Signal processing method, brain-computer system, electronic device and storage medium

Abstract

The application discloses a signal processing method, a brain-computer system, an electronic device and a storage medium. The method comprises the following steps: obtaining a target neural signal of a subject; decoding the target neural signal by using a pre-trained decoder to obtain a decoding result; and controlling a controlled external device to continuously move towards an expected contact target outside a preset area according to the decoding result until the expected contact between the controlled external device and the expected contact target is realized, wherein the controlled external device is an expected control target of the subject, and the controlled external device moves towards the expected contact target according to target motion attributes after contacting a region boundary of the preset area, the target motion attributes being the speed and direction of the controlled external device when contacting the region boundary. The application solves the technical problem in the prior art that the feedback control strategy needs to continuously acquire neural signals to iteratively control the cursor to move to the expected contact target in small steps, thereby causing the cursor control to be inconsistent with the motion habits of the subject, i.e., inconsistent with the neural control principle.

Description

Technical Field

[0001] This invention relates to the field of brain-computer interface technology, and more specifically, to a signal processing method, a brain-computer interface system, an electronic device, and a storage medium. Background Technology

[0002] Intracortical brain-computer interfaces (iBMIs) represent a major breakthrough in neuroengineering, allowing direct control of external devices via brain signals and offering a potential rehabilitation pathway for patients with upper limb paralysis. However, most existing iBMI designs are based on a feedback control model, where subjects adjust their brain signals by observing the real-time position of a cursor (as a virtual arm) to guide it to a target location. While this model achieves a certain degree of connection between brain signals and external devices, it has significant limitations in performance and efficiency, and its control of external devices is not dexterous or natural enough.

[0003] Feedback control requires participants to finely adjust the cursor's movement at every moment, a process inconsistent with natural human behavior. In daily life, when performing hand movements, people do not consciously plan every minute muscle movement; instead, based on goal-oriented thinking, the brain generates control commands to coordinate muscle movements to complete the task. Feedback control, however, forces participants to focus on the process of hand movement rather than the result. This not only increases the complexity of control but also reduces its naturalness and efficiency.

[0004] In summary, the feedback control strategy in the prior art requires the continuous acquisition of neural signals to iteratively control the cursor to move to the desired contact target in small steps. This results in a technical problem where the cursor control does not conform to the subject's movement habits, i.e., it does not conform to the principles of neural control.

[0005] There is currently no effective solution to the above problems. Summary of the Invention

[0006] This invention provides a signal processing method, a brain-computer interface system, an electronic device, and a storage medium to at least solve the technical problem in the prior art where feedback control strategies require continuous acquisition of neural signals to iteratively control the cursor to move to the desired contact target, resulting in cursor control that does not conform to the subject's movement habits, i.e., does not conform to the principles of neural control.

[0007] According to one aspect of the present invention, a signal processing method is provided, comprising: acquiring a target neural signal of a subject; decoding the target neural signal using a pre-trained decoder to obtain a decoding result; and controlling a controlled peripheral device to continuously move towards a desired contact target outside a preset area according to the decoding result, until the controlled peripheral device makes desired contact with the desired contact target, wherein the controlled peripheral device is the desired control target of the subject, and after the controlled peripheral device contacts the boundary of the preset area, it moves towards the desired contact target according to target motion attributes, wherein the target motion attributes are the speed and direction of the controlled peripheral device when it contacts the boundary of the area.

[0008] Optionally, according to the decoding result, controlling the controlled peripheral to continuously move towards the desired contact target outside the preset area until the controlled peripheral makes desired contact with the desired contact target includes: determining a first control command according to the decoding result, wherein the first control command is used to control the movement of the controlled peripheral towards the area boundary of the preset area; generating a second control command in response to the controlled peripheral making contact with the area boundary of the preset area, wherein the second control command includes target motion attributes; and controlling the controlled peripheral to make desired contact with the desired contact target according to the second control command.

[0009] Optionally, controlling the controlled peripheral to continuously move towards the desired contact target outside the preset area according to the decoding result, until the controlled peripheral makes the desired contact with the desired contact target, includes: controlling the controlled peripheral to continuously move towards the desired contact target outside the preset area according to the decoding result, obtaining the positional difference between the controlled peripheral and the desired contact target when the controlled peripheral contacts the boundary of the preset area; judging the target motion attribute according to the positional difference, and obtaining the judgment result; responding to the judgment result that the controlled peripheral cannot make contact with the desired contact target based on the target motion attribute, determining the target motion attribute compensation value according to the positional difference; and compensating the target motion attribute according to the target motion attribute compensation value.

[0010] Optionally, the signal processing method further includes: controlling the controlled peripheral to continuously move towards the desired contact target outside the preset area according to the decoding result and recording the stopping position of the controlled peripheral when it stops moving; in response to the stopping position indicating that the controlled peripheral and the desired contact target have not made the desired contact, determining the area adjustment parameter according to the stopping position; and adjusting the preset area according to the area adjustment parameter.

[0011] Optionally, the pre-training of the decoder includes: acquiring the subject's first neural signal; inputting the first neural signal into the pre-corrected decoder to obtain a first decoding result; controlling the movement of the controlled peripheral device towards the desired contact target outside the preset area according to the first decoding result to obtain a first motion result, wherein the controlled peripheral device is the subject's desired control target, and after the controlled peripheral device contacts the boundary of the preset area, it moves towards the desired contact target with the speed and direction at which the controlled peripheral device contacts the boundary; and adjusting the decoding parameters in the pre-corrected decoder according to the first motion result to obtain a target decoder.

[0012] Optionally, before inputting the first neural signal into the pre-corrected decoder to obtain the first decoding result, the method further includes: controlling the movement of the controlled peripheral device toward the desired contact target multiple times within a first preset time period to obtain multiple movement states of the controlled peripheral device and multiple second neural signals, wherein the second neural signals are generated based on the visual stimulus observed by the subject that the controlled peripheral device moves toward the desired contact target; training the decoder according to the multiple movement states of the controlled peripheral device and the multiple second neural signals to obtain an initialized decoder; and pre-correcting the initialized decoder to obtain a pre-corrected decoder.

[0013] Optionally, the pre-correction of the decoder includes: acquiring multiple third neural signals from the subject within a second preset time period; performing a pre-correction process on the multiple third neural signals to obtain a pre-corrected decoder, wherein the pre-correction process includes: inputting the multiple third neural signals into the decoder to obtain multiple second decoding results; determining multiple control signals based on the multiple second decoding results and a preset signal determination formula; controlling the movement of the controlled peripheral device toward the desired contact target based on the multiple control signals to obtain multiple second motion results; and correcting the decoding parameters in the decoder based on the multiple second motion results to obtain a pre-corrected decoder.

[0014] Optionally, the preset signal determination formula includes an auxiliary vector, auxiliary coefficients, a decoding vector, and decoding coefficients, wherein the auxiliary vector is a vector pointing to the controlled peripheral device, and the decoding vector is the output result of the decoder.

[0015] Optionally, the control signal is determined based on the second decoding result and the preset signal determination formula, including: acquiring the previous motion result, wherein the previous motion result is the motion result corresponding to the third neural signal acquired at the previous acquisition time; adjusting the auxiliary coefficient and the decoding coefficient based on the previous motion result to obtain the adjusted preset signal determination formula; and determining the control signal based on the adjusted preset signal determination formula.

[0016] Optionally, the signal processing method further includes: determining the number of second motion results that meet the standard from a plurality of second motion results, wherein a second motion result that meets a first preset condition is determined to meet the standard; and obtaining a pre-corrected decoder in response to the number of second motion results meeting the second preset condition.

[0017] Optionally, the signal processing method further includes: acquiring multiple fourth neural signals of the subject within a third preset time period; inputting the multiple fourth neural signals into the target decoder to obtain multiple third decoding results; controlling the movement of the controlled peripheral device towards the desired contact target outside the preset area based on the multiple third decoding results to obtain multiple third motion results; and performing usability verification on the target decoder based on the multiple third motion results and preset verification rules to obtain verification results.

[0018] Optionally, the target to be contacted is a dynamic target. Before controlling the movement of the controlled peripheral device toward the target to be contacted outside the preset area according to the first decoding result and obtaining the first motion result, the method further includes: controlling the movement of the target to be contacted according to the preset motion trajectory and the preset motion speed.

[0019] Optionally, the signal processing method further includes: updating the preset motion trajectory and preset motion speed according to preset training requirements.

[0020] Optionally, based on the first motion result, the decoding parameters in the pre-corrected decoder are adjusted to obtain the target decoder, including: in response to the first motion result indicating that the controlled peripheral device failed to intercept the desired contact target, determining the deviation between the controlled peripheral device and the desired contact target; and adjusting the decoding parameters based on the deviation to obtain the target decoder.

[0021] Optionally, the preset area is a circular area.

[0022] Optionally, the signal processing method further includes: acquiring regional adjustment parameters; and adjusting the size and shape of a preset region according to the regional adjustment parameters.

[0023] According to another aspect of the present invention, a brain-computer interface system is also provided, comprising: a neural signal acquisition device for acquiring neural signals of a subject; a decoder for decoding the neural signals to obtain a decoding result; and a feedforward controller for controlling a controlled peripheral device to continuously move towards a desired contact target outside a preset area according to the decoding result, until the controlled peripheral device makes desired contact with the desired contact target, wherein the controlled peripheral device is the desired control target of the subject, and after the controlled peripheral device contacts the boundary of the preset area, it moves towards the desired contact target according to the target motion attributes, wherein the target motion attributes are the speed and direction of the controlled peripheral device when it contacts the boundary of the area.

[0024] According to another aspect of the present invention, an electronic device is also provided, comprising: a memory storing an executable program; and a processor for running the program, wherein the program executes the methods of various embodiments of the present invention during runtime.

[0025] According to another aspect of the present invention, a computer-readable storage medium is also provided, the computer-readable storage medium including a stored executable program, wherein, when the executable program is executed, it controls the device where the computer-readable storage medium is located to perform the methods of various embodiments of the present invention.

[0026] According to another aspect of the present invention, a computer program product is also provided, including a computer program that, when executed by a processor, implements the methods of various embodiments of the present invention.

[0027] According to another aspect of the present invention, a computer program product is also provided, including a non-volatile computer-readable storage medium storing a computer program that, when executed by a processor, implements the methods of various embodiments of the present invention.

