A method and system for motion rehabilitation

Abstract

The application discloses a kind of sports rehabilitation method and system, it is related to the technical field of rehabilitation medical engineering, and its technical scheme main points are: obtaining brain signal and healthy side limb movement signal, and pre-processing, obtain the brain signal and healthy side limb movement signal after pre-processing;Feature is extracted to the brain signal and healthy side limb movement signal after pre-processing, obtain multimodal feature, and the multimodal feature is fused, and obtain characteristic vector sequence;Characteristic vector sequence is input into model, and obtain affected side target control quantity;Affected side target control quantity drives affected side training device, and carries out rehabilitation exercise.The application introduces intention gating mechanism, binds stimulus output with the central motor intention of user, combines healthy side movement mode as physiological template, significantly improves the initiative participation of patient and the timing coupling between brain-muscle, more in line with neural plasticity principle;Realize closed-loop adaptive training, can be through quantitative evaluation affected side execution result, dynamically adjust auxiliary intensity, achieve on-demand assistance.

Description

Technical Field

[0001] This invention relates to the field of rehabilitation medicine engineering technology, and more specifically, to a method and system for sports rehabilitation. Background Technology

[0003] Currently, the mainstream rehabilitation methods for unilateral upper limb and hand dysfunction have obvious limitations: (1) Traditional occupational therapy (OT) / physical therapy (PT): As the basic program for rehabilitation training, the core is to induce neural function remodeling through passive activities, active assistance training, and task-oriented training under the guidance of one-on-one therapists. However, the training effect is highly dependent on the therapist's professional experience and time investment. The duration of a single training session is usually only 30-45 minutes. The daily training repetition dose is far lower than the thousand-times exercise threshold required for neural remodeling. Moreover, the training intensity and repetition are limited by manpower, which cannot cover the training needs of the patient throughout the entire recovery period. It is even more difficult to support long-term, high-frequency training at home outside the hospital. (2) Functional electrical stimulation (FES / NMES): Peripheral motor nerves are stimulated by low-intensity current with preset parameters to induce target muscle contraction and joint movement. It can help patients retain muscle volume and relieve muscle spasms in the early stage of intervention. However, most existing products use a fixed rhythmic periodic stimulation mode or only support simple manual button triggering. The stimulation activation logic is disconnected from the patient's active movement intention. The patient's central nervous system active participation is less than 30%. Long-term use is prone to passive movement dependence, which will inhibit the remodeling of active motor pathways and even aggravate abnormal movement patterns. (3) Mirror therapy or bilateral synergistic training: Based on the mirror neuron mechanism, the motor cortex of the corresponding cerebral cortex on the affected side is activated by the synchronous movement of the healthy side limbs to induce central nervous system function reorganization. However, this type of program only focuses on activation guidance at the central level and lacks quantitative assessment and active intervention of the motor execution end on the affected side. It cannot dynamically adjust the training standards according to the actual motor performance of the affected side. The training process does not form an effective closed loop, and the intervention effect on patients with moderate to severe functional impairment is particularly limited.

[0004] The current rehabilitation techniques still have several common pain points and cannot meet the actual needs of clinical practice and patients: (1) the stimulation trigger is disconnected from the patient's movement intention, making it difficult to promote the reconstruction of the cortex-spinal cord-muscle pathway; (2) there is a lack of healthy-affected side coordination mechanism, and the real movement information of the healthy side is not fully utilized; (3) there is a lack of real-time feedback and adaptive adjustment, and the stimulation parameters are mostly fixed or manually adjusted; (4) the recovery effect of fine hand function is limited, and it is difficult to achieve functional training such as finger separation and thumb opposition.

[0005] In view of this, the present invention proposes a method and system for sports rehabilitation. Summary of the Invention

[0006] The purpose of this invention is to provide a motor rehabilitation method and system that integrates a closed-loop rehabilitation system and method that combines central motor intention, demonstration of the healthy limb, and execution feedback of the affected side, so as to improve the efficiency and quality of upper limb and hand function rehabilitation for hemiplegic patients.

