Feedback-based neuro-rehabilitation device based on intracranial electroencephalography signals

Abstract

The present application belongs to the technical field of physiological electrical signals, and particularly relates to a feedback type neural rehabilitation device based on intracranial electroencephalogram signals, which decodes the electroencephalogram signals into training instructions to control a training peripheral device to execute a training scheme, acquires a signal effectiveness index according to the training instructions, takes the execution result of the training peripheral device as a standard, unifies the evaluation standard of the quality of the electroencephalogram signals on the rehabilitation result, avoids the differences between different subjects and the differences of the same subject at different stages, and then acquires a training effectiveness index according to the rehabilitation index signal, so that the feedback result is more objective, and it is beneficial for the patient or the doctor to make effective regulation and control decisions in time.

Description

Technical Field

[0001] This invention belongs to the field of physiological electrical signal technology, specifically relating to a feedback-based neurorehabilitation device based on intracranial electroencephalogram (EEG) signals. Background Technology

[0002] Brain-Induced Intracranial Cipher Injection (BCI) rehabilitation typically involves patients using a 2D screen (such as a computer screen) as visual cues or feedback in a hospital or rehabilitation center, following a training plan. Intracranial electrodes collect electrophysiological signals from the cerebral cortex, which are then translated into control commands by a signal decoding model. These commands are then executed using training peripherals. Therefore, BCI rehabilitation has relatively low requirements for equipment. For example, for arm movement disorders caused by spinal cord injury, the training peripheral can be a command-controlled, flexible, wearable glove robot worn on the trainee's hand. This provides a basis for patients to conduct rehabilitation training at home and outdoors.

[0003] In related technologies, doctors typically develop training plans based on the patient's condition, including training tasks, training paradigms, and training times. These plans are then sent to the user terminal via online or offline methods. Patients and their families arrange their own training time and location on the user terminal according to their individual circumstances. Doctors then schedule regular measurements of the patient's comprehensive physiological indicators to assess the rehabilitation effect and the suitability of the current training plan. However, during rehabilitation treatment, issues such as incomplete monitoring or untimely handling of safety hazards may arise. For example, patent CN118919042A discloses a multifunctional rehabilitation therapy device. The specification describes a control module that issues alarms or reminders to the user via a voice reminder module. When the data analysis module detects a safety hazard, it automatically adjusts the parameters and modes of the rehabilitation therapy device based on the results of the data analysis module. It supports remote control of the device's operating status and settings to respond to changing treatment needs and real-time data. However, this technical solution does not specifically describe how to address the objective physiological structural differences between different subjects and the state differences of the same subject at different stages. Without the support of large-scale data, the reliability of the analysis results and remote automatic parameter adjustment needs to be verified. Summary of the Invention

[0004] This invention provides a feedback-based neurorehabilitation device based on intracranial electroencephalogram (EEG) signals to verify the reliability of the feedback results.

[0005] To address the aforementioned technical problems, this invention provides a feedback-based neurorehabilitation device, comprising: a first acquisition module, a second acquisition module, a training peripheral, a user terminal, and a management terminal; the first acquisition module acquires intracranial electroencephalogram (EEG) signals; the second acquisition module acquires rehabilitation indicator signals; the management terminal sets a training plan and its expected results and sends them to the user terminal; the user terminal is equipped with a signal decoding module, a first data processing module, a second data processing module, and a real-time feedback module; the signal decoding module decodes the EEG signals into training instructions to control the training peripheral to execute the training plan; the first data processing module obtains signal effectiveness indicators based on the training instructions; the second data processing module obtains training effectiveness indicators based on the rehabilitation indicator signals; the real-time feedback module obtains feedback results based on the signal effectiveness indicators and the training effectiveness indicators, and feeds them back to the management terminal.

[0006] Furthermore, obtaining feedback results includes: a first-level judgment, which determines whether the signal effectiveness index meets the first threshold; if yes, proceed to the second-level judgment; if no, the feedback result is to adjust the device parameters; a second-level judgment, which determines whether the training effectiveness index meets the second threshold; if yes, the feedback result is to maintain the training scheme; if no, the feedback result is to adjust the training scheme.

[0007] Furthermore, the adjustment of device parameters includes: parameter adjustment judgment, setting the assumed value of the parameter adjustment item, and using the EEG signal before parameter adjustment to determine whether the expected result of parameter adjustment meets the requirements; if yes, then proceed to parameter adjustment verification; if no, then report a device fault; parameter adjustment verification, setting the assumed value of the parameter adjustment item and executing the training plan, and judging whether the signal effectiveness index after parameter adjustment meets the first threshold or whether the training effectiveness index after parameter adjustment meets the second threshold; if yes, then set the assumed value as the device parameter; if no, then the device parameter remains unchanged.

[0008] Furthermore, the parameter tuning judgment includes: establishing a parameter tuning evaluation model and setting fixed parameters, tuning terms, and output terms of the model; using the assumed values ​​of the parameter terms as model parameters of the evaluation model, and obtaining the values ​​of the output terms based on the EEG signals before parameter tuning; determining whether the values ​​of the output terms meet a third threshold; wherein the fixed parameters include the channel location and number of implanted electrodes; the output terms are capability evaluation values ​​of the expected parameter tuning results, and the expected parameter tuning results are configured as at least one of the following: signal quality characterization value of effective channels, number of effective channels, and signal effectiveness index.

