Brain-computer interface, brain-computer interface system, signal processing method and storage medium
Through the design of flexible electrode arrays and wireless processors, combined with biocompatible packaging and gene editing, the immune repulsion caused by hard metal electrodes is solved, and the long-term stability and biocompatibility of the brain-computer interface are achieved.
Patent Information
- Application Number
- CN202510513750.7
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-04-23
- Publication Date
- 2025-08-08
AI Technical Summary
The immune rejection caused by hard metal electrodes in existing brain-computer interfaces leads to poor long-term stability.
Using flexible electrode arrays and wireless processors, the flexible electrode arrays include flexible substrates and nanometal coatings, combining biocompatible packaging structures and gene editing technology to reduce immune rejection.
It improves the long-term stability of the brain-computer interface, reduces damage to brain tissue and immune rejection, and enhances biocompatibility.
Smart Images
Figure CN120447729A_ABST
Abstract
Description
Technical Field
[0001] The present application relates to the field of brain-computer interface technology, and in particular to a brain-computer interface, a brain-computer interface system, a signal processing method, and a storage medium. Background Art
[0002] As an emerging neuroscience research tool, brain-computer interface technology has shown great potential in animal experiments and clinical applications. However, the electrodes used in related technologies for brain-computer interfaces are made of hard metal electrodes. These electrodes are very rigid and can easily trigger immune rejection reactions, leading to brain tissue damage and poor long-term stability. Summary of the Invention
[0003] In view of this, the main purpose of the embodiments of the present application is to provide a brain-computer interface, a brain-computer interface system, a signal processing method and a storage medium that can reduce immune rejection reactions.
[0004] To achieve the above objectives, the technical solution of the embodiment of the present application is implemented as follows:
[0005] In a first aspect, an embodiment of the present application provides a brain-computer interface, wherein the brain-computer interface is connected to an external device signal, and the brain-computer interface comprises:
[0006] A flexible electrode array, wherein the flexible electrode array comprises a plurality of flexible electrodes, each of which comprises a flexible substrate and a nano-metal coating located on an outer layer of the flexible substrate;
[0007] A wireless processor is electrically connected to the flexible electrode array and is wirelessly connected to the external device.
[0008] In one embodiment, the flexible substrate is a polyimide substrate; and / or,
[0009] The nano metal coating is a nano gold coating.
[0010] In one embodiment, the flexible electrode array includes 2048 signal channels; and / or,
[0011] The flexible electrode array includes 2048 flexible electrodes integrated in a single array; and / or,
[0012] The thickness of the flexible electrode is less than or equal to 10 μm.
[0013] In one embodiment, the wireless processor includes a signal amplification module, which is electrically connected to the flexible electrode array to amplify the neural signals collected by the flexible electrode array; and / or,
[0014] The wireless processor includes a filtering module, which is electrically connected to the flexible electrode array and is used to filter the neural signals collected by the flexible electrode array; and / or,
[0015] The wireless processor includes a wireless transmission module, which is electrically connected to the flexible electrode array and wirelessly connected to the external device so that the wireless processor and the external device can perform wireless signal transmission.
[0016] In one embodiment, the brain-computer interface further includes a biocompatible packaging structure, and the biocompatible packaging structure covers at least one of the flexible electrode array and the wireless processor.
[0017] In one embodiment, the biocompatible packaging structure includes a neurotrophic factor coating, and the neurotrophic factor coating is coated on the outer surface of the flexible electrode; and / or,
[0018] The biocompatible packaging structure includes a titanium alloy shell and a polyparaxylene coating. The titanium alloy shell has a accommodating cavity, which is used to install the electronic components of the brain-computer interface. The electronic components include the wireless processor. The outer surface of the titanium alloy shell is coated with the polyparaxylene coating.
[0019] In one embodiment, the brain-computer interface further includes a wireless charging module, and the wireless charging module is electrically connected to the wireless processor.
[0020] In one embodiment, the brain-computer interface includes a gene editing layer, which includes a human CD55 protein gene and a human CD46 protein gene, and the gene editing layer is used to replace the Neu5Gc antigen gene and Sda antigen gene of the test subject.