[0007] The above-mentioned technical objective of the present invention is achieved through the following technical solution:

[0008] The first aspect of this invention provides a sports rehabilitation method, comprising the following steps:

[0009] Brain signals and motor signals of the unaffected limbs were acquired and preprocessed to obtain preprocessed brain signals and motor signals of the unaffected limbs.

[0010] Features are extracted from the preprocessed brain signals and the motor signals of the unaffected limbs to obtain multimodal features. The multimodal features are then fused to obtain a feature vector sequence.

[0011] Input the feature vector sequence into the model to obtain the target control quantity for the affected side;

[0012] The target control quantity on the affected side drives the training device on the affected side to perform rehabilitation exercises.

[0013] In conjunction with the first aspect, the present invention further includes the following steps: collecting feedback signals from the affected side during rehabilitation exercises and updating the model.

[0014] In conjunction with the first aspect, the present invention further includes: the brain signals include brain signals related to motor attempts or motor imagery; the healthy limb movement signals include healthy IMU signals and healthy sEMG signals.

[0015] In conjunction with the first aspect, the present invention further includes the following: the preprocessing includes one or more of the following: data cleaning, feature extraction, normalization, and synchronization and alignment of multimodal data.

[0016] In conjunction with the first aspect, the present invention further specifies that the model includes a traditional classification model, a regression model, or a deep learning model.

[0017] In conjunction with the first aspect, the present invention further includes: the affected side training device includes one or more of a rehabilitation robot, an exoskeleton assist, and a functional electrical stimulation (FES) assist.

[0018] A second aspect of the present invention also provides an apparatus / device / system for a sports rehabilitation method and system, comprising a memory, a processor, and a computer program stored in the memory, wherein the processor executes the computer program to implement the steps of the above-described method.

[0019] A third aspect of the present invention also provides a computer-readable storage medium having a computer program / instructions stored thereon, which, when executed by a processor, implement the steps of the above-described method.

[0020] A fourth aspect of the present invention also provides a computer program product, including a computer program / instructions that, when executed by a processor, implement the steps of the above-described method.

[0021] In summary, the present invention has the following beneficial effects:

[0022] (1) Increase the participation of the central nervous system through brain signal intention gating;

[0023] (2) Construct a closed loop of coordination between the healthy side and the affected side to avoid passive stimulation;

[0024] (3) Achieve on-demand assistance and adaptive adjustment to avoid overstimulation;

[0025] (4) It is beneficial for the recovery of fine hand function;

[0026] (5) High safety and strong clinical feasibility;

[0027] In summary, this invention improves the initiative, targeting, and effectiveness of upper limb and hand function rehabilitation for hemiplegic patients without increasing clinical risks. Attached Figure Description

[0028] Figure 1 This is a flowchart of the sports rehabilitation method in an embodiment of the present invention;

[0029] Figure 2 This describes the electromyography (EMG) signal acquisition process of the healthy hand and upper limb in this embodiment of the invention.

[0030] Figure 3 This describes the EEG cap acquisition process in this embodiment of the invention.

[0031] Figure 4 This refers to the EEG cap signal collection in this embodiment of the invention;

[0032] Figure 5 This is the background analysis and processing of electroencephalogram (EEG) signals and electromyogram (EMG) signals from the healthy side in the embodiments of the present invention;

[0033] Figure 6 In this embodiment of the invention, electroencephalogram (EEG) signals and electromyogram (EMG) signals are used to stimulate the affected limb via a converter;

[0034] Figure 7 In this embodiment of the invention, tasks are used to stimulate patients' enthusiasm for exercise, data is collected and fed back in the background, and parameters are automatically adjusted based on the feedback. Detailed Implementation

[0035] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0036] Example 1:

[0037] A type of sports rehabilitation method, such as Figure 1 As shown, it includes the following steps:

[0038] S100. Acquire brain signals and motor signals of the unaffected limbs, and perform preprocessing to obtain preprocessed brain signals and motor signals of the unaffected limbs;

[0039] S200. Extract features from the preprocessed brain signals and the motor signals of the unaffected limbs to obtain multimodal features, and fuse the multimodal features to obtain a feature vector sequence;

[0040] S300. Input the feature vector sequence into the model to obtain the target control quantity for the affected side;

[0041] S400. Drive the target control quantity on the affected side to the training device on the affected side for rehabilitation exercises;

[0042] S500 collects feedback signals from the affected side during rehabilitation exercises and updates the model accordingly.

