection with the rate of change calculation module 240; the rehabilitation control module 260 establishes a control signal connection with the parameter adjustment module 250.

[0039] Among them, the resting signal acquisition module 210 is used to acquire the first EEG signal in the prefrontal cortex region of the patient in a resting state within a first preset time period, so as to provide a stable baseline for the comparison of subsequent activation state signals.

[0040] The first EEG signal serves as the baseline signal for the resting state. The resting state is the patient's primary physiological state of complete muscle relaxation, and this state is maintained for a duration greater than a preset time in a quiet environment to ensure that the patient's primary physiological state remains unchanged throughout the data collection process. The patient's complete muscle relaxation can be verified using the collected EMG signals; if the EMG signal amplitude is less than or equal to 5 μV, the patient is considered to be in a relaxed state, avoiding subjective judgment.

[0041] Both the first preset duration and the preset quiet environment can be selected according to actual conditions. For example, the first preset duration can be selected as 30 seconds or 1 minute to ensure the coverage of stable EEG signal cycles (such as the alpha wave proportion being stable at over 60%), and to avoid short-term fluctuations in EEG signals affecting the reliability of the resting state reference signal. The preset quiet environment can be selected as having no strong electromagnetic interference and an environmental noise level of less than or equal to 40 decibels.

[0042] The activation signal acquisition module 220 is used to acquire the second electroencephalogram (EEG) signal of the prefrontal cortex region within a second preset duration in the patient's activated state.

[0043] The second EEG signal is the activated state signal. The activated state is the second physiological state in which, in a preset quiet environment during passive lower limb rehabilitation training, the patient can be induced to perform passive flexion and extension movements of the lower limbs by the motor execution module because the patient has no voluntary lower limb exertion. It should be noted that the preset environment for measuring the second physiological state is consistent with that of the first physiological state to eliminate EEG signal deviations caused by environmental differences.

[0044] The second preset duration can also be selected according to the actual situation. For example, the second preset duration can be selected as 1~5 seconds. Moreover, the second preset duration is shorter than the first preset duration to capture the dynamic changes of the patient's physiological state in real time, avoid the lag of EEG signals caused by long-term collection, and adapt to the need for real-time adjustment of training parameters.

[0045] It should be noted that the acquisition of the activation status signal is triggered synchronously with the action of the motion execution device. That is, when the motion execution device starts, the acquisition of the activation status signal also begins; when the motion execution device stops, the acquisition of the activation status signal also ends, ensuring that the activation status signal corresponds precisely with the training action.

[0046] In addition, it should be noted that the resting signal acquisition module 210 and the activation signal acquisition module 220 mentioned above are both integrated into the headband acquisition device 110.

[0047] The feature extraction module 230 is used to extract focus features from the first and second EEG signals, respectively. These focus features include alpha waves (α waves) and beta waves (β waves). Alpha waves represent a state of quiet, relaxation, and inactivity without active cognitive activity (e.g., resting with eyes closed). Higher alpha waves indicate more subdued cortical activity and a low-arousal state. Beta waves represent a state of alertness, focus, responsiveness to external stimuli, or slight intention to move (e.g., the brain perceiving limb movement during passive motion). Higher beta waves indicate more active cortical activity and a high-arousal state. In other words, alpha waves and beta waves directly reflect the patient's real-time physiological state (e.g., the transition from relaxation to activation, level of fatigue, etc.).

[0048] In one possible implementation, since the lower limb passive rehabilitation training system provided in this application targets patients in the early and middle stages of rehabilitation, these patients are often unable to generate active movement intentions, that is, they cannot actively express their state through electromyographic signals, movement commands, etc., and may have language impairments or confusion, that is, they cannot subjectively express whether they are fatigued or whether they are adapting to the training intensity. Therefore, acquiring alpha and beta waves in the prefrontal cortex region becomes the only passive state feedback signal that can be stably acquired.

[0049] If the proportion of beta waves increases and the proportion of alpha waves decreases during training, it indicates that the patient's brain is sensing passive movement stimulation and is in an activated state, and can tolerate the current training intensity. If the proportion of alpha waves increases and the proportion of beta waves decreases during training, it indicates that the patient's brain activity is slowing down and may be in a state of fatigue, requiring a reduction in training intensity.

[0050] Therefore, by extracting focus features from the first and second EEG signals, the lower limb passive rehabilitation training system can identify physiological states that patients cannot actively express (such as activation and fatigue), providing a reliable basis for the dynamic adjustment of training parameters. Conversely, if this focus feature extraction is not performed and training relies solely on fixed parameters such as constant speed and constant angle, problems may arise where high-intensity training continues even when the patient is fatigued, or low-intensity training continues even after the patient has adapted. This not only affects the rehabilitation effect but may also cause physical injury. Thus, the lower limb passive rehabilitation training system of this application, through the extraction of focus features, makes the patient's physiological state perceptible and the training parameters of the motion execution device adjustable, ultimately achieving an upgrade of passive rehabilitation from mechanical repetition to intelligent adaptation.

[0051] Then, based on the extracted focus characteristics, the resting state focus index is calculated using the following formula (1). and Activated State Concentration Index .

[0052] , Formula (1) Among them, in the above formula (1) The power of the beta wave in the patient at rest; The power of the alpha wave in the patient at rest; The power of the beta wave in the patient in the activated state; The power of the alpha wave in the patient in the activated state.

[0053] Among them, the resting state concentration index The baseline focus level of a patient in a state of complete relaxation and no active cognitive activity is quantified by the power ratio of beta waves to alpha waves at rest. (Restless Focus Index) The higher the ratio, the higher the level of brain activation during rest. Generally, the resting concentration index... The ratio is at a low level because alpha waves dominate at rest.

[0054] Activated state of focus index The real-time focus level of a patient during motor stimulation is quantified by the power ratio of beta waves to alpha waves in the activated state (i.e., during passive rehabilitation training). Activated State Focus Index The higher the ratio, the more active the patient's brain is in perceiving passive movement, and the more activated their physiological arousal state is when they perceive passive movement stimuli.

[0055] The rate of change calculation module 240 is used to calculate the relative rate of change between active and resting states of a patient based on the resting state attention index and the active state attention index using the following formula (2). .

[0056] Formula (2) Among them, the activation resting relative change rate This index is used to characterize the ratio of change in the focus index during the active state to that during the resting state. Generally, the relative change rate between active and resting states... A positive value indicates that the patient's physiological state during training is activated, and the proportion of β waves increases; if the activation resting relative change rate is positive... A negative value indicates that the patient's physiological state is fatigued or relaxed, with an increased proportion of alpha waves; if the resting relative change rate is activated... A value of zero indicates that the patient's physiological state is stable.

[0057] The parameter adjustment module 250 is used to adjust the training parameters of the motion execution device based on the activation-resting relative rate of change and a preset rate of change threshold. The training parameters include flexion-extension speed and flexion-extension angle range; the preset rate of change threshold can be flexibly selected according to actual rehabilitation needs.

[0058] In one possible implementation, due to the activation of the resting relative rate of change It can reflect the patient's true feedback on the current training intensity, providing a basis for judging whether adjustments to training parameters are needed; and the preset rate of change threshold clarifies the objective boundaries that the lower limb passive rehabilitation training system must follow for adjustments, thus based on the relative rate of change at activation and rest. The automatic adjustment of training parameters of the motion execution device, in conjunction with the preset rate of change threshold, can ensure that the adjustment logic of the lower limb passive rehabilitation training system is consistent, repeatable, and free from subjective bias.

[0059] Furthermore, since this lower limb passive rehabilitation training system is designed for patients in the early to mid-stages of rehabilitation, if the activation of the resting relative change rate is not employed... The automatic adjustment mode based on a preset rate of change threshold, relying solely on fixed parameters for training, is prone to problems such as patients being fatigued but unable to transmit signals, while the lower limb passive rehabilitation training system continues high-intensity training, potentially causing limb injury; or patients adapting to the current intensity but unable to provide feedback, while the lower limb passive rehabilitation training system continues low-intensity training, resulting in low rehabilitation efficiency. Therefore, by activating the resting relative rate of change... Reflecting the patient's real-time physiological state, the system triggers corresponding adjustments based on preset change rate thresholds. This essentially endows the lower limb passive rehabilitation training system with the ability to sense the patient's workload. Without requiring active feedback from the patient, the system can activate the resting relative change rate... The quantitative signal is used to determine the patient's tolerance, and then the training parameters are automatically adapted according to the preset rate of change threshold (if the patient's tolerance is high, the intensity is increased, and if the tolerance is low, the intensity is decreased). In the end, while ensuring the rehabilitation effect, the safety risks caused by information gaps are effectively avoided.