[0009] Furthermore, the parameter tuning judgment also includes: self-tuning judgment and remote tuning judgment executed sequentially; the self-tuning judgment is configured so that the user terminal uses the parameter tuning evaluation model to set the assumed values ​​of the parameter tuning items and performs parameter tuning verification; the parameter tuning items of the self-tuning judgment are configured with usage scenarios, including: hospital, home, rehabilitation center, and outdoors; the remote tuning judgment is configured so that the management terminal uses the parameter tuning evaluation model to set the assumed values ​​of the parameter tuning items and the user terminal performs parameter tuning verification; the parameter tuning items of the remote tuning judgment include: usage scenario, parameters of the signal acquisition module, parameters of the signal decoding module, channel position, and effective frequency band.

[0010] Furthermore, it also includes: a first prompt module located on the user end; and a second prompt module located on the management end. When the feedback result is to adjust device parameters, the first prompt module prompts the user end to stop rehabilitation training and prioritize self-adjustment judgment, while the second prompt module prompts the management end on the progress of self-adjustment judgment. When the feedback result is to maintain the training plan, neither the first nor the second prompt module takes any action. When the feedback result is to adjust the training plan, the first prompt module prompts the user end that a new training plan is needed, while the second prompt module prompts the management end to issue the new training plan to the user end.

[0011] Furthermore, the first data processing module acquires signal validity indicators based on training instructions by: establishing a database of valid instructions, i.e., all instructions in the training instructions that can control the training peripheral to complete the training scheme; acquiring the equivalent value of the instructions, converting the valid instructions and training instructions into corresponding equivalent values, i.e., effective equivalent and training equivalent; and using the extreme value of the effective equivalent as the first threshold.

[0012] The present invention also provides a computer device, including a memory, a processor, and a computer program stored in the memory. The processor executes the computer program to implement the steps of a rehabilitation training control method. The rehabilitation training control method includes: acquiring intracranial electroencephalogram (EEG) signals using a first acquisition module; acquiring rehabilitation indicator signals using a second acquisition module; setting a training plan and its expected results using a management terminal and sending them to a user terminal; decoding the EEG signals into training instructions using a signal decoding module to control training peripherals to execute the training plan; acquiring signal effectiveness indicators using a first data processing module based on the training instructions; acquiring training effectiveness indicators using a second data processing module based on the rehabilitation indicator signals; and acquiring feedback results using a real-time feedback module based on the signal effectiveness indicators and training effectiveness indicators, and feeding them back to the management terminal.

[0013] The present invention also provides a computer-readable storage medium storing a computer program thereon, wherein the computer program, when executed by a processor, implements the steps of a rehabilitation training control method; the rehabilitation training control method includes: acquiring intracranial electroencephalogram (EEG) signals using a first acquisition module; acquiring rehabilitation indicator signals using a second acquisition module; setting a training plan and its expected results using a management terminal and sending them to a user terminal; decoding the EEG signals into training instructions using a signal decoding module to control training peripherals to execute the training plan; acquiring signal effectiveness indicators using a first data processing module based on the training instructions; acquiring training effectiveness indicators using a second data processing module based on the rehabilitation indicator signals; and acquiring feedback results using a real-time feedback module based on the signal effectiveness indicators and training effectiveness indicators, and feeding them back to the management terminal.

[0014] This invention also provides a computer program product, including a computer program that, when executed by a processor, implements the steps of a rehabilitation training control method. The rehabilitation training control method includes: acquiring intracranial electroencephalogram (EEG) signals using a first acquisition module; acquiring rehabilitation indicator signals using a second acquisition module; setting a training plan and its expected results using a management terminal and sending them to a user terminal; decoding the EEG signals into training instructions using a signal decoding module to control training peripherals to execute the training plan; obtaining signal effectiveness indicators using a first data processing module based on the training instructions; obtaining training effectiveness indicators using a second data processing module based on the rehabilitation indicator signals; and obtaining feedback results using a real-time feedback module based on the signal effectiveness indicators and training effectiveness indicators, and feeding them back to the management terminal.

[0015] The beneficial effects of this invention are that the feedback-based neurorehabilitation device decodes EEG signals into training instructions to control the training peripherals to execute the training program. It obtains signal effectiveness indicators based on the training instructions and uses the execution results of the training peripherals as the standard, thus unifying the evaluation standard of EEG signal quality for rehabilitation results. This avoids differences between different subjects and differences between the same subject at different stages. Furthermore, it obtains training effectiveness indicators based on rehabilitation indicator signals, thereby making the feedback results more objective and facilitating timely and effective regulatory decisions by patients or doctors.