[0028] According to another aspect of the present invention, a computer program is also provided, which, when executed by a processor, implements the methods of the various embodiments of the present invention.

[0029] In this embodiment of the invention, the target neural signal of the subject is acquired, and the target neural signal is decoded using a pre-trained decoder to obtain a decoding result. Based on the decoding result, the controlled peripheral device is controlled to continuously move towards the desired contact target outside the preset area until the controlled peripheral device makes desired contact with the desired contact target. The controlled peripheral device is the desired control target of the subject. After the controlled peripheral device contacts the boundary of the preset area, it moves towards the desired contact target according to the target motion attributes, which are the speed and direction of the controlled peripheral device when it contacts the boundary of the area. This invention uses a pre-trained decoder to accurately decode the target neural signals of the subject, transforming abstract neural activity into specific motor commands. The feedforward control mode introduced during the control process ensures that once the controlled peripheral touches the boundary of a preset area, it continues to move towards the desired target in the state at the moment of contact, eliminating the need for motor command control based on decoded neural signals. This achieves the goal of accurately intercepting the desired target based on feedforward control, making the movement of the controlled peripheral more consistent with the subject's movement habits, i.e., conforming to the principles of neural control. This solves the technical problem in existing feedback control strategies that require continuous acquisition of neural signals to iteratively control the cursor to the desired target, resulting in cursor control that does not conform to the subject's movement habits, i.e., does not conform to the principles of neural control. Attached Figure Description

[0030] The accompanying drawings, which are included to provide a further understanding of the invention and form part of this application, illustrate exemplary embodiments of the invention and, together with their description, serve to explain the invention and do not constitute an undue limitation thereof. In the drawings:

[0031] Figure 1 This is a schematic flowchart of a signal processing method provided according to an embodiment of the present invention;

[0032] Figure 2 This is a flowchart illustrating a decoder training process according to an embodiment of the present invention;

[0033] Figure 3 This is a structural example diagram of a brain-computer interface system provided according to an embodiment of the present invention;

[0034] Figure 4 This is a schematic diagram of the structure of a signal processing device according to an embodiment of the present invention. Detailed Implementation

[0035] To enable those skilled in the art to better understand the present invention, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. 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 should fall within the scope of protection of the present invention.

[0036] It should be noted that the terms "first," "second," etc., in the specification, claims, and accompanying drawings of this invention are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of the invention described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover a non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0037] According to an embodiment of the present invention, a signal processing method embodiment is provided. It should be noted that the steps shown in the flowchart in the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions. Furthermore, although a logical order is shown in the flowchart, in some cases, the steps shown or described may be executed in a different order than that shown here.

[0038] Figure 1 This is a method according to an embodiment of the present invention, such as... Figure 1 As shown, the method includes the following steps:

[0039] Step S101: Obtain the target neural signals of the subject.

[0040] In this step, the execution system first captures neural activity signals in the subject's brain related to the movement of the controlled peripheral device. These signals are generated by the brain when planning and executing actions and can be collected through an implanted electrode array, reflecting the subject's anticipation and control intention for the target action.

[0041] Optionally, in some embodiments of the present invention, the controlled peripheral device can be a cursor, robotic arm, wheelchair, vehicle, or other device that interacts with the subject.

[0042] Step S102: The neural signal is decoded using a pre-trained decoder to obtain the decoding result.

[0043] In this step, the pre-trained decoder is a model trained using neural signals and cursor motion data collected in the previous training phase. It can decode the acquired neural signals into specific motion commands. In this stage, the decoder performs real-time decoding based on the acquired neural signals and outputs the motion parameters of the controlled peripheral device, such as direction and speed, as the decoding result.

[0044] Step S103: Based on the decoding result, control the controlled peripheral to continuously move towards the desired contact target outside the preset area until the controlled peripheral makes desired contact with the desired contact target. Here, the controlled peripheral is the desired control target of the subject. After the controlled peripheral contacts the boundary of the preset area, it moves towards the desired contact target according to the target motion attributes. The target motion attributes are the speed and direction of the controlled peripheral when it contacts the boundary of the area.

[0045] In this step, the execution system applies the decoding result obtained in step S102 to the control of the controlled peripheral device, guiding it to continuously move towards a dynamic target outside the preset area. It is important to note that when the controlled peripheral device contacts the boundary of the preset area, it automatically continues to move at the speed and direction at the moment of contact until it touches the desired target. That is, when the controlled peripheral device contacts the boundary of the preset area, it does not need to be controlled by the decoding result. This design achieves feedforward control; that is, after the subject issues a control command, the system automatically completes subsequent actions according to the initial settings of the command, ultimately intercepting the desired target without continuous feedback adjustments, which is more in line with the natural human motion control pattern.

[0046] Understandably, the desired contact goal is the target that the subject expects the controlled peripheral to come into contact with when controlling the controlled peripheral. Desired contact is the contact that the controlled peripheral makes with the desired contact goal that the subject desires, such as grasping or lifting.

[0047] Optionally, the preset area is a virtual area designed as a circular or elliptical ring structure. Its size and position need to be customized according to the specific circumstances of the controlled peripheral device and the target to be contacted, as well as the visual comfort of the subject. For example, for smaller targets to be contacted, the preset area needs to be set relatively compactly to improve precise control; while for targets to be contacted at a distance or at high speed, the preset area needs to be larger to ensure sufficient reaction time and space.

[0048] It should be noted that if the preset area is a two-dimensional area, the plane in which the preset area is located must be the same plane as the plane in which the motion path of the controlled peripheral is located.

[0049] For example, if the preset area is circular, the starting position of the controlled peripheral is selected as the center of the preset area.

[0050] As an optional embodiment, the preset area can be a spherical area, and the starting position of the controlled peripheral device is selected as the center of the preset area.

[0051] It should be noted that the primary function of the preset area is to trigger the feedforward control mode. When the subject touches the boundary of the preset area via a neurally controlled cursor or robotic arm, the control system automatically switches to feedforward control mode. This means that, based on the cursor's speed and direction at the moment of contact, the system automatically projects the cursor or robotic arm towards the dynamic target with pre-set dynamic characteristics. During this period, the subject can no longer adjust the trajectory in real time, thus achieving true feedforward control.

[0052] Optionally, in some embodiments of the present invention, a boundary touch threshold is set for a preset area to ensure that the feedforward mode is only switched when the subject clearly expresses control intention and the controlled peripheral device approaches the desired contact target in a certain direction, continuing to move at the speed and direction at the moment of contact until the desired contact target is touched. When the directional difference between the controlled peripheral device and the desired contact target is greater than the boundary touch threshold, the feedforward mode will not be entered, and the controlled peripheral device will stop moving when it reaches the boundary of the area. The setting of this boundary touch threshold can ensure the safety of the controlled peripheral device in narrow areas.

[0053] Optionally, in some embodiments of the present invention, a safe speed threshold is set for a preset area. Only when the speed at the moment of contact is less than or equal to the safe speed threshold is the device switched to feedforward mode, continuing to move at the speed and direction at the moment of contact until it touches the desired contact target. When the speed at the moment of contact of the controlled peripheral at the boundary of the contact area is less than or equal to the safe speed, the device will not enter feedforward mode; in this case, the controlled peripheral stops moving when it reaches the boundary of the area. This safe contact threshold setting ensures the safety of the controlled peripheral in narrow areas.

[0054] In the above embodiments, the target neural signal of the subject is acquired, and the neural signal is decoded using a pre-trained decoder to obtain the decoding result. Based on the decoding result, the movement of the controlled peripheral is controlled to move towards the desired contact target outside the preset area. The controlled peripheral is the subject's desired control target. After the controlled peripheral contacts the boundary of the preset area, it moves towards the desired contact target with the speed and direction at which it contacts the boundary. This invention, through a pre-trained decoder, can accurately decode the subject's target neural signal, transforming abstract neural activity into specific movement commands. The feedforward control mode introduced in the control process ensures that once the controlled peripheral touches the boundary of the preset area, it continues to move towards the desired contact target in the state at the moment of contact, eliminating the need for movement command control based on the decoded neural signal. This achieves the goal of accurately intercepting the desired contact target based on the feedforward control method, making the movement of the controlled peripheral more consistent with the subject's movement habits, i.e., conforming to the principles of neural control. This solves the technical problem in existing feedback control strategies that require continuous acquisition of neural signals to iteratively control the cursor to move to the desired contact target, resulting in cursor control that does not conform to the subject's movement habits, i.e., does not conform to the principles of neural control.

[0055] Optionally, in step S103, controlling the controlled peripheral to continuously move towards the desired contact target outside the preset area according to the decoding result, until the controlled peripheral makes the desired contact with the desired contact target, may include the following steps:

[0056] Step S1031: Based on the decoding result, determine the first control command, wherein the first control command is used to control the movement of the controlled peripheral device toward the boundary of the preset area.

[0057] In this step, the decoding result contains the subject's intention for the current motion task, such as movement speed and direction. Based on the decoding result, the system generates a first control command to control the movement of the controlled peripheral device towards the boundary of a preset region. Here, the "preset region" refers to a virtual or physical boundary that defines an area in space that triggers feedforward control. The purpose of the first control command is to guide the controlled peripheral device into this boundary, preparing for subsequent feedforward control.

[0058] Step S1032: In response to the controlled peripheral device contacting the boundary of the preset area, a second control command is generated, wherein the second control command includes the target motion attribute.

[0059] During the process, the system monitors the contact between the controlled peripheral device and the preset area boundary in real time. Once contact is detected, it immediately switches to the next control logic. At the instant of contact, the system generates a second control command based on the motion attributes of the controlled peripheral target, i.e., its current state (speed, direction, etc.). The second control command contains all the information required by the feedforward control strategy, ensuring that the controlled peripheral device can approach and contact the target along the optimal path.

[0060] Step S1033: According to the second control instruction, control the controlled peripheral device to make the desired contact with the desired contact target.

[0061] In this step, guided by the second control command, the feedforward controller adjusts the speed and direction of the controlled peripheral device, causing it to move rapidly towards the desired target along a pre-set path. During this stage, the subject no longer needs to actively control the controlled peripheral device, but instead temporarily relinquishes control to the system. Ultimately, the controlled peripheral device makes contact with the dynamic target according to the second control command, completing the predetermined interception task. The contact can be visual (e.g., the cursor aligning with the dynamic target) or physical (e.g., direct contact between a robotic arm and the target).

