Method and system for visualizing monitoring information based on an implantable medical chip

Abstract

The application provides a kind of regulation monitoring information visualization method and system based on implantable medical chip, it is related to data processing technical field, comprising: obtaining original living nerve signal, signal feature abstraction is carried out to original living nerve signal, obtain abstracted signal, can ignore the limitation of original living nerve signal, can carry out feature embedding to original living nerve signal to obtain the first signal feature representation of less sampling point number, the second signal feature representation of less sampling point number is obtained by carrying out feature embedding to abstracted signal, noise is added to the first signal feature representation, noise is added, obtain noise neural signal, because the signal of abstraction ignores the limitation of original living nerve signal, in addition, visual constraint parameter can provide different visual signal parameter, then based on visual constraint parameter, visual neural signal of visual constraint parameter constraint can be accurately obtained, improve the reliability of visual neural signal.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of data processing, and particularly relates to a regulation and monitoring information visualization method and system based on an implantable medical chip. BACKGROUND

[0002] With the rapid development of neuroscience and medical technology, implantable medical chips are increasingly widely used in the field of neural regulation and monitoring. Such chips can collect real-time in vivo neural signals in the human body, providing doctors with a new way to directly observe and analyze neural activity. The original in vivo neural signals can have high complexity and redundancy, and their high sampling point number and random fluctuation characteristics bring great challenges to subsequent signal processing and analysis. Traditional neural signal processing methods rely on specific signal features and pre-set analysis frameworks, which limits their versatility and flexibility for different types of neural signals. In particular, for neural signals from different individuals, different physiological states or different collection conditions, traditional methods are difficult to ensure the consistency and accuracy of the processing results. In addition, traditional visualization methods often focus on the direct display of signals, lack the ability to flexibly adapt to and individually customize the visualization needs of users, and are difficult to meet the needs of medical professionals for in-depth analysis of signals. SUMMARY

[0003] Therefore, the embodiments of the present application at least provide a regulation and monitoring information visualization method and system based on an implantable medical chip. The technical solution of the present application is realized as follows:

[0004] In one aspect, the present application provides a regulation and monitoring information visualization method based on an implantable medical chip, comprising: acquiring an original in vivo neural signal, abstracting signal features of the original in vivo neural signal to obtain an abstracted signal, embedding features of the original in vivo neural signal to obtain a first signal feature representation, embedding features of the abstracted signal to obtain a second signal feature representation; the sampling point number of the original in vivo neural signal is greater than the sampling point number of the first signal feature representation and the second signal feature representation; adding noise to the first signal feature representation to obtain a noise-added neural signal; acquiring a visualization constraint parameter, mapping the visualization constraint parameter to a constraint implicit representation array, mapping the second signal feature representation to a signal implicit representation array, the visualization constraint parameter being used to describe a visualization feature of a visualization display result; removing noise from the noise-added neural signal according to the constraint implicit representation array and the signal implicit representation array to obtain a noise-cleaned transition signal, restoring the noise-cleaned transition signal into a visualization neural signal conforming to a target visualization feature; the sampling point number of the original in vivo neural signal and the visualization neural signal is consistent.

[0005] In another aspect, the present application provides a computer system comprising a memory and a processor, wherein the memory stores a computer program capable of running on the processor, and the processor implements the steps of the above method when executing the program.

[0006] The beneficial effects of the present application include: the embodiment of the present application can ignore the limitation of the original living nerve signal based on obtaining the original living nerve signal, abstract the signal characteristics of the original living nerve signal, obtain the abstracted signal, and different original living nerve signals can be converted into visualized nerve signals, the original living nerve signal can be characterized and embedded to obtain a first signal characteristic representation with a small number of sampling points, the abstracted signal can be characterized and embedded to obtain a second signal characteristic representation with a small number of sampling points, noise can be added to the first signal characteristic representation to obtain a noise-added signal, and the noise-added signal can be mapped into a constraint implicit representation array and a signal implicit representation array based on a focusing strategy, which can be used as a guide array for a target machine learning algorithm, and the target machine learning algorithm can be guided by the guide array to remove the noise-added signal, obtain a noise-cleaning transition signal more related to the abstracted signal and the visualized constraint parameter, and restore the noise-cleaning transition signal into a visualized nerve signal conforming to the visualized characteristics, because the abstracted signal ignores the limitation of the original living nerve signal, and in addition, the visualized constraint parameter can provide different visualized signal parameters, the visualized nerve signal constrained by the visualized constraint parameter can be accurately obtained based on the visualized constraint parameter, and the reliability of the visualized nerve signal is improved. BRIEF DESCRIPTION OF DRAWINGS

[0007] The accompanying drawings, which are incorporated into and form a part of the specification, illustrate embodiments consistent with the present application and, together with the description, serve to explain the principles of the application.

[0008] Figure 1 An implementation flowchart of a method for visualizing regulation and monitoring information based on an implantable medical chip is provided for the embodiment of the present application.

[0009] Figure 2 A hardware entity diagram of a computer system is provided for the embodiment of the present application. DETAILED DESCRIPTION

[0010] The embodiment of the present application provides a method for visualizing regulation and monitoring information based on an implantable medical chip, which can be executed by a processor of a computer system. The computer system can refer to a server, a notebook computer, a tablet computer, a desktop computer, and the like. Figure 1 An implementation flowchart of a method for visualizing regulation and monitoring information based on an implantable medical chip is provided for the embodiment of the present application. Figure 1As shown, the method comprises:

[0011] Step S100: Obtain the original living body neural signal, abstract the signal features of the original living body neural signal to obtain an abstracted signal, embed the features of the original living body neural signal to obtain a first signal feature representation, and embed the features of the abstracted signal to obtain a second signal feature representation; the number of sampling points of the original living body neural signal is greater than the number of sampling points of the first signal feature representation and the second signal feature representation.

[0012] In step S100, the computer system first connects to the organism through the implantable medical chip and collects the original living body neural signal in real time. These signals can be composed of a series of electrical signals, reflecting the electrical activity of neurons in the organism. For example, in brain-machine interface (BMI) applications, an implantable electrode array can record local field potentials (LFP) or spike activity in the cerebral cortex region.

[0013] Before feature abstraction, the computer system can preprocess the original living body neural signal to remove noise and interference and improve signal quality. This can include steps such as filtering, detrending, baseline correction, etc. For example, a band-pass filter can be used to remove low-frequency drift and high-frequency noise, and only the frequency band related to neural activity (such as 0.1 Hz to 300 Hz) is retained.