[0016] Other features and advantages of the invention will be set forth in the following description, and will be apparent in part from the description, or may be learned by practicing the invention. To make the foregoing objects, features, and advantages of the invention more apparent and understandable, preferred embodiments are described in detail below with reference to the accompanying drawings. Attached Figure Description

[0017] To more clearly illustrate the specific embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the specific embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained from these drawings without creative effort.

[0018] Figure 1 This is a flowchart of rehabilitation training regulation methods.

[0019] Figure 2 This is a schematic diagram of the principle of a neurorehabilitation device.

[0020] Figure 3 This is a schematic diagram of a digital neural pathway.

[0021] Figure 4 This is a flowchart illustrating the process of obtaining feedback results. Detailed Implementation

[0022] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions 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, 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.

[0023] See Figures 1-4 This embodiment provides a feedback-based neurorehabilitation device, including: a first acquisition module, a second acquisition module, a training peripheral device, a user terminal and a management terminal that communicate with each other; the first acquisition module acquires intracranial electroencephalogram (EEG) signals; the second acquisition module acquires rehabilitation indicator signals; the management terminal sets a training plan and its expected results and sends them to the user terminal; the user terminal is equipped with a signal decoding module, a first data processing module, a second data processing module, and a real-time feedback module; the signal decoding module decodes the EEG signals into training instructions to control the training peripheral device to execute the training plan; the first data processing module obtains signal effectiveness indicators based on the training instructions; the second data processing module obtains training effectiveness indicators based on the rehabilitation indicator signals; the real-time feedback module obtains feedback results based on the signal effectiveness indicators and training effectiveness indicators, and feeds them back to the management terminal.

[0024] In some embodiments, see Figure 2 and Figure 3The feedback-based neurorehabilitation device includes structural components, hardware circuits, and software (such as signal decoding modules, first data processing modules, second data processing modules, real-time feedback modules, etc., and their parameters, with algorithms for each module running on a processor or host computer). Specifically, it consists of electrode sensors, a lower-level machine (including an in-body machine and an external machine), a host computer (including a user terminal and a management terminal, and their corresponding human-computer interaction system), and training peripherals. The rehabilitation training control method and its corresponding software run on the lower-level and host computers. Electroencephalogram (EEG) signals are collected by the implanted electrode sensors, amplified, and converted into digital electrical signals by an analog-to-digital converter before being transmitted to the in-body machine. The in-body machine then sends the signals to the external machine, which in turn sends them to the user terminal (used by the patient) and the management terminal (used by the doctor). After acquiring the EEG signals, the user terminal decodes them into training instructions using the signal decoding module, controlling the training peripherals to execute the training plan, thus forming a digital neural pathway. The first acquisition module can be understood as the electrode sensor, which collects raw analog EEG signals. The internal unit includes an amplifier module (hardware analog circuit) that amplifies the original analog EEG signal; and an analog-to-digital conversion module (hardware) that converts the amplified analog EEG signal into a digital signal at a fixed sampling rate. The external unit (MCU / ARM / DSP hardware + embedded software) stores and forwards the digital signal. The host computer (PC + PC software or mobile phone + APP software) processes the digital signal and utilizes a human-computer interaction system to coordinate with the UI workflow of rehabilitation training control methods; a display module presents the UI interface, showing various assessment results and feedback results, allowing users to operate the UI to set device parameters or communicate online. The second acquisition module can be understood as an electromyography (EMG) sensor, which can collect EMG signals from the affected area to characterize training effectiveness indicators. Initially, an initial threshold for the training effectiveness indicators can be set based on experience or historical values. EMG signals greater than or equal to the initial threshold are used as effective rehabilitation indicators, and the second threshold is obtained based on the effective rehabilitation indicator assessment.

[0025] Doctors comprehensively assess the extent of damage to the affected area based on scales and other objective physiological indicators, and combine this with MRI images of the patient's brain to locate relevant brain functional areas (such as the motor cortex) to determine the channel location, number, and specific type of electrodes to be implanted. Based on the patient's basic condition and experience, doctors will set the equipment parameters and provide an initial training plan and expected results, as well as the scale's follow-up cycle, to evaluate the rehabilitation effect and periodically adjust the training plan.

[0026] During the training and rehabilitation process, a human-computer interaction device (HCI) is installed on the user terminal, which includes an interface for inputting device parameters, an operation interface for the first data processing module, an operation interface for the second data processing module, an operation interface for the real-time feedback module, and a display interface. The patient trains according to medical instructions, and the lower-level machine sends the patient's EEG signals and training data to the user terminal and the management terminal. On the user terminal, the signal decoding module decodes the EEG signals into training instructions to control the training peripherals to execute the training plan. The operation interface of the first data processing module is used to set the definition parameters of valid instructions. For example, a valid instruction is defined as a training instruction whose training result meets the expected result. The number or content of instructions acquired in each training session may be different, so valid instructions can be normalized and converted into corresponding equivalent values. Initially, an initial equivalent threshold for valid instructions can be set based on experience or historical values. Training instructions that are greater than or equal to the initial equivalent threshold or meet the expected result are considered valid instructions and entered into the database of valid instructions to update the initial equivalent threshold to obtain the first threshold. Similarly, the training equivalent corresponding to the training instruction is obtained to characterize the effectiveness of the EEG signal. Obviously, if during the training process, even if the quality of the EEG signal and the corresponding instruction after decoding fluctuate due to other factors, as long as the training equivalent meets the first threshold, the instruction is considered valid and acceptable. That is, the fluctuation does not affect the EEG signal and its decoding process, nor does it affect the training peripheral from executing the training plan.

