Exercise intention recognition method and system based on brain waves and exercise auxiliary equipment

Abstract

The invention is suitable for the technical field of brain wave data processing, and provides a brain wave-based motion intention recognition method and system and a motion assisting device.The method comprises the steps of brain wave collection, brain electric power extraction, time sliding window numerical value change amplitude change analysis, motion intention level judgment and a voting decision mechanism; the brain wave features related to the motion intention of the user can be efficiently grabbed, and then accurate recognition of the motion intention is achieved. The method has the potential of providing more reliable and more efficient control signals for motion auxiliary equipment or rehabilitation machines based on the brain-computer interface, the user can recover related motion functions more effectively, and the life quality of the user is improved. Meanwhile, a new thought and a new method are opened up for research in the fields of brain wave processing and motion intention recognition.

Description

Technical Field

[0001] This invention belongs to the field of brainwave data processing technology, and particularly relates to a method, system and motion assist device for recognizing motor intentions based on brainwaves. Background Technology

[0002] Brain-computer interface (BCI) technology belongs to the cutting-edge field of interdisciplinary integration of neuroscience and engineering. In recent years, it has shown great potential and value in assisting neurorehabilitation, especially in the rehabilitation of neurological functions caused by stroke and motor nerve injury. This technology, by analyzing motor-related electroencephalogram (EEG) data in the brain, has opened up a new rehabilitation pathway for patients with motor dysfunction by directly connecting the brain to rehabilitation training equipment. However, BCI technology still faces a series of pressing technical challenges in its further promotion and practical application.

[0003] In order to obtain brain activity information, the EEG acquisition module in current brain-computer interface systems often uses eight-channel, sixteen-channel, or even more acquisition methods. While this method has its advantages, it increases system complexity and cost, and generates a large amount of redundant EEG data that is irrelevant to motor intention analysis, making subsequent data processing and analysis difficult.

[0004] In brain-computer interfaces (BCIs), EEG-motor intention analysis algorithms are crucial for converting brain signals into actions performed by external devices. The complexity of these algorithms stems primarily from the inherent complexity of brain activity, resulting in a non-simplistic correspondence between brain waves and motor intentions. Furthermore, due to individual differences, existing EEG-motor intention recognition algorithms have limited generalization capabilities, making it difficult to meet the diverse needs of users in neurorehabilitation scenarios.

[0005] In brain-computer interface-assisted rehabilitation training, especially in the field of hand function rehabilitation, real-time feedback is of great significance to user experience and rehabilitation outcomes. However, existing EEG data-motor intention recognition systems, which employ data-driven (e.g., deep learning) technologies, often suffer from delays in data processing and algorithm execution, making it difficult to achieve real-time intention recognition and feedback. This impacts rehabilitation outcomes as well as user satisfaction and acceptance.

[0006] Based on this, this application proposes a method, system, and motion assist device for recognizing motor intentions based on electroencephalograms (EEGs). Summary of the Invention

[0007] The purpose of this invention is to provide a method, system, and motion assistive device for recognizing motor intentions based on electroencephalograms, in order to solve the technical problems existing in the prior art mentioned in the background.

[0008] The present invention is implemented as follows: a method for recognizing motor intentions based on electroencephalograms, the method comprising:

[0009] Acquire preprocessed brainwave data from multiple brainwave acquisition channels, wherein the preprocessing method includes at least signal filtering, noise reduction, and amplification;

[0010] Extract the power data of the EEG data from each EEG acquisition channel within a preset characteristic frequency range to obtain the EEG power and value of several characteristic frequency ranges most relevant to the user's movement intention;

[0011] The rate of change of brainwave power and value within the sliding window period for each of the characteristic frequency ranges is calculated by means of a sliding time window, and the rate of change of power is used to characterize the degree of oscillation of brainwave power.

[0012] Based on the preset rules for the correspondence between power change rate and motor intention level, the EEG power and value of the characteristic frequency range of each channel are classified into levels, and the user's final motor intention level is determined and output through the majority voting principle.

[0013] Another objective of this invention is to provide a brainwave-based motion intention recognition system, comprising a brainwave data acquisition terminal and a signal processing terminal, wherein:

[0014] The EEG data acquisition terminal is used to acquire EEG data from multiple preprocessed EEG acquisition channels. The preprocessing method includes at least signal filtering, noise reduction, and amplification.

[0015] The signal processing terminal includes:

[0016] The power extraction module is used to extract the power data of the brainwave data of each brainwave acquisition channel within a preset characteristic frequency range, so as to obtain the brainwave power and value of several characteristic frequency ranges that are most relevant to the user's movement intention.

[0017] The power change rate calculation module calculates the power change rate of the brainwave power and value in each characteristic frequency range within the sliding window period through a sliding time window. The power change rate is used to characterize the degree of oscillation of brainwave power.

[0018] The motion intention recognition module is used to classify the EEG power and value of each channel based on the preset correspondence rules between the power change rate and the motion intention level, and to determine and output the user's final motion intention level through the majority voting principle.

[0019] Another objective of this invention is to provide a motion assist device, including an end effector and the aforementioned motion intention recognition system, wherein the end effector and the motion intention recognition system communicate via wired or wireless means.

[0020] Another objective of this invention is to provide a computer device, including a memory and a processor, wherein the memory stores a computer program, and when the computer program is executed by the processor, the processor performs the steps of the brainwave-based motion intention recognition method.

[0021] Another objective of this invention is to provide a computer-readable storage medium storing a computer program, which, when executed by a processor, causes the processor to perform the steps of the brainwave-based motion intention recognition method.