[0043] In step S100, the brain signal is the electroencephalogram (EEG) signal; the healthy limb motion signal includes the kinematic signal of the healthy inertial measurement unit (IMU) and the healthy surface electromyography (sEMG) signal. The EEG, IMU, and sEMG acquisition devices are synchronously acquired through a unified clock or hardware triggering, and all data are timestamped for subsequent alignment.

[0044] like Figure 3 and Figure 4 As shown, the S110 EEG signal acquisition (brain signal sensor)

[0045] EEG data is acquired through an EEG acquisition system, which includes EEG electrodes (wet or dry electrodes), an amplifier, and an analog-to-digital converter.

[0046] Electrode arrangement can use a 10-20 or 10-10 system;

[0047] Preferably, the motor cortex-related leads are covered, including but not limited to C3, C4, Cz and adjacent leads;

[0048] The preferred sampling rate is 250–1000 Hz.

[0049] S111 healthy side sEMG signal acquisition (one of the healthy side motion signal sensors)

[0050] Healthy side sEMG is obtained via surface electromyography electrodes / patches:

[0051] Electrodes are placed on the surface of target muscle groups on the healthy side that are related to rehabilitation movements (e.g., muscles related to upper limb flexion and extension, wrist and finger movements; or muscles related to lower limb gait).

[0052] Data is acquired using a bipolar or multi-channel method;

[0053] The preferred sampling rate is 1000–2000 Hz.

[0054] S112 Healthy Side IMU Signal Acquisition (Second Healthy Side Motion Signal Sensor)

[0055] The healthy side IMU acquires data through an inertial measurement unit fixed to the healthy side limb segment. The IMU includes at least a three-axis accelerometer and a three-axis gyroscope (optional magnetometer).

[0056] The IMU is fixed to a key limb segment on the healthy side (e.g., upper limb: upper arm, forearm; lower limb: thigh, calf, dorsum of foot).

[0057] The preferred sampling rate is 50–200 Hz.

[0058] In step S100, because EEG suffers from power frequency interference, electrooculography / electromyography contamination, motion artifacts, and baseline drift; sEMG suffers from power frequency interference, motion artifacts, and electrode contact variations; and IMU suffers from high-frequency jitter and gyroscope zero-bias drift, and because the three have different sampling rates, this embodiment performs preprocessing on the acquired signals to obtain stable inputs that can be used for subsequent decoding, mapping, and control. The preprocessing includes one or more of the following: data cleaning, feature extraction, normalization, and synchronization and alignment of multimodal data, as detailed below:

[0059] S121 EEG Pretreatment

[0060] Perform one or more of the following processes on the EEG signal:

[0061] Bandpass filter: 0.5–40 Hz or 1–45 Hz;

[0062] Power frequency notch: 50 Hz (or 60 Hz) notch;

[0063] DC removal and trend removal: Removing baseline drift;

[0064] Heavy reference: average reference or CAR (common average reference);

[0065] Artifact suppression: Threshold rejection of bad segments and / or ICA separation and removal of blinking, eye movement and obvious electromyographic contamination components are employed;

[0066] Segmentation: Segment by sliding window (window length 200–1000 ms, step size 50–200 ms);

[0067] Standardization: Perform z-score standardization or segment-based normalization on each channel.

[0068] S122 healthy side sEMG pretreatment

[0069] Perform one or more of the following processing on the healthy side's sEMG signal:

[0070] Bandpass filter: 20–450 Hz (or 10–500 Hz);

[0071] Power frequency notch: 50 Hz (or 60 Hz) notch;

[0072] Full-wave rectification: take the absolute value;

[0073] Envelope extraction: Low-pass the rectified signal at 2–10 Hz or use a 50–200 ms sliding RMS.