[0062] Understandably, based on the above steps, the decoder first generates and executes a first control command to guide the controlled peripheral device into a preset area. Then, upon contact with the boundary of the preset area, the system automatically activates a feedforward control mode, achieving precise contact with the dynamic target according to a second control command. This control process not only improves the interception efficiency and success rate of dynamic targets but also reduces the cognitive load on the subject during the control process, making the control process more natural and intuitive.

[0063] Optionally, in step 103, controlling the controlled peripheral to continuously move towards the desired contact target outside the preset area based on the decoding result, until the controlled peripheral makes the desired contact with the desired contact target, may include the following steps:

[0064] Step S1034: Based on the decoding result, control the controlled peripheral to continuously move towards the desired contact target outside the preset area, and obtain the positional difference between the controlled peripheral and the desired contact target when the controlled peripheral contacts the boundary of the preset area.

[0065] In this step, the system generates control commands based on the neural signal results parsed by the decoder, guiding the controlled peripheral device to move towards a dynamic target outside the preset area. Simultaneously, the system records the positional difference between the controlled peripheral device and the desired contact target in real time, providing data support for subsequent judgment and compensation. Positional difference refers to the spatial distance between the controlled peripheral device and the target, and is an important indicator for evaluating control effectiveness and predicting the likelihood of contact.

[0066] Optionally, the preset area is a specific area set by the system for the controlled peripheral device to trigger the feedforward control mode. Its boundary is a certain distance from the target motion trajectory, so that the controlled peripheral device has enough space and time to adjust its motion state.

[0067] Step S1035: Based on the positional differences, the target motion attributes are judged to obtain the judgment result.

[0068] Specifically, when the controlled peripheral device comes into contact with the boundary of the preset area, the system will, based on the real-time positional differences and the target's motion attributes (such as velocity, direction, acceleration, etc.), make a judgment: whether the controlled peripheral device, based on the current target motion attributes, can continue to move and make contact with the desired target. The judgment result reflects the degree of matching between the current control strategy and the actual motion state of the dynamic target, and is the basis for deciding whether to activate the compensation mechanism. If the judgment result shows that control based on the current target motion attributes cannot enable the controlled peripheral device to successfully make contact with the target, the system will proceed to the next step and perform compensation adjustments.

[0069] Step S1036: In response to the judgment result indicating that the controlled peripheral device cannot contact the desired contact target based on the target motion attribute, the target motion attribute compensation value is determined based on the position difference.

[0070] Specifically, when contact cannot be achieved based on the current strategy, the system calculates compensation values ​​for the target's motion attributes based on the current positional differences. Positional differences provide direct information about control deviations; by analyzing this information, the system can estimate the degree and direction to which the target's motion attributes (such as speed and direction) need adjustment to achieve contact. Determining the compensation values ​​for the target's motion attributes is a crucial calculation process. It comprehensively considers the current state of the controlled peripheral, the target's motion trend, and the preset control objectives to calculate the most suitable compensation parameters to correct the deviations in the control strategy.

[0071] Step S137: Compensate the target motion attributes according to the target motion attribute compensation value.

[0072] In this step, after obtaining the target motion attribute compensation value, the system will make corresponding compensation adjustments to the target motion attribute. The controlled peripheral moves based on the compensated target motion attribute, which can better match the motion trajectory of the controlled peripheral, thereby increasing the possibility of contact between the two.

[0073] Optionally, in some embodiments of the present invention, this compensation process can be dynamic, that is, the system continuously adjusts as the controlled peripheral approaches the desired contact target to correct the control parameters in real time, so as to ensure that even if the target motion properties are unstable or the prediction is inaccurate, the controlled peripheral can approach and contact the target as closely as possible.

[0074] Understandably, based on the above steps, through real-time positional difference analysis and target motion attribute compensation, the system can dynamically adjust the control strategy, overcome the uncertainty of the moving target, and improve the accuracy and success rate of contact between the controlled peripheral device and the dynamic target, thereby significantly enhancing the practical value and user experience of brain-computer interfaces in complex environments.

[0075] Optionally, the signal processing method further includes:

[0076] Step S104: Based on the decoding result, control the controlled peripheral device to continuously move towards the desired contact target outside the preset area and record the stopping position of the controlled peripheral device when it stops moving.

[0077] Specifically, based on the decoding results, the system controls the controlled peripheral device to continuously move towards the desired contact target (i.e., the dynamic target). The preset area is a key reference point in the control system; when the controlled peripheral device comes into contact with this area, it triggers feedforward control. When the controlled peripheral device stops moving, the system records its stopping position.

[0078] Step S105: In response to the stop position indicating that the controlled peripheral device has not made the desired contact with the target, the area adjustment parameters are determined based on the stop position.

[0079] Specifically, if the stopping position indicates that the controlled peripheral device failed to successfully contact the dynamic target, this may involve factors such as insufficient control precision by the subject or errors in the system's prediction of the target's motion. In this case, the region adjustment parameters can be determined based on the stopping position. These parameters can be the size, shape, and position of a preset region. By adjusting these parameters, the probability of the controlled peripheral device contacting the target can be optimized. For example, if the subject frequently stops when approaching the target but not making complete contact, the system may increase the preset region to give the subject more room for error; conversely, if the subject's control is overly conservative and does not fully utilize the preset region, the system may decrease the preset region to encourage more precise control.

[0080] Step S106: Adjust the preset area according to the area adjustment parameters.

[0081] After determining the regional adjustment parameters, the system will make corresponding adjustments to the preset region to adapt to the control characteristics of the subject and the motion characteristics of the dynamic target.

[0082] Alternatively, this adjustment can be dynamic, that is, gradually fine-tuned through multiple trials until the best combination of parameters is found.

[0083] As an optional implementation, the pre-training of the decoder includes the following steps:

[0084] Step S1021: Obtain the subject's first neural signal;

[0085] During the training of the decoder, it is necessary to acquire the first neural signals generated by the subject during the training phase. The first neural signals reflect the subject's motor intentions and goal-oriented information during the training phase.

[0086] Step S1022: Input the first neural signal into the pre-corrected decoder to obtain the first decoding result;

[0087] In this step, the pre-corrected decoder refers to a decoding model that has been initially adjusted to match the specific neural signal characteristics of the subject. By inputting the acquired first neural signal into the pre-corrected decoder, the decoder can convert the first neural signal into executable motor commands, i.e., the first decoding result. The first decoding result contains the subject's expectations of the motion parameters such as direction and speed of the controlled peripheral device.

[0088] Step S1023: Based on the first decoding result, control the controlled peripheral device to move toward the desired contact target outside the preset area to obtain the first motion result. Here, the controlled peripheral device is the subject's desired control target. After the controlled peripheral device contacts the boundary of the preset area, it moves toward the desired contact target with the speed and direction at which it contacts the boundary.

[0089] In this step, the preset area is a virtual ring-shaped area, which includes the controlled peripheral device. The target to be contacted is outside the preset area. The controlled peripheral device starts moving from the preset area. When it moves to the boundary of the preset area, the feedforward control mode is triggered. That is, the controlled peripheral device will continue to move at the speed and direction at the boundary, moving towards the target to be contacted. The movement process can be completed without further neural signal input from the subject, thus obtaining the first movement result.

[0090] Step S1024: Based on the first motion result, adjust the decoding parameters in the pre-corrected decoder to obtain the target decoder.

[0091] During training, after completing an interception attempt to reach the desired target, the system evaluates the effectiveness of the initial motion result, such as whether the target was successfully intercepted and whether the motion trajectory matches expectations. Based on these evaluation results, the decoder parameters are further adjusted to improve its accuracy and control over decoding neural signals.

[0092] Optionally, the process can be performed iteratively, training the pre-corrected decoder iteratively, and by continuously fine-tuning the decoding parameters, a target decoder that can decode and execute the subject's motion intentions more efficiently and accurately is finally obtained, laying the foundation for achieving more natural target interception.

[0093] As an optional implementation, before inputting the first neural signal into the pre-corrected decoder to obtain the first decoding result, the method further includes:

[0094] Step S1011: Within a first preset time period, the controlled peripheral device is controlled to move toward the desired contact target multiple times to obtain multiple movement states of the controlled peripheral device and multiple second neural signals. The second neural signals are generated based on the visual stimulus observed by the subject that the controlled peripheral device moves toward the desired contact target.

[0095] In this step, the first preset time period refers to the initial stage specifically used for decoder training, and the controlled peripheral device is a cursor, robotic arm, or other similar device that moves under brain control. During this stage, the subject receives visual feedback by observing the controlled peripheral device moving towards the desired contact target, generating a second neural signal corresponding to the actual movement intention. The system will repeatedly record the second neural signal and the actual movement state corresponding to the controlled peripheral device, forming a preliminary training dataset. This data will be used to initialize the decoder, enabling it to initially recognize and decode the subject's movement intention.

[0096] Step S1012: Based on the motion states of multiple controlled peripherals and multiple second neural signals, train the decoder to obtain the initialized decoder;

[0097] With the training data collected in step S1011, the decoder will be trained using machine learning algorithms to establish a mapping relationship between neural signals and the actual motion state of the controlled peripheral device. This stage of decoder training is fundamental. Through continuous learning and adjustment, the decoder gradually develops an understanding of the second neural signals and can convert them into motion commands, thus initially realizing the decoding of the subject's neural signals. After training is completed, the decoder initialization is finished, preparing it for subsequent pre-correction and practical application.

[0098] For example, the initialization process involves another subject observing a cursor moving towards a target, while the system simultaneously collects the cursor's motion state and the subject's neural signals at the current moment. After 2-3 minutes, these two types of data are used to train the decoder.

[0099] Step S1013: Perform pre-correction on the initialized decoder to obtain the pre-corrected decoder.

[0100] The pre-correction phase aims to enable the initialized decoder to more accurately match and decode the subject's neural signal characteristics. During this phase, the decoder's performance is fine-tuned through the subject's actual operation in feedback control mode. The system compares the motion commands output by the decoder with the actual motion effects the subject desires, and adjusts the parameters in the decoder accordingly to reduce bias and improve decoding accuracy and efficiency. The pre-corrected decoder will be more stable and accurate, better serving subsequent target interception tasks.