[0060] The rehabilitation control module 260 is used to send control commands to the motion execution device according to the adjusted training parameters, so as to drive the patient to complete the passive rehabilitation training of the lower limbs.

[0061] In one possible implementation, the rehabilitation control module 260 receives the adjusted flexion-extension speed and flexion-extension angle range, and converts them into digital control commands (such as pulse width modulation signals PWM) to ensure that the new training parameters can be recognized by the motion execution device. The control box is electrically connected to the drive component in the motion execution device (such as an RS485 interface) and transmits the generated control commands to the drive component in real time. After receiving the control commands, the motor in the drive component starts according to the control commands. Through the hinges between the drive component and the passive moving rod and the fixed drive rod, the passive moving rod rotates around the hinge point with the drive component. The end of the passive moving rod away from the drive component is fixed to the patient's lower leg with a medical strap, so that the rotation of the passive moving rod will directly drive the patient's lower leg to perform flexion-extension movements. At the same time, the fixed drive rod is fixed to the patient's thigh with a medical strap, providing stable support for the lower limb and ensuring that the thigh position does not shift during lower leg flexion-extension, thereby realizing passive rehabilitation training of the patient's lower limb.

[0062] It should be noted that the aforementioned feature extraction module 230, rate of change calculation module 240, parameter adjustment module 250, and rehabilitation control module 260 are all integrated into the control box 120.

[0063] For example, in the initial training phase: the patient wears a headband acquisition device, the fixed drive rod is connected to the preset bed, and the lower limbs are bound; the resting signal acquisition module 210 acquires the first EEG signal for 30-60 seconds, and then the feature extraction module 230 calculates the resting state attention index, using the resting state attention index as a benchmark.

[0064] Rehabilitation training phase: The rehabilitation control module 260 sends initial training parameters, the motion execution device initiates passive flexion and extension movements, the activation signal acquisition module 220 acquires the second EEG signal every 1-5 seconds, then the feature extraction module 230 calculates the active state attention index, and the rate of change calculation module 240 calculates the relative rate of change between active and resting states in real time based on the active state attention index and the resting state attention index. The training parameters are dynamically optimized through the parameter adjustment module 250, and finally the rehabilitation control module 260 updates the control commands to drive the motion execution device to perform actions.

[0065] It should be noted that the above rehabilitation training process is repeated until the preset training time ends (the preset training time can generally be set to 10-30 minutes through the control box). No real-time intervention from medical staff is required throughout the process, thus achieving automated and individualized passive rehabilitation training.

[0066] The lower limb passive rehabilitation training system of this application can be composed of a resting signal acquisition module, an activation signal acquisition module, a feature extraction module, a rate of change calculation module, a parameter adjustment module, and a rehabilitation control module. First, the resting signal acquisition module is used to acquire the first EEG signal of the prefrontal cortex region within a first preset duration in the patient's resting state. The activation signal acquisition module is used to acquire the second EEG signal of the prefrontal cortex region within a second preset duration in the patient's activated state. By acquiring EEG signals corresponding to two physiological states in the same preset quiet environment, there is no need for the patient to actively generate movement intentions. Even with blurred consciousness, effective signals can be generated by passively maintaining the physiological state. Simultaneously, the same environment avoids EEG signal distortion, solving the problem of failed movement intention signal capture from the source and improving the adaptability of the lower limb passive rehabilitation training system to different patients. Second, the feature extraction module is used to extract focus features from the first and second EEG signals respectively, so that the extracted focus features can truly reflect the alpha and beta brain waves of the patient's physiological state, rather than brain wave features dependent on movement intentions. Then, based on the extracted focus features, the resting state focus index and the activated state focus index are calculated respectively, transforming the abstract EEG signals into... The system offers quantifiable indicators, addressing the analytical challenge of extracting features related to motor intent. Furthermore, the rate of change calculation module calculates the relative rate of change between the patient's active and resting states based on the resting state attention index and the active state attention index. This quantifies the fluctuation of the patient's current state relative to their resting state baseline, allowing for objective assessment without patient feedback and filling the gap in existing technologies that cannot quantify the state of unconscious patients. Simultaneously, the parameter adjustment module adjusts the training parameters of the motor execution device based on the relative rate of change between the active and resting states and a preset rate of change threshold. The training parameters are adjusted based on the relative rate of change at rest and the preset rate of change threshold. This allows for automatic output of control commands, dynamically adapting to the patient's real-time physiological state. This eliminates reliance on preset programs, preventing discomfort caused by muscle strength fluctuations and tolerance differences in the early stages, reducing the risk of joint injury and muscle soreness, and improving the safety and comfort of rehabilitation training. Finally, the rehabilitation control module sends control commands to the motion execution device based on the adjusted training parameters, guiding the patient to complete passive lower limb rehabilitation training. This ensures smooth and controllable passive flexion and extension movements, meeting the needs of early and mid-stage rehabilitation patients with limited joint mobility and requiring precise training.Therefore, the lower limb passive rehabilitation training system of this application is realized by relying on the hardware of the lower limb passive rehabilitation training device. The data transmission between each module adopts a standardized interface, with low transmission latency and high stability. Moreover, the lower limb passive rehabilitation training system operates automatically throughout the entire process. Data collection, analysis, parameter adjustment, and execution do not require active operation by the patient, which also reduces the real-time intervention cost of medical staff. It can also be adapted to various scenarios such as bedside and wheelchair, making it convenient for patients to use for rehabilitation at home or in the hospital. It completely fills the application gap of existing brain control technology in early rehabilitation scenarios and truly realizes safe and precise passive rehabilitation for early paralyzed patients with blurred consciousness and inability to generate motor intentions.

[0067] Optionally, continue to refer to Figure 3 The feature extraction module 230 may include a preprocessing submodule 231 and a feature extraction submodule 232.

[0068] Among them, the preprocessing submodule 231 establishes data transmission connections with the resting signal acquisition module 210 and the activation signal acquisition module 220; the focus feature extraction submodule 232 establishes data transmission connections with the preprocessing submodule 231.

[0069] The preprocessing submodule 231 is used to preprocess the first EEG signal and the second EEG signal respectively, and output the preprocessed resting signal and the preprocessed activation signal.

[0070] In one possible implementation, the original first and second EEG signals inevitably contain various interference signals, which directly lead to distortion in subsequent focus feature extraction. These interference signals can include at least environmental interference, physiological artifacts, and signal noise. Directly using the original EEG signals containing interference to extract focus features can result in errors in calculating the power of alpha and beta waves, thus distorting the focus index and ultimately causing parameter adjustment deviations, such as misjudging patient fatigue as an active state, leading to inappropriate training intensity. Therefore, by preprocessing the first and second EEG signals, the preprocessed resting and activated signals are removed, retaining only effective EEG signal components such as alpha and beta waves.

[0071] The feature extraction submodule 232 is used to perform focused feature extraction on the preprocessed resting signal and the preprocessed activation signal, respectively.

[0072] In one possible implementation, since the preprocessed resting and activation signals are essentially still time-series EEG data, they cannot directly reflect the patient's physiological state activation level. This prevents the lower limb passive rehabilitation training system from determining whether the patient is in a relaxed (resting) or activated state from electrical signals (such as voltage values). Without extracting alpha and beta waves, two features strongly correlated with attention and arousal, it is impossible to calculate the attention index, let alone determine the patient's physiological state through the relative change rate between activation and resting states. Ultimately, this leads to a lack of basis for adjusting training parameters, and the lower limb passive rehabilitation training system degenerates into fixed-parameter training. Therefore, it is necessary to extract attention features from the preprocessed resting and activation signals separately. This transforms the massive time-series EEG data into a small number of features with clear physiological significance, turning abstract signals into interpretable physiological state indicators. This establishes a correlation between physiological state and EEG signals, providing direct data for subsequent calculations of the resting and activated state attention indices.