[0021] A second aspect of an embodiment of the present application provides a brain-computer interface system, which includes an external device and any of the above-mentioned brain-computer interfaces, wherein the external device is connected to the wireless processor by wireless signals.
[0022] In one embodiment, a first signal transmission path passing through the wireless processor is formed between the flexible electrode array and the external device, and the first signal transmission path is used for transmitting the neural signals collected by the flexible electrode array to the external device after being processed by the wireless processor; and / or,
[0023] A second signal transmission path passing through the wireless processor is formed between the external device and the flexible electrode array. The second signal transmission path is used for the external device to output a command signal to the wireless processor, and for the wireless processor to output a stimulation signal to the flexible electrode array according to the command signal.
[0024] A third aspect of the embodiments of the present application provides a signal processing method, which is used in any of the above-mentioned brain-computer interface systems, and is characterized in that the signal processing method includes the following steps:
[0025] The flexible electrode array collects neural signals from the test subject;
[0026] The wireless processor receives and processes the neural signal collected by the flexible electrode array to output a corresponding wireless signal;
[0027] The external device trains an LSTM model according to the wireless signal output by the wireless processor;
[0028] The external device performs signal recognition on the wireless signal output by the wireless processor according to the LSTM model.
[0029] In one embodiment, the external device trains the LSTM model according to the wireless signal output by the wireless processor, specifically including:
[0030] The external device labels the wireless signal output by the wireless processor;
[0031] The external device obtains a model training sample according to the wireless signal and the signal labeling result;
[0032] The external device trains the LSTM model according to the model training sample.
[0033] In one embodiment, the external device labels the wireless signal output by the wireless processor, specifically including:
[0034] The external device labels the wireless signal output by the wireless processor based on at least one of the visual information and physiological information of the test subject.
[0035] In one embodiment, the external device labels the wireless signal output by the wireless processor, specifically including:
[0036] The external device performs frequency domain feature annotation on the wireless signal having power spectrum density changes in different frequency bands; and / or,
[0037] The external device performs time domain feature annotation on the wireless signal with peak value changes; and / or,
[0038] The external device performs time domain feature annotation on the wireless signal with variance change.
[0039] In one embodiment, the external device labels the wireless signal output by the wireless processor, specifically including:
[0040] When the test subject performs a specific action, the external device marks the moment when the test subject's action begins as a start time point, marks the moment when the test subject's action ends as an end time point, and marks the wireless signal output by the wireless processor between the start time point and the end time point.
[0041] A fourth aspect of an embodiment of the present application provides a storage medium, which stores computer-executable instructions. The computer-executable instructions can be executed by an external device to implement the steps of any of the above-mentioned signal processing methods.
[0042] The embodiments of the present application provide a brain-computer interface, a brain-computer interface system, a signal processing method and a storage medium. The brain-computer interface includes a flexible electrode array and a wireless processor, the flexible electrode array includes a plurality of flexible electrodes, and the flexible electrode includes a flexible substrate and a nano-metal coating located on the outer layer of the flexible substrate. In other words, by adopting a flexible substrate for the flexible electrode, the rigidity of the flexible electrode can be greatly reduced, and the softness of the flexible electrode can be made low, and its softness can be as close to the brain tissue as possible, thereby greatly reducing the immune rejection reaction, reducing brain tissue damage, and thus improving the long-term stability of the brain-computer interface. Moreover, by providing a nano-metal coating on the outer layer of the flexible substrate, it is possible to greatly improve the biocompatibility of the flexible electrode while facilitating the flexible electrode to collect neural signals, further reducing the immune rejection reaction, and thus further improving the long-term stability of the brain-computer interface. BRIEF DESCRIPTION OF THE DRAWINGS
[0043] Figure 1 This is a schematic diagram of the coordination relationship between the brain-computer interface and the test object in one embodiment of the present application;
[0044] Figure 2 This is a schematic diagram of the coordination relationship between the brain-computer interface and the test subject in another embodiment of the present application;
[0045] Figure 3 for Figure 1 Schematic diagram of the coordination between the various structures of the brain-computer interface and the test subject;
[0046] Figure 4 for Figure 3 Schematic diagram of the structure of the wireless processor;
[0047] Figure 5 for Figure 1 Schematic diagram of the biocompatible packaging structure of the brain-computer interface;
[0048] Figure 6 This is a schematic diagram of the coordination relationship between the brain-computer interface and external devices;
[0049] Figure 7 This is a flowchart of a signal processing method according to an embodiment of the present application.