[0101] As an optional implementation, the decoder pre-correction includes the following steps:

[0102] Step S1014: Acquire multiple third neural signals from the subject within the second preset time period;

[0103] Step S1015: For multiple third neural signals, a pre-correction process is performed to obtain a pre-corrected decoder. The pre-correction process includes: inputting multiple third neural signals into the decoder to obtain multiple second decoding results; determining multiple control signals based on the multiple second decoding results and a preset signal determination formula; controlling the movement of the controlled peripheral device towards the desired contact target based on the multiple control signals to obtain multiple second motion results; and correcting the decoding parameters in the decoder based on the multiple second motion results to obtain the pre-corrected decoder.

[0104] Specifically, the second preset time period refers to the time period corresponding to the pre-correction process after the decoder initialization. The purpose of this stage is to further learn from the neural signals generated by the subjects in real tasks, so that the decoder's performance is closer to the individual differences and actual control needs of the subjects. Multiple third neural signals are neural activity information generated by the brain during the process of the subjects attempting to control the controlled peripheral device to intercept the desired contact target. When acquiring multiple third neural signals, the pre-correction process is performed by traversing multiple third neural signals to obtain the pre-corrected decoder.

[0105] The pre-correction process includes: First, multiple third neural signals are sequentially input into the decoder to obtain second decoding results. Each time a third neural signal is input, the decoder attempts to decode the subject's brain signals and outputs corresponding control commands, resulting in multiple second decoding results used to guide the movement of the controlled peripheral device. Then, a control signal is determined based on the second decoding results and a preset formula. The preset formula is used to generate an auxiliary control signal in conjunction with the decoding results. The calculated control signal is applied to drive the controlled peripheral device towards the desired contact target. The system records the movement state of the controlled peripheral device, including speed, direction, and acceleration, to obtain second motion results, which are used to evaluate the immediate effect of the decoder and control strategy. The final step of pre-correction is to correct the decoder parameters based on the second motion results. This is an iterative optimization process. The system analyzes the deviation between the second motion results and the desired motion, adjusts the parameter weights within the decoder, and gradually narrows the gap between the decoding results and the actual motion until a satisfactory performance standard is achieved.

[0106] Understandably, by optimizing the decoder with multiple third neural signals, the pre-corrected decoder will have better performance, be able to decode the subject's neural signals more accurately, and achieve effective control over external devices.

[0107] As an optional implementation, the preset signal determination formula includes an auxiliary vector, auxiliary coefficients, a decoding vector, and decoding coefficients, wherein the auxiliary vector is a vector pointing to the controlled peripheral device, and the decoding vector is the output result of the decoder.

[0108] An auxiliary vector is a fixed directional vector that always points towards the intercepted target, providing the subject with intuitive target guidance. During training and control, the auxiliary vector acts as a "navigator," providing directional guidance even when the subject's neural signal decoding is not entirely accurate, helping the controlled peripheral device move more directly towards the dynamic target. The presence of the auxiliary vector allows the subject to focus on neural control rather than overly complex motion planning; it is particularly important for improving task completion rates, especially in the early stages of pre-correction.

[0109] The auxiliary coefficient is an adjustable parameter used to control the contribution ratio of the auxiliary vector in the total control signal. In the pre-correction phase, the auxiliary coefficient is initially set relatively high, meaning the auxiliary vector has a strong influence on the direction of motion of the controlled peripheral. As the decoder performance improves and the subject becomes more familiar with the system's control logic, the auxiliary coefficient gradually decreases, eventually dropping to zero. This means that in the later stages of pre-correction, the subject's control will primarily rely on their own neural signals, while the influence of the auxiliary vector will weaken until it disappears completely. At this point, the subject can independently achieve precise control of the controlled peripheral through the decoded vector.

[0110] The decoded vector is the output of the decoder after decoding the subject's neural signals. It represents the subject's expected movement of the controlled peripheral device, including but not limited to direction and speed. By learning and matching the association between neural signals and actual movement commands, the decoder can convert complex neural activity into specific control commands. As the decoder's pre-correction improves, the accuracy and stability of the decoded vector will gradually increase, becoming the main basis for the subject to control the controlled peripheral device.

[0111] Decoding coefficients are used to adjust the weight of the decoding vector in the overall control signal. Unlike auxiliary coefficients, decoding coefficients increase as the pre-correction process progresses, reflecting the increasing dominance of the subject's neural signals in the actual motion control of the controlled peripheral. An increase in decoding coefficients means that the influence of the decoding vector on motion commands gradually increases until it ultimately dominates the entire control process, reflecting the subject's mastery of the system and the improved decoding ability of the decoder.

[0112] Understandably, the preset signal determination formula constructs a dynamic control signal by integrating auxiliary vectors and decoding vectors, as well as their corresponding auxiliary coefficients and decoding coefficients. This control signal guides the motion of the controlled peripheral device to the desired contact with the target, while allowing the role of the subject's neural signals in the control to gradually increase until a feedforward control based entirely on neural signals is achieved, which greatly enhances the system's application potential and user experience in complex tasks such as target interception.

[0113] As an optional implementation, determining the control signal based on the second decoding result and the preset signal determination formula includes: acquiring the previous motion result, wherein the previous motion result is the motion result corresponding to the third neural signal acquired at the previous acquisition time; adjusting the auxiliary coefficient and the decoding coefficient based on the previous motion result to obtain the adjusted preset signal determination formula; and determining the control signal based on the adjusted preset signal determination formula.

[0114] Obtaining the previous motion result: In the pre-correction process, after decoding each third neural signal and completing a control attempt based on the decoding result, the system records the result of that control, i.e., the previous motion result. It includes key information such as the motion trajectory, speed, acceleration of the controlled peripheral, and the degree of proximity to the dynamic target or whether the interception was successful.

[0115] Based on the previous exercise result, the auxiliary coefficient and decoding coefficient are adjusted to obtain the adjusted preset signal determination formula. After evaluating the previous exercise result, the system will adjust the auxiliary coefficient and decoding coefficient through a pre-set adjustment strategy. The adjustment goal of the auxiliary coefficient is usually to reduce its value, which means reducing the proportion of the auxiliary vector in the total control signal, thereby encouraging the subject to rely more on its own neural signals for control. Conversely, the adjustment of the decoding coefficient tends to increase its value, thereby enhancing the influence of the decoding vector in the total control signal. The dynamic adjustment of these two coefficients reflects the improvement of the subject's skill level and the optimization of the decoder's decoding ability. The ultimate goal is to enable the subject to control the controlled peripheral device through pure neural signals without relying on auxiliary signals, achieving efficient and natural target interception.

[0116] Based on the adjusted preset signal determination formula, the control signal is determined: once the auxiliary coefficients and decoding coefficients are appropriately adjusted, the system uses the updated preset signal determination formula to decode the new round of third neural signals and generate the next control signal. Adjusting the preset signal determination formula is essentially an optimization of the control strategy; that is, while retaining the guiding role of the auxiliary vector, it gives the decoding vector greater weight, making the control signal closer to the subject's true movement intention, thereby improving the accuracy and flexibility of control. In the pre-correction stage of the decoder, this control signal will directly guide the movement of the controlled peripheral, enabling it to more effectively track or intercept targets, ultimately reflected in the improvement of the second movement result.

[0117] The purpose of decoder correction is to achieve bidirectional adaptation between the decoder and the subject's neural signals. For example, during the correction phase, the subject's control over the cursor equals the auxiliary vector * auxiliary coefficient + decoding vector * decoding coefficient. The auxiliary vector is a vector that always points to the target, the auxiliary coefficient is a constant that can be adjusted during the task and controlled by the system based on the subject's performance, and the decoding vector is the actual output of the decoder. The decoding coefficient, similar to the auxiliary coefficient, can also be controlled. During correction, the system simultaneously collects the corrected control result and the current neural signals to correct the decoder. The corrected control result maintains the magnitude of the control over the cursor unchanged, but directs the control towards the target. During correction, the system adjusts the auxiliary coefficient and the decoding coefficient, eventually reducing the auxiliary coefficient to 0 and increasing the decoding coefficient to 1 or higher, finally pausing correction as needed.

[0118] As an optional implementation, the signal processing method further includes: determining the number of second motion results that meet the standard from a plurality of second motion results, wherein a second motion result that meets a first preset condition is determined to meet the standard; and obtaining a pre-corrected decoder in response to the number of second motion results meeting the second preset condition.

[0119] During the pre-correction phase, subjects control a peripheral device (such as a cursor or robotic arm) to move towards a dynamic target via neural signals. The system records the movement result of each attempt, defining it as the second movement result. This includes specific data such as the movement trajectory, speed, and proximity to the target of the controlled peripheral device.

[0120] The first preset condition defines a successful target interception operation. That is, when the movement result of the controlled peripheral device reaches or exceeds a preset performance indicator, the movement result is considered satisfactory. These indicators include, but are not limited to, the accuracy of the movement trajectory and the reasonableness of the movement speed, depending on the application scenario and task requirements.

[0121] The system filters out movement results that meet the first preset condition from all collected second movement results and calculates the number of times the target is met. The number of times the target is met reflects the number of times the subject can successfully control the controlled peripheral device to complete the task by adjusting neural signals, and is an important basis for determining whether the expected goal has been achieved in the pre-correction stage.

[0122] The second precondition is the standard for completion of the pre-correction phase, which is manifested in the number of successful attempts reaching a specific value or percentage, indicating that the neural signal decoding between the subject and the decoder has reached a stable and efficient interactive state. For example, the pre-correction phase requires that the success rate of the subject's operation reach more than 80% within one minute, or that the number of successful attempts remain stable at a high level in multiple consecutive attempts.

[0123] When the number of successful tests meets the second preset condition, it indicates that the decoder has completed pre-correction, and its accuracy and efficiency in decoding neural signals have met the design requirements. At this point, the pre-corrected decoder will be retained and used for subsequent target interception tasks, ensuring that the subject can control the controlled peripheral device in a more natural and precise way, and achieve interaction and control of the dynamic environment.

[0124] As an optional implementation, the signal processing method further includes:

[0125] Step S1025: Acquire multiple fourth neural signals from the subject within the third preset time period;

[0126] The third preset time period refers to the period after the decoder pre-correction is completed and the feedforward mode training to obtain the target decoder is initially completed, and then the verification is carried out. During the third preset time period, the subject will continue to perform the target interception task in feedforward control mode, generating multiple fourth neural signals.