[0073] The lower limb passive rehabilitation training system of this application comprises a feature extraction module consisting of a preprocessing submodule and a feature extraction submodule. The preprocessing submodule preprocesses the first and second EEG signals respectively, outputting preprocessed resting signals and preprocessed activation signals to improve the accuracy of subsequent focus feature extraction and avoid power calculation errors of α or β waves caused by interference. Simultaneously, it indirectly ensures the calculation accuracy of the two types of focus indices and the relative change rate between activation and resting states, providing an unbiased decision-making basis for the parameter adjustment module and mitigating the risk of misadjustment of training parameters due to EEG signal distortion from the source. The feature extraction submodule extracts focus features from the preprocessed resting signals and preprocessed activation signals respectively. Through focus feature extraction, the lower limb passive rehabilitation training system can understand the patient's passive physiological state feedback, providing a basis for subsequent dynamic adjustment of training parameters and achieving an upgrade from mechanical repetition to intelligent adaptation. Therefore, this application first eliminates EEG signal interference through preprocessing to ensure the basic reliability of subsequent analysis. Then, it extracts meaningless EEG signals through attention features and transforms them into alpha and beta wave features that can reflect the patient's activation or fatigue state, bridging the gap between the original EEG signals and rehabilitation needs. This provides accurate data for attention index calculation, activation-resting relative change rate analysis, and parameter adjustment. Ultimately, it enables the lower limb passive rehabilitation training system to dynamically adapt the training intensity according to the patient's real-time physiological state, ensuring the training safety of patients in the early and middle stages of rehabilitation and improving rehabilitation efficiency.

[0074] Optionally, in Figure 3 On this basis, Figure 4 A schematic diagram of the structure of a lower limb passive rehabilitation training system provided in this application embodiment. Figure 2 .like Figure 4As shown, the preprocessing submodule 231 includes: a baseline drift removal unit 231-1, an artifact removal unit 231-2, and a filtering unit 231-3.

[0075] Specifically, the baseline drift removal unit 231-1 establishes a data transmission connection with the resting signal acquisition module 210 and the activation signal acquisition module 220; the artifact removal unit 231-2 establishes a data transmission connection with the baseline drift removal processing unit 231-1; and the filtering processing unit 231-3 establishes a data transmission connection with the artifact removal unit 231-2.

[0076] Among them, the baseline drift removal unit 231-1 is used to perform baseline drift removal processing on the first EEG signal and the second EEG signal respectively, and output the first resting signal and the first activation signal.

[0077] In one possible implementation, the original first and second EEG signals may experience slow baseline drift due to factors such as patient breathing, slight body movement, and changes in electrode contact pressure. This baseline drift manifests as an overall vertical shift of the EEG signal along the time axis. This drift is not a valid component of the EEG signal; if left untreated, it can lead to amplitude distortion and misinterpretation during subsequent artifact removal and filtering. Therefore, the baseline drift removal unit performs baseline drift removal processing on the first and second EEG signals respectively using a preset linear trend removal method (such as least squares fitting of the signal baseline and then subtracting the baseline trend from the original signal) or a preset high-pass filtering method, ultimately outputting a first resting signal and a first activation signal with stable baselines.

[0078] The method for baseline drift removal can be selected according to the actual situation. For example, the baseline drift removal method can be a preset linear trend removal method (such as fitting the signal baseline using the least squares method and then subtracting the baseline trend from the original signal) or a preset high-pass filtering method.

[0079] The artifact removal unit 231-2 is used to perform artifact removal processing on the first resting signal and the first activation signal respectively, and output the second resting signal and the second activation signal.

[0080] In one possible implementation, physiological interference artifacts remain in the first resting signal and the first activation signal. These artifacts are generated by non-EEG activity and typically have amplitudes 5-10 times that of alpha and beta waves, directly drowning out effective brainwaves. These physiological interference artifacts can include at least: electrooculography (EOG) artifacts, electromyography (EMG) artifacts, and electrocardiogram (ECG) artifacts.

[0081] If these physiological interference artifacts are retained, the subsequently extracted alpha and beta wave features will actually be artifact signals, leading to complete distortion of the attention index calculation. Therefore, the artifact removal unit uses preset independent component analysis (ICA) or preset threshold methods to process the first resting signal and the first activation signal respectively, and outputs a second resting signal and a second activation signal with the main physiological interference artifacts removed.

[0082] The method for artifact removal can be selected according to the actual situation. For example, artifact removal can be performed using preset independent component analysis or preset threshold method.

[0083] The filtering processing unit 231-3 is used to filter the second resting signal and the second activation signal respectively, and output the pre-processed resting signal and the pre-processed activation signal.

[0084] In one possible implementation, since environmental interference and external noise interference still exist in the second silent signal and the second active signal, the environmental interference includes at least power frequency interference and electromagnetic radiation. Furthermore, since this application only needs to focus on the characteristics, namely α waves and β waves, a bandpass filter can be set according to the frequency characteristics of α waves and β waves, such as the frequency of α waves being 8-12Hz and the frequency of β waves being 13-30Hz. This extracts the 8-30Hz frequency band, accurately retaining the α wave and β wave frequency bands, and filtering out external noise below 8Hz and above 30Hz.

[0085] It should be noted that the above preprocessing must first perform baseline drift removal, then artifact removal, and finally filtering. Baseline drift is a low-frequency, slowly changing interference that directly affects the accuracy of artifact removal in separating EEG signals; therefore, it must be removed first to ensure data stability. Furthermore, the frequencies of artifacts such as eye movement and electromyography (EMG) signals may overlap with the target EEG frequency; artifact removal first prevents the accidental deletion of valid EEG components during filtering. Filtering, as the final step, accurately removes remaining low-frequency noise and high-frequency interference, ultimately preserving the target EEG signal.

[0086] The preprocessing submodule of the lower limb passive rehabilitation training system of this application can be composed of a baseline drift removal unit, an artifact removal unit, and a filtering unit. The baseline drift removal unit is used to perform baseline drift removal processing on the first EEG signal and the second EEG signal respectively, outputting a first resting signal and a first activation signal to ensure that the true fluctuations of alpha and beta waves can be clearly presented, avoiding deviations in attention feature extraction caused by drift, and providing a data basis for the subsequent artifact removal unit. The artifact removal unit is used to perform artifact removal processing on the first resting signal and the first activation signal respectively, outputting a second resting signal. The system consists of a first resting signal and a second activation signal, ensuring that the proportions of alpha and beta waves are relatively high, providing a data foundation for subsequent focus feature extraction. A filtering unit is used to filter the second resting signal and the second activation signal respectively, outputting pre-processed resting signal and pre-processed activation signal to retain only alpha and beta waves, completely eliminating environmental interference and out-of-band noise. This allows subsequent focus feature extraction to proceed without distinguishing between effective components and interference, improving the computational efficiency of the feature extraction module and ensuring that the focus index can be updated in real time to adapt to the needs of dynamic adjustment of training parameters.

[0087] Optionally, continue to refer to Figure 4 The aforementioned feature extraction submodule 232 includes: an alpha brainwave extraction unit 232-1 and a beta brainwave extraction unit 232-2.

[0088] Among them, the alpha brainwave extraction unit 232-1 establishes a data transmission connection with the preprocessing submodule 231; the beta brainwave extraction unit 232-2 establishes a data transmission connection with the preprocessing submodule 231.

[0089] Because there is a fixed difference in the physiological frequency bands of alpha brain waves (α waves) and beta brain waves (β waves), each extraction module achieves accurate extraction through preset frequency bands, and strictly follows the setting that the first preset frequency band is less than the second preset frequency band.

[0090] Both the first and second preset frequency bands can be selected according to actual conditions. For example, the first preset frequency band can be set to 8-12Hz; the second preset frequency band can be set to 13-30Hz.

[0091] Among them, the alpha brainwave extraction unit 232-1 is used to extract features from the preprocessed resting signal and the preprocessed activation signal using the first preset frequency band, and output the resting alpha brainwave and the activation alpha brainwave.