[0027] Converting instructions to equivalent values ​​involves: if the actual instruction matches the expectation, it is recorded as a valid instruction equivalent; if the actual instruction does not match the expectation (or mostly does not match), it is recorded as an invalid instruction equivalent. Differences between instructions and intermediate cases can be temporarily disregarded. Finally, instruction accuracy is calculated based on the ratio of valid instruction equivalents to total instruction equivalents.

[0028] In some embodiments, the characteristic values ​​of electromyographic signals only qualitatively characterize whether the training is effective and do not represent the actual rehabilitation effect. Moreover, the feedback results are for sudden abnormalities during the training cycle and, in principle, do not affect the scale review cycle. This is because patients need to go to the hospital or rehabilitation center regularly for a more comprehensive scale review to assess the rehabilitation effect of the affected area and discuss whether to adjust the training plan based on the patient's subjective and objective indicators. Examples of tests include clinical scales (Fugl-Meyer upper limb score), standardized functional assessment (ARAT scale), muscle strength testing, and electrophysiological indicators (sEMG, SEP).

[0029] When the patient is training, see Figure 4The user terminal acquires EEG signals and decodes them into training instructions, which control the training peripherals to execute the training plan. The first data processing module converts the training instructions into training equivalents and performs a first-level judgment, determining whether the training equivalents meet a first threshold. If yes, it proceeds to a second-level judgment; otherwise, the real-time feedback module adjusts the device parameters as feedback to the management terminal. The second data processing module acquires the feature values ​​of the EMG signals and performs a second-level judgment, determining whether the feature values ​​of the EMG signals meet a second threshold. If yes, the real-time feedback module maintains the training plan as feedback to the management terminal; otherwise, the real-time feedback module adjusts the training plan as feedback to the management terminal.

[0030] Clearly, different feedback results imply different handling methods on both the user and management sides, and also represent the direction of communication between doctors and patients. For example, if the feedback result is to adjust equipment parameters, the user needs to stop rehabilitation training first to avoid obtaining inaccurate data that could affect the doctor's misjudgment of subsequent training plans. It's also necessary to further determine whether the patient can adjust the parameters themselves or if the doctor must adjust them—a rigorous issue. It's crucial to facilitate quick handling of emergencies on the user side while preventing unprofessional personnel from arbitrarily adjusting the training equipment parameters. Therefore, it's necessary to determine if self-adjustment is allowed. If the patient is allowed to adjust the parameters themselves, the user only needs to verify the adjustment results and synchronize the parameters to the management side; the user can handle this entirely without the time and cost of communication between the patient and doctor, and it allows the user to address some abnormal situations more promptly. If the patient is not allowed to adjust the parameters themselves, the management side needs to design an adjustment plan, with the user cooperating in verification; online or offline methods can be used. For example, if the feedback result is to adjust the training plan, the doctor can prioritize issuing the new training plan online to the user. Since the data is accurate, only the rehabilitation effect of the training plan is debatable; therefore, before the new training plan is issued, it does not affect the patient's continued implementation of the original training plan. For example, if the feedback indicates that the training plan should be maintained, then no new changes or communication are needed between the doctor and the patient.

[0031] In some embodiments, see Figure 4 The user-side or management-side interface includes a parameter adjustment interface, which can be accessed actively or passively prompted. For example, when the feedback indicates parameter adjustment, the device will prompt whether to proceed with parameter adjustment judgment, and based on the judgment result, whether to proceed with parameter adjustment verification. A pop-up dialog box will automatically appear on the user-side or management-side interface, asking the operator to confirm whether to proceed with parameter adjustment verification. After clicking "confirm," the operator can adjust the device parameters according to the assumed values ​​and then obtain signal validity indicators. The user-side interface will also pop up a dialog box asking the operator to confirm whether to save the adjusted device parameters. Clicking "confirm" sets the assumed values ​​as the device parameters. Clicking "exit" will not save the assumed values, and the device parameters will be considered as not having been adjusted.

[0032] Optionally, the parameter tuning judgment includes: firstly, entering self-tuning judgment, i.e., judging whether the parameter tuning item meets the self-tuning conditions; if yes, the user terminal performs parameter tuning verification; if no, then entering remote tuning judgment; remote tuning judgment, i.e., judging whether the parameter tuning item meets the remote tuning conditions; if yes, the management terminal and the user terminal jointly perform parameter tuning verification; if no, then a device fault is reported, and the management terminal handles it offline. Here, self-tuning judgment and remote tuning judgment can be understood as, with the model's parameters fixed, changing the model's parameter tuning items and seeing whether the output items recalculated by the software based on the same EEG signal data (before parameter tuning) improve; in fact, it verifies the improvement of the theoretical result and does not require executing a training scheme. Parameter tuning verification can be understood as, after theoretical verification, using the EEG signal data or EMG signal data after parameter tuning to verify the improvement of the actual training effect, which requires executing a training scheme.