[0022] The beneficial effects of this invention are as follows: The method described in the embodiments of this invention, through EEG acquisition, extraction of EEG power, analysis of changes in the amplitude of time sliding window values, determination of motor intention level, and a voting decision mechanism, can efficiently capture EEG data features related to the user's motor intention, thereby achieving accurate recognition of motor intention. It has the potential to provide a more reliable and efficient control scheme for brain-computer interface-based motion assistive devices or rehabilitation machines, helping users to directly communicate with rehabilitation equipment through brain-computer interface devices to assist in completing expected movements, thereby effectively restoring motor nerve function. At the same time, it also opens up new ideas and methods for research in the fields of EEG data processing and motor intention recognition. Attached Figure Description

[0023] Figure 1 This is an application environment diagram of a motion intention recognition method based on electroencephalograms provided in an embodiment of the present invention;

[0024] Figure 2 A flowchart of a motion intention recognition method based on electroencephalograms provided in an embodiment of the present invention;

[0025] Figure 3 This is a flowchart illustrating the process of acquiring preprocessed electroencephalogram (EEG) data from multiple EEG acquisition channels, as provided in an embodiment of the present invention.

[0026] Figure 4 A flowchart for obtaining EEG power values ​​of several characteristic frequency ranges most relevant to the user's movement intention, provided as an embodiment of the present invention;

[0027] Figure 5 A flowchart for calculating the rate of change of power of EEG power and value in each characteristic frequency range within the sliding window period, provided by an embodiment of the present invention;

[0028] Figure 6 A flowchart for determining and outputting the user's final motion intention level based on the majority voting principle, as provided in this embodiment of the invention;

[0029] Figure 7 A structural block diagram of a motion intention recognition system based on electroencephalograms provided in an embodiment of the present invention;

[0030] Figure 8 This is a structural block diagram of the EEG data acquisition terminal provided in an embodiment of the present invention;

[0031] Figure 9 This is a structural block diagram of the power extraction module provided in an embodiment of the present invention;

[0032] Figure 10 This is a structural block diagram of the power change rate calculation module provided in an embodiment of the present invention;

[0033] Figure 11 This is a structural block diagram of the motion intent recognition module provided in an embodiment of the present invention;

[0034] Figure 12 This is a block diagram of the internal structure of a computer device in one embodiment. Detailed Implementation

[0035] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention.

[0036] It is understood that the terms "first," "second," etc., used in this application may be used herein to describe various elements, but unless otherwise specified, these elements are not limited by these terms. These terms are used only to distinguish one element from another. For example, without departing from the scope of this application, a first script may be referred to as a second script, and similarly, a second script may be referred to as a first script.

[0037] Figure 1 This is an application environment diagram of a brainwave-based motion intention recognition method provided in an embodiment of the present invention, such as... Figure 1 As shown, this application environment includes an EEG data acquisition terminal 100 and a signal processing terminal 200, etc.

[0038] In this embodiment of the invention, the EEG data acquisition terminal 100 is used to acquire raw EEG data and perform simple preprocessing, etc. The signal processing terminal 200 is used to perform in-depth processing of the EEG to identify the user's movement intentions, etc.

[0039] In this embodiment of the invention, the EEG data acquisition terminal 100 and the signal processing terminal 200 can communicate via a network or a USB interface, and the invention does not impose any limitations on this.

[0040] like Figure 2 As shown, in one embodiment, a method for recognizing motor intentions based on brainwaves is proposed, which may specifically include the following steps:

[0041] Step S100: Obtain brainwave data from multiple preprocessed brainwave acquisition channels. The preprocessing method includes at least signal filtering, noise reduction, and amplification.

[0042] In this embodiment of the invention, since the acquired raw brain waves are quite complex and cover multiple frequency components, the raw brain waves are processed by signal filtering, noise reduction and amplification to ensure that interference signals are filtered out for the raw brain wave data, thereby providing a high-quality, stable and reliable signal foundation for subsequent processing.

[0043] As one embodiment of the present invention, the signal filtering, denoising and amplification processes can be set to at least once, depending on the actual situation. The present invention does not impose specific limitations here.

[0044] Step S200: Extract the power data of the EEG data of each EEG acquisition channel within a preset characteristic frequency range to obtain the EEG power and value of several characteristic frequency ranges most relevant to the user's movement intention.

[0045] In this embodiment of the invention, although the EEG data undergoes preprocessing, it remains relatively complex, containing multiple frequency components. To highlight key features related to motor intention, this embodiment retains power data within the characteristic frequency range highly correlated with motor intention, facilitating subsequent motor intention recognition.

[0046] Step S300: Calculate the rate of change of brainwave power and value within the sliding window period for each of the characteristic frequency ranges. This rate of change characterizes the degree of oscillation in brainwave power. The magnitude of the rate of change directly reflects the activity level of brainwaves within a specific time range. A higher rate of change indicates more intense fluctuations in brainwave activity, and vice versa, providing a quantitative basis for subsequent motion intention analysis based on brainwave data.

[0047] Because brainwaves are constantly changing, this invention employs a sliding time window technique for dynamic signals to effectively capture these changes. By analyzing the amplitude changes in the power and values ​​of characteristic brainwave frequencies in real time, the activity level of brainwaves within a range of motion-related characteristic frequencies can be effectively captured, reflecting the strength of motor intention changes in relevant brain regions, thus achieving the goal of motor intention recognition through brainwave data analysis.

[0048] Step S400: Based on the preset correspondence rules between power change rate and motor intention level, the EEG power and value of the characteristic frequency range of each channel are classified into levels. The user's final motor intention level is determined and output through majority voting. In fact, different intensity levels correspond to different degrees of motor intention. The motor intention M resolved from each characteristic frequency range of each EEG channel can be defined as a vote. Therefore, N EEG channels generate M×N votes.