[0092] In one possible implementation, since alpha waves reflect the patient's level of relaxation, they dominate in the resting state and decrease in the proportion of alpha waves as brain arousal increases in the activated state. Without extracting alpha waves separately, it is impossible to quantify the patient's baseline resting state and changes in relaxation during training, and it is also impossible to subsequently determine the patient's physiological state by comparing alpha waves with beta waves. Furthermore, the frequency bands of the preprocessed resting signal and the preprocessed activated signal are 8-30Hz. Therefore, the alpha wave extraction unit 232-1 uses a first preset bandpass filter to lock the 8-12Hz frequency band, further eliminating overlap interference with the beta wave frequency band, and outputs resting alpha waves and activated alpha waves.

[0093] The beta brainwave extraction unit 232-2 is used to extract features from the preprocessed resting signal and the preprocessed activation signal using a second preset frequency band, and output resting beta brainwaves and activation beta brainwaves.

[0094] In one possible implementation, since beta waves reflect the patient's activation level, the proportion of beta waves increases with the patient's enhanced perception of motor stimuli in the activated state, while the proportion of beta waves is low in the resting state. Without extracting beta waves separately, the patient's activation level cannot be quantified, and the subsequent relative change rate between activation and resting states becomes unreliable, making it impossible for the lower limb passive rehabilitation training system to determine whether the training intensity is appropriate. Furthermore, the preprocessed resting signal and preprocessed activated signal operate in the 8-30Hz frequency band. Therefore, the beta wave extraction unit 232-2 uses a first preset bandpass filter to lock the 13-30Hz frequency band, eliminating overlap interference with the alpha wave frequency band, and outputs resting beta waves and activated beta waves.

[0095] It should be noted that the alpha brainwave extraction unit 232-1 and the beta brainwave extraction unit 232-2 extract alpha and beta waves from the resting and activated states, respectively. If the activated alpha wave power is less than the resting alpha wave power, and the activated beta wave power is greater than the resting beta wave power, it indicates that the patient has transitioned from relaxation to activation and has high tolerance. If the activated alpha wave power is greater than the resting alpha wave power, and the activated beta wave power is less than the resting beta wave power, it indicates that the patient has transitioned from relaxation to further relaxation (or fatigue), and the intensity needs to be reduced. This allows the lower limb passive rehabilitation training system to understand the patient's passive physiological state, bridging the gap in active feedback that patients cannot provide during the early and middle stages of rehabilitation.

[0096] The feature extraction submodule of the lower limb passive rehabilitation training system of this application can be composed of an alpha brainwave extraction unit and a beta brainwave extraction unit. The alpha brainwave extraction unit is used to extract features from the preprocessed resting signal and the preprocessed activated signal using a first preset frequency band, outputting resting alpha brainwaves and activated alpha brainwaves. The beta brainwave extraction unit is used to extract features from the preprocessed resting signal and the preprocessed activated signal using a second preset frequency band, outputting resting beta brainwaves and activated beta brainwaves. The first preset frequency band is less than the second preset frequency band. Therefore, this application uses two independent alpha brainwave extraction units and beta brainwave extraction units to extract alpha and beta brainwaves under different physiological states, providing a data foundation for subsequent attention index analysis.

[0097] Optionally, in Figure 4 On this basis, Figure 5 A schematic diagram of the structure of a lower limb passive rehabilitation training system provided in this application embodiment. Figure 3 .like Figure 5 As shown, the alpha brainwave extraction unit 232-1 includes: a discrete processing unit 23211, a weighted processing unit 23212, a Fourier transform unit 23213, a power spectral density calculation unit 23214, and a brainwave determination unit 23215.

[0098] Specifically, the discrete processing unit 23211 establishes a data transmission connection with the preprocessing submodule 231; the weighted processing unit 23212 establishes a data transmission connection with the discrete processing unit 23211; the Fourier transform unit 23213 establishes a data transmission connection with the weighted processing unit 23212; the power spectral density calculation unit 23214 establishes a data transmission connection with the Fourier transform unit 23213; and the brainwave determination unit 23215 establishes a data transmission connection with the power spectral density calculation unit 23214.

[0099] The discrete processing unit 23211 is used to perform discrete processing on the preprocessed resting signal and the preprocessed activation signal according to a preset unit time window, and output the first discrete resting signal and the first discrete activation signal.

[0100] The preset unit time window can be selected according to the actual situation. For example, the preset unit time window can be the same as the second preset duration setting, such as setting the preset unit time window to 1-5 seconds.

[0101] In one possible implementation, since the preprocessed resting signal and preprocessed activation signal are continuous-time EEG signals, while subsequent Fourier transforms and power spectral density calculations are based on discrete signals, directly processing the continuous EEG signal would lead to computational limitations due to the infinite amount of data and could easily introduce time truncation errors. Therefore, the discrete processing unit 23211 truncates and samples the continuous EEG signal through a preset unit time window, outputting a first discrete resting signal and a first discrete activation signal composed of discrete sampling points.

[0102] The weighted processing unit 23212 is used to multiply the first discrete resting signal and the first discrete activation signal by a preset Hanning window function, respectively, and output the first weighted resting signal and the first weighted activation signal.

[0103] The preset Hanning window function w(n) is a window function with smooth edges, which allows the amplitude of the discrete signal to transition slowly to zero at the window edge (rather than being abruptly truncated), thereby significantly reducing spectral leakage. It can be expressed by the following formula (3): w(n)=0.5-0.5×cos(2πn / (N-1)) Formula (3) Where N is the total window length, i.e., the total number of sampling points. n is the index (serial number) of the nth sampling point in the window function, and its value range is 0≤n≤N-1.

[0104] In one possible implementation, since the discrete signal is obtained by truncating continuous EEG signals, and EEG signal truncation leads to spectral leakage—that is, alpha waves originally concentrated at a certain frequency will spread to adjacent frequencies in the frequency domain—it becomes impossible to accurately locate the true frequency range of the alpha waves, and may even overlap with the adjacent beta wave spectrum. Therefore, the weighting processing unit 23212 multiplies the first discrete resting signal and the first discrete activation signal point by point with a preset Hanning window function w(n) to smoothly weight the EEG signals, and outputs the first weighted resting signal and the first weighted activation signal.

[0105] For example, the weighted EEG signal x_w(n) can be represented by the following formula (4): x_w(n)=x(n) w(n) formula (4) Where x(n) is the activation signal.

[0106] The Fourier transform unit 23213 is used to perform a preset Fourier transform on the first weighted resting signal and the first weighted activated signal respectively, and output the first resting frequency domain sequence and the first activated frequency domain sequence.

[0107] Among them, the Preset Fourier Transform (FFT) is an efficient time-frequency conversion algorithm that can convert discrete signals in the time domain into frequency domain sequences corresponding to frequency amplitudes.

[0108] In one possible implementation, since the preprocessed resting signal and preprocessed activation signal, as well as the first weighted resting signal and the first weighted activation signal, are all in the time domain, only the changes in EEG signals (such as voltage) over time can be observed, and it is impossible to directly identify which frequency components have a high proportion. Without converting to the frequency domain, it is impossible to locate and extract the frequency features of the alpha wave. Therefore, the Fourier transform unit 23213 performs a preset fast Fourier transform on the first weighted resting signal and the first weighted activation signal to convert the discrete signals in the time domain into frequency domain sequences corresponding to the frequency amplitudes, and outputs the first resting frequency domain sequence and the first activation frequency domain sequence.

[0109] For example, performing an FFT on the weighted signal x_w(n) yields a frequency domain sequence X(k) in complex form, which can be calculated using the following formula (5).

[0110] X(k)=FFT[x_w(n)] Formula (5) Where the magnitude |X(k)| of X(k) represents the amplitude at the corresponding frequency, and the phase information is negligible. k=0,1,...,N-1.

[0111] The power spectral density calculation unit 23214 is used to calculate the first resting frequency domain sequence and the first active frequency domain sequence respectively, and output the first resting power spectral density and the first active power spectral density.

[0112] In one possible implementation, since the frequency domain sequence can only reflect the amplitude of each frequency, while the physiological characteristics of alpha waves need to be quantified by power, the power spectral density (PSD) can reflect the power per unit frequency, directly reflecting the energy proportion of a certain frequency band (such as 8-12Hz), and is the main indicator for judging whether alpha waves are dominant. If only the amplitude is used, the intensity of alpha waves cannot be accurately characterized, because the amplitude and power have a square relationship, and the power may differ significantly when the amplitudes are similar. Therefore, the power spectral density calculation unit 23214 calculates the power spectral density of each frequency based on the first resting frequency domain sequence and the first activated frequency domain sequence using the following formula (6), and then outputs the first resting power spectral density and the first activated power spectral density.