[0127] Step S1026: Input multiple fourth neural signals into the target decoder to obtain multiple third decoding results;

[0128] The target decoder refers to a pre-corrected and feedforward-trained decoding model that can more accurately decode the subject's neural signals and understand the subject's motor intentions. When the subject's fourth neural signal is input into the target decoder, the decoder will output multiple third decoding results based on the pre-trained neural signal decoding model. These results contain interpretations of the subject's motor commands, such as specific control instructions for direction and speed.

[0129] Step S1027: Based on multiple third decoding results, control the movement of the controlled peripheral device toward the desired contact target outside the preset area to obtain multiple third motion results;

[0130] After obtaining the third decoding result, the system will drive the controlled peripheral device (such as a cursor or robotic arm) to perform corresponding movements in an attempt to intercept a dynamic target located outside a preset area, thereby obtaining multiple third motion results. These multiple third motion results reflect the actual motion state of the controlled peripheral device after executing the control commands converted from the subject's neural signals and whether the target can be successfully intercepted.

[0131] Step S1028: Based on multiple third motion results and preset verification rules, perform usability verification on the target decoder to obtain the verification result.

[0132] The pre-defined validation rules are a set of standards used to evaluate the effectiveness of the target decoder and the subject's control ability. These rules can include multiple dimensions such as the accuracy of the motion trajectory, the success rate of target interception, and motion smoothness. The system compares multiple third-party motion results with these rules to determine whether the target decoder can stably and accurately convert the subject's neural signals into effective control actions, thereby successfully intercepting the dynamic target. The validation results will determine whether the target decoder can be formally put into feedforward control mode to achieve more natural and smooth target interaction, and whether further fine-tuning and optimization are needed.

[0133] As an optional implementation, the target to be contacted is a dynamic target. Before controlling the movement of the controlled peripheral device toward the target to be contacted outside the preset area according to the first decoding result and obtaining the first motion result, the method further includes: controlling the movement of the target to be contacted according to the preset motion trajectory and the preset motion speed.

[0134] If training is required to intercept dynamic targets, the system will control the target to move before the subject begins to interact. The preset trajectory and speed can be determined based on the specific training task and application context. For example, the dynamic target may rotate at a constant speed along a circular path, or move at a random but known speed and direction on a two-dimensional plane. These preset parameters provide the basis for the system to control the movement of the dynamic target, ensuring that the target moves as intended within the preset area, providing the subject with a stable dynamic interaction object.

[0135] The purpose of this step is to create a realistic dynamic target environment in which subjects can practice and test their control skills. It also provides the decoder with real-time feedback on the movement of the dynamic target, helping the decoder to learn and adapt to the movement characteristics of the dynamic target, so as to more effectively process and predict the neural signals generated by the subjects when facing the dynamic target.

[0136] As an optional implementation, the signal processing method further includes: updating the preset motion trajectory and preset motion speed according to preset training requirements.

[0137] Preset training requirements are preset goals based on the training plan and the development of the subject's control ability. For example, in the early stages of training, the preset motion trajectory is defined as a simple and regular circular path, and the preset motion speed is relatively slow, such as 15° / second. This design is to allow the subject to gradually become familiar with the system and establish a basic control mode.

[0138] As subjects gain a deeper understanding of the system and improve their control skills, the preset motion parameters should be adjusted accordingly to introduce more challenges and further enhance control skills. For example, when subjects can stably intercept the target at lower speeds, the training requirements shift to increasing the target's speed or changing the complexity of the trajectory, such as adding irregularly shaped trajectories or introducing multi-directional abrupt changes in the motion path.

[0139] The system can monitor the subject's control performance and success rate, and determine whether the training goal under the current motion parameters has been achieved based on this data. Once achieved, the system will automatically or the experimenter will manually adjust the preset motion trajectory and preset motion speed to introduce a new level of difficulty. This adjustment is usually done gradually, for example, from 15° / second to 30° / second, then to 60° / second, and finally to 120° / second, ensuring the continuity and controllability of the training.

[0140] As an optional implementation, the target decoder is obtained by adjusting the decoding parameters in the pre-corrected decoder based on the first motion result, including: in response to the first motion result indicating that the controlled peripheral device failed to intercept the desired contact target, determining the deviation between the controlled peripheral device and the intercepted target; and adjusting the decoding parameters based on the deviation to obtain the target decoder.

[0141] When the system detects that the controlled peripheral device has failed to successfully intercept a moving target, it calculates the deviation between the two. This deviation includes, but is not limited to, distance deviation, time deviation, and velocity deviation, depending on the nature of the motion task and the system design. Determining this deviation provides a quantitative basis for subsequent parameter adjustments, ensuring that the adjustments are targeted and effective.

[0142] Decoding parameters are the variables and coefficients in the decoder model used to decode the subject's neural signals and convert them into specific control commands. They directly affect the decoder's decoding efficiency and control accuracy. Parameter tuning based on deviation is a closed-loop learning process. The system analyzes the causes of deviation and adjusts the decoding parameters accordingly, such as changing the ratio of auxiliary coefficients to decoding coefficients or fine-tuning the parameter matrix of the linear model. The goal of this process is to minimize the deviation, allowing the controlled peripheral device to respond more accurately and quickly to the subject's neural signals, thus improving the success rate of intercepting dynamic targets. After this adjustment and optimization, the final target decoder will have higher decoding accuracy and control stability, better supporting the subject's interactive tasks in complex dynamic environments.

[0143] As an optional implementation, the signal processing method further includes: acquiring region adjustment parameters; and adjusting the size and shape of a preset region according to the region adjustment parameters.

[0144] The region adjustment parameters are used to indicate changes in the size and shape of the preset region. After obtaining the region adjustment parameters, the system can adjust the size and shape of the preset region according to the obtained parameters.

[0145] It's important to note that in practical applications of the decoder, a smaller preset area means that the subject must control the controlled peripheral more precisely to trigger the feedforward control mode. This increases the task's difficulty but also promotes improvement in the subject's control precision. Conversely, a larger preset area provides the subject with greater error tolerance, which is beneficial for initial learning and exploration. The shape of the area can be adjusted based on the dynamic target's motion pattern and the subject's control habits. For example, if the dynamic target moves along a circle, the preset area can be designed as a fan-shaped or arc-shaped region tangent to the target's trajectory, allowing the subject to trigger feedforward control at a specific angle. If the target's trajectory is more complex, the shape of the preset area will also change accordingly to match the subject's most effective control strategy.

[0146] Optionally, refer to Figure 2In some embodiments of the present invention, during the correction phase of the MC decoder (motor cortex decoder), after receiving neural signals generated from the posterior parietal cortex (PPC) and motor cortex (MC) in the brain regions, the decoder uses the ReFIT-KF (Recalibrated Feedback Intention Trained Kalman Filter) algorithm to decode the neural signals, decoding cursor control information including speed and direction. The cursor movement based on the control information is displayed on the peripheral device. After seeing the cursor movement, auxiliary vector, and decoding vector displayed on the peripheral device, the subject generates visual feedback to the brain regions. The subject can then generate neural signals again based on the visual feedback, decode them based on the adjusted decoder, and control the cursor according to the decoding result, completing the feedback correction process.

[0147] During the training phase of feedforward intracortical brain-computer interfaces (iBMIs), after receiving neural signals from the posterior parietal cortex (PPC) and motor cortex (MC) in the brain regions, the decoder uses the ReFIT-KF (Recalibrated FeedbackIntention Trained Kalman Filter) algorithm to decode the neural signals, extracting cursor control information including velocity and direction. The peripheral device displays the cursor's movement based on this control information. When the cursor touches a ring, a feedforward control command is generated to control the cursor. After seeing the cursor movement displayed on the peripheral device, the subject generates visual feedback to the brain regions. Based on this visual feedback, the subject can then generate neural signals again to control the cursor, iteratively executing the above steps until the training objective is achieved.

[0148] It should be noted that, Figure 2 In this context, 'cs' refers to the central sulcus in the cerebral cortex, and 'as' refers to the arcuate sulcus in the cerebral cortex.

[0149] Through the above description of the embodiments, those skilled in the art can clearly understand that the methods according to the above embodiments can be implemented by means of software plus necessary general-purpose hardware platforms. Of course, they can also be implemented by hardware, but in many cases the former is a better implementation method. Based on this understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product is stored in a storage medium (such as ROM / RAM, magnetic disk, optical disk) and includes several instructions to cause a terminal device (which may be a mobile phone, computer, server, or network device, etc.) to execute the methods described in the various embodiments of the present invention.

[0150] Reference Figure 3 This invention also provides a brain-computer interface system, comprising: a neural signal acquisition unit 201 for acquiring neural signals of a subject; a decoder 202 for decoding the neural signals to obtain a decoding result; and a feedforward controller 203 for controlling a controlled peripheral device to continuously move towards a desired contact target outside a preset area according to the decoding result, until the controlled peripheral device makes desired contact with the desired contact target, wherein the controlled peripheral device is the subject's desired control target, and after the controlled peripheral device contacts the boundary of the preset area, it moves towards the desired contact target according to the target motion attributes, wherein the target motion attributes are the speed and direction of the controlled peripheral device when it contacts the boundary of the area.

[0151] The brain-computer interface system (BCI) is designed as an integrated control platform. Its core components include a neural signal acquisition unit 201, a decoder 202, and a feedforward controller 203. These components work together to achieve efficient interaction between the subject and a dynamic target. The following is a detailed explanation of the functions and workflows of each component:

[0152] A neural signal acquisition device serves as the interface between the brain-computer interface (BCI) system and the subject's brain. Its primary task is to monitor and acquire neural signals generated by the subject during task execution in real time. These signals contain the subject's intentions and planning for the motor task and are fundamental data for the entire BCI system's operation. Neural signal acquisition devices can employ invasive or non-invasive signal acquisition techniques. The most suitable acquisition method should be selected based on needs and availability to ensure signal quality and acquisition efficiency.

[0153] The decoder is a key component that converts acquired neural signals into understandable control commands. The decoder uses an algorithm and model, such as a machine learning model, to decode the neural signals and obtain the decoding result. The decoding result reflects the subject's current motor intention, such as the expected direction and speed of cursor movement. The performance of the decoder directly affects the system's understanding of the subject's intention and the accuracy of command generation.