[0074] Amplitude normalization: Normalize according to MVC or normalize according to the mean and variance of the training phase;

[0075] Action start detection: Detect action onset based on the rising edge of the envelope (threshold = resting mean + N times standard deviation).

[0076] S123 Healthy IMU Preprocessing

[0077] Perform one or more of the following processing on the healthy side IMU signal:

[0078] Low-pass filtering: Low-pass for acceleration / angular velocity (e.g., 0–6 Hz upper limb, 0–10 Hz gait scenarios);

[0079] Zero bias correction: Estimating and correcting the gyroscope's zero bias using the stationary segment;

[0080] Attitude calculation: Attitude angles / quaternions are obtained by fusing complementary filtering or Kalman filtering;

[0081] Kinematic calculations: Calculate indices such as angular velocity modulus, attitude change, amplitude of movement, and rhythm;

[0082] Event detection: Detect action segments and periodic boundaries using peak angular velocity and periodic features.

[0083] S124 Multimodal Synchronization and Alignment

[0084] Synchronization methods: hardware triggering (TTL), the same master clock, or a unified timestamp protocol to achieve synchronous acquisition of EEG / IMU / sEMG;

[0085] Alignment method: Resample the IMU and sEMG to a unified time axis, or align the EEG segments with "healthy side action onset / peak event" as the anchor point;

[0086] Data cleaning: Remove saturated segments, missing segments, and strong artifact segments; interpolate short missing segments if necessary.

[0087] In step S200, features are extracted from the preprocessed EEG, healthy side sEMG, and healthy side IMU, and multimodal input is constructed respectively:

[0088] EEG: Time-domain statistics, power spectral density, time-frequency energy, power variation in the μ / β band, CSP, etc.

[0089] sEMG: RMS, MAV, WL, ZC, SSC, median frequency in the frequency domain, peak envelope / integral, etc.;

[0090] IMU: Attitude angle, angular velocity magnitude, acceleration magnitude, amplitude of motion, peak velocity, period parameters, etc.;

[0091] The multimodal features are then concatenated according to a unified time window or fused through an attention / gating mechanism to form a feature vector sequence for decoding.

[0092] In step S300, based on multimodal features, a mapping relationship of "brain signal - healthy side action - affected side target action" is established to obtain the affected side target control quantity:

[0093] Decoding targets: motion category, joint angle trajectory, velocity / acceleration curve, muscle activation pattern, or joint torque command;

[0094] Model format: Traditional classification, regression models, or deep learning models (such as CNN / LSTM / Transformer) can be used to output the target control quantity on the affected side;

[0095] Mirror mapping: Based on the posture / joint trajectory obtained from the healthy side IMU, combined with the patient's range of motion and safety threshold on the affected side, the target trajectory on the affected side is obtained; at the same time, the amplitude or assistance level of the target trajectory is modulated using EEG to represent "motor intention intensity / attention level".

[0096] In step S400, the affected-side training device is driven according to the affected-side target control quantity output in S300, including at least one of the following embodiments:

[0097] Implementation Method 1: Robot / Exoskeleton Assistance

[0098] Control the exoskeleton or rehabilitation robot on the affected side to output joint angle / speed / torque assistance, so that the affected side can perform the target action; and set safety constraints (maximum angle, maximum speed, maximum torque, collision detection and emergency stop).

[0099] Implementation Method 2: Functional Electrical Stimulation (FES) Assistance (Optional)

[0100] like Figure 6 As shown, the muscle activation patterns obtained from decoding are mapped to FES stimulation parameters (pulse width, frequency, current amplitude, duty cycle, channel combination), and the corresponding muscle groups on the affected side are synchronously stimulated to generate / enhance movement; similarly, stimulation upper limits and skin impedance monitoring are set to ensure safety.

[0101] Implementation Method 3: Robot + FES Collaboration (Optional)

[0102] The robot provides trajectory and stability, while FES provides active muscle engagement. By allocating weights, it achieves "more work for the capable" and improves the training effect of neuroplasticity.