[0113] PSD(k)=(2×|X(k)|²) / (Fs×N×W) Formula (6) Where |X(k)|² is the square of the amplitude (energy) of the FFT result; Fs is the sampling frequency; N is the total window length; and W is the energy correction coefficient of the window function.

[0114] For example, in the Fast Fourier Transform (FFT), there is a clear correspondence between the index k of the frequency domain sequence X(k) and the actual frequency f(k), specifically f(k) = k × (Fs / N), where Fs is the sampling frequency and N is the number of FFT points (i.e., the total number of sampling points after signal discretization). In this lower limb passive rehabilitation training system, the sampling frequency Fs = 256 Hz and the number of FFT points N = 256. Therefore, the frequency resolution of the system is Fs / N = 1 Hz, which means that for every increase of 1 in the index k, the corresponding actual frequency increases by 1 Hz.

[0115] Based on this, the target frequency range for alpha brainwaves (α waves) is 8-12Hz, corresponding to indices k=8 to k=12 in the FFT output (k=8 corresponds to 8Hz, k=12 corresponds to 12Hz). To calculate the total power of α waves for subsequent concentration index calculations, the power spectral density (PSD) at each frequency point within the 8-12Hz band needs to be summed, i.e., total α wave power P_α=ΣPSD(k), where k ranges from 8 to 12.

[0116] The brainwave determination unit 23215 is used to determine the resting alpha brainwave and the activated alpha brainwave based on the first resting power spectral density and the first activated power spectral density.

[0117] In one possible implementation, resting alpha brainwaves and activated alpha brainwaves are ultimately output based on a first resting power spectral density and a first activated power spectral density.

[0118] It should be noted that the structure of the beta brainwave extraction unit is completely identical to that of the alpha brainwave extraction unit, also including a discrete processing unit, a weighted processing unit, a Fourier transform unit, a power spectral density calculation unit, and a brainwave determination unit. Its processing flow for preprocessed resting signals and preprocessed activated signals (from discretization and Hanning window weighting to Fourier transform, power spectral density calculation, and then brainwave feature determination) is also exactly the same as that of the alpha brainwave extraction unit, except that it focuses on the 13-30Hz target frequency band corresponding to beta brainwaves (β waves) in the brainwave determination stage. Given the high degree of consistency in the overall processing logic, the specific operational details of each subunit will not be elaborated upon here.

[0119] The alpha brainwave extraction unit in the lower limb passive rehabilitation training system of this application can be composed of a discrete processing unit, a weighted processing unit, a Fourier transform unit, a power spectral density calculation unit, and a brainwave determination unit. The discrete processing unit is used to discretely process the pre-processed resting signal and the pre-processed activation signal according to a preset unit time window, outputting a first discrete resting signal and a first discrete activation signal; this converts the continuous EEG signal into a discrete sequence that meets the requirements of subsequent frequency domain analysis, providing an operable data format for the entire alpha wave extraction process. The weighted processing unit is used to multiply the first discrete resting signal and the first discrete activation signal by a preset Hanning window function, outputting a first weighted resting signal and a first weighted activation signal. The system includes: a Fourier transform unit, which performs preset Fourier transforms on the first weighted resting signal and the first weighted activated signal respectively, outputting a first resting frequency domain sequence and a first activated frequency domain sequence to transform the EEG signal from the time dimension to the frequency dimension, allowing the target frequency band of the alpha wave to emerge from the complex signal; a power spectral density calculation unit, which calculates the first resting frequency domain sequence and the first activated frequency domain sequence respectively, outputting a first resting power spectral density and a first activated power spectral density, clearly defining the power value of each frequency within the target frequency band of the alpha wave, providing a quantitative basis for determining whether the alpha wave is dominant; and a brainwave determination unit, which determines the resting alpha brainwave and the activated alpha brainwave based on the first resting power spectral density and the first activated power spectral density. Therefore, this application can guarantee the data accuracy of the alpha wave, ensuring the accurate and reliable calculation of subsequent attention index and the relative change rate between resting and activated states.

[0120] Optionally, the aforementioned preset rate of change thresholds include: a positive threshold, a stable threshold, and a negative threshold. The positive threshold, stable threshold, and negative threshold can all be selected according to the actual situation. For example, the positive threshold is usually set to ≥1, indicating that the patient's activation state is significantly higher than their resting state. At this time, the patient is in a state of high concentration, their brain is actively perceiving passive movement, their physical tolerance is high, and they have the conditions to increase training intensity. The stability threshold is usually set between 0 and 1, which means that the patient's active state and resting state are basically matched. At this time, the patient's state is stable, the patient is in a moderate state of focus, the current training parameters match the tolerance, and no adjustment is needed. The negative threshold is usually set to ≤0, which means that the patient's activation state is significantly lower than that of the resting state. At this time, the patient is in a low-focus state, the patient's brain activity tends to be slow, and the patient may be in a state of fatigue. The training intensity needs to be reduced to avoid risks.

[0121] Continue to refer to Figure 3 The parameter adjustment module 250 mentioned above includes: adjustment module 251.

[0122] Among them, the adjustment module 251 is used to receive the active resting relative change rate transmitted by the change rate calculation module. .

[0123] When the resting relative change rate is activated When the positive threshold is greater than or equal to, it indicates that the patient is in a highly activated state, meaning that the patient's brain is actively perceiving passive movement of the lower limbs, with a high proportion of beta waves and a low proportion of alpha waves. The body has a high tolerance for the current training intensity, and there may even be problems with insufficient intensity and low rehabilitation efficiency. At this time, it is necessary to stimulate nerves and muscles by increasing the intensity to accelerate the rehabilitation process. Therefore, the speed increase and angle expansion parameter adjustment commands can be output to increase the lower limb flexion and extension speed of the motion execution device by a preset speed increase value to avoid joint discomfort caused by sudden speed changes. At the same time, the flexion and extension angle range can be expanded by a preset angle expansion value to gradually expand the patient's joint range of motion. The preset speed increase value and preset angle expansion value can be selected according to the actual situation. For example, the preset speed increase value can be selected as 0.5~1cm / s, and the preset angle expansion value can be selected as 5°~10°.

[0124] When the resting relative change rate is activated When the threshold is reached, it indicates that the patient is in a state of optimal adaptation, meaning the difference between the activated and resting states is small. The current training parameters (speed, angle) are highly matched with the patient's tolerance, ensuring that rehabilitation is neither inefficient due to excessively low intensity nor fatigued due to excessively high intensity. At this point, the output command to maintain the current parameters aims to avoid training fluctuations caused by frequent adjustments, allowing the patient to continuously receive regular stimulation at an appropriate intensity, which aligns better with the gradual principle of passive rehabilitation.

[0125] When the resting relative change rate is activated When the value is less than or equal to the negative threshold, it indicates that the patient is in a state of low activation or fatigue. The patient's brain has reduced perception of passive movement, with a high proportion of alpha waves and a low proportion of beta waves. The body may no longer be able to tolerate the current intensity. If the original parameters are maintained, it may lead to risks such as muscle soreness and joint strain. At this time, the intensity needs to be reduced, but the adjustment strategy must take into account both safety and rehabilitation goals. Output deceleration and angle parameter adjustment commands to reduce the lower limb flexion and extension speed of the motion execution device by a preset deceleration value from the current flexion and extension speed, while keeping the flexion and extension angle range unchanged.

[0126] The preset deceleration value can be selected according to the actual situation. For example, the preset deceleration value can be selected as 0.5cm / s.

[0127] It should be noted that the speed of flexion and extension is the main factor affecting the exercise load. Deceleration can quickly reduce the burden on muscles and joints, and the adjustment is smooth and less likely to cause discomfort. Keeping the range of flexion and extension angles unchanged is because patients in the early and middle stages have limited joint mobility. A sudden reduction in the flexion and extension angle may lead to a worsening of joint stiffness, while a sudden increase in the flexion and extension angle may exceed the joint's tolerance range. Therefore, keeping the angle stable can both avoid joint damage and maintain the current joint mobility training goal.