[0050] Description of Reference Numerals
[0051] 10. Brain-computer interface; 11. Flexible electrode array; 111. Flexible electrode; 12. Wireless processor; 13. Biocompatible packaging structure; 20. External device. DETAILED DESCRIPTION
[0052] In the description of the embodiments of this application, the technical terms "first," "second," "third," etc. are used only to distinguish different objects and should not be understood to indicate or imply relative importance or implicitly specify the quantity, specific order, or primary and secondary relationship of the indicated technical features. In the description of the embodiments of this application, the meaning of "plurality" is two or more, unless otherwise specifically defined.
[0053] References herein to "embodiments" mean that a particular feature, structure, or characteristic described in connection with the embodiments may be included in at least one embodiment of the present application. The appearance of this phrase in various places in the specification does not necessarily refer to the same embodiment, nor does it constitute an independent or alternative embodiment that is mutually exclusive of other embodiments. It is understood, both explicitly and implicitly, by those skilled in the art that the embodiments described herein may be combined with other embodiments.
[0054] An embodiment of the present application provides a brain-computer interface 10, see Figure 2 、 Figure 3 and Figure 6 The brain-computer interface 10 is connected to the external device 20 by signal, and the brain-computer interface 10 includes a flexible electrode array 11 and a wireless processor 12.
[0055] See also Figure 1 The flexible electrode array 11 includes a plurality of flexible electrodes 111 , and the flexible electrode 111 includes a flexible substrate and a nano-metal coating located on the outer layer of the flexible substrate.
[0056] See also Figure 4 The wireless processor 12 is electrically connected to the flexible electrode array 11 , and the wireless processor 12 is wirelessly connected to the external device 20 .
[0057] Another embodiment of the present application provides a brain-computer interface 10 system, see Figure 3 and Figure 6 The brain-computer interface 10 system includes an external device 20 and the brain-computer interface 10 described in any embodiment of the present application, and the external device 20 is wirelessly connected to the wireless processor 12.
[0058] Specifically, the brain-computer interface 10 can establish a connection between a test object (such as a human or animal brain) and an external device 20, thereby enabling information exchange between the brain and the device. For ease of description, this application uses the application of the brain-computer interface 10 to a pig brain (test object) as an example.
[0059] In fact, the external device 20 is located outside the test subject. The brain-computer interface 10 of this application is an invasive brain-computer interface 10. The flexible electrode array 11 and the wireless processor 12 are both used to be installed inside the brain of the test subject to realize signal transmission between the test subject and the external device 20.
[0060] The flexible electrode array 11 can fit snugly on the surface of the test subject's brain or be inserted into brain tissue, adapting to the brain's complex shape and physiological activity, thereby reducing mechanical damage to brain tissue. The flexible electrode array 11, comprising multiple flexible electrodes 111, can record the electrical activity signals of multiple neurons at high density, quickly and accurately outlining brainstem nuclei in the form of a "heat map," providing a rich and accurate source of neural signal data for the brain-computer interface 10.
[0061] The flexible electrode 111 includes a flexible substrate and a nano-metal coating coated on the outer surface of the flexible substrate, thereby greatly improving the softness of the flexible electrode 111 to reduce immune rejection reactions.
[0062] It should be noted that the flexible substrate is a flexible substrate material, and its specific material type is not limited. For example, the flexible substrate is a polyimide substrate, thereby greatly improving the softness of the flexible electrode 111.
[0063] The specific type of the nanometal coating can be set according to actual conditions. For example, the nanometal coating is a nanogold coating, which can enable the flexible electrode 111 to better collect neural signals and improve the biocompatibility of the flexible electrode 111.
[0064] It should be noted that the number of flexible electrodes 111 and the number of channels of the flexible electrode array 11 can be determined according to actual conditions.