[0014] Signal feature abstraction is the core step in step S100, which aims to extract useful features from the complex raw signal for subsequent processing and analysis, while ignoring unimportant details. This process can include the following procedures: the computer system first identifies action potentials (or called spikes) in the raw signal, which are direct manifestations of neuronal firing. For example, a threshold detection method can be set to identify signal segments that exceed the set threshold as potential action potentials. For each identified action potential, the system calculates its waveform feature values, such as peak amplitude, half-width, rise time, etc. In order to reduce the impact of noise on feature extraction, the computer system smoothes the raw signal or the identified action potentials. This can be achieved by moving average filtering, median filtering or Gaussian filtering, etc. The smoothed signal (called smoothed signal) will be smoother, which is beneficial for subsequent feature calculation. Then, the system calculates the time gradient and amplitude gradient of the smoothed signal, i.e. the rate of change of the signal with time and the rate of change of the amplitude. Based on these gradient information, the system can further perform local extreme value detection to more accurately locate the starting and ending points of the action potential. For example, if a signal segment shows a rapid change from positive to negative in the time gradient, and the amplitude gradient also changes significantly, then this segment may be an action potential. After identifying all important action potentials, the system performs feature abstraction based on the waveform features, time gradients and amplitude gradients of the action potentials, etc. This step may involve simplifying complex waveforms into a series of feature vectors, each containing multiple feature values such as peak amplitude, width, shape parameters of action potentials, etc. These feature vectors constitute the abstracted signal, which is more concise than the original signal, but retains enough information for subsequent analysis.

[0015] On the basis of feature abstraction, the computer system further performs feature embedding on the original live neural signals and the abstracted signals to generate first signal feature representations and second signal feature representations. Feature embedding maps high-dimensional raw data to a low-dimensional space while trying to preserve important information in the raw data. For the original live neural signals, the system uses target machine learning algorithms (such as autoencoders, variational autoencoders, etc.) for feature embedding. These algorithms learn how to effectively compress high-dimensional raw signals into low-dimensional first signal feature representations through unsupervised learning. For example, an autoencoder model may contain an encoder network and a decoder network, the encoder compresses the raw signal into a low-dimensional vector (first signal feature representation), and the decoder tries to reconstruct the original signal from the vector. The model is trained by minimizing the reconstruction error. Similarly, the system performs feature embedding on the abstracted signals to generate second signal feature representations. Since the abstracted signals have been preliminarily feature extracted and simplified, their feature embedding process may be more direct and efficient. For example, linear transformation, principal component analysis (PCA) or other dimensionality reduction techniques can be directly used to generate second signal feature representations. Through the feature embedding process, the number of sampling points of the original live neural signals and the abstracted signals is significantly reduced. This is because feature embedding aims to preserve the main features of the data while removing redundant and noisy information. For example, the original signal may contain millions of sampling points, while the first signal feature representation and the second signal feature representation after feature embedding may only have a few hundred or a few thousand elements. This reduction in the number of sampling points not only reduces the computational complexity of subsequent processing, but also improves the running efficiency of the algorithm.

[0016] Step S100 successfully transforms complex original live neural signals into more easily processed and visualized first signal feature representations and second signal feature representations through three sub-steps of signal preprocessing, feature abstraction and feature embedding. This process not only reduces the redundancy and noise of the data, but also preserves the key information.

[0017] Step S200: Adding noise to the first signal feature representation to obtain a noisy neural signal.

[0018] In step S200, the computer system can choose to add different types of noise. Common noise types include Gaussian noise, salt and pepper noise, multiplicative noise, etc. Each type of noise has its specific distribution characteristics and application scenarios. Taking Gaussian noise as an example, the computer system can add noise to the first signal feature representation according to the following steps:

[0019] First, determine the intensity of the noise to be added, which can be controlled by a noise scaling factor (such as σ). The magnitude of the noise scaling factor determines the amount of noise added to the signal, and it can be adjusted according to the specific application scenario and experimental requirements. For example, a smaller σ value can be set to simulate slight interference, or a larger σ value can be set to simulate severe interference. Based on the selected noise type and intensity, the computer system generates a noise vector with the same dimensions as the first signal feature representation. For Gaussian noise, random numbers conforming to a Gaussian distribution can be generated using a random number generator and filled into the noise vector. The generated noise vector is then added element-wise to the first signal feature representation (for multiplicative noise, this is done by multiplying element-wise and then adding the original signal). In this way, each element will be subject to a certain degree of noise interference.

[0020] The first signal feature representation after noise addition processing (now called the noisy neural signal) will be used in subsequent steps, such as mapping of visualization constraint parameters, noise removal, and signal restoration. The presence of the noisy neural signal makes the entire processing flow closer to real-world application scenarios, improving the model's practicality and reliability.

[0021] For example, suppose the first signal feature is represented as a vector x=[x1,x2,…,x] of length 100. 100 ], where each element x i This represents the projection of the original living neural signal onto a certain feature dimension. The computer system selects Gaussian noise as the noise type and sets the noise scaling factor σ=0.1. Then, it generates a Gaussian noise vector n=[n1,n2,…,n] of the same length 100. 100 ], where each element n i The noise vector is randomly drawn from the standard normal distribution N(0,1). Finally, the noise vector n is multiplied by the noise scaling factor σ and added to the first signal feature representation x to obtain the noisy neural signal y=x+0.1·n.

[0022] Step S200 simulates real-world interference factors by adding noise to the first signal feature representation. This process enhances the robustness and generalization ability of subsequent processing steps, making the entire visualization process more closely resemble practical application scenarios. By appropriately selecting the noise type and intensity, the computer system can flexibly control the characteristics of the noisy neural signals to meet the needs of different experiments and applications.

[0023] Step S300: Obtain the visualization constraint parameters, map the visualization constraint parameters to an array of implicit constraint representations, and map the second signal feature representation to an array of implicit signal representations. The visualization constraint parameters are used to describe the visualization features of the visualization display results.

[0024] The visualization constraint parameters are a set of parameters that reflect the user's visualization needs, which can be represented by tags. In a computer system, these parameters can exist in the form of key-value pairs, JSON objects, or data structures in a specific format. For example, the user may specify a color mapping scheme (such as "display red for high action potential amplitude and blue for low") or specify a dynamic display method for the signal (such as "signal intensity changes over time in the form of a waveform").

[0025] The computer system first parses these visualization constraint parameters, understands the user's intentions, and converts them into internal representations. This conversion process may involve validity verification, type conversion, unit unification, and other operations to ensure that subsequent steps can be correctly processed.

[0026] In order to effectively apply the visualization constraint parameters to subsequent signal processing and visualization generation processes, the computer system maps these parameters to constraint implicit representation arrays. This process can be achieved through a constraint embedding branch component in the target machine learning algorithm (such as a deep learning model).