[0154] Based on the decoding results provided by the decoder, the feedforward controller controls the controlled peripheral device (such as a cursor or robotic arm) to move towards the desired contact target outside the preset area. In traditional feedback control modes, the subject needs to adjust their control strategy in real time to cope with the dynamic changes of the target. In this invention, when the controlled peripheral device contacts the boundary of the preset area, the feedforward controller automatically and continuously drives the controlled peripheral device towards the desired contact target in a feedforward manner similar to natural motion, based on the speed and direction at the time of contact, without requiring real-time input from the subject. This mechanism improves the coherence and efficiency of control, reduces the need for real-time attention from the subject, and also simulates the natural motion control mode of the human body, improving the naturalness and fluency of operation.

[0155] Understandably, the workflow of the entire brain-computer interface system is as follows: The neural signal acquisition unit continuously acquires neural signals from the subject's brain and transmits them to the decoder for decoding. Based on the decoding results of the neural signals, the decoder generates corresponding control commands, which are then executed by the feedforward controller to control the movement of the controlled peripheral device. When the controlled peripheral device touches the boundary of a preset area, the feedforward controller will automatically adopt a feedforward control strategy to guide the controlled peripheral device to move rapidly with the speed and direction at the time of contact until it intercepts the target, thereby achieving efficient and natural control in a dynamic environment.

[0156] This invention also provides a signal processing apparatus for implementing the above embodiments and preferred embodiments; details already described will not be repeated. As used below, the term "module" can refer to a combination of software and / or hardware that performs a predetermined function. Although the apparatus described in the following embodiments is preferably implemented in software, hardware implementation, or a combination of software and hardware, is also possible and contemplated.

[0157] Figure 4 This is a structural block diagram of a signal processing apparatus according to one embodiment of the present invention, such as... Figure 4 As shown, the device includes: an acquisition module 301 for acquiring the target neural signal of the subject; a decoding module 302 for decoding the neural signal using a pre-trained decoder to obtain a decoding result; and a control module 303 for controlling a controlled peripheral device to continuously move towards a desired contact target outside a preset area according to the decoding result, until the controlled peripheral device makes desired contact with the desired contact target. The controlled peripheral device is the subject's desired control target. After the controlled peripheral device contacts the boundary of the preset area, it moves towards the desired contact target according to target motion attributes, where the target motion attributes are the speed and direction of the controlled peripheral device when it contacts the boundary of the area.

[0158] Embodiments of this application also provide an electronic device, including: a memory storing an executable program; and a processor for running the program, wherein the program executes the methods in various embodiments of the present invention during runtime.

[0159] Optionally, in this embodiment, the processor in the above-described electronic device can be configured to run an executable program to perform the following steps:

[0160] Step S101: Obtain the target neural signals of the subject.

[0161] Step S102: The neural signal is decoded using a pre-trained decoder to obtain the decoding result.

[0162] Step S103: Based on the decoding result, control the controlled peripheral to continuously move towards the desired contact target outside the preset area until the controlled peripheral makes desired contact with the desired contact target. Here, the controlled peripheral is the desired control target of the subject. After the controlled peripheral contacts the boundary of the preset area, it moves towards the desired contact target according to the target motion attributes. The target motion attributes are the speed and direction of the controlled peripheral when it contacts the boundary of the area.

[0163] Optionally, the processor in the aforementioned electronic device may be configured to run an executable program to perform the following steps: determining a first control instruction based on the decoding result, wherein the first control instruction is used to control the movement of the controlled peripheral device toward the boundary of a preset area; generating a second control instruction in response to the controlled peripheral device contacting the boundary of the preset area, wherein the second control instruction includes target motion attributes; and controlling the controlled peripheral device to make desired contact with the desired target based on the second control instruction.

[0164] Optionally, the processor in the aforementioned electronic device may be configured to run an executable program to perform the following steps: based on the decoding result, controlling the controlled peripheral to continuously move towards the desired contact target outside the preset area, and obtaining the positional difference between the controlled peripheral and the desired contact target when the controlled peripheral contacts the boundary of the preset area; based on the positional difference, judging the target motion attribute and obtaining the judgment result; in response to the judgment result indicating that the controlled peripheral cannot contact the desired contact target based on the target motion attribute, determining the target motion attribute compensation value based on the positional difference; and compensating the target motion attribute based on the target motion attribute compensation value.

[0165] Optionally, the processor in the aforementioned electronic device may be configured to run an executable program to perform the following steps: based on the decoding result, controlling the controlled peripheral to continuously move towards the desired contact target outside the preset area and recording the stopping position of the controlled peripheral when it stops moving; in response to the stopping position indicating that the controlled peripheral and the desired contact target have not made the desired contact, determining the area adjustment parameter based on the stopping position; and adjusting the preset area based on the area adjustment parameter.

[0166] Optionally, the processor in the aforementioned electronic device can be configured to run an executable program to perform the following steps: acquiring a first neural signal from the subject; inputting the first neural signal into a pre-corrected decoder to obtain a first decoding result; controlling the movement of a controlled peripheral device toward a desired contact target outside a preset area according to the first decoding result to obtain a first movement result, wherein the controlled peripheral device is the subject's desired control target, and after the controlled peripheral device contacts the boundary of the preset area, it moves toward the desired contact target at the speed and direction at which the controlled peripheral device contacts the boundary; and adjusting the decoding parameters in the pre-corrected decoder according to the first movement result to obtain a target decoder.

[0167] Optionally, the processor in the aforementioned electronic device can be configured to run an executable program to perform the following steps: within a first preset time period, repeatedly controlling the movement of the controlled peripheral device toward the desired contact target to obtain multiple movement states of the controlled peripheral device and multiple second neural signals, wherein the second neural signals are generated based on the visual stimulus observed by the subject that the controlled peripheral device moves toward the desired contact target; training the decoder according to the multiple movement states of the controlled peripheral device and the multiple second neural signals to obtain an initialized decoder; and pre-correcting the initialized decoder to obtain a pre-corrected decoder.

[0168] Optionally, the processor in the aforementioned electronic device can be configured to run an executable program to perform the following steps: acquiring multiple third neural signals from the subject within a second preset time period; performing a pre-correction process for the multiple third neural signals to obtain a pre-corrected decoder, wherein the pre-correction process includes: inputting the multiple third neural signals into the decoder to obtain multiple second decoding results; determining multiple control signals based on the multiple second decoding results and a preset signal determination formula; controlling the movement of the controlled peripheral device toward the desired contact target based on the multiple control signals to obtain multiple second motion results; and correcting the decoding parameters in the decoder based on the multiple second motion results to obtain a pre-corrected decoder.

[0169] Optionally, the preset signal determination formula includes an auxiliary vector, auxiliary coefficients, a decoding vector, and decoding coefficients, wherein the auxiliary vector is a vector pointing to the controlled peripheral device, and the decoding vector is the output result of the decoder.

[0170] Optionally, the processor in the aforementioned electronic device may be configured to run an executable program to perform the following steps: acquiring the previous motion result, wherein the previous motion result is the motion result corresponding to the third neural signal acquired at the previous acquisition time; adjusting the auxiliary coefficient and the decoding coefficient according to the previous motion result to obtain the adjusted preset signal determination formula; and determining the control signal according to the adjusted preset signal determination formula.

[0171] Optionally, the processor in the above-mentioned electronic device may be configured to run an executable program to perform the following steps: determining the number of second motion results that meet the standard from a plurality of second motion results, wherein a second motion result that meets a first preset condition is determined to meet the standard; and obtaining a pre-corrected decoder in response to the number of second motion results meeting the second preset condition.

[0172] Optionally, the processor in the aforementioned electronic device may be configured to run an executable program to perform the following steps: acquiring multiple fourth neural signals of the subject within a third preset time period; inputting the multiple fourth neural signals into the target decoder to obtain multiple third decoding results; controlling the movement of the controlled peripheral device towards the desired contact target outside the preset area based on the multiple third decoding results to obtain multiple third motion results; and performing usability verification on the target decoder based on the multiple third motion results and preset verification rules to obtain verification results.

[0173] Optionally, the processor in the aforementioned electronic device may be configured to run an executable program to perform the following steps: controlling the movement of the target to be contacted according to a preset motion trajectory and a preset motion speed.

[0174] Optionally, the processor in the aforementioned electronic device can be configured to run an executable program to perform the following steps: updating the preset motion trajectory and preset motion speed according to preset training requirements.

[0175] Optionally, the processor in the aforementioned electronic device may be configured to run an executable program to perform the following steps: in response to a first motion result indicating that the controlled peripheral device failed to intercept the desired contact target, determining the amount of deviation between the controlled peripheral device and the desired contact target; and adjusting the decoding parameters according to the amount of deviation to obtain a target decoder.

[0176] Optionally, the preset area is a circular area.

[0177] Optionally, the processor in the aforementioned electronic device may be configured to run an executable program to perform the following steps: obtaining region adjustment parameters; and adjusting the size and shape of a preset region according to the region adjustment parameters.

[0178] Embodiments of this application also provide a computer-readable storage medium including a stored executable program, wherein, when the executable program is running, it controls the device where the computer-readable storage medium is located to perform the methods of various embodiments of the present invention.

[0179] Optionally, in this embodiment, the executable program can be configured to store an executable program for performing the following steps:

[0180] Step S101: Obtain the target neural signals of the subject.

[0181] Step S102: The neural signal is decoded using a pre-trained decoder to obtain the decoding result.

[0182] Step S103: Based on the decoding result, control the controlled peripheral to continuously move towards the desired contact target outside the preset area until the controlled peripheral makes desired contact with the desired contact target. Here, the controlled peripheral is the desired control target of the subject. After the controlled peripheral contacts the boundary of the preset area, it moves towards the desired contact target according to the target motion attributes. The target motion attributes are the speed and direction of the controlled peripheral when it contacts the boundary of the area.

[0183] Optionally, the executable program can be configured to store an executable program for performing the following steps: determining a first control instruction based on the decoding result, wherein the first control instruction is used to control the movement of the controlled peripheral device toward the boundary of a preset area; generating a second control instruction in response to the controlled peripheral device contacting the boundary of the preset area, wherein the second control instruction includes target motion attributes; and controlling the controlled peripheral device to make desired contact with the desired contact target according to the second control instruction.

[0184] Optionally, the executable program can be configured to store an executable program for performing the following steps: based on the decoding result, controlling the controlled peripheral to continuously move towards the desired contact target outside the preset area, obtaining the positional difference between the controlled peripheral and the desired contact target when the controlled peripheral contacts the boundary of the preset area; judging the target motion attribute based on the positional difference, and obtaining the judgment result; in response to the judgment result indicating that the controlled peripheral cannot contact the desired contact target based on the target motion attribute, determining the target motion attribute compensation value based on the positional difference; and compensating the target motion attribute based on the target motion attribute compensation value.