[0065] For example, the flexible electrode array 11 includes 2048 signal channels, thereby enabling better signal analysis with high temporal and spatial resolution.
[0066] For another example, the flexible electrode array 11 includes 2048 flexible electrodes 111 integrated in a single array, thereby enabling better signal capture at the single neuron level.
[0067] The flexible electrode array 11 can achieve three-dimensional multi-directional distribution and can cover the motor cortex and olfactory center (key functional areas of the pig brain).
[0068] The specific size of the flexible electrode 111 can be set according to actual conditions.
[0069] For example, the thickness of the flexible electrode 111 is less than or equal to 10 μm, such as 5 μm or 10 μm, thereby facilitating the collection of neural signals and reducing damage to brain tissue.
[0070] The wireless processor 12 is electrically connected to the flexible electrode array 11 and is also wirelessly connected to the external device 20. Thus, information transmission between the flexible electrode 111 and the external device 20 can be achieved.
[0071] In fact, the wireless processor 12 can pre-process the neural signals collected by the flexible electrode array 11 (such as amplification, filtering, digitization, etc.), perform preliminary signal analysis and feature extraction, and transmit the processed data to the external device 20 through wireless communication technology. At the same time, it can also receive instructions from the external device 20 to realize remote control and parameter adjustment of the brain-computer interface 10.
[0072] For example, a first signal transmission path passing through the wireless processor 12 is formed between the flexible electrode array 11 and the external device 20. The first signal transmission path is used for transmitting the neural signals collected by the flexible electrode array 11 to the external device 20 after being processed by the wireless processor 12. This enables the brain-computer interface 10 to better output the collected neural signals.
[0073] Specifically, the flexible electrode array 11 collects neural signals (electrical activity signals) from neurons or neural tissue of the test subject (such as a pig brain) and transmits them to the wireless processor 12 through the connection line between the flexible electrode array 11 and the wireless processor 12. The wireless processor 12 performs preprocessing operations (such as amplification, filtering, digitization, etc.) on the input neural signals, and performs feature extraction and analysis. The processed data is transmitted to the external device 20 in the form of a wireless signal through the wireless processor 12. On the external device 20, the data is further analyzed, decoded, and visualized to understand the relationship between the neural activity pattern of the test subject and the behavior or physiological phenomenon. The above-mentioned signal transmission path is the first signal transmission path.
[0074] For another example, a second signal transmission path is formed between the external device 20 and the flexible electrode array 11, passing through the wireless processor 12. This second signal transmission path is used for the external device 20 to output a command signal to the wireless processor 12, which then causes the wireless processor 12 to output a corresponding stimulation signal to the flexible electrode array 11 based on the command signal. This effectively enables the input of external signals to the test subject.
[0075] Specifically, the external device 20 sends a command signal to the wireless processor 12 via wireless communication. After receiving the command signal, the wireless processor 12 sends a corresponding stimulation signal to the flexible electrode array 11, thereby implementing intervention and regulation on the test object.
[0076] The specific size of the wireless processor 12 can be set according to actual conditions.
[0077] For example, the wireless processor 12 is a micro wireless processor 12 , the diameter of which is less than or equal to 23 mm and the thickness of which is less than or equal to 8 mm.
[0078] The brain-computer interface 10 of the embodiment of the present application includes a flexible electrode array 11 and a wireless processor 12. The flexible electrode array 11 includes a plurality of flexible electrodes 111. The flexible electrode 111 includes a flexible substrate and a nano-metal coating located on the outer layer of the flexible substrate. In other words, by using a flexible substrate for the flexible electrode 111, the rigidity of the flexible electrode 111 can be greatly reduced, and the softness of the flexible electrode 111 can be made low, and its softness can be as close to the brain tissue as possible, thereby greatly reducing the immune rejection reaction and reducing brain tissue damage, thereby improving the long-term stability of the brain-computer interface 10. In addition, by providing a nano-metal coating on the outer layer of the flexible substrate, it is possible to greatly improve the biocompatibility of the flexible electrode 111 while facilitating the flexible electrode 111 to collect neural signals, further reducing the immune rejection reaction, thereby further improving the long-term stability of the brain-computer interface 10.