[0027] The constraint embedding branch component is a network structure specifically designed to process visualization constraint parameters. It receives visualization constraint parameters as input and, through a series of nonlinear transformations (such as fully connected layers, convolutional layers, activation functions, etc.), embeds these parameters into a high-dimensional space to generate constraint implicit representation arrays. This array is a numerical vector, and each element represents the contribution or influence of the visualization constraint parameter in a specific dimension. For example, assuming the user specifies a color mapping scheme as a visualization constraint parameter, the constraint embedding branch component can convert this scheme into a vector containing multiple color coding values, each corresponding to a specific signal intensity range or feature category. In this way, in subsequent signal processing and visualization generation processes, the computer system can determine how to map different parts of the signal to different colors according to this vector.

[0028] At the same time, the computer system also maps the second signal feature representation to the signal implicit representation array. This step is similar to the generation process of the constraint implicit representation array, but the input data is different. The signal implicit representation array is a further abstraction and compression of the second signal feature representation, which retains the main feature information of the signal while removing redundant and noisy parts.

[0029] The generation of the signal implicit representation array can also rely on a component in the target machine learning algorithm (which may be different from the constraint embedding branch component). This component receives the second signal feature representation as input and, through a series of transformation operations (such as the encoding part of the autoencoder, dimension reduction algorithms, etc.), maps it to a low-dimensional implicit space to generate the signal implicit representation array.

[0030] For example, assume there is a convolutional neural network (CNN) based target machine learning algorithm that contains a constraint embedding branch component and a signal embedding component. In step S300:

[0031] The user specifies a color mapping scheme as "action potential amplitude between 0-50 μV is displayed as blue, 50-100 μV as green, and 100 μV and above as red". These parameters are encoded into a vector containing color coding values, for example [blue_code, green_code, red_code], where each color coding value is a three-dimensional vector (corresponding to the red, green, and blue components in the RGB color space). The constraint embedding branch component receives this color coding vector as input and maps it to a higher-dimensional implicit representation vector (e.g., 128-dimensional) through a series of convolutional layers, pooling layers, and fully connected layers, etc. Each element in this vector contains some abstract information about the color mapping scheme, but no longer directly corresponds to a specific color coding value. At the same time, the signal embedding component receives the second signal feature representation (assume it is a 64-dimensional vector) as input and maps it to a lower-dimensional implicit representation vector (e.g., 32-dimensional) through the encoding part of an autoencoder or similar dimension reduction algorithm. This vector compactly represents the main features of the original living nerve signal.

[0032] Step S300 provides the necessary input data for the subsequent noise removal and signal restoration steps by mapping the visualization constraint parameters and the second signal feature representation into constraint implicit representation arrays and signal implicit representation arrays, respectively. These two implicit representation arrays not only contain the user's desired visualization feature information, but also retain the main feature information of the signal, enabling the computer system to remove noise while preserving the original characteristics of the signal, ultimately generating a visualized nerve signal that meets the desired visualization characteristics.

[0033] Step S400: According to the constraint implicit representation array and the signal implicit representation array, the noise removal of the noisy nerve signal is performed to obtain a noise cleaned transitional signal, and the noise cleaned transitional signal is restored to a visualized nerve signal that meets the target visualization characteristics; the sampling point number of the original living nerve signal and the visualized nerve signal is consistent.

[0034] In step S400, the goal of noise removal is to retain the main features of the signal while removing as much noise as possible. This process can rely on the denoising component in the target machine learning algorithm or related algorithms. Taking a deep learning model as an example, the computer system can use a neural network containing a denoising layer to implement noise removal. The denoising layer can be a convolutional layer, an autoencoder layer, or a specially designed denoising network layer. In the training process, these layers learn how to recover the clean signal from the noisy signal.

[0035] In implementation, the computer system provides the constraint implicit representation array and the signal implicit representation array as conditional inputs or additional information to the denoising network. These implicit representation arrays are integrated into the denoising process in some way (e.g., concatenation, weighted sum, or conditional input) to guide the network on how to remove noise while preserving information related to the visualization constraints and signal features.

[0036] For example, assume the denoising network is a conditional generative adversarial network (cGAN) based architecture, where the generator (G) is responsible for generating a noise-cleaned transitional signal from the noisy neural signal, and the discriminator (D) is responsible for distinguishing whether the generated signal is close enough to the real one and complies with the visualization constraints. In this case, the constraint implicit representation array and the signal implicit representation array can be provided as conditional inputs to the generator, influencing the style and content of the generated signal.

[0037] Through the denoising process, the computer system generates a noise-cleaned transitional signal. This signal is the result of the noisy neural signal after noise removal, which removes most of the noise components while preserving the main features of the signal and the user-desired visualization features.

[0038] The generation of the noise-cleaned transitional signal is an iterative or optimization process, which may involve multiple forward and backward propagations. In each iteration, the denoising network generates a candidate noise-cleaned signal based on the current input (noisy neural signal, constraint implicit representation array, signal implicit representation array) and the current network parameters. Then, the computer system evaluates the quality of this candidate signal through some loss function (e.g., mean square error loss, adversarial loss, feature matching loss, etc.), and updates the network parameters according to the gradient information of the loss function.

[0039] After obtaining the noise-cleaned transitional signal, the computer system further restores it to a visualized neural signal that complies with the target visualization features. This process may involve signal reconstruction, feature mapping, color mapping, etc. First, the computer system converts the noise-cleaned transitional signal from the implicit space back to the original signal space. This can be achieved through a decoder or inverse transformation process, which is the opposite of the feature embedding process in step S100. The decoder or inverse transformation function gradually recovers the sampling points and feature dimensions of the original signal based on the information in the noise-cleaned transitional signal. Then, the computer system further processes the restored signal according to the visualization constraint parameters, such as applying color mapping, adjusting dynamic effects, etc., to generate the final visualized neural signal. For example, if the visualization constraint parameters specify a color mapping scheme, the computer system will select appropriate colors for rendering based on the signal intensity or feature values in the noise-cleaned transitional signal.

[0040] For example, assume the noisy neural signal is a one-dimensional array of length N, representing noisy neural activity collected at T time points. The constraint implicit representation array is a vector of length M, containing visualization constraint information such as color mapping, dynamic effects, etc.; the signal implicit representation array is a vector of length K, summarizing the main characteristics of the signal.

[0041] During the denoising process, the computer system uses a cGAN-based denoising network. The generator G receives the noisy neural signal, the constraint implicit representation array, and the signal implicit representation array as input, and generates a noise-cleared signal of length N. The discriminator D is responsible for evaluating the quality of the generated signal and optimizing the generator G through adversarial training.

[0042] During the training process, the computer system defines a composite loss function L, which combines the mean square error loss (used to evaluate the accuracy of signal reconstruction) and the adversarial loss (used to encourage the generation of signals that meet the visualization constraints):

[0043] L=λ1·MSE(x clean ,G(x noisy ,c,z))+λ2·log(1-D(G(x noisy ,c,z)));

[0044] Where x clean is the true clean signal (available during the training phase), x noisy is the noisy neural signal, c is the constraint implicit representation array, z is the signal implicit representation array, and λ1 and λ2 are hyperparameters that balance the weights of different loss terms.

