Brain-computer interface assisted patient gait training system

Abstract

The technical scheme of the invention discloses a brain-computer interface assisted patient gait training system. The brain-computer interface system is combined to provide the accuracy of the brain motion intention, the infrared optical motion capture system is adopted to capture reflective identification points on the human body through infrared rays emitted by the lens, and accurate and real-time motion information is collected and generated. The gait of a standing or walking training patient under the suspended sky rail is tracked and measured in real time, and the track of a three-dimensional space point, the motion posture of a rigid body and the action of a human body are recorded as real visual feedback of a lower limb training brain-computer interface of the dyskinesia patient. It is guaranteed that when the brain of a patient outputs a motion intention, input from the real motion ability of the patient can be obtained, peripheral electrical stimulation or epidural electrical stimulation is combined as tactile feedback, and therefore precise closed-loop regulation and control of the brain on the output of a motion control signal of the paralyzed lower limb are promoted.

Description

Technical Field

[0001] The present invention relates to a brain-computer interface-assisted gait training system for patients suspended on a ceiling rail based on real-time feedback from infrared optical motion capture mirrors, which is used to accelerate the recovery of lower limb function in patients with lower limb movement disorders caused by stroke, spinal cord injury or other neurological diseases. Background Art

[0002] Whether it is vascular disease, trauma, infection or ischemia and hypoxia, all factors may cause damage to the central nervous system. According to statistics, worldwide, stroke has become one of the three major diseases that increase national mortality rates, and the total number of SCI patients has exceeded 20 million. Whether it is hemiplegia after brain injury or paraplegia after spinal cord injury, lower limb motor dysfunction is a key factor that affects the patient's quality of life and reduces the patient's ability to live independently. At present, the clinical treatment for lower limb motor dysfunction mainly includes rehabilitation techniques, exoskeleton robots, central and peripheral nerve regulation technology, etc. Although the current rehabilitation treatment methods can improve the patient's limb hemiplegia to a certain extent, the recovery of most patients gradually slows down and gradually enters a plateau period 3-6 months after the onset of the disease. The complete recovery of motor function after damage to the nervous system is still a difficult problem in clinical practice.

[0003] However, the prior art has the following deficiencies:

[0004] ① Insufficient feedback authenticity: It often relies on virtual scenes (such as computer-animated lower limb movements), which cannot accurately reflect the patient's actual movement trajectory, resulting in a disconnect between virtual and actual movements and poor immersion;

[0005] ② Imperfect closed-loop control: Lack of accurate feedback on the patient's actual movement completion makes it difficult to form an effective closed loop of "brain intention-peripheral action-feedback correction", limiting the effect of inducing neuroplasticity;

[0006] ③ Lack of evaluation parameters: It is impossible to provide real-time, quantitative kinematic parameters (such as joint angles and step length symmetry), which is not conducive to training effect evaluation and program adjustment.

[0007] Therefore, there is an urgent need for a training system that can provide real dynamic feedback and perfect closed-loop control to break through the existing technical bottleneck and accelerate the recovery of patients' lower limb function. Summary of the Invention

[0008] The technical problem to be solved by the present invention is that the existing lower limb exercise rehabilitation training system has unrealistic virtual feedback, low decoding accuracy and imperfect closed-loop control.

[0009] In order to solve the above technical problems, the technical solution of the present invention is to disclose a brain-computer interface assisted patient gait training system, which is characterized by comprising:

[0010] The overhead rail suspension weight-reduction module is used to provide patients with adjustable weight-reduction support to maintain balance and safety during standing and walking;

[0011] The EEG signal acquisition and decoding module is used to collect EEG signals from the patient's motor cortex area and extract lower limb movement intention parameters through algorithm decoding;

[0012] Infrared optical motion capture module, used to capture the three-dimensional spatial coordinates and motion trajectory of key nodes of the patient's lower limbs in real time and generate kinematic data;

[0013] A mirror feedback generation module is used to construct a three-dimensional mirror model of the patient's lower limbs based on the kinematic data of the infrared optical motion capture module, and generate multimodal feedback information after comparing it with a preset standard gait model;

[0014] The electrical stimulation feedback module is used to output electrical stimulation signals to assist the patient in completing lower limb movements based on the decoding results of the EEG signal acquisition and decoding module and the comparison results of the mirror feedback generation module;

[0015] The training control module is connected to the ceiling rail suspension weight reduction module, the EEG signal acquisition and decoding module, the infrared optical motion capture module, the mirror feedback generation module and the electrical stimulation feedback module respectively, and dynamically adjusts the training parameters based on the EEG decoding results, motion capture data and feedback information.