[0079] The wireless processor 12 can receive the neural signals collected from the flexible electrode array 11 and process the neural signals.
[0080] For example, the wireless processor 12 includes a signal amplification module, which is electrically connected to the flexible electrode array 11 to amplify the neural signals collected by the flexible electrode array 11. In this way, the neural signals can be enhanced to a processable range for easy processing and analysis.
[0081] For example, the wireless processor 12 includes a filtering module electrically connected to the flexible electrode array 11 for filtering the neural signals collected by the flexible electrode array 11. This allows for frequency filtering and noise suppression of the collected raw neural signals, thereby extracting effective information and optimizing signal quality.
[0082] For another example, the wireless processor 12 includes a wireless transmission module, which is electrically connected to the flexible electrode array 11 and wirelessly connected to the external device 20, so that the wireless processor 12 and the external device 20 can perform wireless signal transmission. This facilitates the transmission of neural signals processed by the wireless processor 12 to the external device 20 via wireless signals, and also facilitates the wireless processor 12 to receive command signals from the external device 20.
[0083] In one embodiment, please refer to Figure 5 The brain-computer interface 10 further includes a biocompatible packaging structure 13 , which covers at least one of the flexible electrode array 11 and the wireless processor 12 .
[0084] Specifically, the biocompatible packaging structure 13 is a structure made of a material with good biocompatibility, which can isolate the internal components from the surrounding brain tissue, protect the internal components from erosion and physical damage by biological fluids, ensure normal operation, and extend the service life of the device. At the same time, it has good mechanical properties, can adapt to the physiological environment of the test subject, reduce pressure and stimulation on the brain tissue, and reduce the occurrence of inflammatory response.
[0085] In fact, the biocompatible packaging structure 13 may cover only the flexible electrode array 11 , or only the wireless processor 12 , or may cover both the flexible electrode array 11 and the wireless processor 12 .
[0086] It should be noted that the specific type of the biocompatible packaging structure 13 can be set according to actual conditions.
[0087] For example, the biocompatible packaging structure 13 includes a neurotrophic factor coating (such as Brain-Derived Neurotrophic Factor, BDNF), which is coated on the outer surface of the flexible electrode 111. This can promote the fusion of nerve tissue and the electrode interface.
[0088] For example, the biocompatible packaging structure 13 includes a titanium alloy shell and a parylene coating. The titanium alloy shell has a cavity for mounting the electronic components of the brain-computer interface 10, including the wireless processor 12. The outer surface of the titanium alloy shell is coated with the parylene coating. This can further reduce foreign body reactions.
[0089] It is understandable that, in addition to accommodating the wireless processor 12 , the accommodating cavity may also accommodate other electronic components according to actual conditions.
[0090] In one embodiment, the brain-computer interface 10 further includes a wireless charging module, which is electrically connected to the wireless processor 12. Thus, an external power source can be used to charge the wireless processor 12 via the wireless charging module.
[0091] In a specific embodiment, the wireless processor 12 supports Bluetooth 5.0 and near-field wireless power supply, the wireless charging module frequency is 13.56 MHz, the battery life is ≥ 72 hours, and the charging efficiency is ≥ 85%.
[0092] In one embodiment, the brain-computer interface 10 includes a gene editing layer, which includes human CD55 protein genes and human CD46 protein genes. The gene editing layer is used to replace the Neu5Gc antigen gene and Sda antigen gene of the test subject, thereby reducing immune rejection reactions.
[0093] Specifically, the gene editing layer can edit the genes of the test subject (such as a pig) to express some proteins or factors that can reduce the immune response in the test subject, so as to reduce the recognition and attack of the test subject's immune system on the implanted brain-computer interface 10, reduce the probability of immune rejection reaction, improve the long-term stability and biocompatibility of the device in the test subject, and realize the long-term effective operation of the brain-computer interface 10.
[0094] In a specific embodiment, the test subject is the Zhongke Oger DPF medical donor pig model. By knocking out the Neu5Gc and Sda antigen genes in the pig brain tissue and transferring the human CD55 / CD46 protein genes, the immune rejection reaction can be reduced.