[0045] By minimizing the loss function L, the computer system continuously optimizes the denoising network until it generates a visualized neural signal that meets the target visualization characteristics. Ultimately, this signal not only removes the noise in the original noisy neural signal, but also accurately reflects the user's visualization needs, providing medical professionals with intuitive and clear neural activity monitoring results.

[0046] As an implementation, in step S100, an original living body neural signal is obtained, and signal feature abstraction is performed on the original living body neural signal to obtain an abstracted signal. Specifically, it can include:

[0047] Step S110: Obtain waveform feature values corresponding to x action potentials in the original living body neural signal, and perform smoothing operation on the original living body neural signal according to the waveform feature values to obtain a smoothed signal; x≥1;

[0048] Step S120: Obtain the time axis change rate and amplitude axis change rate corresponding to the x action potentials in the smoothed signal respectively, determine the intensity and angle corresponding to the x action potentials in the smoothed signal according to the time axis change rate and amplitude axis change rate, perform local extreme value detection on the smoothed signal based on the intensity and angle, and obtain an extreme value nerve signal;

[0049] Step S130: Generate an abstract signal according to the time axis change rate and amplitude axis change rate corresponding to the x action potentials in the extreme value nerve signal.

[0050] Step S100 is the basis of the entire implantable medical chip-based regulation and monitoring information visualization method, and its core task is to extract key features from the original living nerve signal, providing basic data for subsequent signal processing and visualization. In this embodiment, step S100 is refined into three specific steps: S110, S120 and S130, each step carrying a specific data processing task. The following is a detailed explanation of these steps, combined with examples for illustration.

[0051] In step S110, the computer system captures the waveform features of action potentials from the original living nerve signal, and performs smoothing processing on the signal based on these features to reduce noise and interference and improve signal quality.

[0052] Specifically, the computer system first analyzes the original living nerve signal, which can be a series of voltage values changing over time, reflecting the activity of neurons. By setting appropriate threshold values, the system can identify action potentials in the signal, i.e. those parts of the voltage values that deviate significantly from the baseline level. These action potentials are a direct manifestation of neuron firing and are crucial for subsequent feature extraction. For each identified action potential, the computer system calculates its waveform feature values. These feature values may include peak amplitude, half-width (i.e. the time required for the action potential to rise from the baseline to the peak and then fall back to the baseline), rise time, fall time, etc. These feature values quantify the shape and size of the action potential and are an important basis for subsequent smoothing processing.

[0053] Based on the calculated waveform feature values, the computer system performs smoothing processing on the original live neural signal. The purpose of smoothing processing is to reduce high-frequency noise and random fluctuations in the signal, making the signal smoother and facilitating subsequent feature extraction and analysis. Common smoothing methods include moving average filtering, median filtering, Gaussian filtering, etc. For example, a suitable window size (such as 5 or 7 sampling points) can be selected, and the signal values within the window are averaged or median calculated to obtain the smoothed signal values. This process will iterate through the entire original signal until a complete smoothed signal is generated. For example, assume that the original live neural signal is an array containing 1000 sampling points, with a sampling rate of 10 kHz. By setting the threshold to -50 μV (relative to the baseline level), the system identifies 50 action potentials. For each action potential, the system calculates its peak amplitude, half-width, rise time, and fall time, and other waveform feature values. Subsequently, a moving average filter with a window size of 7 sampling points is selected to smooth the original signal. The processed smoothed signal will be used as input for the next step of analysis.

[0054] After obtaining the smoothed signal, step S120 further analyzes the time gradient and amplitude gradient (i.e., the rate of change of the time axis and the rate of change of the amplitude axis) in the signal, based on which the intensity and angle of the action potential are determined, and local extremum detection is performed, so as to more accurately locate the features of the action potential.

[0055] The computer system first calculates the time gradient and amplitude gradient of each sampling point in the smoothed signal. The time gradient reflects the rate of change of the signal value with respect to time, while the amplitude gradient reflects the rate of change of the signal value with respect to amplitude. These gradient information is crucial for understanding the dynamic characteristics and local features of the signal. Gradient calculation can be achieved through difference method, i.e., calculating the difference between adjacent sampling points to approximate the gradient value.

[0056] Based on the calculated gradient information, the computer system can further determine the intensity and angle of each action potential. Here, "intensity" can be understood as the amplitude or energy level of the action potential, while "angle" may refer to the angle between the gradient vector and a certain reference direction, used to describe the shape or directional characteristics of the action potential. However, in practical applications, the term "angle" may not be directly applicable to the field of neural signal processing, and here it is more to illustrate the multidimensionality of gradient information in describing signal features. In actual operation, more attention may be paid to the absolute value or directionality (such as positive or negative) of the gradient, rather than the specific angle value.

[0057] Next, using the gradient information and intensity information, the computer system performs local extremum detection. Local extremum refers to a point in the signal whose value is greater than its neighboring points (for maxima) or less than its neighboring points (for minima). In neural signal processing, action potentials can manifest as local maxima in the signal. By comparing the value of each sample point with its neighboring points, the system can identify all local maxima points, which correspond to the peak positions of action potentials.

[0058] For example, continuing with the smoothed signal in step S110. Suppose the smoothed signal is an array of length 1000. The computer system first calculates the temporal gradient and amplitude gradient for each sample point in the array. Then, based on the gradient information, it determines the "intensity" (simplified here as peak amplitude) and "directionality" (judged by comparing the signs of neighboring gradient values) of each action potential. Next, the system performs local extremum detection by comparing each sample point with its two neighboring points to identify local maxima points. Finally, the system obtains an array of indices containing the peak positions of action potentials, which correspond to the local maxima points in the smoothed signal.

[0059] In step S130, the computer system generates an abstracted signal based on the temporal gradient and amplitude gradient information in the extremum neural signal (i.e., the signal containing local extremum points). The abstracted signal is a simplified representation of the original signal that retains the main features while ignoring the minor details.

[0060] Based on the extremum neural signal, the computer system further analyzes the temporal gradient and amplitude gradient of each action potential. These gradient information reflects the dynamic characteristics and shape features of action potentials. However, due to the complexity and high dimensionality of the original signal, directly processing all gradient information can be both time-consuming and inefficient. Therefore, the system needs to select those features that are most valuable for subsequent analysis and perform dimensionality reduction to reduce computational load. There are many methods for dimensionality reduction, such as Principal Component Analysis (PCA), Linear Discriminant Analysis (LDA), t-SNE, etc. However, in the field of neural signal processing, since the features of action potentials are relatively simple and clear (such as peak amplitude, half-width, etc.), it may be more inclined to use feature selection-based methods rather than complex dimensionality reduction algorithms. The system can select a few key features (such as peak amplitude and half-width) to represent each action potential.