[0128] The lower limb passive rehabilitation training system of this application has a preset rate of change threshold, which can be composed of a positive threshold, a stable threshold, and a negative threshold. The parameter adjustment module can be composed of an adjustment module. The adjustment module is used to receive the activation-resting relative rate of change transmitted by the rate of change calculation module. When the activation-resting relative rate of change is greater than or equal to the positive threshold, it outputs speed-up and angle-expansion parameter adjustment commands to increase the lower limb flexion-extension speed of the motion execution device by a preset speed-up value on the current flexion-extension speed, and at the same time expand the flexion-extension angle range by a preset angle-expansion value. When the activation-resting relative rate of change is at the stable threshold, it outputs a command to keep the current parameters unchanged, so as to keep the current training parameters unchanged and avoid repeated changes in flexion-extension speed and flexion-extension angle due to small state fluctuations, thereby improving the patient's training experience and reducing the mechanical wear of the motion execution device. When the activation-resting relative rate of change is less than or equal to the negative threshold, it outputs deceleration and angle-keeping parameter adjustment commands to reduce the lower limb flexion-extension speed of the motion execution device by a preset deceleration value on the current flexion-extension speed, while keeping the flexion-extension angle range unchanged, thereby reducing the load and avoiding the stimulation of the joints by angle fluctuations, thus ensuring the training safety of patients in the early and middle stages of rehabilitation. Therefore, this application dynamically adjusts training parameters through threshold judgment, which allows the training intensity to follow the patient's state in real time. Without the patient's active participation, it can achieve a fully automatic closed loop from physiological state changes and command output to parameter adjustment. This reduces the intervention cost for medical staff and avoids rehabilitation risks caused by information gaps, thus achieving individualized and precise rehabilitation.

[0129] Optionally, continue to refer to Figure 3 The parameter adjustment module 250 also includes a pause / restart submodule 252.

[0130] Among them, the pause / restart submodule 252 establishes a data transmission connection with the adjustment module 251.

[0131] The pause / restart submodule 252 is used to continuously receive the relative change rate at rest during activation. When parameter adjustment module 250 outputs deceleration and hold angle parameter adjustment commands, pause / restart submodule 252 starts monitoring to avoid over-protection; if the relative change rate of the active-resting parameters corresponds to a preset number of preset unit time windows... If all values ​​are less than or equal to the negative threshold, then the pause / restart submodule 252 sends a pause-and-restart control command to the rehabilitation control module.

[0132] The preset unit time window is the time interval for each acquisition of the rate of change, which is usually consistent with the activation signal acquisition cycle (e.g., 2 seconds / window) to ensure the timeliness of status monitoring. The preset unit time window can be selected according to the actual situation, and the preset unit time window can be selected as 3-5.

[0133] The preset quantity can be selected according to the actual situation. For example, the preset quantity can be selected as 3.

[0134] In one possible implementation, the activation-resting relative change rate is satisfied when a preset number (e.g., 3) of consecutive preset unit time windows (e.g., 3) are met. When all values ​​are less than or equal to the negative threshold, the pause / restart submodule 252 determines that the patient is in a state of extreme fatigue, and slowing down alone cannot alleviate the condition. It then sends a pause-restart control command to the rehabilitation control module 260.

[0135] It should be noted that the pause-and-restart control command does not permanently stop training, but rather pauses briefly (e.g., 10-20 seconds) to allow the patient to recover, and then automatically restarts training. For example, after receiving the pause-and-restart control command, the rehabilitation control module 260 immediately controls the motion execution device to stop the lower limb flexion and extension movements, giving the patient's muscles and nerves a short rest time and reducing the burden caused by extreme fatigue. When the pause time ends, the lower limb passive rehabilitation training system automatically restarts training with decelerated parameters to avoid a sudden increase in intensity after restarting and to prevent fatigue from recurring, while ensuring the continuity of rehabilitation training.

[0136] The parameter adjustment module in the lower limb passive rehabilitation training system of this application can also be composed of a pause and restart submodule. The pause and restart submodule establishes a data transmission connection with the rate of change calculation module. The pause and restart submodule is used to continuously receive the relative rate of change at rest. When the parameter adjustment module outputs deceleration and angle parameter adjustment commands, if the relative rate of change at rest corresponding to a preset number of preset unit time windows is less than or equal to a negative threshold, the pause and restart submodule sends a pause and restart control command to the rehabilitation control module. Thus, the pause and restart submodule of this application triggers pause by judging continuous fatigue, which is equivalent to adding a secondary protection of pause and rest in addition to deceleration. That is, when deceleration cannot relieve fatigue, the load input is directly cut off by a short pause, allowing muscles and joints to rest immediately, fundamentally avoiding hidden damage caused by over-fatigue, and is the last line of defense for the safety of patients in the early and middle stages of training.

[0137] Optionally, this application also provides an example of a control box 120. Figure 6 This is a schematic diagram of a control box provided in an embodiment of this application. Figure 6 As shown, the control box 120 may include a processor 121 and a memory 122.

[0138] The memory 122 stores machine-executable instructions that can be executed by the processor 121. When the control box 120 is running, these machine-executable instructions are executed. The processor 121 communicates with the memory 122 via a bus. The processor 121 can execute these machine-executable instructions to implement a passive rehabilitation training method for the lower limbs.

[0139] The memory 122, processor 121, and bus components are electrically connected directly or indirectly to enable data transmission or interaction. For example, these components can be electrically connected to each other via one or more communication buses or signal lines. The mobile storage device includes at least one software function module that can be stored in the memory 122 as software or firmware or embedded in the operating system (OS) of a computer device. The processor 121 is used to execute the executable modules stored in the memory 122, such as the software function modules and computer programs included in the lower limb passive rehabilitation training method using mobile storage media.

[0140] The memory 122 may be, but is not limited to, random access memory (RAM), read-only memory (ROM), programmable read-only memory (PROM), erasable programmable read-only memory (EPROM), electrically erasable programmable read-only memory (EEPROM), etc.

[0141] The control box 120 can be selected according to actual needs. For example, the control box 120 can be selected as a physical control device, and the physical control device integrates a display screen to display attention index, training duration, number of flexion and extension repetitions, etc. The control box 120 also has software or APP (application) that can execute lower limb passive rehabilitation training methods.

[0142] The passive rehabilitation training method for lower limbs provided in this application embodiment can be executed by the processor in the control box 120. The passive rehabilitation training method for lower limbs provided in this application embodiment will be explained further below. Figure 7 This is a flowchart illustrating a passive rehabilitation training method for the lower limbs provided in an embodiment of this application. Figure 7As shown, this method is applied to the aforementioned lower limb passive rehabilitation training system 200, and the method may include: S310. Acquire the first EEG signal of the prefrontal cortex region within a first preset duration while the patient is at rest.

[0143] Among them, the first EEG signal is the resting state reference signal.

[0144] The resting state is the first physiological state in which the patient's whole body muscles are relaxed, and the duration of maintaining the first physiological state in a preset quiet environment is longer than the first preset duration.

[0145] In one possible approach, in a pre-set quiet environment, the patient is allowed to enter a first physiological state of full-body muscle relaxation, and this state must be maintained for more than a first pre-set duration before the first electroencephalogram (EEG) signal of the prefrontal cortex region is collected for the first pre-set duration.

[0146] It should be noted that the patient must remain in a resting state for more than the first preset time before data collection. This is to allow the patient to fully relax and avoid the tension that comes with entering a quiet environment. This ensures that the collected signal reflects the true baseline state of relaxation. The prefrontal cortex is chosen because the alpha and beta waves in the prefrontal cortex are the most prominent and can accurately reflect the patient's state.

[0147] S320: Acquire the second EEG signal of the prefrontal cortex region within a second preset duration in the patient's activated state.

[0148] The second EEG signal is the activation state signal.

[0149] The activated state is the second physiological state in which the patient is in a preset quiet environment and is in the process of passive rehabilitation training of the lower limbs, and is driven by the motion execution module to perform passive flexion and extension movements of the lower limbs.

[0150] The second preset duration is much shorter than the first preset duration.

[0151] In one possible implementation, in a pre-set quiet environment, the patient is placed in a passive lower limb rehabilitation training process, and the motor execution module drives the patient's lower limbs to perform passive flexion and extension movements, while simultaneously collecting a second EEG signal from the prefrontal cortex region for a second pre-set duration.