[0016] Preferably, the overhead rail suspension weight reduction module includes a track, a sliding device arranged on the track, a weight reduction drive unit fixedly connected to the sliding device, and an adjustable sling fixedly connected to the weight reduction drive unit, wherein: the adjustable sling is fixed to the patient's pelvis and torso; the weight reduction drive unit can adjust the weight reduction ratio to 0-50% of body weight, providing weight reduction support of 0-50% of body weight.

[0017] Preferably, the EEG signal acquisition and decoding module includes an EEG acquisition device and a decoding unit connected to the EEG acquisition device, wherein: the EEG acquisition device adopts a 32-lead or above non-invasive scalp EEG cap; the decoding unit adopts a deep learning algorithm to extract features of μ rhythm and β rhythm related to motor imagery, and decode lower limb movement intention parameters including step direction, expected step length and force intensity.

[0018] Preferably, the electrodes of the non-invasive scalp EEG cap are distributed to cover the motor cortex corresponding to the C3, C4, and Cz leads in the international 10-20 system, with a sampling frequency of ≥500 Hz and a built-in anti-motion artifact algorithm.

[0019] Preferably, the infrared optical motion capture module includes 8-12 infrared cameras, reflective marking points and a data processing unit connected to the infrared cameras, wherein: the reflective marking points are affixed to the bony landmarks of the patient's lower limbs, including the bilateral anterior superior iliac spines, greater trochanter of the femur, lateral femoral condyle, lateral tibial condyle, lateral malleolus, heel and toe; the data processing unit generates kinematic parameters including joint angles, step length, step width, and center of gravity displacement trajectory in real time based on the data obtained by the infrared camera.

[0020] Preferably, the mirror feedback generation module includes a three-dimensional modeling unit, a comparison unit and a feedback output unit, wherein: the three-dimensional modeling unit generates a 1:1 scale three-dimensional mirror model of the patient's lower limbs in a virtual environment based on kinematic data; the comparison unit compares the actual joint angle with the standard gait model and calculates the deviation value; the feedback output unit outputs multimodal feedback through vision and hearing.

[0021] Preferably, the electrical stimulation feedback module includes surface electrodes or implantable electrodes used for severely injured patients and a stimulation control unit, wherein: the surface electrodes are attached to target muscle groups including the quadriceps femoris and tibialis anterior; the implantable electrodes are spinal epidural electrodes; the stimulation control unit outputs electrical stimulation, and the stimulation is triggered when the movement intention decoding match is ≥80% or the movement deviation is ≤10°, thereby strengthening the "intention-action" association by inducing muscle contraction.

[0022] Preferably, the adjustment parameters of the training control module include: overhead rail suspension weight reduction ratio, electrical stimulation intensity, feedback form and training difficulty.

[0023] The present invention combines a brain-computer interface system to provide accuracy in brain movement intentions. It uses an infrared optical motion capture system to capture reflective markers on the human body through infrared rays emitted by the lens, and collects and generates accurate, real-time motion information. The gait of patients undergoing standing or walking training under the overhead rail is tracked and measured in real time, and the trajectory of three-dimensional space points, the motion posture of rigid bodies, and the movements of the human body are recorded. This serves as real visual feedback for the brain-computer interface of lower limb training for patients with movement disorders, ensuring that the patient's brain can obtain input from their own real motor ability when outputting movement intentions. Combined with peripheral electrical stimulation or epidural electrical stimulation as tactile feedback, this promotes precise closed-loop regulation of the brain's output of movement control signals for paralyzed lower limbs.

[0024] Compared with the existing technical solutions, the present invention has the following beneficial effects:

[0025] (1) Feedback authenticity and immersion: Infrared optical motion capture is used to generate a real-time three-dimensional mirror image of the patient's lower limbs, solving the problem of disconnection between virtual scenes and actual movements. Patients can directly observe their own movement deviations and improve their training initiative.