[0049] In practical applications, after extracting the power change rate of EEG waves across multiple characteristic frequency ranges, this invention employs intensity level classification and a multi-channel, multi-frequency voting mechanism for intent recognition. This achieves accurate recognition of motor intentions, further enhancing the accuracy and robustness of intent recognition.

[0050] The method described in this invention, through EEG acquisition, extraction of EEG power, analysis of changes in the amplitude of time-sliding window values, determination of motor intention level, and a voting decision mechanism, can efficiently capture EEG data features related to the user's motor intention, thereby achieving accurate recognition of motor intention. It has the potential to provide a more reliable and efficient control scheme for brain-computer interface-based motion assistive devices or rehabilitation machines, helping users to directly communicate with rehabilitation equipment through brain-computer interface devices to assist in completing expected movements, thereby effectively restoring motor nerve function. Simultaneously, it also opens up new ideas and methods for research in the fields of EEG data processing and motor intention recognition.

[0051] In one embodiment, such as Figure 3 As shown, step S100 may specifically include the following steps:

[0052] Step S101: Obtain raw brainwave data by using electrodes set at predetermined locations and a multi-channel EEG acquisition device.

[0053] In this embodiment of the invention, the multi-channel EEG acquisition device has four acquisition channels, and the electrodes are arranged reasonably according to the user's actual needs.

[0054] In this embodiment of the invention, the electrodes include a data acquisition electrode, a grounding electrode, and a reference electrode. The data acquisition electrode is arranged in regions C3 and C4 of the EEG International Standard 10-20 system, the grounding electrode is arranged in region C2 of the EEG International Standard 10-20 system, and the reference electrode is arranged in region A1 or A2 of the EEG International Standard 10-20 system. More preferably, one data acquisition electrode is arranged in region C3 or C4 on the healthy side of the user's brain, and three data acquisition electrodes are arranged on the affected side of the user's brain (region C3 or C4), with the three data acquisition electrodes arranged around the center point of region C3 or C4.

[0055] In one embodiment of the present invention, taking hand rehabilitation for stroke patients as an example, these patients typically exhibit unilateral motor dysfunction symptoms. Since brain and limb control are contralateral—that is, the left brain controls the right side of the limbs, and the right brain controls the left side of the limbs (for example, if the right hand has motor dysfunction, the affected side is on the right, and the healthy side is on the left)—a data acquisition electrode is placed in the C3 or C4 region of the brain hemisphere where the user controls the healthy hand. Three data acquisition electrodes are placed symmetrically in the brain hemisphere where the user controls the affected hand, and the three electrodes are arranged around the center point of the C3 or C4 region. z Place a ground electrode, and place a reference electrode on A1 or A2 (corresponding to the left or right earlobe). This electrode arrangement allows for the collection of bilateral brainwave data closely related to motor intention, laying the foundation for subsequent motor intention analysis.

[0056] For the four-channel EEG acquisition device, the embodiments of the present invention take into account the basic principles of EEG acquisition and practical application requirements, especially for the specific scenario of hand function motor intention recognition.

[0057] Neuroscience research indicates that specific areas of the brain are closely related to hand control and motor intentions, such as the precentral gyrus (the main area of ​​the motor cortex), the parietal lobe (involved in sensory processing and spatial localization), and parts of the frontal lobe (involved in higher motor planning and decision-making). Since electroencephalograms (EEGs) are electrical signals generated by the activity of neurons in the brain, these signals can be non-invasively recorded using electrodes on the scalp. Traditional EEG acquisition systems typically include multiple channels (such as 8, 16, or even more) to comprehensively cover different areas of the brain and obtain rich information on neural activity. However, not all channel information is necessary.

[0058] When using multi-channel EEG acquisition equipment, increasing the number of channels may introduce more redundant EEG data because EEG waves propagate through the skull and scalp, but the intensity of the electrical signal attenuates with the distance traveled. When the electrodes are not located in the central region where the signal originates, the obtained signal is attenuated, resulting in a lower signal-to-noise ratio. Therefore, acquiring more EEG data does not necessarily provide more help in analyzing motor intentions.

[0059] In this embodiment of the invention, the number of channels is limited to four, and the electrode positions C3, C4, and C5 are aligned with the 10-20 international standard EEG acquisition electrode locations. Z Collaborative approaches can improve signal quality by moving closer to the signal-generating area. Furthermore, this reduces redundant EEG data while ensuring signal availability, which is particularly important for achieving real-time intent recognition.

[0060] In addition, reducing the number of channels can simplify the hardware configuration of the system, making it lighter and easier to wear, which is an important advantage for users in long-term monitoring or daily application scenarios.

[0061] Step S102: Filter and denoise the raw EEG data to retain the EEG in the frequency range related to the motor intention.

[0062] Step S103: Amplify the brainwaves to increase the signal amplitude.

[0063] In one embodiment of the present invention, the EEG data acquisition terminal 100 (including electrodes and a four-channel EEG acquisition device) uses the Texas Instruments (TI) ADS1299 chip as its core component. This chip integrates four acquisition channels and internally possesses key components such as a high-performance 24-bit delta-sigma ADC, a programmable gain amplifier (PGA), an internal reference voltage source, and an onboard oscillator.

[0064] This chip features a unique filtering algorithm that effectively removes high-frequency noise above 100Hz and low-frequency drift below 0.1Hz. After noise reduction, the signal is amplified internally to enhance its amplitude.

[0065] In one embodiment, such as Figure 4 As shown, step S200 may specifically include the following steps:

[0066] Step S201: Perform a Fast Fourier Transform on the EEG data collected from each channel to convert the EEG data from the time domain to the frequency domain and analyze the power of EEG at different frequencies.