[0103] In step S500, feedback signals from the affected side are collected during training (optional: affected side IMU / encoder, torque sensor, affected side sEMG, pain / fatigue score, etc.) to update and iterate the model, forming a closed loop:

[0104] Deviation calculation: deviation between the actual trajectory and the target trajectory on the affected side, and deviation in muscle involvement;

[0105] Adaptive adjustment: Adjusts the assist level, FES intensity, and trajectory gain in real time based on the deviation;

[0106] Model updates: Online calibration or periodic retraining is used to reduce the impact of cross-day drift and electrode position changes.

[0107] Example 2:

[0108] The present invention also provides an apparatus / device / system for a sports rehabilitation method and system, comprising a memory, a processor, and a computer program stored in the memory, characterized in that the processor executes the computer program to implement the steps of the above-described method.

[0109] The present invention also provides a computer-readable storage medium having a computer program / instructions stored thereon, which, when executed by a processor, implement the steps of the above-described method.

[0110] The present invention also provides a computer program product, including a computer program / instructions that, when executed by a processor, implement the steps of the above-described method.

[0111] The various embodiments in this specification are described in a progressive manner, with each embodiment focusing on its differences from other embodiments. Similar or identical parts between embodiments can be referred to interchangeably. For the apparatus disclosed in the embodiments, since it corresponds to the method disclosed in the embodiments, the description is relatively simple; relevant parts can be referred to in the method section.

[0112] Those skilled in the art will further recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, computer software, or a combination of both. To clearly illustrate the interchangeability of hardware and software, the components and steps of the various examples have been generally described in terms of functionality in the foregoing description. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0113] The steps of the methods or algorithms described in conjunction with the embodiments disclosed herein can be implemented directly by hardware, a software module executed by a processor, or a combination of both. The software module can be located in random access memory (RAM), main memory, read-only memory (ROM), electrically programmable ROM, electrically erasable programmable ROM, registers, hard disk, removable disk, CD-ROM, or any other form of storage medium known in the art.

[0114] Finally, it should be noted that in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, 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. Without further limitations, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or apparatus that includes said element.

[0115] The technical solutions 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 above embodiments are only for the purpose of helping to understand the methods and core ideas of this application. At the same time, for those skilled in the art, there will be changes in the specific implementation methods and application scope based on the ideas of this application. Therefore, the content of this specification should not be construed as a limitation of this application.

Claims

1. A sports rehabilitation method, characterized by: Includes the following steps: Brain signals and motor signals of the unaffected limbs were acquired and preprocessed to obtain preprocessed brain signals and motor signals of the unaffected limbs. Features are extracted from the preprocessed brain signals and the motor signals of the unaffected limbs to obtain multimodal features. The multimodal features are then fused to obtain a feature vector sequence. Input the feature vector sequence into the model to obtain the target control quantity for the affected side; The target control quantity on the affected side drives the training device on the affected side to perform rehabilitation exercises.

2. The sports rehabilitation method according to claim 1, characterized in that: It also includes the following steps: collecting feedback signals from the affected side during rehabilitation exercises and updating the model.

3. The sports rehabilitation method according to claim 1, characterized in that: The brain signals include brain signals related to motor attempts or motor imagery; the healthy limb movement signals include healthy IMU signals and healthy sEMG signals.

4. The sports rehabilitation method according to claim 1, characterized in that: The preprocessing includes one or more of the following: data cleaning, feature extraction, normalization, and synchronization and alignment of multimodal data.

5. A sports rehabilitation method according to claim 1, characterized in that: The models include traditional classification, regression models, or deep learning models.

6. A sports rehabilitation method according to claim 1, characterized in that: The affected side training device includes one or more of the following: a rehabilitation robot, an exoskeleton, and FES (functional electrical stimulation).

7. A sports rehabilitation device / equipment / system, comprising a memory, a processor, and a computer program stored in the memory, characterized in that, The processor executes the computer program to implement the steps of the method according to any one of claims 1-6.

8. A computer-readable storage medium having a computer program / instructions stored thereon, characterized in that, When the computer program / instructions are executed by the processor, they implement the steps of the method described in any one of claims 1-6.

9. A computer program product comprising a computer program / instructions, characterized in that, When the computer program / instructions are executed by the processor, they implement the steps of the method described in any one of claims 1-6.

Images