[0026] (2) Closed-loop efficiency: Deep learning algorithms are used to improve the accuracy of EEG intention decoding, and combined with electrical stimulation feedback to strengthen the "intention-action" association, improve brain-periphery closed-loop regulation, and accelerate the induction of neural plasticity;

[0027] (3) Personalization: The training control module dynamically adjusts the weight loss ratio, stimulation intensity, and difficulty to suit the functional level of different patients;

[0028] (4) Quantitative evaluation and program optimization: Provide accurate kinematic parameters such as joint angles and step length symmetry to provide data support for efficacy evaluation and training program adjustment, and avoid subjective judgment bias. BRIEF DESCRIPTION OF THE DRAWINGS

[0029] Figure 1 This is a flow chart of the brain-computer interface-assisted gait training system for patients suspended under a ceiling rail based on infrared optical motion capture mirror real-time feedback of the present invention.

[0030] Figure 2 This is a module schematic diagram of the brain-computer interface-assisted gait training system for patients suspended under a ceiling rail based on infrared optical motion capture and real-time feedback.

[0031] Figure 3 This is a flowchart of the implementation of the brain-computer interface-assisted patient gait training system under ceiling rail suspension based on real-time feedback of infrared optical motion capture mirror. DETAILED DESCRIPTION

[0032] Below in conjunction with specific embodiment, further set forth the present invention.Should be understood that these embodiments are only used to illustrate the present invention and are not used in limiting the scope of the present invention.In addition, should be understood that after reading the content taught by the present invention, those skilled in the art can make various changes or modifications to the present invention, and these equivalent forms fall equally within the scope limited by the appended claims of the application.

[0033] An embodiment of the present invention discloses a brain-computer interface-assisted gait training system for patients under ceiling rail suspension based on real-time feedback from infrared optical motion capture mirrors. By integrating ceiling rail suspension, EEG decoding, infrared motion capture, multimodal feedback and electrical stimulation assistance, a precise, immersive, closed-loop training system is constructed to improve rehabilitation efficacy. Specifically, it includes the various modules described below.

[0034] The overhead rail suspension weight loss module is used to provide patients with adjustable weight loss support to maintain balance and safety during standing and walking, and further includes a track, a sliding device, an adjustable sling and a weight loss drive unit. The adjustable sling is fixed to the patient's pelvis and trunk. By adjusting the sling tension, the patient's standing balance is maintained, the weight load on the lower limbs is reduced, and excessive fatigue is avoided. It is suitable for patients with different muscle strength levels. The weight loss drive unit is connected to the adjustable sling and can be moved along the track through a sliding device. In an embodiment of the present invention, the weight loss drive unit can adjust the weight loss ratio to 0-50% of body weight, and can provide weight loss support of 0-50% of body weight.

[0035] The EEG signal acquisition and decoding module is used to collect EEG signals from the patient's motor cortex area and extract the lower limb movement intention parameters through algorithm decoding. It further includes an EEG acquisition device and a decoding unit connected to the EEG acquisition device. The EEG acquisition device uses a 32-lead or above non-invasive scalp EEG cap, and the electrode distribution covers the C3 (left motor cortex), C4 (right motor cortex), Cz (central motor area) and other leads in the international 10-20 system. The sampling frequency is ≥500Hz (in this embodiment, the sampling frequency is 500Hz), and the anti-motion artifact algorithm (such as ICA independent component analysis) is built-in. The decoding unit uses a deep learning algorithm (for example: LSTM - long short-term memory network) to extract features of the μ rhythm (8-12Hz) and β rhythm (13-30Hz) related to motor imagery, and decode the lower limb movement intention parameters, including the direction of walking, expected step length and force intensity. The decoding delay is ≤100ms, and the accuracy rate is stable after training ≥85%.

[0036] The infrared optical motion capture module is used to capture the three-dimensional spatial coordinates and motion trajectories of the key nodes of the patient's lower limbs in real time and generate kinematic data. It further includes 8-12 infrared cameras (in the embodiment of the present invention, 10 OptiTrack series infrared cameras are used, distributed in a circle in a 4m×4m training area), reflective marking points and a data processing unit connected to the infrared cameras. The reflective marking points are affixed to the bony landmarks of the patient's lower limbs, including the bilateral anterior superior iliac spines, greater trochanter of the femur, lateral femoral condyle, lateral tibial condyle, lateral malleolus, heel and toe (in the embodiment of the present invention, 14 reflective marking points are affixed to the bony landmarks of the lower limbs, including the bilateral anterior superior iliac spines, greater trochanter of the femur, lateral femoral condyle, lateral tibial condyle, lateral malleolus, heel and toe, to ensure full capture of the three-dimensional movements of the hip joint, knee joint and ankle joint). The sampling frequency of the infrared camera is ≥120Hz, and the positioning accuracy is ≤0.5mm (in the embodiment of the present invention, the sampling frequency of the infrared camera is 120Hz, and the positioning accuracy is 0.3mm). The data processing unit generates kinematic parameters such as joint angles (such as hip flexion and extension angles, knee flexion and extension angles, ankle plantar flexion / dorsiflexion angles), step length, step width, and center of gravity displacement trajectory in real time based on the data obtained by the infrared camera.