[0067] In this embodiment of the invention, a Fast Fourier Transform (FFT) can be used to convert brainwave data from the time domain to the frequency domain, allowing for the analysis of its frequency components. This operation leverages its advantages in speed and flexibility, significantly improving the recognition of motor intentions from brainwaves.

[0068] Step S202: Analyze the power distribution of EEG data at different frequencies to obtain the power and value of EEG in several characteristic frequency ranges that are most relevant to the user's movement intention.

[0069] In this embodiment of the invention, the power of brain waves at different frequencies is calculated, and the μ-wave band (frequency range of 8–13 Hz) and β-wave band (frequency range of 14–30 Hz) appearing in the sensorimotor region are selected as characteristic frequencies related to motor intention, and the above frequency range is set as the characteristic frequency range. Numerous experiments and data show that these characteristic frequencies are highly correlated with motor intention in brain waves related to hand function motor intention.

[0070] In one embodiment, such as Figure 5 As shown, step S300 may specifically include the following steps:

[0071] Step S301: Set the current time as the starting point of the calculation time window. The brainwave power and value data collected in the time period closest to the starting point of the time window are the brainwave power time series data (the time period closest to the starting point of the time window can be understood as, assuming the current time is t0, then the closest 5 seconds are t0-5, t0-4, t0-3, t0-2, and t0-1). Calculate its average value as the power benchmark value within the time window.

[0072] In this embodiment of the invention, the length of the time window is set to L, and the sum of EEG power and value data P of the time period closest to the current time t0 within the time window is extracted. t As historical data of EEG power time series, assuming the historical data within the time window is... The power reference value P_avg can then be expressed as:

[0073]

[0074] Step S302: Calculate the ratio of the instantaneous power at the current time point t to the power reference value as the power change rate.

[0075] In this embodiment of the invention, the power change rate characterizes the degree of deviation of the instantaneous power at the current time from the average value of the time window, reflecting the degree of change in signal power. Instantaneous power change rate R t The formula can be expressed as:

[0076] R t =Pt / P_avg

[0077] Where P t This represents the instantaneous power and value at the current time point t.

[0078] In one embodiment of the present invention, the length of the sliding time window is L = 4 seconds, and the sliding step size is t. step The sampling interval of the time window on the data is determined, and the sliding step is set to be less than the length of the time window, for example, t. step = 1 second. Of course, in practical applications, the length of the sliding time window can be set according to the actual needs of the usage scenario to ensure that key changes in clear and stable EEG waves can be captured.

[0079] The power of brainwaves varies over time across different frequency ranges. This invention analyzes the rate of change R of power and values ​​in motion-related frequency bands (e.g., the 8–13 Hz range). In this embodiment, the collected brainwave data falls within the frequency range of motion intention-related features, and the start frequency f of the motion intention-related features is defined. bg and cutoff frequency f cut Define the frequency f to be calculated. bg ~f cut The power value is the brainwave value collected at a certain point in time at f bg ~f cut The sum of power values ​​in the frequency range is P fbg～fcut ,Right now:

[0080]

[0081] The sum of power values ​​of brain waves in the characteristic frequency range related to movement is obtained as the sum of power values ​​of the characteristic frequency range at time t0. For example, one of the characteristic frequencies is the motor cortex μ oscillation frequency range of 8Hz to 13Hz.

[0082] Power and sum calculations are performed using a sliding time window technique, with the sliding window step size t set. step = 1 second. Calculate the brainwaves sequentially over time:

[0083] t-(L-1)×t step ...t-3×t step , t-2×t step , t-1×t step , t-0×t stepThe sum of power values ​​within the characteristic frequency range of 8–13 Hz and 15–25 Hz is collected. For example, if the frequency range is 8–13 Hz, the step size is 1 second, the current time is t0, and the sliding time window length is L = 4 seconds, the sum of power values ​​of brain waves within the frequency range of 8–13 Hz in the (t0–3) second period is 20 mW, the sum of power values ​​of brain waves within the frequency range of 8–13 Hz in the (t0–2) second period is 21 mW, the sum of power values ​​of brain waves within the frequency range of 8–13 Hz in the (t0–1) second period is 22 mW, and the sum of power values ​​of brain waves within the frequency range of 8–13 Hz in the t0 second period is 23 mW.

[0084] Calculate the amplitude R of the change in power and value of EEG waves related to motor intention at time t. t Taking the frequency range of 8–13 Hz as an example, the length of the sliding time window is L = 4 seconds. The sum of the power values ​​in the first 3 seconds (t0–3, t0–2, t0–1) is used as the time series data of the EEG sliding time window. The power reference value of the sliding time window in the frequency range of 8–13 Hz at time t0 is calculated, i.e., P. avg =(20+21+22) / 3=21mW as the power reference value. For example, the power sum P at the current time t0. t0 =23mW, the ratio R to the time window power reference value t0 R t0 =23 / 21=1.095. This ratio is used to characterize the power change of EEG in a specified frequency range, providing a key quantitative basis for subsequent motor intention recognition and related analysis.

[0085] In practical applications, the sliding time window technology of this invention can process dynamic signals in a timely manner, avoiding signal processing delays caused by complex algorithms, thereby ensuring that it can respond to motion intentions in real time. Compared with existing technologies, it has a significant advantage in real-time performance.

[0086] In one embodiment, such as Figure 6 As shown, step S400 may specifically include the following steps:

[0087] Step S401: Based on the correspondence between the preset power change rate value and the motion intention level classification threshold, the frequency range of each characteristic frequency of each channel is statistically analyzed, and the power change rate that meets the set threshold range is used as the motion intention vote.