[0152] It should be noted that the duration of the active state is shorter because the patient's state (activated or fatigued) changes dynamically during training. Shorter acquisition duration can reflect the current state in real time and avoid signal mismatch due to longer acquisition duration. The environment is consistent with the resting state to eliminate the interference of environmental differences on EEG signals and ensure the effectiveness of the comparison between resting and active signals.

[0153] S330. Focus features are extracted from the first EEG signal and the second EEG signal respectively, and the resting state focus index and the active state focus index are calculated based on the extracted focus features.

[0154] Among these, the characteristics of focus include alpha brainwaves and beta brainwaves.

[0155] In one possible implementation, the first and second EEG signals are preprocessed (e.g., baseline drift removal, artifact removal, and filtering) to extract two types of attention features: alpha waves (α waves, 8-12Hz) and beta waves (β waves, 13-30Hz).

[0156] Then, based on the extracted alpha and beta wave features, the resting state concentration index was calculated. and Activated State Concentration Index (As in formula (1) above). This transforms abstract EEG signals into quantifiable and comparable indices.

[0157] Among them, alpha waves represent relaxation, are dominant in the resting state, and have a low index; beta waves represent activation, are dominant in training, and have a high index; the index calculated by the power ratio of β / α can intuitively reflect the degree of activation of patients in the two states, providing data for subsequent state comparison.

[0158] S340. Calculate the relative change rate between active and resting states of the patient based on the resting state attention index and the active state attention index.

[0159] Among them, the activation resting relative change rate It is used to characterize the ratio of the change in the active state concentration index to the resting state concentration index.

[0160] In one possible implementation, based on the resting state concentration index and Activated State Concentration Index The activation-resting relative change rate is calculated using the above formula (2). This method converts the absolute difference between the resting state attention index and the active state attention index into a relative change ratio, avoiding misjudgments due to differences in baseline indices among different patients. For example, patient A's resting state attention index... The active state concentration index is 0.5. If the score is 0.6, then the absolute difference between the two is 0.1; Patient B's resting state attention index A score of 1.0 indicates an activated state of focus. If the absolute difference between the two is 0.1, then the relative change rate is +20%, which better reflects the same level of activation and makes the physiological state judgment more individualized.

[0161] S350: Adjust the training parameters of the motion execution device based on the activation-resting relative rate of change and the preset rate of change threshold.

[0162] The training parameters include: flexion-extension speed and flexion-extension angle range.

[0163] In one possible implementation, since the preset rate of change threshold is divided into a positive threshold, a stable threshold, and a negative threshold, the activation resting relative rate of change is used as the basis for calculation. The matching relationship with the preset rate of change threshold is used to adjust the flexion and extension speed and flexion and extension angle range of the motion execution device.

[0164] For example, when activating the resting relative rate of change If the value is greater than or equal to the positive threshold, the patient is in a highly activated state, and commands for increasing speed and expanding angle parameters are adjusted to enhance training intensity; if the activation resting relative change rate is higher... If the patient's physiological state is within a stable threshold, the system will adapt and output a command to maintain the current parameters, thus keeping the current training parameters unchanged; if the resting relative change rate is activated... If the threshold is ≤ negative, the patient is prone to fatigue, prompting the output to decelerate and maintain angle parameter adjustments to reduce the physical load. This transforms the patient's physiological state into training parameters that the motion execution device can execute, avoiding the drawbacks of traditional fixed-parameter training (such as maintaining high intensity when fatigued and low intensity after adaptation). This allows the training intensity to dynamically adapt to the patient's state, ensuring both rehabilitation efficiency and mitigating safety risks.

[0165] S360 sends control commands to the motion execution device based on the adjusted training parameters, driving the patient to complete passive rehabilitation training of the lower limbs.

[0166] In one possible implementation, the adjusted training parameters are converted into digital instructions (such as PWM drive signals) that can be recognized by the motion execution device. These instructions are then transmitted to the motion execution device via a communication link (such as an RS485 interface), driving the drive components (such as a motor) within the device to move a passive motion lever. Since the passive motion lever is secured to the patient's lower leg with medical straps and supported by the fixed thigh, it assists the patient in completing standardized passive flexion and extension movements of the lower limbs. Simultaneously, angle and speed sensors calibrate the movement accuracy in real time. This ensures that the dynamically adjusted training parameters translate into precise rehabilitation movements. For patients in the early to mid-stages of rehabilitation (those with insufficient muscle strength and unable to move actively), standardized training can be completed without conscious effort, avoiding joint stiffness and muscle atrophy caused by the inability to move actively. It also ensures a complete match between the movements and the adjusted parameters, achieving a closed loop from training parameter adjustment to movement adaptation.

[0167] The passive rehabilitation training method for lower limbs provided in this application is applied to a passive rehabilitation training system for lower limbs, including: acquiring a first electroencephalogram (EEG) signal within a first preset duration in the prefrontal cortex region of the patient in a resting state; the resting state is a first physiological state in which the patient's whole body muscles are relaxed, and the duration of maintaining the first physiological state in a preset quiet environment is greater than the first preset duration; acquiring a second EEG signal within a second preset duration in the prefrontal cortex region of the patient in an activated state, wherein the second preset duration is less than the first preset duration; the activated state is a second physiological state in which the patient, in a preset quiet environment, is in the process of passive rehabilitation training for lower limbs, and is driven by the motion execution module to perform passive flexion and extension movements of the lower limbs; and respectively processing the first EEG signal and the second EEG signal. The signal is used to extract focus features, and resting state focus index and active state focus index are calculated based on the extracted focus features. Focus features include alpha brain waves and beta brain waves. The relative change rate between active and resting states is calculated based on the resting state focus index and the active state focus index. The relative change rate between active and resting states represents the proportion of change in the active state focus index relative to the resting state focus index. The training parameters of the motion execution device are adjusted based on the relative change rate between active and resting states and a preset change rate threshold. The training parameters include flexion-extension speed and flexion-extension angle range. Based on the adjusted training parameters, control commands are sent to the motion execution device to guide the patient through passive lower limb rehabilitation training. Therefore, this application achieves a fully automatic response from physiological state changes to control command output through automatic EEG signal perception, automatic physiological state judgment, and automatic adjustment of training parameters, without requiring active patient participation. This reduces medical intervention costs, avoids rehabilitation risks caused by information asymmetry, and achieves personalized and precise rehabilitation.

[0168] Optionally, embodiments of this application provide a computer-readable storage medium storing a computer program. When the computer program is executed by a processor, it implements the various processes of the passive rehabilitation training method for the lower limbs in the above-described method embodiments and achieves the same technical effect. To avoid repetition, it will not be described again here.

[0169] Optionally, this application embodiment also provides a computer program product, which stores a computer program. When the computer program is executed by a processor, it implements each process of the passive rehabilitation training method for the lower limbs in the above method embodiment and can achieve the same technical effect. To avoid repetition, it will not be described again here.

[0170] Those skilled in the art will understand that embodiments of this application can be provided as methods, systems, or computer program products. Therefore, this application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this application can take the form of a computer program product implemented on one or more computer-usable storage media containing computer-usable program code.

[0171] It should be understood that the phrase "one embodiment" or "an embodiment" throughout the specification means that a specific feature, structure, or characteristic related to the embodiment is included in at least one embodiment of this application. Therefore, "in one embodiment" or "in an embodiment" appearing throughout the specification does not necessarily refer to the same embodiment. Furthermore, these specific features, structures, or characteristics can be combined in any suitable manner in one or more embodiments. Those skilled in the art should also understand that the embodiments described in the specification are all optional embodiments, and the actions and modules involved are not necessarily essential to this application. The above-mentioned multiple embodiments are not necessarily multiple independent embodiments; they are divided into multiple embodiments only to highlight different technical features in different embodiments. Those skilled in the art should understand that the above-mentioned multiple embodiments can also be combined arbitrarily.

[0172] In the various embodiments of this application, it should be understood that the sequence number of each process does not necessarily imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of this application.

[0173] The passive rehabilitation training system, method, and apparatus for the lower limbs provided in this application have been described in detail above. Specific examples have been used to illustrate the principles and implementation methods of this application. The descriptions of the embodiments above are only intended to help understand the method and core ideas of this application. It should be noted that those skilled in the art can make various improvements and modifications to this application without departing from its principles, and these improvements and modifications also fall within the protection scope of the claims of this application.