[0037] The mirror feedback generation module is used to construct a three-dimensional mirror model of the patient's lower limbs based on the kinematic data of the infrared optical motion capture module, and generate multimodal feedback information after comparing it with the preset standard gait model. It further includes a three-dimensional modeling unit, a comparison unit and a feedback output unit. The three-dimensional modeling unit generates a 1:1 scale three-dimensional mirror model of the patient's lower limbs in a virtual environment based on the kinematic data, and the movement synchronization delay is ≤50ms. The comparison unit compares the actual joint angle with the standard gait model (a database of healthy people of the same age group) and calculates the deviation value. The feedback output unit outputs multimodal feedback through visual and auditory means, including color labeling, numerical quantization and voice prompts. In the embodiment of the present invention: the color labeling can be: green for deviation ≤5°, yellow for 5°-15°, and red for >15°; the numerical quantization can be: displaying the specific deviation value and parameters such as step length and step frequency; the voice prompt can be: playing guiding words in real time.

[0038] The electrical stimulation feedback module is used to output electrical stimulation signals to assist patients in completing lower limb movements based on the decoding results of the EEG signal acquisition and decoding module and the comparison results of the mirror feedback generation module, and further includes surface electrodes or implantable electrodes and a stimulation control unit. The surface electrodes are attached to target muscle groups such as the quadriceps femoris (knee extension) and tibialis anterior (ankle dorsiflexion), or for severely injured patients, implantable electrodes are used, and the implantable electrodes are spinal epidural electrodes. The stimulation control unit outputs electrical stimulation, and the stimulation parameters are: intensity 10-30mA (within the patient's tolerance range), pulse width 100-300μs (in the embodiment of the present invention, the pulse width can be set to 200μs). Stimulation is triggered when the motor intention decoding match is ≥80% or the movement deviation is ≤10°, and the "intention-action" association is strengthened by inducing muscle contraction.

[0039] The training control module is connected to the overhead rail suspension weight reduction module, the EEG signal acquisition and decoding module, the infrared optical motion capture module, the mirror feedback generation module, and the electrical stimulation feedback module. It dynamically adjusts training parameters based on EEG decoding results, motion capture data, and feedback information. The training control module's adjustment parameters include: the overhead rail suspension weight reduction ratio (dynamically adjusted based on balance state), electrical stimulation intensity (graded and adjusted based on movement target achievement), feedback form (switching visual / audio weighting according to the training stage), and training difficulty (increasing step length / angle requirements based on decoding accuracy). For example: Weight reduction ratio: Appropriately reduce by 5-10% when balance is stable, down to a minimum of 10%; Electrical stimulation intensity: Reduce intensity when movement target achievement is high, and increase it when it is low; Training difficulty: Increase step length requirements or introduce terrain changes when decoding accuracy is ≥90%; Closed-loop logic: Update control instructions every 10ms to ensure smooth and adaptable training.

[0040] Combine Figure 1 、 3The application implementation process of a brain-computer interface-assisted gait training system for patients suspended under a ceiling rail based on infrared optical motion capture and real-time feedback disclosed in an embodiment of the present invention specifically includes the following steps:

[0041] S1. The patient wears EEG acquisition equipment and reflective markers, and obtains initial weight loss support through the ceiling rail suspension weight loss module;

[0042] S2: The system starts, the infrared optical motion capture module begins to collect lower limb motion data, and the mirror feedback generation module synchronously builds a real-time three-dimensional mirror;

[0043] S3: The patient first performs basic imagery training without movement requirements to familiarize himself with the feedback. On this basis, the patient either performs lower limb gait training based on the standard gait or performs autonomous control training, and the real-time EEG signal acquisition and decoding module outputs movement intention parameters in real time;

[0044] S4, the training control module combines the movement intention parameters and the action deviation of the mirror feedback, and outputs auxiliary stimulation through the electrical stimulation feedback module to induce muscle contraction;

[0045] S5: The patient adjusts their movements based on multimodal feedback, while the system records training data including intention-action matching and joint angle range.

[0046] S6. Adjust the weight loss ratio, electrical stimulation intensity, and training difficulty based on training data;

[0047] S7. Create a training report.