[0088] In one embodiment of the present invention, the step size t of the sliding time window is [not specified]. step Both are set to 1 second, and the movement intention is analyzed through EEG data. In fact, the time window length L and step size t... step It can be set according to specific scenarios. In one embodiment of the present invention, the power change rate R is... tLess than threshold A, for example, setting threshold A = 0.3, it is determined as a weak-level motion intention; setting threshold B for medium-level motion intention, for example, setting threshold B = 1.3, when A < R t < B, for example, when 0.3 < R t< < 1.3, it is determined as a medium-level motion intention; setting threshold C for strong-level motion intention, for example, setting threshold C = 1.3, R t > threshold C, for example, R t > 1.3, it is determined as a strong-level motion intention. For example, when R t0 = 1.095, it is determined that the moment t0 belongs to the medium-level motion intention.

[0089] Step S402, determine the user's final motion intention level through the majority voting principle and output it.

[0090] In the embodiments of the present invention, multiple characteristic frequencies of each channel participate in the voting for motion intention determination. The motion intention levels of each characteristic frequency in each acquisition channel are counted, and the majority voting principle is used to determine the final motion intention level. The majority voting principle is to select the level that appears most frequently in all votes as the final motion intention determination result. In the embodiments of the present invention, N EEG channels generate M×N votes. When comprehensively considering the voting results of different channels and characteristic frequencies, weighted voting or other decision-making mechanisms can be introduced for determination to improve the robustness and generalization ability of the intention recognition algorithm. For example, in combination with the feedback results of different users, the weights of characteristic frequencies matching different users can also be adjusted, or other adaptive algorithms can be used for decision fusion.

[0091] In the practical application of the embodiments of the present invention, taking its application in related motion assistance devices as an example, the user can, according to the prompt information sent by the motion assistance device and in coordination with the画面displayed by the motion assistance device, in the brain-computer interface rehabilitation assistance device, achieve corresponding expected rehabilitation training motions according to the strength of the motion intention, such as guiding hand rehabilitation training actions. In this process, the user actively mobilizes their own motor nervous system to prompt the hand to start performing the training. At the same time, the brain wave data acquisition end 100 will continuously obtain the user's brain wave data. The signal processing end 200 will then process these brain wave data and start to analyze the motion intention and its intensity of the user's hand. Once the analysis result shows that the motion intention intensity reaches different preset strong, medium, and weak levels, the signal processing end 200 will immediately send control instructions corresponding to different intensities to the execution mechanism in the motion assistance device, and the execution mechanism will provide necessary assistance to help the user successfully complete the expected rehabilitation training actions.

[0092] Such as Figure 7As shown, in one embodiment, a brainwave-based motion intention recognition system is provided, which may specifically include a brainwave data acquisition terminal 100 and a signal processing terminal 200, wherein:

[0093] The EEG data acquisition terminal 100 is used to acquire EEG data obtained from multiple preprocessed EEG acquisition channels. The preprocessing method includes at least signal filtering, noise reduction and amplification.

[0094] The signal processing terminal 200 includes a power extraction module 201, a power change rate calculation module 202, and a motion intention recognition module 203, wherein:

[0095] The power extraction module 201 is used to extract the power data of the brainwave data of each brainwave acquisition channel within a preset characteristic frequency range, so as to obtain the brainwave power and value of several characteristic frequency ranges that are most relevant to the user's movement intention.

[0096] The power change rate calculation module 202 is used to calculate the power change rate of the brainwave power and value in each characteristic frequency range within the sliding window period through a sliding time window. The power change rate is used to characterize the degree of oscillation of brainwave power.

[0097] The motion intention recognition module 203 is used to classify the EEG power and value of each channel's characteristic frequency range based on a preset rule corresponding to the power change rate and motion intention level, and to determine and output the user's final motion intention level through a majority voting principle.

[0098] In this embodiment of the invention, the EEG data acquisition terminal 100 uses the Texas Instruments (TI) ADS1299 chip as its core component. This chip integrates four acquisition channels and internally possesses key components such as a high-performance 24-bit delta-sigma ADC, a programmable gain amplifier (PGA), an internal reference voltage source, and an onboard oscillator.

[0099] The EEG data acquisition terminal 100 also includes a power management module and a data transmission interface, consisting of a rechargeable battery and an external power interface, and equipped with a power conversion circuit and existing low-power design. The battery can provide a stable DC power supply, and when an external power supply is connected, it can be switched through the power conversion circuit to ensure stable power supply to the EEG data acquisition terminal 100 in different environments, while reducing power consumption to extend battery life.

[0100] The data transmission interface can use a USB interface to transmit data with external devices, ensuring high-speed and stable data transmission efficiency. Of course, while ensuring high-speed and stable data transmission efficiency, data transmission can also be achieved wirelessly, such as through 5G / Wi-Fi, etc., but this embodiment of the invention does not specifically limit the specific methods used.

[0101] In one embodiment of the present invention, the signal processing terminal 200 employs a high-performance industrial control computer as the core device for data processing. The industrial control computer's processor can be an Intel Core i7 series, which possesses powerful computing speed and excellent multi-threaded processing capabilities, enabling efficient handling of complex data processing tasks. The industrial control computer's memory configuration can be 16GB / DDR4, which can meet the requirements for caching large amounts of data and fast read / write operations during EEG processing. Its hard drive can be a 1TB solid-state drive (SSD), which provides extremely fast data storage and read speeds, thereby ensuring the efficiency and stability of the entire data processing process and meeting the stringent requirements for data storage and retrieval during EEG acquisition, processing, and analysis.

[0102] like Figure 8 As shown, in one embodiment, the EEG data acquisition terminal 100 includes an acquisition unit 101, a signal processing unit 102, and a signal amplification unit 103, wherein:

[0103] The acquisition unit 101 is used to acquire raw brainwave data through electrodes set at predetermined positions and a multi-channel brainwave acquisition device.

[0104] The signal processing unit 102 is used to filter and denoise the raw EEG data in order to retain the EEG in the frequency range related to the motor intention.