[0033] Optionally, the self-tuning judgment includes: the user terminal sets a self-tuning evaluation model, and sets fixed parameters, self-tuning parameter items, and output items for the model; the user terminal inputs the assumed values ​​of the self-tuning parameter items into the self-tuning evaluation model to obtain the values ​​of the output items; and determines whether the values ​​of the output items meet a third threshold. The fixed parameters include the channel location and number of implanted electrodes. The output item is a capability evaluation value of the expected tuning result, which is configured as at least one of the following: signal quality characterization value of effective channels, number of effective channels, and signal effectiveness index. The self-tuning parameter items are configured according to usage scenarios, including: hospital, home, rehabilitation center, and outdoors. The third threshold can be obtained based on empirical and historical values. Specifically, the user terminal saves the adjusted parameter configuration for each actual usage scenario. Users are only allowed to switch between scenarios to achieve the purpose of device parameter tuning, but are not allowed to directly tune the device.

[0034] During rehabilitation, patients' main activities and living spaces are concentrated in hospitals, homes, rehabilitation centers, and outdoors. Different scenarios may lead to different signal quality issues, requiring the selection of different signal processing methods and decoding models based on the scenario to filter high-quality signals and obtain accurate decoding results. This can improve the compatibility between the patient's environment and the rehabilitation equipment, thereby improving signal effectiveness indicators. Obviously, patients or end-user operators have the ability and convenience to distinguish these usage scenarios. When the usage scenario is adjusted, if the output indicators improve, it indicates that the self-tuned parameters are beneficial and can proceed to parameter tuning verification; if the output indicators worsen or remain unchanged, it indicates that the self-tuned parameters are ineffective, and there is no need to modify the equipment parameters or proceed to parameter tuning verification.

[0035] Typically, in hospital environments, the dense concentration of medical equipment generates strong, constant 50Hz (depending on grid voltage, it may be 60Hz in regions like the US and Canada) power frequency and its harmonic interference. This necessitates processing algorithms with deep and precise notch filtering capabilities, requiring stronger filtering of the power frequency and its harmonics during signal processing. Therefore, when the self-tuning parameter is set to hospital settings, the rehabilitation equipment will automatically add preprocessing methods targeting power frequency and its harmonic interference during the signal preprocessing stage.

[0036] In a home setting, the environment is complex and uncontrollable. Non-stationary, broadband interference such as electromyography artifacts and motion artifacts caused by the patient's spontaneous or passive activities become the main noise sources, placing higher demands on the stability of the model and requiring the selection of a more robust signal decoding model. Therefore, when the self-tuning parameters are set to home, the rehabilitation device will automatically select a more suitable signal decoding module.

[0037] In special settings such as rehabilitation centers, various types of noise are generally controllable. However, the operational skill level of relevant personnel should be considered (as the internal system configuration is usually not adjusted). Real-time alerts should be provided for any sudden abnormal artifacts to allow for timely adjustments to the external environment and maintain stability in the current environment. Therefore, when the self-adjustment parameter is set to "rehabilitation center," the rehabilitation equipment will automatically add preprocessing methods for abnormal artifacts during the signal preprocessing stage.

[0038] Optionally, the remote tuning judgment includes: the user terminal sets a remote tuning evaluation model, and sets the model's fixed parameters, remote tuning parameter items, and output items; the management terminal inputs the assumed values ​​of the remote tuning parameter items online into the remote tuning parameter item evaluation model to obtain the output items; and determines whether the output items meet a third threshold. The fixed parameters include the channel location and number of implanted electrodes. The output items are capability evaluation values ​​of the expected tuning results, configured as at least one of the following: signal quality characterization value of effective channels, number of effective channels, and signal effectiveness index. The remote tuning parameter items include: usage scenario, parameters of the signal acquisition module, parameters of the signal decoding module, channel location of effective channels, and effective frequency band. After surgery, because the EEG signal sensor is implanted outside the patient's dura mater, the EEG electrodes and leads may be damaged, fail, or interfered with, causing changes in the effective signal ratio of each channel's signal decoding. When the effective signal ratio continuously decreases, the channel performance degradation status is determined; when the signal effectiveness index is below the first threshold, the doctor and engineer are prompted to intervene in the system's decoding model, training new decoding model parameters in the backup channel. Generally, rehabilitation training selects electrodes with a high degree of matching with the current patient. For example, the effective channel location represents the matching degree between the implantation site and the brain functional area of ​​the affected area, and the effective frequency band represents the proportion of effective data. Matching different signal acquisition modules and signal decoding modules according to the effective data can better adapt to existing specific interference, thereby improving the signal quality characterization value and the number of channels in the effective channel. Obviously, remote parameter tuning is more suitable for professionals, such as doctors. After remote parameter tuning, if the output indicators improve, it means that the remotely tuned parameters are beneficial and can be validated. If the output indicators worsen or remain unchanged, it means that the remotely tuned parameters are ineffective or the results do not meet expectations, and there is no need to validate them. In this case, offline processing by the management side is required to further investigate the cause.