[0095] In one embodiment, the brain-computer interface 10 system further includes a surgical robot that matches the brain-computer interface 10 , and the surgical robot can be used to implant the flexible electrode array 11 into a target area of a test subject.
[0096] In a specific embodiment, the surgical robot uses laser-assisted drilling (aperture ≤ 2 mm) to accurately locate the brain area, with an implantation depth error of ≤ 50 μm and an operation time of ≤ 30 minutes.
[0097] In a specific embodiment, the surgical robot is equipped with optical navigation and vascular avoidance algorithms to achieve minimally invasive skull opening (aperture ≤ 2 mm) and precise electrode implantation (positioning accuracy ± 50 μm).
[0098] Another embodiment of the present application provides a signal processing method. Figure 7 The signal processing method is used for the brain-computer interface 10 system described in any embodiment of the present application, and the signal processing method includes the following steps:
[0099] Step S1: The flexible electrode array 11 collects neural signals from the test subject.
[0100] Step S2: The wireless processor 12 receives and processes the neural signals collected by the flexible electrode array 11 to output corresponding wireless signals.
[0101] Step S3: the external device 20 trains the LSTM model according to the wireless signal output by the wireless processor 12 .
[0102] Step S4: the external device 20 performs signal recognition on the wireless signal output by the wireless processor 12 according to the LSTM model.
[0103] Specifically, steps S1 to S3 are a process in which the external device 20 collects neural signals through the flexible electrode array 11 and the wireless processor 12 transmits the neural signals to train the LSTM model.
[0104] Step S4 is a process in which the external device 20 identifies the wireless signal subsequently output by the wireless processor 12 according to the trained LSTM model.
[0105] For example, after the LSTM model is trained, neural signals are collected, processed and output through the brain-computer interface 10, so that the LSTM model can identify the current action (such as running) or physical condition (such as pain) of the test subject (pig) based on the output wireless signal.
[0106] In the multimodal neural decoding algorithm, the LSTM neural network (Long Short-Term Memory) includes a cell state, a forget gate, an input gate, and an output gate.
[0107] The cell state is responsible for transmitting information between different time steps, maintaining a relatively stable memory of the entire time series and used for storing long-term information. The forget gate generates a ft value between 0 and 1 based on the previous hidden state ht-1 and the current input xt. A ft value close to 0 indicates that the corresponding information is discarded, while a value close to 1 indicates that the information is retained. Its calculation formula is ft = σ(Wf·(ht-1,xt)+bf). Wf is the weight matrix, bf is the bias term, and σ is the Sigmoid function. The input gate (InputGate) consists of a Sigmoid layer and a Tanh layer. The Sigmoid layer generates an it value, which is used to control the extent to which new information is written; the Tanh layer generates a new candidate value Ct. Ultimately, the new cell state Ct is calculated as Ct = ft*Ct-1+it*Ct, where Ct-1 is the cell state at the previous moment. The output gate first obtains an ot value through the Sigmoid layer, then uses the Tanh function to process the cell state, and finally outputs ht=ot*tanh(Ct), where ht is the hidden state at the current moment, that is, the output of LSTM.
[0108] An LSTM neural network consists of two unidirectional LSTMs operating in opposite directions. At each time instant t, input is fed simultaneously to both LSTMs. One forward LSTM processes the input sequence in chronological order, while the other backward LSTM processes the input sequence in reverse chronological order. The final output is determined by the combined outputs of these two unidirectional LSTMs.
[0109] Constructing a test subject's motion intention prediction model through the LSTM neural network can enable the brain-computer interface 10 system to better recognize neural signals.
[0110] In one embodiment, the external device 20 trains the LSTM model based on the wireless signal output by the wireless processor 12, specifically including the following steps:
[0111] The external device 20 labels the wireless signal output by the wireless processor 12 .
[0112] The external device 20 obtains model training samples based on the wireless signal and the signal labeling result.
[0113] The external device 20 trains the LSTM model according to the model training sample.
[0114] Specifically, by labeling wireless signals to obtain model training samples, and feeding the model training samples to the LSTM model, the LSTM model can learn the characteristics of neural signals and recognize neural signals.