[0105] The signal amplification unit 103 is used to amplify the brain waves to enhance the signal amplitude.

[0106] like Figure 9 As shown, in one embodiment, the power extraction module 201 includes a frequency domain conversion unit 2011 and a power analysis unit 2012, wherein:

[0107] The frequency domain conversion unit 2011 is used to perform a fast Fourier transform on the EEG data acquired from each channel, converting the EEG data from the time domain to the frequency domain.

[0108] The power analysis unit 2012 is used to analyze the power distribution of brain waves at different frequencies to obtain the power and value of brain waves in several characteristic frequency ranges that are most relevant to the user's movement intention.

[0109] like Figure 10 As shown, in one embodiment, the power change rate calculation module 202 includes a reference value calculation unit 2021 and a ratio calculation unit 2022, wherein:

[0110] The reference value calculation unit 2021 is used to set the current time as the starting point of the calculation time window, and the brain wave power and value data collected in the time period closest to the starting point of the time window are brain wave power time series data, and the average value is calculated as the power reference value within the time window.

[0111] The ratio calculation unit 2022 is used to calculate the ratio of the instantaneous power sum at the current time point to the power reference value as the power change rate.

[0112] like Figure 11 As shown, in one embodiment, the motion intent recognition module 203 includes an intent voting unit 2031 and a result output unit 2032, wherein:

[0113] The intention voting unit 2031 is used to calculate each characteristic frequency range of each channel according to the correspondence between the preset power change rate value and the motion intention level classification threshold, so as to use the power change rate that meets the set threshold range as the motion intention vote.

[0114] The result output unit 2032 is used to determine and output the user's final motion intention level through the majority voting principle.

[0115] In one embodiment, the electrodes include a acquisition electrode, a ground electrode, and a reference electrode. The acquisition electrode is arranged in regions C3 and C4 of the EEG International Standard 10-20 system, the ground electrode is arranged in region C2 of the EEG International Standard 10-20 system, and the reference electrode is arranged in region A1 or A2 (left or right earlobe) of the EEG International Standard 10-20 system. More preferably, one acquisition electrode is arranged in region C3 or C4 of the brain hemisphere where the user controls the healthy hand, and three acquisition electrodes are arranged symmetrically in the brain hemisphere where the user controls the affected hand, with the three acquisition electrodes arranged around the center point of region C3 or C4.

[0116] In one embodiment, the multi-channel EEG acquisition device has four acquisition channels.

[0117] In one embodiment, a motion assistive device is provided, including an end effector and the motion intention recognition system, wherein the end effector communicates with the motion intention recognition system via a wired or wireless means.

[0118] In this embodiment of the invention, the end effector primarily assists the user in executing commands issued by the motion intention recognition system. Taking hand rehabilitation as an example, the essence of the end effector is to assist the user's hand in performing corresponding actions, such as finger bending and grasping. End effectors are widely disclosed in the prior art, such as hand exoskeletons and mechanical gloves, and this embodiment of the invention does not specifically limit them. Once the motion intention recognition system identifies the user's hand movement intention, it issues corresponding commands to the end effector to assist the user in completing the corresponding action.

[0119] In one embodiment, the motion assistive device further includes a human-computer interaction interface and peripherals.

[0120] In one embodiment of the present invention, the human-computer interaction interface may include a touch screen, a voice interaction module, etc., and peripherals may include a keyboard, a mouse, etc. The touch screen intuitively displays the user's rehabilitation training progress and effect in a graphical way. In practical application, the touch screen will display clear instructions and animations to guide the patient to perform specific rehabilitation exercise training tasks, and will also display the patient's brainwave characteristics and motor intention recognition results in real time.

[0121] For the voice interaction module, voice can guide patients and provide feedback. In practical applications, the voice prompts in this embodiment of the invention can provide clear and concise instructions, telling the user what to do next. When the user completes a motor imagery task, the voice prompts will provide real-time encouraging feedback, such as "Well done" or "Keep it up." This encouraging feedback can enhance the user's confidence and motivation, and improve the effectiveness of rehabilitation training.

[0122] Figure 12 An internal structural diagram of a computer device in one embodiment is shown. Specifically, this computer device may be... Figure 1 The signal processing terminal 200 in the middle. For example... Figure 12 As shown, the computer device includes a processor, memory, network interface, input device, and display screen connected via a system bus. The memory includes non-volatile storage media and internal memory. The non-volatile storage media stores an operating system and may also store computer programs. When executed by the processor, these programs enable the processor to implement a brainwave-based motion intention recognition method. The internal memory may also store computer programs, which, when executed by the processor, enable the processor to implement the brainwave-based motion intention recognition method. The display screen can be an LCD screen or an e-ink screen. The input device can be a touch layer covering the display screen, buttons, a trackball, or a touchpad mounted on the computer device's casing, or an external keyboard, touchpad, or mouse.

[0123] Those skilled in the art will understand that Figure 12 The structure shown is merely a block diagram of a portion of the structure related to the present application and does not constitute a limitation on the computer device to which the present application is applied. Specific computer devices may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements.

[0124] In one embodiment, the brainwave-based motion intention recognition system provided in this application can be implemented as a computer program, which can be implemented in the form of, for example... Figure 12 The system runs on the computer device shown. The computer device's memory can store the various program modules that make up this brainwave-based motion intention recognition system, for example... Figure 7 The diagram shows an EEG data acquisition terminal 100 and a signal processing terminal 200. The computer program, comprised of various program modules, causes the processor to execute the steps in the EEG-based motion intention recognition methods of the various embodiments of this application described in this specification. For example, Figure 12 The computer device shown can be used as follows Figure 7 The EEG data acquisition terminal 100 in the EEG-based motion intention recognition system shown executes step S100.