Claims

1. A brain-computer interface assisted patient gait training system, characterized in that: include: The overhead rail suspension weight-reduction module is used to provide patients with adjustable weight-reduction support to maintain balance and safety during standing and walking; The EEG signal acquisition and decoding module is used to collect EEG signals from the patient's motor cortex area and extract lower limb movement intention parameters through algorithm decoding; Infrared optical motion capture module, used to capture the three-dimensional spatial coordinates and motion trajectory of key nodes of the patient's lower limbs in real time and generate kinematic data; A mirror feedback generation module is used to construct a three-dimensional mirror model of the patient's lower limbs based on the kinematic data of the infrared optical motion capture module, and generate multimodal feedback information after comparing it with a preset standard gait model; The electrical stimulation feedback module is used to output electrical stimulation signals to assist the patient in completing lower limb movements based on the decoding results of the EEG signal acquisition and decoding module and the comparison results of the mirror feedback generation module; The training control module is connected to the ceiling rail suspension weight reduction module, the EEG signal acquisition and decoding module, the infrared optical motion capture module, the mirror feedback generation module and the electrical stimulation feedback module respectively, and dynamically adjusts the training parameters based on the EEG decoding results, motion capture data and feedback information.

2. A brain-computer interface assisted patient gait training system as claimed in claim 1, characterized in that: The overhead rail suspension weight reduction module includes a track, a sliding device arranged on the track, a weight reduction drive unit fixedly connected to the sliding device, and an adjustable sling fixedly connected to the weight reduction drive unit, wherein: the adjustable sling is fixed to the patient's pelvis and torso; the weight reduction drive unit can adjust the weight reduction ratio to 0-50% of body weight, providing weight reduction support of 0-50% of body weight.

3. The brain-computer interface assisted patient gait training system according to claim 1, characterized in that: The EEG signal acquisition and decoding module includes an EEG acquisition device and a decoding unit connected to the EEG acquisition device, wherein: the EEG acquisition device adopts a 32-lead or above non-invasive scalp EEG cap; the decoding unit adopts a deep learning algorithm to extract features of μ rhythm and β rhythm related to motor imagery, and decode lower limb movement intention parameters including step direction, expected step length and force intensity.

4. A brain-computer interface assisted patient gait training system as claimed in claim 3, characterized in that: The electrodes of the non-invasive scalp EEG cap are distributed over the motor cortex corresponding to the C3, C4, and Cz leads in the international 10-20 system, with a sampling frequency of ≥500 Hz and a built-in anti-motion artifact algorithm.

5. The brain-computer interface assisted patient gait training system according to claim 1, characterized in that: The infrared optical motion capture module includes 8-12 infrared cameras, reflective marking points and a data processing unit connected to the infrared cameras, wherein: the reflective marking points are affixed to the bony landmarks of the patient's lower limbs, including the bilateral anterior superior iliac spines, greater trochanter of the femur, lateral femoral condyle, lateral tibial condyle, lateral malleolus, heel and toe; the data processing unit generates kinematic parameters including joint angles, step length, step width, and center of gravity displacement trajectory in real time based on the data obtained by the infrared camera.

6. The brain-computer interface assisted patient gait training system according to claim 1, characterized in that: The mirror feedback generation module includes a three-dimensional modeling unit, a comparison unit and a feedback output unit, wherein: the three-dimensional modeling unit generates a 1:1 three-dimensional mirror model of the patient's lower limbs in a virtual environment based on kinematic data; the comparison unit compares the actual joint angles with the standard gait model and calculates the deviation value; the feedback output unit outputs multimodal feedback through vision and hearing.

7. The brain-computer interface assisted patient gait training system according to claim 1, characterized in that: The electrical stimulation feedback module includes surface electrodes or implantable electrodes used for severely injured patients and a stimulation control unit, wherein: the surface electrodes are attached to target muscle groups including the quadriceps femoris and tibialis anterior; the implantable electrodes are spinal epidural electrodes; the stimulation control unit outputs electrical stimulation, and the stimulation is triggered when the movement intention decoding match is ≥80% or the movement deviation is ≤10°, thereby strengthening the "intention-action" association by inducing muscle contraction.

8. The brain-computer interface assisted patient gait training system according to claim 1, characterized in that: The adjustment parameters of the training control module include: the weight reduction ratio of the overhead rail suspension, the intensity of electrical stimulation, the feedback form and the training difficulty.

Images