[0039] Generally, remote parameter adjustment is limited to remote fine-tuning. If the adjustment is too large, it is recommended that the user come to the hospital for a follow-up visit, especially in cases of obvious abnormalities such as changes in the effective channel or a significant decrease in signal effectiveness.

[0040] For example, the process for determining the effectiveness of remote parameter tuning of the effective frequency band is as follows: Before verification, the patient wears the external device and attaches its coil to the internal device. The patient also wears the training peripheral, such as a pneumatic handpiece. The external device, software, and pneumatic handpiece are then turned on to complete the configuration. The calibration paradigm is then initiated, lasting approximately 10 minutes in total. During this time, the patient needs to try to imagine clenching or relaxing their fist according to the on-screen prompts to execute the training plan. The software records the EEG at the corresponding stage and uses this data for calibration.

[0041] Three patients are selected. First, the software recommends characteristic frequency bands of EEG signals based on each patient's historical data or the doctor's experience, as assumed values ​​for the parameter tuning terms. The output values ​​are obtained according to the remote parameter tuning evaluation model, and it is determined whether they meet the third threshold. The output values ​​represent the parameter capabilities; here, the F1 score, a weighted average of precision and recall (positive and negative results), is selected to evaluate the classifier's performance. A value less than 0.85 suggests re-recommending effective frequency bands; 0.85-0.90 allows for optimization of effective frequency bands; and a value greater than 0.90 allows for direct determination of effective frequency bands. In the interface, drag to select the optimal classification frequency band between high and low frequencies, adjust it to a suitable output value, and complete the parameter tuning judgment.

[0042] Before parameter tuning, each patient was assigned a general frequency band for motor imagery: low-frequency activity 5-35Hz (mu rhythm and beta rhythm), and high-frequency activity 60-140Hz (high-gamma broadband activity). Calculations were performed under these conditions, and the values ​​of the untuned output terms are shown in Table 1.

[0043] Table 1 Output Items Without Parameter Tuning

[0044]

[0045] Remote parameter tuning was performed on the effective frequency band for each patient, and individualized frequency bands were selected. The results are shown in Table 2 below.

[0046] Table 2 Effective frequency band after parameter tuning

[0047]

[0048] The expected results of the parameter tuning were obtained using this frequency band, and the values ​​of the output terms after parameter tuning are shown in Table 3 below.

[0049] Table 3 Output items after parameter tuning

[0050]

[0051] The test results in Tables 1 and 3 show that the mean accuracy for the three FIM patients without parameter tuning was 0.87 ± 0.09, while the mean accuracy after parameter tuning was 0.94 ± 0.03 ≥ 0.85. Both scenarios achieved usable levels, but the expected results after parameter tuning were significantly improved.

[0052] After parameter tuning assessment, parameter tuning verification is performed. Each patient performs motor imagery to achieve brain-controlled pneumatic and manual movement, and the validity of the signal after parameter tuning is verified to meet the first threshold or the validity of the training after parameter tuning to meet the second threshold.

[0053] In some embodiments, the neurorehabilitation device further includes: a first prompting module located at the user end; and a second prompting module located at the management end. When the feedback result is to adjust device parameters, the first prompting module prompts the user end to stop rehabilitation training and prioritize self-adjustment judgment, while the second prompting module prompts the management end on the progress of self-adjustment judgment. When the feedback result is to maintain the training plan, neither the first nor the second prompting module takes any action. When the feedback result is to adjust the training plan, the first prompting module prompts the user end that a new training plan is needed, while the second prompting module prompts the management end to issue a new training plan to the user end.

[0054] In summary, after training is performed on the user end, when anomalies occur, a step-by-step judgment approach can be used to handle different situations. This allows for immediate verification of the reliability of each feedback result and provides specific handling methods to resolve anomalies promptly. For example, in the first-level judgment, the first data processing module can be used to investigate the impact of signal quality on decoding instructions. The signal validity index is used to prioritize the assessment of device reliability, leading to parameter tuning judgment. Priority is given to verifying device parameter tuning performed by the user end, followed by determining whether the management end and user end should jointly perform parameter tuning verification, or whether the management end should perform offline parameter tuning. Finally, parameter tuning is assumed and verified; only after successful verification are the assumed values ​​set as device parameters. Similarly, in the second-level judgment, the second data processing module can be used to investigate the suitability of the training program for rehabilitation effects. The training validity index is used to assess the suitability of the training program, thus providing feedback on whether adjustments to the training program are necessary. Combining periodic review with real-time feedback, integrating the feedback results of step-by-step judgment with subsequent processing, and combining tuning with verification, leaves the decision-making power to doctors or patients. This approach ensures the objectivity and reliability of feedback results while improving the timeliness and effectiveness of regulatory decisions.