[0061] After selecting the key features, the computer system combines these features into a new array or vector to represent each action potential. This new representation is the abstracted signal. Compared with the original live neural signal, the abstracted signal has lower dimensionality and simpler structure, but it still contains enough information to reflect the main features of the original signal.

[0062] The specific form of the abstracted signal depends on the selected features and the requirements of the subsequent analysis. For example, if the system is mainly interested in the amplitude information of the action potentials without caring about their shape details, the abstracted signal can simply be an array containing the peak amplitude of all the action potentials. If the system also needs to consider the duration or shape features of the action potentials, the abstracted signal can contain more dimensions of features. Optionally, to improve the efficiency and stability of the subsequent processing, the computer system can perform standardization or normalization on the abstracted signal. Standardization is to scale the signal values into a certain fixed range (such as 0 to 1 or -1 to 1), while normalization is to adjust the mean and variance of the signal to conform to a certain statistical distribution (such as normal distribution). These processing steps help to reduce the impact of the differences between different signals on the subsequent analysis.

[0063] For example, suppose in step S120, the computer system has identified all the local maximum points in the smoothed signal and calculated the time gradient and amplitude gradient corresponding to these points. Now, the system selects the peak amplitude and half-width as the key features to represent each action potential. For each action potential, the system extracts its peak amplitude and half-width values and combines these values into a two-dimensional array (or vector). Assuming that a total of 50 action potentials are identified, the final abstracted signal will be a 50x2 array or list set.

[0064] Further, for the convenience of subsequent processing, the system can perform normalization on this abstracted signal. For example, the peak amplitude can be normalized to the range of 0 to 1 (achieved by subtracting the minimum value and dividing by the difference between the maximum and minimum values), while the half-width may not need to be normalized (as the value of the half-width can vary within a relatively stable range). The normalized abstracted signal will be more suitable for the input of a machine learning model or further signal processing tasks.

[0065] As an implementation, step S130, generating an abstracted signal according to the time axis change rate and the amplitude axis change rate corresponding to the x action potentials in the extreme value neural signal, can include:

[0066] Step S131: obtaining a first reference value and a second reference value, determining the change rate corresponding to the x action potentials in the extreme value neural signal according to the time axis change rate and the amplitude axis change rate corresponding to the x action potentials in the extreme value neural signal;

[0067] Step S132: determining a significant action potential in the x action potentials of the extreme value neural signal, the change rate of the significant action potential being greater than or equal to the first reference value;

[0068] Step S133: Determine ambiguous action potentials in the x action potentials of the extreme neural signal, the change rate of the ambiguous action potential is greater than or equal to the second reference value and less than the first reference value;

[0069] Step S134: Determine the ambiguous action potential adjacent to the significant action potential as a candidate action potential, and generate an abstract signal according to the significant action potential and the candidate action potential.

[0070] In step S131, the computer system sets two reference values, the first reference value and the second reference value, which will be used for subsequent classification of action potentials. The selection of the reference value can be based on prior knowledge, experimental data or expert opinion, aiming to distinguish action potentials of different significance.

[0071] Specifically, the first reference value and the second reference value can be preset according to experimental design or historical data analysis results. For example, in the setting of the first reference value, a certain percentile (such as 90%) of the absolute value of the time gradient or amplitude gradient of the action potential can be considered as a threshold, and the action potential exceeding this threshold is considered to have higher significance. Similarly, the second reference value can be set to a lower percentile (such as 70%) to distinguish those ambiguous action potentials whose change rate is not as high as the significant action potential, but is still higher than the random noise level. For each action potential in the extreme neural signal, the computer system calculates its comprehensive change rate according to its time axis change rate and amplitude axis change rate (i.e. time gradient and amplitude gradient). Here, the comprehensive change rate can be a certain weighted sum or Euclidean distance of the time gradient and amplitude gradient, depending on the relative importance of the action potential features. For example, if the time gradient is more important for distinguishing action potentials, it can be given a higher weight.

[0072] For example, assume that the extreme neural signal contains 100 action potentials, and the time gradient and amplitude gradient of each action potential have been calculated in the previous step. The computer system sets the first reference value to be greater than 0.5V / ms (assuming the unit) and the second reference value to be 0.2V / ms. For each action potential, the weighted sum of its time gradient and amplitude gradient is calculated as the comprehensive change rate, and the weight is determined according to the experimental design or expert opinion, for example, the time gradient weight is 0.7 and the amplitude gradient weight is 0.3.

[0073] After determining the reference values and calculating the change rates, in step S132, the computer system then identifies those action potentials whose change rates exceed the first reference value, i.e. significant action potentials.

[0074] In particular, for each action potential in the extreme neural signal, the computer system compares its integrated rate of change to a first reference value. If the integrated rate of change of the action potential is greater than or equal to the first reference value, the action potential is labeled as a significant action potential. All segments of the signal that are labeled as significant action potentials are stored for later processing. These significant action potentials can represent the most significant, most meaningful portion of the neural activity. For example, continuing the previous example, the computer system iterates through the 100 action potentials and labels as significant action potentials those action potentials with an integrated rate of change greater than or equal to 0.5 V / ms. Assuming that 20 action potentials satisfy this condition, these action potentials are stored in a dedicated array for later analysis.

[0075] After identifying the significant action potentials, the computer system determines those action potentials whose rates of change are between the first reference value and a second reference value, i.e., ambiguous action potentials. These action potentials, while not as significant as the significant action potentials, can still contain useful information.

[0076] In particular, for the remaining portion of the extreme neural signal that is not labeled as a significant action potential, the computer system performs a comparison filter again. This time, the integrated rate of change of each action potential is compared to a second reference value. If the integrated rate of change of the action potential is greater than or equal to the second reference value and less than the first reference value, the action potential is labeled as an ambiguous action potential.

[0077] All segments of the signal that are labeled as ambiguous action potentials are likewise stored. These ambiguous action potentials can be important in some cases to understanding the complexity of the neural activity. For example, in the previous example, the computer system continues to iterate through the remaining 80 segments of the signal that are not labeled as significant action potentials. Action potentials with an integrated rate of change between 0.2 V / ms and 0.5 V / ms are labeled as ambiguous action potentials. Assuming that 30 action potentials satisfy this condition, these action potentials are stored in another array.

[0078] After determining the significant action potentials and the ambiguous action potentials, the computer system needs to generate an abstracted signal from these action potentials. This process can involve merging and simplifying the significant action potentials and the ambiguous action potentials adjacent to them.