[0185] Optionally, the executable program can be configured to store an executable program for performing the following steps: based on the decoding result, controlling the controlled peripheral to continuously move towards the desired contact target outside the preset area and recording the stopping position of the controlled peripheral when it stops moving; in response to the stopping position indicating that the controlled peripheral and the desired contact target have not made the desired contact, determining the area adjustment parameter based on the stopping position; and adjusting the preset area based on the area adjustment parameter.

[0186] Optionally, the executable program can be configured to store an executable program for performing the following steps: acquiring a first neural signal from the subject; inputting the first neural signal into a pre-corrected decoder to obtain a first decoding result; controlling the movement of a controlled peripheral device toward a desired contact target outside a preset area according to the first decoding result to obtain a first movement result, wherein the controlled peripheral device is the subject's desired control target, and after the controlled peripheral device contacts the boundary of the preset area, it moves toward the desired contact target at the speed and direction at which the controlled peripheral device contacts the boundary; and adjusting the decoding parameters in the pre-corrected decoder according to the first movement result to obtain a target decoder.

[0187] Optionally, the executable program can be configured to store an executable program for performing the following steps: within a first preset time period, controlling the movement of the controlled peripheral device toward the desired contact target multiple times to obtain multiple movement states of the controlled peripheral device and multiple second neural signals, wherein the second neural signals are generated based on the visual stimulus observed by the subject that the controlled peripheral device moves toward the desired contact target; training the decoder according to the multiple movement states of the controlled peripheral device and the multiple second neural signals to obtain an initialized decoder; and pre-correcting the initialized decoder to obtain a pre-corrected decoder.

[0188] Optionally, the executable program can be configured to store an executable program for performing the following steps: acquiring multiple third neural signals from the subject within a second preset time period; performing a pre-correction process for the multiple third neural signals to obtain a pre-corrected decoder, wherein the pre-correction process includes: inputting the multiple third neural signals into the decoder to obtain multiple second decoding results; determining multiple control signals based on the multiple second decoding results and a preset signal determination formula; controlling the movement of the controlled peripheral device toward the desired contact target based on the multiple control signals to obtain multiple second motion results; and correcting the decoding parameters in the decoder based on the multiple second motion results to obtain a pre-corrected decoder.

[0189] Optionally, the preset signal determination formula includes an auxiliary vector, auxiliary coefficients, a decoding vector, and decoding coefficients, wherein the auxiliary vector is a vector pointing to the controlled peripheral device, and the decoding vector is the output result of the decoder.

[0190] Optionally, the executable program can be configured to store an executable program for performing the following steps: acquiring the previous motion result, wherein the previous motion result is the motion result corresponding to the third neural signal acquired at the previous acquisition time; adjusting the auxiliary coefficient and decoding coefficient according to the previous motion result to obtain the adjusted preset signal determination formula; and determining the control signal according to the adjusted preset signal determination formula.

[0191] Optionally, the executable program can be configured to store an executable program for performing the following steps: determining the number of second motion results that meet the standard from a plurality of second motion results, wherein a second motion result that meets a first preset condition is determined to meet the standard; and obtaining a pre-corrected decoder in response to the number of second motion results meeting the second preset condition.

[0192] Optionally, the executable program can be configured to store an executable program for performing the following steps: acquiring multiple fourth neural signals of the subject within a third preset time period; inputting the multiple fourth neural signals into the target decoder to obtain multiple third decoding results; controlling the movement of the controlled peripheral device towards the desired contact target outside the preset area according to the multiple third decoding results to obtain multiple third movement results; and performing usability verification on the target decoder according to the multiple third movement results and preset verification rules to obtain verification results.

[0193] Optionally, the executable program can be configured to store an executable program for performing the following steps: controlling the movement of the target to be contacted according to a preset motion trajectory and a preset motion speed.

[0194] Optionally, the executable program can be configured to store an executable program for performing the following steps: updating the preset motion trajectory and preset motion speed according to preset training requirements.

[0195] Optionally, the executable program can be configured to store an executable program for performing the following steps: in response to a first motion result indicating that the controlled peripheral device failed to intercept the desired contact target, determining the amount of deviation between the controlled peripheral device and the desired contact target; and adjusting the decoding parameters according to the amount of deviation to obtain a target decoder.

[0196] Optionally, the preset area is a circular area.

[0197] Optionally, the executable program can be configured to store an executable program for performing the following steps: obtaining region adjustment parameters; and adjusting the size and shape of a preset region according to the region adjustment parameters.

[0198] Embodiments of this application also provide a computer program product, including a computer program that, when executed by a processor, implements the methods of various embodiments of the present invention.

[0199] Optionally, in this embodiment, the computer program, when executed by the processor, performs the following steps:

[0200] Step S101: Obtain the target neural signals of the subject.

[0201] Step S102: The neural signal is decoded using a pre-trained decoder to obtain the decoding result.

[0202] Step S103: Based on the decoding result, control the controlled peripheral to continuously move towards the desired contact target outside the preset area until the controlled peripheral makes desired contact with the desired contact target. Here, the controlled peripheral is the desired control target of the subject. After the controlled peripheral contacts the boundary of the preset area, it moves towards the desired contact target according to the target motion attributes. The target motion attributes are the speed and direction of the controlled peripheral when it contacts the boundary of the area.

[0203] Optionally, when the above computer program is executed by the processor, it performs the following steps: determining a first control instruction based on the decoding result, wherein the first control instruction is used to control the movement of the controlled peripheral device toward the boundary of a preset area; generating a second control instruction in response to the controlled peripheral device contacting the boundary of the preset area, wherein the second control instruction includes target motion attributes; and controlling the controlled peripheral device to make desired contact with the desired target based on the second control instruction.

[0204] Optionally, when the above computer program is executed by the processor, it performs the following steps: based on the decoding result, it controls the controlled peripheral device to continuously move towards the desired contact target outside the preset area, and obtains the positional difference between the controlled peripheral device and the desired contact target when the controlled peripheral device contacts the boundary of the preset area; based on the positional difference, it judges the target motion attribute and obtains the judgment result; in response to the judgment result indicating that the controlled peripheral device cannot contact the desired contact target based on the target motion attribute, it determines the target motion attribute compensation value based on the positional difference; and it compensates the target motion attribute based on the target motion attribute compensation value.

[0205] Optionally, when the above computer program is executed by the processor, it performs the following steps: based on the decoding result, it controls the controlled peripheral to continuously move towards the desired contact target outside the preset area and records the stopping position of the controlled peripheral when it stops moving; in response to the stopping position indicating that the controlled peripheral and the desired contact target have not made the desired contact, it determines the area adjustment parameter based on the stopping position; and adjusts the preset area based on the area adjustment parameter.

[0206] Optionally, when the above computer program is executed by the processor, it performs the following steps: acquiring the subject's first neural signal; inputting the first neural signal into the pre-corrected decoder to obtain a first decoding result; controlling the controlled peripheral device to move towards the desired contact target outside the preset area according to the first decoding result to obtain a first motion result, wherein the controlled peripheral device is the subject's desired control target, and after the controlled peripheral device contacts the boundary of the preset area, it moves towards the desired contact target at the speed and direction at which the controlled peripheral device contacts the boundary; and adjusting the decoding parameters in the pre-corrected decoder according to the first motion result to obtain a target decoder.

[0207] Optionally, when the above computer program is executed by the processor, it performs the following steps: within a first preset time period, it controls the movement of the controlled peripheral device toward the desired contact target multiple times, thereby obtaining multiple movement states of the controlled peripheral device and multiple second neural signals, wherein the second neural signals are generated based on the visual stimulus observed by the subject that the controlled peripheral device moves toward the desired contact target; based on the multiple movement states of the controlled peripheral device and the multiple second neural signals, it trains the decoder to obtain an initialized decoder; and pre-corrects the initialized decoder to obtain a pre-corrected decoder.

[0208] Optionally, when the above computer program is executed by the processor, it performs the following steps: within a second preset time period, acquiring multiple third neural signals from the subject; performing a pre-correction process for the multiple third neural signals to obtain a pre-corrected decoder, wherein the pre-correction process includes: inputting the multiple third neural signals into the decoder to obtain multiple second decoding results; determining multiple control signals based on the multiple second decoding results and a preset signal determination formula; controlling the movement of the controlled peripheral device toward the desired contact target based on the multiple control signals to obtain multiple second motion results; and correcting the decoding parameters in the decoder based on the multiple second motion results to obtain a pre-corrected decoder.

[0209] Optionally, the preset signal determination formula includes an auxiliary vector, auxiliary coefficients, a decoding vector, and decoding coefficients, wherein the auxiliary vector is a vector pointing to the controlled peripheral device, and the decoding vector is the output result of the decoder.

[0210] Optionally, when the above computer program is executed by the processor, it performs the following steps: acquiring the previous motion result, wherein the previous motion result is the motion result corresponding to the third neural signal acquired at the previous acquisition time; adjusting the auxiliary coefficient and the decoding coefficient according to the previous motion result to obtain the adjusted preset signal determination formula; and determining the control signal according to the adjusted preset signal determination formula.

[0211] Optionally, when the above computer program is executed by the processor, it performs the following steps: determining the number of second motion results that meet the standard from a plurality of second motion results, wherein a second motion result that meets a first preset condition is determined to meet the standard; in response to the number of second motion results meeting the second preset condition, obtaining a pre-corrected decoder.

[0212] Optionally, when the above computer program is executed by the processor, it performs the following steps: within a third preset time period, acquire multiple fourth neural signals of the subject; input the multiple fourth neural signals into the target decoder to obtain multiple third decoding results; based on the multiple third decoding results, control the movement of the controlled peripheral device toward the desired contact target outside the preset area to obtain multiple third movement results; based on the multiple third movement results and preset verification rules, perform usability verification on the target decoder to obtain verification results.

[0213] Optionally, when the above computer program is executed by the processor, it performs the following steps: controlling the movement of the target to be contacted according to the preset motion trajectory and preset motion speed.

[0214] Optionally, when the above computer program is executed by the processor, it performs the following steps: updating the preset motion trajectory and preset motion speed according to the preset training requirements.