[0055] In some embodiments, a computer device is also provided, including 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 rehabilitation training modulation method. See Figure 1 and Figure 4 The rehabilitation training control method includes: acquiring intracranial electroencephalogram (EEG) signals using a first acquisition module; acquiring rehabilitation indicator signals using a second acquisition module; setting a training plan and its expected results using a management terminal and sending them to the user terminal; decoding the EEG signals into training instructions using a signal decoding module to control the training peripherals to execute the training plan; acquiring signal effectiveness indicators using a first data processing module based on the training instructions; acquiring training effectiveness indicators using a second data processing module based on the rehabilitation indicator signals; and acquiring feedback results using a real-time feedback module based on the signal effectiveness indicators and training effectiveness indicators, and feeding them back to the management terminal.

[0056] The processor can be a central processing unit (CPU), an ASIC, or one or more integrated circuits configured to implement embodiments of the present invention. In specific implementations, if the memory and processor are implemented independently, they can be interconnected via a bus to communicate with each other. The bus can be an Industry Standard Architecture (ISA) bus, a PCI bus, or an Extended Industry Standard Architecture (EISA) bus, etc. The bus can be categorized as an address bus, a data bus, a control bus, etc. If the memory and processor are integrated onto a single chip, they can communicate with each other through an internal interface.

[0057] In some embodiments, a computer-readable storage medium is also provided, on which a computer program is stored, which, when executed by a processor, implements the steps of the rehabilitation training modulation method. See Figure 1 and Figure 4 The rehabilitation training control method includes: acquiring intracranial electroencephalogram (EEG) signals using a first acquisition module; acquiring rehabilitation indicator signals using a second acquisition module; setting a training plan and its expected results using a management terminal and sending them to the user terminal; decoding the EEG signals into training instructions using a signal decoding module to control the training peripherals to execute the training plan; acquiring signal effectiveness indicators using a first data processing module based on the training instructions; acquiring training effectiveness indicators using a second data processing module based on the rehabilitation indicator signals; and acquiring feedback results using a real-time feedback module based on the signal effectiveness indicators and training effectiveness indicators, and feeding them back to the management terminal.

[0058] The storage media include: USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks or optical disks, and other media capable of storing program code.

[0059] In some embodiments, a computer program product is also provided, including a computer program that, when executed by a processor, implements the steps of a rehabilitation training modulation method. See [link to product description]. Figure 1 and Figure 4The rehabilitation training control method includes: acquiring intracranial electroencephalogram (EEG) signals using a first acquisition module; acquiring rehabilitation indicator signals using a second acquisition module; setting a training plan and its expected results using a management terminal and sending them to the user terminal; decoding the EEG signals into training instructions using a signal decoding module to control the training peripherals to execute the training plan; acquiring signal effectiveness indicators using a first data processing module based on the training instructions; acquiring training effectiveness indicators using a second data processing module based on the rehabilitation indicator signals; and acquiring feedback results using a real-time feedback module based on the signal effectiveness indicators and training effectiveness indicators, and feeding them back to the management terminal.

[0060] The computer program may include program code, which includes computer operation instructions and may be stored in a computer-readable storage medium. Based on this understanding, when the technical solution of the present invention, in essence, or the part that contributes to the prior art, or a part of the technical solution, can be implemented in the form of a software product or sold or used as an independent product, the computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) or processor to execute all or part of the steps of the methods described in the various embodiments of the present invention.

[0061] In the several embodiments provided in this application, it should be understood that the disclosed systems, apparatuses, and methods can be implemented in other ways. The apparatus embodiments described above are merely illustrative. For example, the division of units is only a logical functional division, and there may be other division methods in actual implementation. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed.

[0062] Based on the above-described preferred embodiments of the present invention, and through the above description, those skilled in the art can make various changes and modifications without departing from the technical concept of the present invention. That is, the technical scope of the present invention is not limited to the contents of the specification.

Claims

1. A feedback-based neurorehabilitation device, characterized in that, include: The system consists of a first data acquisition module, a second data acquisition module, training peripherals, a user terminal, and a management terminal. The first acquisition module acquires intracranial electroencephalogram (EEG) signals; The second acquisition module acquires rehabilitation indicator signals; The management terminal sets the training plan and its corresponding expected results and sends them to the user terminal. The user terminal is equipped with a signal decoding module, a first data processing module, a second data processing module, and a real-time feedback module; The signal decoding module decodes the EEG signals into training instructions to control the training peripherals to execute the training program; The first data processing module obtains the signal validity index according to the training instructions; The second data processing module obtains training effectiveness indicators based on rehabilitation indicator signals; The real-time feedback module obtains feedback results based on signal validity indicators and training validity indicators, and then the user terminal sends the feedback results to the management terminal.

2. The feedback-type neurorehabilitation device according to claim 1, characterized in that, The feedback results obtained include: The first-level judgment determines whether the signal validity index meets the first threshold; if yes, it proceeds to the second-level judgment; if no, the feedback result is to adjust the equipment parameters. The second-level judgment determines whether the training effectiveness index meets the second threshold; if yes, the feedback is to maintain the training plan; if no, the feedback is to adjust the training plan.