[0079] In particular, for each significant action potential, the computer system checks whether there are ambiguous action potentials adjacent to it. If there are, these ambiguous action potentials are considered as candidate action potentials. The purpose of this step is to preserve the information of ambiguous action potentials that can be functionally associated with the significant action potentials.

[0080] For each significant action potential and its set of candidate action potentials, the computer system selects a series of key features to represent them. These features can include peak amplitude, half-width, rise time, fall time, and relative position with respect to the significant action potential, etc. Then, based on these features, the system generates an abstracted representation to replace the original set of action potentials. This abstracted representation can be a vector or an array containing the key feature values.

[0081] Finally, the computer system combines all the significant action potentials and their corresponding abstracted representations to construct the final abstracted signal. This signal is a lower-dimensional representation that contains the main features of the original signal, facilitating subsequent processing and analysis.

[0082] For example, in the previous example, the computer system iterates through each significant action potential and checks its neighboring action potentials for the presence of ambiguous action potentials. For each significant action potential and its set of candidate ambiguous action potentials (if any), the system selects peak amplitude, half-width, and relative position as key features. Then, for each significant action potential, it generates a vector containing these feature values as an abstracted representation. Finally, these vectors are combined in order to form a new array or list as the abstracted signal output.

[0083] Suppose there are 20 significant action potentials identified, each with an average of 1 to 2 candidate ambiguous action potentials adjacent to it. For each significant action potential and its candidate set, the system generates a vector containing 3 feature values (peak amplitude, half-width, relative position index). The final output abstracted signal will be an array or list of 60 to 120 feature values, depending on the number of candidate ambiguous action potentials.

[0084] As an implementation, in step S100, the feature embedding of the original living body neural signal to obtain the first signal feature representation can include:

[0085] Step S140: loading the original living body neural signal into a target machine learning algorithm; the target machine learning algorithm includes a latent space mapping branch component, and the latent space mapping branch component includes a feature representation operator;

[0086] Step S150: performing feature extraction on the original living body neural signal based on the feature representation operator to obtain a center array and a dispersion array of the original living body neural signal;

[0087] Step S160: randomly sampling in the center array and the dispersion array of the original living body neural signal to obtain a latent space center array and a latent space dispersion array, and generating the first signal feature representation according to the latent space center array and the latent space dispersion array.

[0088] In step S140, the computer system loads the collected raw bio- neural signals as input data into the target machine learning algorithm. The target machine learning algorithm is a pre-designed and trained model specifically for processing neural signal data. This algorithm contains multiple components, among which the latent space mapping branch component is the key part to complete the feature embedding task.

[0089] Before loading the raw signals, some preprocessing operations such as denoising, filtering, standardization, etc. can be performed to ensure the quality and consistency of the input data. However, the specific content of these preprocessing steps depends on the characteristics of the signal and the requirements of the algorithm, and may vary depending on the application scenario.

[0090] The appropriate machine learning algorithm is selected according to the task requirements. In the field of neural signal processing, commonly used algorithms include auto encoder, variational auto encoder (VAE), and variants of generative adversarial network (GAN) such as conditional GAN. These algorithms can learn effective feature representations from data through unsupervised learning or supervised learning.

[0091] The preprocessed raw bio- neural signals are loaded as input data into the target machine learning algorithm. This may involve formatting the data into a format acceptable to the algorithm and setting appropriate parameters and data flow.

[0092] For example, suppose a model based on variational auto encoder is chosen as the target machine learning algorithm. The model contains an encoder and a decoder, where the encoder is used to map the raw signal to the latent space (i.e. perform feature embedding), and the decoder attempts to reconstruct the raw signal from the latent space. When loading the raw bio- neural signals, format them as a two-dimensional array, where rows represent different time points and columns represent signal values recorded by different channels or electrodes. Then, pass this array as input to the encoder part of the variational auto encoder.

[0093] After the raw bio- neural signals are loaded into the target machine learning algorithm, feature representation operators are used to extract the features of the signals. Feature representation operators can be one or more neural network layers in the algorithm, which generate useful feature representations by learning the internal structure of the data.

[0094] Specifically, the raw signal is forward propagated through the feature representation operator. In this process, the signal data undergoes a series of linear transformations (such as fully connected layers or convolutional layers) and nonlinear activation functions (such as ReLU, sigmoid, etc.), gradually abstracting high-level features of the signal.

[0095] Under the action of the feature representation operator, the original signal is transformed into a series of feature vectors or feature maps. These features can include statistical information of the signal (such as mean, variance), time-frequency characteristics, spatial distribution, etc. In this embodiment, the central array (μ) and the dispersion array (σ) are mentioned, which refer to the mean and variance representation of the signal in the latent space, for subsequent probability distribution modeling.

[0096] For example, continuing with the variational autoencoder as an example. In the encoder, the original signal is first passed through one or more convolutional layers to extract local features (such as edges, textures, etc.), and then through a pooling layer to reduce the resolution of the feature map and increase the receptive field. Next, the feature map is converted to a one-dimensional vector through a fully connected layer, which is the preliminary representation of the central array (μ). At the same time, the encoder will also generate another one-dimensional vector as the preliminary representation of the dispersion array (σ), which describes the uncertainty or diversity of the features in the latent space. These two vectors will be used in the subsequent steps to generate the latent space central array and the latent space dispersion array.

[0097] After obtaining the central array (μ) and the dispersion array (σ) of the original living neural signal, the next step is to randomly extract (or called sample, decode) on the basis of these arrays to generate the first signal feature representation. This process can involve sampling the latent space distribution or generating feature vectors according to deterministic rules.

[0098] Specifically, in a generative model such as a variational autoencoder, the latent space can be modeled as a Gaussian distribution (or other types of probability distribution), whose mean and variance are given by the encoder (i.e. the central array μ and the dispersion array σ). In this step, one or more latent space points can be randomly sampled from this distribution as feature representations. This method can capture the diversity in the latent space and help generate diverse signal representations.

[0099] In some cases, in order to simplify the calculation or meet certain requirements, it is also possible to choose not to randomly sample but to directly use the central array (μ) as the feature representation (i.e. ignoring the information of the dispersion array σ). This method, although sacrificing diversity, can obtain a deterministic and repeatable feature representation.

[0100] Whether through random sampling or deterministic mapping, the latent space points obtained will be further processed to generate the first signal feature representation. This can involve mapping the latent space points back to the dimension space of the original signal or converting them into another form of feature vector.