[0115] It should be noted that the specific manner in which the external device 20 labels the wireless signal output by the wireless processor 12 can be determined according to actual conditions.
[0116] For example, the external device 20 labels the wireless signal output by the wireless processor 12, specifically including:
[0117] The external device 20 combines at least one of the visual information and physiological information of the test subject to label the wireless signal output by the wireless processor 12. In this way, the accuracy of the LSTM model can be improved.
[0118] For example, the external device 20 may label the neural signals in the corresponding state according to the visual information (such as external appearance) of the test subject.
[0119] The external device 20 may also label the state of the neural signal in the corresponding state according to the physiological information (such as physiological indicators) of the test subject.
[0120] The external device 20 may also combine visual information annotation with neural signal for fusion annotation.
[0121] The external device 20 may also combine physiological information annotation with neural signal annotation for fusion annotation.
[0122] In one embodiment, the external device 20 labels the wireless signal output by the wireless processor 12, specifically including:
[0123] External device 20 performs frequency domain feature annotation on wireless signals that exhibit variations in power spectrum density across different frequency bands. In other words, external device 20 performs frequency domain feature annotation on neural signals that exhibit significant variations in features such as power spectrum density across different frequency bands. This improves the accuracy of the LSTM model.
[0124] In one embodiment, the external device 20 labels the wireless signals output by the wireless processor 12. Specifically, the external device 20 labels the time-domain features of wireless signals that exhibit peak value variations. In other words, the external device 20 labels the time-domain features of neural signals that exhibit significant peak value variations. This improves the accuracy of the LSTM model.
[0125] In one embodiment, the external device 20 labels the wireless signals output by the wireless processor 12. Specifically, the external device 20 labels the time-domain features of wireless signals that exhibit variance changes. In other words, the external device 20 labels the time-domain features of neural signals that exhibit significant variance changes. This improves the accuracy of the LSTM model.
[0126] In one embodiment, the external device 20 labels the wireless signal output by the wireless processor 12, specifically including:
[0127] When the test subject performs a specific action, the external device 20 marks the start time of the test subject's action as the start time point and the end time point as the end time point. It also annotates the wireless signals output by the wireless processor 12 between the start and end time points. This improves the accuracy of the LSTM model.
[0128] Yet another embodiment of the present application provides a storage medium storing computer-executable instructions. The computer-executable instructions can be executed by an external device 20 to implement the steps of the signal processing method described in any embodiment of the present application.
[0129] In the description of this application, the descriptions with reference to the terms "in one embodiment", "in some embodiments", "in a specific embodiment", or "exemplary" etc. mean that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the embodiments of the present application. In this application, the schematic representations of the above terms do not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described can be combined in a suitable manner in any one or more embodiments or examples. In addition, those skilled in the art can combine different embodiments or examples described in this application and features of different embodiments or examples without contradiction.
[0130] The above are merely preferred embodiments of the present application and are not intended to limit the present application. Persons skilled in the art will readily appreciate that various modifications and variations are possible. Any modifications, equivalent substitutions, or improvements made within the spirit and principles of the present application are intended to be within the scope of protection of the present application.
Claims
1. A brain-computer interface, characterized in that: The brain-computer interface is connected to an external device signal, and the brain-computer interface includes: A flexible electrode array, wherein the flexible electrode array comprises a plurality of flexible electrodes, each of which comprises a flexible substrate and a nano-metal coating located on an outer layer of the flexible substrate; A wireless processor is electrically connected to the flexible electrode array and is wirelessly connected to the external device.
2. The brain-computer interface according to claim 1, characterized in that The flexible substrate is a polyimide substrate; and / or, The nano metal coating is a nano gold coating.
3. The brain-computer interface according to claim 1, characterized in that The flexible electrode array includes 2048 signal channels; and / or, The flexible electrode array includes 2048 flexible electrodes integrated in a single array; and / or, The thickness of the flexible electrode is less than or equal to 10 μm.