[0215] Optionally, when the above computer program is executed by the processor, it performs the following steps: in response to the first motion result indicating that the controlled peripheral device failed to intercept the desired contact target, it determines the deviation between the controlled peripheral device and the desired contact target; based on the deviation, it adjusts the decoding parameters to obtain the target decoder.

[0216] Optionally, the preset area is a circular area.

[0217] Optionally, when the above computer program is executed by the processor, it performs the following steps: obtaining region adjustment parameters; and adjusting the size and shape of the preset region according to the region adjustment parameters.

[0218] Embodiments of this application also provide a computer program product, including a non-volatile computer-readable storage medium for storing a computer program that, when executed by a processor, implements the methods in various embodiments of the present invention.

[0219] Embodiments of this application also provide a computer program that, when executed by a processor, implements the methods described in the various embodiments of the present invention.

[0220] In the above embodiments of the present invention, the descriptions of each embodiment have different focuses. For parts not described in detail in a certain embodiment, please refer to the relevant descriptions of other embodiments.

[0221] In the several embodiments provided in this application, it should be understood that the disclosed technical content can be implemented in other ways. The device embodiments described above are merely illustrative; for example, the division of units can be a logical functional division, and in actual implementation, there may be other division methods. For instance, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the displayed or discussed mutual coupling, direct coupling, or communication connection may be through some interfaces; the indirect coupling or communication connection between units or modules may be electrical or other forms.

[0222] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0223] Furthermore, the functional units in the various embodiments of the present invention can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit.

[0224] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This 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.) to execute all or part of the steps of the methods described in the various embodiments of the present invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, read-only memory (ROM), random access memory (RAM), portable hard drives, magnetic disks, or optical disks.

[0225] The above description is only a preferred embodiment of the present invention. It should be noted that for those skilled in the art, several improvements and modifications can be made without departing from the principle of the present invention, and these improvements and modifications should also be considered within the scope of protection of the present invention.

Claims

1. A signal processing method, characterized in that, include: Acquire the target neural signals of the subjects; The target neural signal is decoded using a pre-trained decoder to obtain a decoding result. The pre-trained decoder is a model trained using neural signals and cursor motion data collected in the early training phase. The decoding result is used to represent the motion parameters of the controlled peripheral. Based on the decoding result, the controlled peripheral device is controlled to continuously move towards the desired contact target outside the preset area until the controlled peripheral device makes desired contact with the desired contact target. Here, the controlled peripheral device is the desired control target of the subject. After the controlled peripheral device contacts the boundary of the preset area, it moves towards the desired contact target according to the target motion attributes. The target motion attributes are the speed and direction of the controlled peripheral device when it contacts the boundary of the area. The pre-training of the decoder includes: acquiring a first neural signal from the subject; inputting the first neural signal into a pre-corrected decoder to obtain a first decoding result; controlling a controlled peripheral device to move towards a desired contact target outside a preset area based on the first decoding result to obtain a first motion result, wherein the controlled peripheral device is the subject's desired control target, and after the controlled peripheral device contacts the boundary of the preset area, it moves towards the desired contact target at the speed and direction at which it contacts the boundary; and adjusting the decoding parameters in the pre-corrected decoder based on the first motion result to obtain a target decoder.

2. The signal processing method according to claim 1, characterized in that, The step of controlling the controlled peripheral device to continuously move towards the desired contact target outside the preset area based on the decoding result, until the controlled peripheral device makes desired contact with the desired contact target, includes: Based on the decoding result, a first control command is determined, wherein the first control command is used to control the controlled peripheral device to move toward the boundary of a preset area; In response to the controlled peripheral device contacting the boundary of the preset area, a second control command is generated, wherein the second control command includes the target motion attribute; According to the second control command, the controlled peripheral device is controlled to make desired contact with the desired contact target.

3. The signal processing method according to claim 1, characterized in that, The step of controlling the controlled peripheral device to continuously move towards the desired contact target outside the preset area based on the decoding result, until the controlled peripheral device makes desired contact with the desired contact target, includes: Based on the decoding result, the controlled peripheral device is controlled to continuously move towards the desired contact target outside the preset area, and the positional difference between the controlled peripheral device and the desired contact target is obtained when the controlled peripheral device contacts the area boundary of the preset area. Based on the positional differences, the motion attributes of the target are determined, and a determination result is obtained; In response to the judgment result indicating that the controlled peripheral device cannot reach the desired target based on the target motion attribute, a target motion attribute compensation value is determined based on the position difference; The target motion attribute is compensated based on the target motion attribute compensation value.

4. The signal processing method according to claim 1, characterized in that, Also includes: Based on the decoding result, the controlled peripheral device is controlled to continuously move towards the desired contact target outside the preset area, and the stopping position of the controlled peripheral device when it stops moving is recorded. In response to the stop position indicating that the controlled peripheral device has not made the desired contact with the desired contact target, the area adjustment parameters are determined based on the stop position; The preset region is adjusted according to the region adjustment parameters.

5. The signal processing method according to claim 1, characterized in that, Before inputting the first neural signal into the pre-corrected decoder to obtain the first decoding result, the method further includes: Within a first preset time period, the controlled peripheral device is repeatedly controlled to move toward the desired contact target, resulting in multiple controlled peripheral device movement states and multiple second neural signals. The second neural signals are generated based on the visual stimulus observed by the subject that the controlled peripheral device moves toward the desired contact target. Based on the motion states of the multiple controlled peripherals and the multiple second neural signals, the decoder is trained to obtain the initialized decoder; The initialized decoder is pre-corrected to obtain the pre-corrected decoder.

6. The signal processing method according to claim 5, characterized in that, The pre-correction of the decoder includes: During a second preset time period, multiple third neural signals of the subject are acquired; For multiple third neural signals, a pre-correction process is performed to obtain a pre-corrected decoder, wherein the pre-correction process includes: The plurality of third neural signals are input into the decoder to obtain a plurality of second decoding results; Based on the multiple second decoding results and the preset signal determination formula, multiple control signals are determined. The preset signal determination formula includes an auxiliary vector, an auxiliary coefficient, a decoding vector, and a decoding coefficient. The auxiliary vector is a fixed direction vector that always points to the intercepted target, and the decoding vector is the output result of the decoder. Based on the multiple control signals, the controlled peripheral device is controlled to move toward the desired contact target, resulting in multiple second motion results; Based on the multiple second motion results, the decoding parameters in the decoder are corrected to obtain the pre-corrected decoder.

7. The signal processing method according to claim 6, characterized in that, The step of determining the control signal based on the second decoding result and the preset signal determination formula includes: Obtain the previous motion result, wherein the previous motion result is the motion result corresponding to the third neural signal obtained at the previous acquisition time; Based on the previous motion result, the auxiliary coefficient and the decoding coefficient are adjusted to obtain the adjusted preset signal determination formula; The control signal is determined according to the formula based on the adjusted preset signal.

8. The signal processing method according to claim 6, characterized in that, Also includes: The number of second motion results that meet the standard is determined from multiple second motion results, wherein the second motion result that meets the first preset condition is determined to meet the standard; In response to the number of qualified items meeting the second preset condition, a pre-corrected decoder is obtained.

9. The signal processing method according to claim 1, characterized in that, Also includes: During a third preset time period, multiple fourth neural signals of the subject are acquired; The multiple fourth neural signals are input into the target decoder to obtain multiple third decoding results; Based on the multiple third decoding results, the controlled peripheral device is controlled to move towards the desired contact target outside the preset area, thereby obtaining multiple third motion results; Based on the multiple third motion results and preset verification rules, the usability of the target decoder is verified to obtain the verification result.

10. The signal processing method according to claim 1, characterized in that, The desired contact target is a dynamic target. Before controlling the controlled peripheral device to move towards the desired contact target outside the preset area based on the first decoding result, and obtaining the first motion result, the method further includes: The desired contact target is controlled to move according to a preset motion trajectory and a preset motion speed.

11. The signal processing method according to claim 10, characterized in that, Also includes: Update the preset movement trajectory and preset movement speed according to the preset training requirements.

12. The signal processing method according to claim 1, characterized in that, The step of adjusting the decoding parameters in the pre-corrected decoder based on the first motion result to obtain the target decoder includes: In response to a first motion result indicating that the controlled peripheral device failed to intercept the desired contact target, the deviation between the controlled peripheral device and the desired contact target is determined. The decoding parameters are adjusted based on the deviation to obtain the target decoder.

13. The signal processing method according to claim 1, characterized in that, The preset area is a circular area.

14. The signal processing method according to claim 1, characterized in that, Also includes: Obtain the regional adjustment parameters; Adjust the size and shape of the preset region according to the region adjustment parameters.

15. A brain-computer interface system, characterized in that, include: A neural signal acquisition device is used to acquire neural signals from the subject. A decoder is used to decode the neural signals to obtain a decoding result, wherein the decoder is a model trained by the neural signals and cursor motion data collected in the early training phase, and the decoding result is used to represent the motion parameters of the controlled peripheral. A feedforward controller is used to control the controlled peripheral device to continuously move towards a desired contact target outside a preset area based on the decoding result, until the controlled peripheral device makes desired contact with the desired contact target. The controlled peripheral device is the desired control target of the subject. After the controlled peripheral device contacts the boundary of the preset area, it moves towards the desired contact target based on the target motion attributes, which are the speed and direction of the controlled peripheral device when it contacts the boundary of the area. The pre-training of the decoder includes: acquiring a first neural signal from the subject; inputting the first neural signal into a pre-corrected decoder to obtain a first decoding result; controlling a controlled peripheral device to move towards a desired contact target outside a preset area based on the first decoding result to obtain a first motion result, wherein the controlled peripheral device is the subject's desired control target, and after the controlled peripheral device contacts the boundary of the preset area, it moves towards the desired contact target at the speed and direction at which it contacts the boundary; and adjusting the decoding parameters in the pre-corrected decoder based on the first motion result to obtain a target decoder.

16. An electronic device, characterized in that, include: Memory, which stores executable programs; A processor for running the program, wherein the program, when running, performs the method according to any one of claims 1 to 14.

17. A computer-readable storage medium, characterized in that, The computer-readable storage medium includes a stored executable program, wherein, when the executable program is executed, it controls the device on which the storage medium is located to perform the method according to any one of claims 1 to 14.

18. A computer program product, characterized in that, Includes a computer program that, when executed by a processor, implements the method according to any one of claims 1 to 14.

Images