[0101] For example, the process of latent space sampling is illustrated by taking the variational autoencoder as an example. Suppose that the center array μ and the dispersion array σ output by the encoder represent the mean and standard deviation vector of a Gaussian distribution in the latent space, respectively. To generate the first signal feature representation, a point z can be randomly sampled from this Gaussian distribution. Specifically, this can be achieved using the "reparameterization trick": first, a noise ε is sampled from the standard normal distribution N(0, I), and then z = μ + ε * exp(0.5 * σ 2 ) is calculated. Here, exp(0.5 * σ 2 ) is used to convert σ into the standard deviation form (because the encoder can output the logarithmic form of the variance, i.e., log(σ 2 )). The resulting z is the first signal feature representation, which is located in the latent space and captures the key features of the original living neural signal.

[0102] It should be noted that, in actual applications, σ can be constrained to ensure its non-negativity (such as using a soft plus function, etc.) for stability and numerical convenience. In addition, techniques such as batch normalization and layer normalization can be used to optimize the generation process of the feature representation to speed up the training and inference process.

[0103] As an implementation, the first signal feature representation is generated in a latent space mapping branch component in the target machine learning algorithm, and the latent space mapping branch component includes a normalization operator and a pooling operator. Based on this, in step S200, the first signal feature representation is added with noise to obtain a noisy neural signal, which can include:

[0104] Step S210: Obtain a noise distribution and an iteration number s, and perform a normalization operation on the waveform feature values of the action potential in the first signal feature representation based on the normalization operator and the noise distribution; s ≥ 1;

[0105] Step S220: Based on the pooling operator and the iteration number s, the normalized first signal feature representation is added with noise s times to obtain a noisy neural signal, wherein the noise distribution represents the confidence distribution information of the added noise, and the iteration number s represents the degree of adding noise to the first signal feature representation according to the noise distribution.

[0106] Step S210: Obtain the noise distribution and the iteration number, and perform a normalization operation

[0107] In this step, the computer system obtains the noise distribution and the iteration number s, and uses these parameters to perform a normalization operation on the first signal feature representation, preparing for the subsequent noise addition process.

[0108] Noise distribution is a parameter that describes the type, intensity, and probability distribution of noise in time or space. In practical applications, the choice of noise distribution depends on various factors, such as application scenarios, signal characteristics, and desired simulation effects. Common noise distributions include Gaussian distribution, uniform distribution, Laplace distribution, etc. The parameters of noise distribution (such as mean and variance) can be obtained in various ways. One way is to determine it by analyzing historical data or prior knowledge. For example, if it is known that the noise level of neural signals under certain experimental conditions follows a Gaussian distribution with a mean of 0 and a variance of σ², these parameters can be directly used as the parameters of the noise distribution. Another method is to determine it through experimental design, that is, artificially introducing different types of noise during the experiment and observing its effect on signal quality, so as to select appropriate noise distribution and parameters.

[0109] The number of iterations s is a hyperparameter that controls the degree of noise addition. It determines the number of times or the intensity of noise addition to the first signal feature representation. The selection of the number of iterations s also depends on various factors, such as the signal-to-noise ratio of the signal, the complexity of the noise, and the tolerance of the subsequent processing steps to the noise, etc. In practical applications, the number of iterations s can be determined by cross-validation, grid search, and other optimization algorithms. Specifically, noise can be added to the first signal feature representation under different numbers of iterations, and the performance changes of subsequent processing steps (such as noise removal, signal restoration, etc.) can be observed. By comparing the performance indicators (such as signal reconstruction error, visualization effect, etc.) under different numbers of iterations, the number of iterations that makes the performance of subsequent processing steps optimal is selected as the value of s, which is not limited.

[0110] After determining the noise distribution and the number of iterations, the computer system normalizes the action potential waveform feature values in the first signal feature representation using the standardization operator. The purpose of standardization is to scale the feature values to a unified scale, so as to facilitate subsequent noise addition and signal processing.

[0111] In the standardization operation, the mean and standard deviation of all action potential waveform feature values in the first signal feature representation can be calculated first; then each feature value is subtracted from the mean and divided by the standard deviation to obtain the standardized feature value. The feature values processed in this way will have zero mean and unit variance, thus eliminating the scale difference between different features.

[0112] For example, assume that the first signal feature representation is a vector f containing multiple action potential waveform feature values, each representing a certain attribute of an action potential (such as peak amplitude, half-width, etc.). First, the computer system calculates the mean μ and standard deviation σ of the vector f. Then, for each feature value f i in the vector f, its standardized value f i '=(f i- μ) / σ. The normalized vector f' obtained after this processing will serve as input for the subsequent noise addition process.

[0113] At the same time, suppose that the noise distribution has been determined to be Gaussian N(0, σ²_noise) through some means such as analyzing historical data or experimental design, where σ²_noise is the variance of the noise. The number of iterations s is then determined to be 3 through an optimization algorithm such as cross-validation.

[0114] Based on step S210, the computer system performs noise addition operation on the normalized first signal feature representation using a pooling operator and a number of iterations s to generate a noisy neural signal.

[0115] The pooling operator is a downsampling module that aggregates feature values in local regions to reduce the number of parameters and computational load while preserving important feature information. In the noise addition process, the pooling operator can be used to control the distribution characteristics of noise in space or time. One possible implementation is to perform some form of weighted summation or convolution operation between the noise distribution and the first signal feature representation in space or time to achieve local or global noise addition.

[0116] Based on the normalized first signal feature representation obtained in step S210 and the noise distribution parameters (such as mean, variance, etc.), the computer system begins to perform iterative noise addition operations. In each iteration, a certain number of noise samples are randomly drawn from the noise distribution and added to the normalized first signal feature representation. Through multiple iterations, the strength and complexity of the noise can be gradually enhanced to more realistically simulate the noise conditions in the actual environment.

[0117] It should be noted that after each iteration, the signal after adding noise can be processed (such as re-normalization, limiting noise intensity, etc.) to ensure that it is still within a reasonable range and meets the requirements of subsequent processing steps. For example, continuing the previous example, suppose that in step S210, the normalized first signal feature representation vector f' and noise distribution parameters (such as Gaussian distribution N(0, σ²_noise)) have been obtained. Now enter step S220 to perform iterative noise addition operations.

[0118] First, perform the first iteration: randomly draw a noise vector n1 of the same length as vector f' from Gaussian distribution N(0, σ²_noise); then add noise vector n1 to vector f' to obtain the noise-added vector f'_1 = f' + n1. If you need to limit the noise intensity, you can scale the noise vector before adding (such as multiplying by a scaling factor less than 1).

[0119] A second and third iteration is then performed: the above process is repeated to generate noise vectors n2 and n3 respectively and add them to the vector f'_1 to obtain vectors f'_2 and f'_3 respectively. The final noisy neural signal obtained after three iterations can be represented as f'_3 or a vector that is further processed as needed.

[0120] In practical applications, the cumulative effect of noise can be considered, i.e. the noise intensity can gradually increase during multiple iterations, leading to signal distortion. In order to avoid this, the signal can be processed (such as limiting the noise intensity, adjusting the noise distribution parameters, etc.) after each iteration to control the cumulative effect of noise.