[0125] In one embodiment, a computer device is provided, the computer device including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to perform the following steps:

[0126] Step S100: Obtain brainwave data from multiple preprocessed brainwave acquisition channels. The preprocessing method includes at least signal filtering, noise reduction, and amplification.

[0127] Step S200: Extract the power data of the EEG data of each EEG acquisition channel within a preset characteristic frequency range to obtain the EEG power and value of several characteristic frequency ranges most relevant to the user's movement intention.

[0128] Step S300: Calculate the rate of change of power of brainwave power and value in each of the characteristic frequency ranges within the sliding window period using a sliding time window. The rate of change of power is used to characterize the degree of oscillation of brainwave power.

[0129] Step S400: Based on the preset correspondence rules between power change rate and movement intention level, the EEG power and value of the characteristic frequency range of each channel are classified into levels, and the user's final movement intention level is determined and output through the majority voting principle.

[0130] In one embodiment, a computer-readable storage medium is provided, on which a computer program is stored, which, when executed by a processor, causes the processor to perform the following steps:

[0131] Step S100: Obtain brainwave data from multiple preprocessed brainwave acquisition channels. The preprocessing method includes at least signal filtering, noise reduction, and amplification.

[0132] Step S200: Extract the power data of the EEG data of each EEG acquisition channel within a preset characteristic frequency range to obtain the EEG power and value of several characteristic frequency ranges most relevant to the user's movement intention.

[0133] Step S300: Calculate the rate of change of power of brainwave power and value in each of the characteristic frequency ranges within the sliding window period using a sliding time window. The rate of change of power is used to characterize the degree of oscillation of brainwave power.

[0134] Step S400: Based on the preset correspondence rules between power change rate and movement intention level, the EEG power and value of the characteristic frequency range of each channel are classified into levels, and the user's final movement intention level is determined and output through the majority voting principle.

[0135] It should be understood that although the steps in the flowcharts of the various embodiments of the present invention are shown sequentially according to the arrows, these steps are not necessarily executed in the order indicated by the arrows. Unless explicitly stated herein, there is no strict order restriction on the execution of these steps, and they can be executed in other orders. Moreover, at least some steps in the various embodiments may include multiple sub-steps or multiple stages. These sub-steps or stages are not necessarily completed at the same time, but can be executed at different times. The execution order of these sub-steps or stages is not necessarily sequential, but can be performed alternately or in turn with other steps or at least a portion of the sub-steps or stages of other steps.

[0136] Those skilled in the art will understand that all or part of the processes in the above embodiments can be implemented by a computer program instructing related hardware. The program can be stored in a non-volatile computer-readable storage medium, and when executed, it can include the processes of the embodiments described above. Any references to memory, storage, databases, or other media used in the embodiments provided in this application can include non-volatile and / or volatile memory. Non-volatile memory can include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memory can include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in various forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), dual data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link DRAM (SLDRAM), RAMbus direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and RAMbus dynamic RAM (RDRAM), etc.

[0137] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0138] The embodiments described above are merely illustrative of several implementations of the present invention, and while the descriptions are specific and detailed, they should not be construed as limiting the scope of the present invention. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of the present invention, and these modifications and improvements all fall within the scope of protection of the present invention. Therefore, the scope of protection of this patent should be determined by the appended claims.

[0139] The above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A method for recognizing a motion intention based on brain waves, characterized by, The method comprises: acquiring electroencephalogram data obtained from a plurality of preprocessed electroencephalogram acquisition channels, the preprocessing mode at least including signal filtering, denoising and amplification; extracting power data of the electroencephalogram data of each electroencephalogram acquisition channel in a preset characteristic frequency range to obtain a plurality of characteristic frequency range electroencephalogram power sum values most relevant to the user's motion intention; calculating the power change rate of each characteristic frequency range electroencephalogram power sum value in a sliding window period through a sliding time window, the power change rate being used to represent the oscillation degree of the electroencephalogram power; based on a preset corresponding rule of the power change rate and the motion intention level, grading the characteristic frequency range electroencephalogram power sum value of each channel, determining the final motion intention level of the user through the majority voting principle and outputting.

2. The brainwave-based motion-intention recognition method according to claim 1, characterized by, The step of acquiring electroencephalogram data obtained from a plurality of preprocessed electroencephalogram acquisition channels, the preprocessing mode at least including signal filtering, denoising and amplification, specifically comprises: acquiring raw electroencephalogram data through electrodes arranged at predetermined positions and a multi-channel electroencephalogram acquisition device; filtering and denoising the raw electroencephalogram data to retain electroencephalogram in a frequency range related to motion intention; amplifying the electroencephalogram to enhance signal amplitude. 3.The brainwave-based motion-intention recognition method according to claim 1, wherein The step of extracting power data of the electroencephalogram data of each electroencephalogram acquisition channel in a preset characteristic frequency range to obtain a plurality of characteristic frequency range electroencephalogram power sum values most relevant to the user's motion intention, specifically comprises: performing fast Fourier transform on the electroencephalogram data collected by each channel to convert the electroencephalogram data from time domain to frequency domain; analyzing the power distribution of the electroencephalogram data at different frequencies to obtain the power sum values of a plurality of characteristic frequency range electroencephalogram most relevant to the user's motion intention. 4.The brainwave-based motion-intention recognition method according to claim 1, wherein, The step of calculating the power change rate of each characteristic frequency range electroencephalogram power sum value in a sliding window period through a sliding time window, specifically comprises: setting the current time as the starting point of the calculation time window, and the electroencephalogram power sum value data collected in the time period closest to the starting point of the time window as the electroencephalogram power time series data, calculating the average value thereof as the power reference value in the time window; calculating the ratio of the instantaneous power sum value at the current time point to the power reference value as the power change rate. 5.The brainwave-based motion-intention recognition method according to claim 1, wherein, The step of determining the final motion intention level of the user through the majority voting principle and outputting, specifically comprises: according to the preset corresponding relationship between the power change rate value and the motion intention level division threshold, counting each characteristic frequency range of each channel to meet the power change rate in the set threshold range as the motion intention vote; determining the final motion intention level of the user through the majority voting principle and outputting. 6.The brainwave-based motion-intention recognition method according to claim 2, wherein The electrodes comprise acquisition electrodes, a grounding electrode and a reference electrode, the acquisition electrodes are arranged in the C3 and C4 regions of the international 10-20 system for electroencephalogram, the grounding electrode is arranged in the Cz region of the international 10-20 system for electroencephalogram, and the reference electrode is arranged in the A1 or A2 region of the international 10-20 system for electroencephalogram. 7.The brainwave-based motion-intention recognition method according to claim 3, wherein A collection electrode is arranged in the C3 or C4 region of the brain hemisphere of the user's control hand, and three collection electrodes are arranged in the symmetrical position of the brain hemisphere of the user's affected hand, and the three collection electrodes are arranged in a manner surrounding the center point of the C3 or C4 region.