4. The brain-computer interface according to any one of claims 1 to 3, characterized in that: The wireless processor includes a signal amplification module, which is electrically connected to the flexible electrode array to amplify the neural signals collected by the flexible electrode array; and / or, The wireless processor includes a filtering module, which is electrically connected to the flexible electrode array and is used to filter the neural signals collected by the flexible electrode array; and / or, The wireless processor includes a wireless transmission module, which is electrically connected to the flexible electrode array and wirelessly connected to the external device so that the wireless processor and the external device can perform wireless signal transmission.
5. The brain-computer interface according to any one of claims 1 to 3, characterized in that: The brain-computer interface further includes a biocompatible packaging structure, which covers at least one of the flexible electrode array and the wireless processor.
6. The brain-computer interface according to any one of claims 1 to 3, characterized in that: The biocompatible packaging structure includes a neurotrophic factor coating, and the neurotrophic factor coating is coated on the outer surface of the flexible electrode; and / or, The biocompatible packaging structure includes a titanium alloy shell and a polyparaxylene coating. The titanium alloy shell has a accommodating cavity, which is used to install the electronic components of the brain-computer interface. The electronic components include the wireless processor. The outer surface of the titanium alloy shell is coated with the polyparaxylene coating.
7. The brain-computer interface according to any one of claims 1 to 3, characterized in that: The brain-computer interface also includes a wireless charging module, which is electrically connected to the wireless processor.
8. The brain-computer interface according to any one of claims 1 to 3, characterized in that: The brain-computer interface includes a gene editing layer, which includes a human CD55 protein gene and a human CD46 protein gene. The gene editing layer is used to replace the Neu5Gc antigen gene and the Sda antigen gene of the test subject.
9. A brain-computer interface system, characterized in that: The brain-computer interface system includes an external device and the brain-computer interface according to any one of claims 1 to 8, and the external device is connected to the wireless processor by wireless signals.
10. The brain-computer interface system according to claim 9, characterized in that: A first signal transmission path passing through the wireless processor is formed between the flexible electrode array and the external device, and the first signal transmission path is used for transmitting the neural signals collected by the flexible electrode array to the external device after being processed by the wireless processor; and / or, A second signal transmission path passing through the wireless processor is formed between the external device and the flexible electrode array. The second signal transmission path is used for the external device to output a command signal to the wireless processor, and for the wireless processor to output a stimulation signal to the flexible electrode array according to the command signal.
11. A signal processing method, used in the brain-computer interface system according to claim 9 or 10, characterized in that: The signal processing method comprises the following steps: The flexible electrode array collects neural signals from the test subject; The wireless processor receives and processes the neural signal collected by the flexible electrode array to output a corresponding wireless signal; The external device trains an LSTM model according to the wireless signal output by the wireless processor; The external device performs signal recognition on the wireless signal output by the wireless processor according to the LSTM model.
12. The signal processing method according to claim 11, characterized in that: The external device trains the LSTM model according to the wireless signal output by the wireless processor, specifically including: The external device labels the wireless signal output by the wireless processor; The external device obtains a model training sample according to the wireless signal and the signal labeling result; The external device trains the LSTM model according to the model training sample.
13. The signal processing method according to claim 12, wherein: The external device labels the wireless signal output by the wireless processor, specifically including: The external device labels the wireless signal output by the wireless processor based on at least one of the visual information and physiological information of the test subject.
14. The signal processing method according to claim 12 or 13, characterized in that: The external device labels the wireless signal output by the wireless processor, specifically including: The external device performs frequency domain feature annotation on the wireless signal having power spectrum density changes in different frequency bands; and / or, The external device performs time domain feature annotation on the wireless signal with peak value changes; and / or, The external device performs time domain feature annotation on the wireless signal with variance change.
15. The signal processing method according to claim 12 or 13, characterized in that: The external device labels the wireless signal output by the wireless processor, specifically including: When the test subject performs a specific action, the external device marks the moment when the test subject's action begins as a start time point, marks the moment when the test subject's action ends as an end time point, and marks the wireless signal output by the wireless processor between the start time point and the end time point.
16. A storage medium, characterized in that The storage medium stores computer-executable instructions, and the computer-executable instructions can be executed by an external device to implement the steps of the signal processing method according to any one of claims 11 to 15.