Claims

1. A passive rehabilitation training system for the lower limbs, characterized in that, include: The resting signal acquisition module is used to acquire the first electroencephalogram (EEG) signal in the prefrontal cortex region of the patient during a first preset duration in a resting state. The resting state is the first physiological state in which the patient's whole body muscles are relaxed, and the duration of maintaining the first physiological state in a preset quiet environment is greater than the first preset duration. The activation signal acquisition module is used to acquire the second electroencephalogram (EEG) signal of the prefrontal cortex region within a second preset duration in the patient's activated state; wherein the second preset duration is less than the first preset duration; the activated state is the second physiological state in which the patient is in the preset quiet environment, is in the process of passive rehabilitation training of the lower limbs, and is driven by the motion execution module to perform passive flexion and extension movements of the lower limbs. The feature extraction module is used to extract focus features from the first EEG signal and the second EEG signal respectively, and to calculate the resting state focus index and the active state focus index based on the extracted focus features; the focus features include: alpha brain waves and beta brain waves; The rate of change calculation module is used to calculate the relative rate of change between active and resting states of the patient based on the resting state attention index and the active state attention index; the relative rate of change between active and resting states is used to characterize the proportion of change of the active state attention index relative to the resting state attention index. The parameter adjustment module is used to adjust the training parameters of the motion execution device according to the activated resting relative rate of change and the preset rate of change threshold; the training parameters include: flexion-extension speed and flexion-extension angle range; The rehabilitation control module is used to send control commands to the motion execution device according to the adjusted training parameters, so as to drive the patient to complete the passive rehabilitation training of the lower limbs.

2. The system according to claim 1, characterized in that, The feature extraction module includes: a preprocessing submodule and a feature extraction submodule; The preprocessing submodule is used to preprocess the first EEG signal and the second EEG signal respectively, and output the preprocessed resting signal and the preprocessed activation signal. The feature extraction submodule is used to extract focus features from the preprocessed resting signal and the preprocessed activation signal, respectively.

3. The system according to claim 2, characterized in that, The preprocessing submodule includes: a baseline drift removal unit, an artifact removal unit, and a filtering unit; The baseline drift removal unit is used to perform baseline drift removal processing on the first EEG signal and the second EEG signal respectively, and output a first resting signal and a first activation signal. The artifact removal unit is used to perform artifact removal processing on the first resting signal and the first activation signal respectively, and output the second resting signal and the second activation signal. The filtering unit is used to filter the second resting signal and the second activation signal respectively, and output the preprocessed resting signal and the preprocessed activation signal.

4. The system according to claim 2, characterized in that, The feature extraction submodule includes: an alpha brainwave extraction unit and a beta brainwave extraction unit; The alpha brainwave extraction unit is used to extract features from the preprocessed resting signal and the preprocessed activation signal using a first preset frequency band, and output resting alpha brainwave and activation alpha brainwave. The beta brainwave extraction unit is used to extract features from the preprocessed resting signal and the preprocessed activation signal using a second preset frequency band, and output resting beta brainwaves and activation beta brainwaves; wherein the first preset frequency band is smaller than the second preset frequency band.

5. The system according to claim 4, characterized in that, The alpha brainwave extraction unit includes: a discrete processing unit, a weighted processing unit, a Fourier transform unit, a power spectral density calculation unit, and a brainwave determination unit; The discrete processing unit is used to perform discrete processing on the preprocessed resting signal and the preprocessed activation signal according to a preset unit time window, and output the first discrete resting signal and the first discrete activation signal. The weighting processing unit is used to multiply the first discrete resting signal and the first discrete activation signal by a preset Hanning window function, respectively, and output the first weighted resting signal and the first weighted activation signal. The Fourier transform unit is used to perform a preset Fourier transform on the first weighted resting signal and the first weighted activated signal respectively, and output the first resting frequency domain sequence and the first activated frequency domain sequence. The power spectral density calculation unit is used to calculate the first resting frequency domain sequence and the first active frequency domain sequence respectively, and output the first resting power spectral density and the first active power spectral density. The brainwave determination unit is used to determine the resting alpha brainwave and the activated alpha brainwave based on the first resting power spectral density and the first activated power spectral density.

6. The system according to claim 1, characterized in that, The preset rate of change threshold includes: a positive threshold, a stable threshold, and a negative threshold; The parameter adjustment module includes: an adjustment module; The adjustment module is used to receive the activation-resting relative change rate transmitted by the change rate calculation module; when the activation-resting relative change rate is greater than or equal to the positive threshold, it outputs the speed-up and angle-expansion parameter adjustment instructions to increase the lower limb flexion-extension speed of the motion execution device by a preset speed-up value on the current flexion-extension speed, and at the same time expand the flexion-extension angle range by a preset angle-expansion value. When the activation-resting relative change rate is at the stability threshold, an instruction to maintain the current parameters is output to keep the current training parameters unchanged. When the relative change rate at rest is less than or equal to the negative threshold, a deceleration and angle parameter adjustment command is output to reduce the lower limb flexion and extension speed of the motion execution device by a preset deceleration value from the current flexion and extension speed, while keeping the flexion and extension angle range unchanged.

7. The system according to claim 6, characterized in that, The parameter adjustment module further includes: a pause / restart submodule; The pause / restart submodule is used to continuously receive the relative change rate between activation and rest. When the parameter adjustment module outputs the deceleration and hold angle parameter adjustment instructions, if the relative change rate of the activation and resting states corresponding to a preset number of consecutive preset unit time windows is less than or equal to the negative threshold, then the pause restart submodule sends a pause and restart control instruction to the rehabilitation control module.

8. A passive rehabilitation training method for the lower limbs, characterized in that, The lower limb passive rehabilitation training system described in any one of claims 1 to 7 comprises: Acquire the first EEG signal in the prefrontal cortex region of the patient in a resting state within a first preset duration; the resting state is the first physiological state in which the patient's whole body muscles are relaxed, and the duration of maintaining the first physiological state in a preset quiet environment is greater than the first preset duration. Acquire a second EEG signal within a second preset duration in the prefrontal cortex region of the patient in an activated state, wherein the second preset duration is less than the first preset duration; the activated state is the second physiological state in which the patient is in a preset quiet environment, and is in the process of passive rehabilitation training of the lower limbs, and is driven by the motor execution module to perform passive flexion and extension movements of the lower limbs. Focus features were extracted from the first and second EEG signals respectively, and the resting state focus index and the active state focus index were calculated based on the extracted focus features; the focus features included: alpha brain waves and beta brain waves; Based on the resting state attention index and the activated state attention index, the relative change rate of the patient's activated state attention index is calculated; the relative change rate of the activated state attention index is used to characterize the proportion of change of the activated state attention index relative to the resting state attention index. The training parameters of the motion execution device are adjusted based on the activated resting relative rate of change and a preset rate of change threshold; the training parameters include: flexion-extension speed and flexion-extension angle range. Based on the adjusted training parameters, control commands are sent to the motion execution device to guide the patient to complete passive rehabilitation training of the lower limbs.

9. A passive rehabilitation training device for the lower limbs, characterized in that, The lower limb passive rehabilitation training device integrates the lower limb passive rehabilitation training system according to any one of claims 1 to 7, and the lower limb passive rehabilitation training device includes: a headband acquisition device, a control box, and a motion execution device; the motion execution device includes: a fixed drive rod, a drive assembly, and a passive moving rod. The headband data acquisition device is communicatively connected to the control box; The control box is electrically connected to the drive assembly; One end of the drive assembly is hinged to the fixed drive rod, and the other end of the drive assembly is hinged to the passive movable rod to form a three-bar linkage structure. The fixed drive rod is used to be fixedly connected to the preset nursing equipment. The end of the passive moving rod away from the drive assembly is provided with a medical strap for fixing the patient's lower leg; the other end of the fixed drive rod away from the drive assembly is provided with a medical strap for fixing the patient's thigh.

10. The apparatus according to claim 9, characterized in that, The control box includes: Memory containing executable program code; and the processor coupled to the memory; The processor calls the executable program code stored in the memory to execute the lower limb passive rehabilitation training method as described in claim 8.

Images