8. The brainwave-based motion-intention recognition method according to any one of claims 2 to 7, characterized by, The collection channels of the multi-channel brain wave collection device are four.

9. A brainwave-based motion-intention recognition system, characterized by, The brain wave data collection device comprises a brain wave data collection end and a signal processing end. The brain wave data collection end is used for acquiring brain wave data obtained through pre-processing of a plurality of brain wave collection channels, and the pre-processing mode at least comprises signal filtering, denoising and amplification. The signal processing end comprises: A power extraction module is configured to extract power data of brain wave data of each brain wave collection channel in a preset characteristic frequency range, so as to obtain several characteristic frequency range brain wave power sum values most relevant to the user's motion intention. A power change rate calculation module is configured to calculate a power change rate of each characteristic frequency range brain wave power sum value in a sliding window period through a sliding time window, and the power change rate is used to represent the oscillation degree of the brain wave power. A motion intention recognition module is configured to divide the characteristic frequency range brain wave power sum values of each channel according to a preset corresponding rule of the power change rate and the motion intention grade, and determine the final motion intention grade of the user through the majority voting principle and output. 10.The brainwave-based motion-intention recognition system according to claim 9, wherein, The brain wave data collection end comprises: A collection unit is configured to acquire original brain wave data through electrodes arranged at predetermined positions and a multi-channel brain wave collection device; A signal processing unit is configured to perform filtering and denoising processing on the original brain wave data, so as to retain brain waves in a frequency range related to the motion intention; A signal amplification unit is configured to perform amplification processing on the brain waves, so as to enhance the signal amplitude. 11.The brainwave-based motion-intention recognition system of claim 9, wherein The power extraction module comprises: A frequency domain conversion unit is configured to perform fast Fourier transform on the brain wave data collected by each channel, and convert the brain wave data from time domain to frequency domain; A power analysis unit is configured to analyze the power distribution of the brain wave data at different frequencies, and obtain several characteristic frequency range brain wave power sum values most relevant to the user's motion intention. 12.The brainwave-based motion-intention recognition system of claim 9, wherein The power change rate calculation module comprises: A reference value calculation unit is configured to set the current time as the starting point of the calculation time window, and the brain wave power sum value data collected in the time period closest to the starting point of the time window as the brain wave power time sequence data, and calculate the average value thereof as the power reference value in the time window; A ratio calculation unit is configured to calculate the ratio of the instantaneous power sum value at the current time point to the power reference value as the power change rate. 13.The brainwave-based motion-intention recognition system of claim 9, wherein The motion intention recognition module comprises: An intention voting unit is configured to count each characteristic frequency range of each channel according to a preset corresponding relationship between the power change rate value and the motion intention grade division threshold, and regard the power change rate satisfying the set threshold range as the motion intention vote; A result output unit is configured to determine the final motion intention grade of the user through the majority voting principle and output. 14.The brainwave-based motion-intention recognition system according to claim 10, wherein The electrodes include a collection electrode, a grounding electrode and a reference electrode, the collection electrode is arranged in C3 and C4 area of the international 10-20 system of electroencephalogram, the grounding electrode is arranged in Cz area of the international 10-20 system of electroencephalogram, and the reference electrode is arranged in A1 or A2 area of the international 10-20 system of electroencephalogram. 15.The brainwave-based motion-intention recognition system of claim 14, wherein One collection electrode is arranged in C3 or C4 area of the brain hemisphere of the user's control side hand, three collection electrodes are arranged in the symmetrical position of the brain hemisphere of the user's affected side hand, and the three collection electrodes are arranged in a manner of surrounding the center point of C3 or C4 area.

16. The brainwave-based motion intent recognition system according to any one of claims 9 to 15, wherein The collection channels of the multi-channel electroencephalogram collection device are four.

17. A motion assist apparatus comprising an end effector, characterized by, The motion intention recognition system of any one of claims 9-16, wherein the end effector communicates with the motion intention recognition system through wired or wireless communication.

18. The motion assist device according to claim 17, characterized by The human-computer interaction interface and the peripheral device are further included.

19. A computer device, comprising: The memory stores a computer program, and the computer program is executed by the processor to make the processor execute the steps of the motion intention recognition method based on electroencephalogram in any one of claims 1-8.

20. A computer-readable storage medium, characterized in that, The computer readable storage medium stores a computer program, and the computer program is executed by the processor to make the processor execute the steps of the motion intention recognition method based on electroencephalogram in any one of claims 1-8.

Images