3. The feedback-type neurorehabilitation device according to claim 2, characterized in that, The adjustment of equipment parameters includes: The parameter tuning process involves setting assumed values ​​for the parameter tuning items and using the EEG signals before tuning to determine whether the expected tuning results meet the requirements. If yes, the process proceeds to parameter tuning verification; otherwise, a device malfunction is reported. For parameter tuning verification, set the assumed values ​​of the parameter tuning items and execute the training scheme. Determine whether the signal effectiveness index after parameter tuning meets the first threshold or whether the training effectiveness index after parameter tuning meets the second threshold. If yes, set the assumed values ​​to the device parameters; otherwise, leave the device parameters unchanged.

4. The feedback-type neurorehabilitation device according to claim 3, characterized in that, The parameter tuning judgment includes: Establish a parameter tuning and evaluation model, and set the model's fixed parameters, tuning terms, and output terms; The assumed values ​​of the parameter terms are used as the model parameters for evaluating the model, and the values ​​of the output terms are obtained based on the EEG signals before parameter tuning. Determine whether the value of the output item satisfies the third threshold; where The fixed parameters include the channel location and number of channels for the implanted electrodes; The output item is a capability assessment value of the expected parameter tuning result, which is configured as at least one of the following: signal quality characterization value of effective channels, number of effective channels, and signal effectiveness index.

5. The feedback-type neurorehabilitation device according to claim 4, characterized in that, The parameter adjustment judgment also includes: self-adjustment judgment and remote adjustment judgment executed sequentially; The self-tuning judgment is configured so that the user terminal uses the parameter tuning evaluation model to set the assumed values ​​of the parameter tuning items and perform parameter tuning verification; the parameter tuning items of the self-tuning judgment are configured with usage scenarios, including: hospital, home, rehabilitation center, and outdoors; The remote tuning judgment is configured so that the management end uses the tuning evaluation model to set the assumed values ​​of the tuning items and the user end performs the tuning verification; the tuning items of the remote tuning judgment include: usage scenario, parameters of the signal acquisition module, parameters of the signal decoding module, channel position, and effective frequency band.

6. The feedback-type neurorehabilitation device according to claim 5, characterized in that, The first notification module is located on the user's end. The second notification module is located on the management end; When the feedback result is to adjust the equipment parameters, the first prompt module prompts the user to stop rehabilitation training and prioritize the self-adjustment judgment, while the second prompt module prompts the management end to show the progress of the self-adjustment judgment. When the feedback result is to maintain the training scheme, the first prompt module does not process it, and the second prompt module does not process it either. When the feedback result is to adjust the training plan, the first prompt module prompts the user to need a new training plan, and the second prompt module prompts the management to issue the new training plan to the user.

7. The feedback-type neurorehabilitation device according to claim 2, characterized in that, The first data processing module obtains signal validity indicators based on training instructions, including: Establish a database of valid instructions, that is, training instructions whose training results meet the expected results are considered valid instructions; Obtain the equivalent value of the instruction, and convert the effective instruction and the training instruction into their corresponding equivalent values, namely the effective equivalent and the training equivalent; The extreme value of the effective equivalent is used as the first threshold.

8. A computer device 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 rehabilitation training control method; The rehabilitation training regulation methods include: Intracranial electroencephalogram (EEG) signals were acquired using the first acquisition module. The second acquisition module is used to collect rehabilitation indicator signals; The management terminal is used to set up the training plan and its expected results, and then distributed them to the user terminal. The signal decoding module is used to decode EEG signals into training instructions to control the training peripherals to execute the training program; The first data processing module is used to obtain signal validity indicators according to training instructions; The second data processing module is used to obtain training effectiveness indicators based on rehabilitation indicator signals; The real-time feedback module obtains feedback results based on signal validity indicators and training validity indicators, and then sends them back to the management terminal.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the steps of the rehabilitation training control method. The rehabilitation training regulation methods include: Intracranial electroencephalogram (EEG) signals were acquired using the first acquisition module. The second acquisition module is used to collect rehabilitation indicator signals; The management terminal is used to set up the training plan and its expected results, and then distributed them to the user terminal. The signal decoding module is used to decode EEG signals into training instructions to control the training peripherals to execute the training program; The first data processing module is used to obtain signal validity indicators according to training instructions; The second data processing module is used to obtain training effectiveness indicators based on rehabilitation indicator signals; The real-time feedback module obtains feedback results based on signal validity indicators and training validity indicators, and then sends them back to the management terminal.

10. A computer program product, comprising a computer program, characterized in that, When the computer program is executed by the processor, it implements the steps of the rehabilitation training control method. The rehabilitation training regulation methods include: Intracranial electroencephalogram (EEG) signals were acquired using the first acquisition module. The second acquisition module is used to collect rehabilitation indicator signals; The management terminal is used to set up the training plan and its expected results, and then distributed them to the user terminal. The signal decoding module is used to decode EEG signals into training instructions to control the training peripherals to execute the training program; The first data processing module is used to obtain signal validity indicators according to training instructions; The second data processing module is used to obtain training effectiveness indicators based on rehabilitation indicator signals; The real-time feedback module obtains feedback results based on signal validity indicators and training validity indicators, and then sends them back to the management terminal.

Images