[0121] As an implementation, in step S300, mapping the visual constraint parameters into the constraint implicit representation array can include:

[0122] Step S310: loading the visual constraint parameters into the target machine learning algorithm, wherein the target machine learning algorithm includes a constraint embedding branch component; the constraint embedding branch component includes a feature embedding operator, a weight allocation operator and a fusion operator;

[0123] Step S320: performing feature embedding on the visual constraint parameters based on the feature embedding operator to obtain y constraint array representations; y>1;

[0124] Step S330: generating y constraint array representations corresponding to the focusing weight respectively based on the weight allocation operator, and adjusting the constraint array representation based on the focusing weight to obtain y to-be-used focusing arrays;

[0125] Step S340: performing weight fusion on the y to-be-used focusing arrays based on the fusion operator to obtain the constraint implicit representation array.

[0126] In the embodiments of the present application, step S300 converts the user-defined visual constraint parameters into an internal representation form that can be processed by the computer system, i.e. the constraint implicit representation array.

[0127] In step S310, the computer system loads the visual constraint parameters provided by the user or external system into the target machine learning algorithm. These visual constraint parameters can include but are not limited to color mapping schemes, dynamic effect settings, view angle selection, label style, etc., which together define the user's desired visual display effect.

[0128] The target machine learning algorithm is pre-designed and trained, and is a model specially used for processing visual constraint parameters and generating constraint implicit representation arrays. The algorithm can include multiple branch components, and the constraint embedding branch component is one of the core parts. The constraint embedding branch component is responsible for mapping the visual constraint parameters into the implicit space of the algorithm for subsequent processing.

[0129] The constraint embedding branch component is further subdivided into a feature embedding operator, a weight assignment operator, and a fusion operator. These operators work together to convert complex visualization constraint parameters into concise and efficient constraint implicit representation arrays. For example, suppose a user defines a set of visualization constraint parameters, which includes a color mapping scheme (mapping signal intensity to red, yellow, and green colors), a dynamic effect setting (color gradient display when signal intensity changes), a view angle selection (top-down perspective), and a label style (displaying specific numerical values at signal peaks). The computer system loads these visualization constraint parameters into the target machine learning algorithm in the form of structured data (e.g., a JSON object).

[0130] After the visualization constraint parameters are loaded into the target machine learning algorithm, they are subjected to feature embedding using the feature embedding operator. Feature embedding is the process of converting raw data (in this case, visualization constraint parameters) into a numerical representation that is more easily processed by algorithms.

[0131] The feature embedding operator maps the visualization constraint parameters into a high-dimensional feature space by analyzing their internal structure and semantic information. In this feature space, similar visualization constraint parameters will be mapped to close locations, facilitating subsequent weight assignment and fusion operations.

[0132] For example, taking the color mapping scheme as an example, the feature embedding operator may first convert color names (red, yellow, green) into RGB color codes (e.g., red corresponds to [255, 0, 0], yellow corresponds to [255, 255, 0], and green corresponds to [0, 255, 0]). Then, for the dynamic effect setting (e.g., color gradient), the feature embedding operator may further generate a series of intermediate color codes to represent each state in the color gradient process. These color codes will be embedded as part of the feature vector in the feature space.

[0133] For other visualization constraint parameters such as view angle selection and label style, the feature embedding operator will also use similar methods to process them, converting them into numerical feature vectors. Finally, for a given set of visualization constraint parameters, the feature embedding operator will generate y constraint array representations (where y is the number or category of constraint parameters), each containing the feature information of the constraint parameter. The feature embedding operator can be an encoder.

[0134] After obtaining y constraint array representations, step S330 uses the weight assignment operator to generate corresponding focus weights and adjust the constraint array representations. The purpose of focus adjustment is to adjust the influence weight of constraint parameters in subsequent processing according to their importance and relevance.

[0135] The weight assignment operator is an attention module that determines the importance of each constraint parameter by analyzing the similarities and differences between the constraint array representations. For constraint parameters that are more critical or highly correlated with other constraint parameters, the weight assignment operator assigns them higher weights; otherwise, it assigns them lower weights.

[0136] For example, suppose three constraint array representations are generated in step S320: color mapping scheme A, dynamic effect setting B, and view angle selection C. The weight assignment operator first analyzes the similarities and differences between these three constraint array representations. For example, if color mapping scheme A and dynamic effect setting B are highly correlated in visual effects (e.g., color gradient needs to rely on a specific color mapping), the weight assignment operator can assign them higher mutual weights.

[0137] Then, the weight assignment operator adjusts the focus of the constraint array representations according to these weight information. The specific operations may include weighted sum of feature values in each array representation, application of nonlinear transformation, or adjustment of the direction of feature vectors, etc. After focus adjustment, the y focus arrays to be used will highlight the key features and reduce redundant information.

[0138] After obtaining the y focus arrays to be used, step S340 uses a fusion operator to weight fuse these arrays to generate the final constraint implicit representation array. The purpose of weight fusion is to integrate the effects of multiple constraint parameters into a unified internal representation form.

[0139] The fusion operator is a computational network layer that generates the constraint implicit representation array by considering the weights of each focus array to be used and their interactions. For example, continuing the previous example, suppose three focus arrays are obtained in step S330: focused color mapping scheme A', focused dynamic effect setting B', and focused view angle selection C'. The fusion operator first reads the weight information of each focus array. These weights may be determined by the weight assignment operator in step S330 or obtained through other mechanisms (such as user input).

[0140] Then, the fusion operator performs fusion operations on the focus arrays according to these weight information. The specific operations include, for example, weighted sum of feature values in each focus array, application of specific fusion functions (such as weighted average, max pooling, etc.), or inference using deep learning models, etc. The final generated constraint implicit representation array will be a comprehensive representation form that contains the effects of all visualization constraint parameters, which will serve as input data for subsequent steps (such as noise removal, signal restoration, and visualization rendering, etc.).

[0141] In practical applications, each sub-step in step S300 can be adjusted and optimized according to specific requirements and algorithm design. For example, the feature embedding operator can use different types of neural network structures (such as fully connected networks, convolutional neural networks, autoencoders, etc.) to achieve more efficient feature extraction; the weight assignment operator can use attention mechanisms or graph neural networks to capture complex relationships between constraint parameters; and the fusion operator can combine domain knowledge and optimization algorithms to ensure the accuracy and robustness of the fusion results.

[0142] As an implementation, the noise-added neural signal is obtained based on s times of noise addition to the first signal feature representation, and the first signal feature representation is obtained in a hidden space mapping branch component in the target machine learning algorithm, and the hidden space mapping branch component includes a residual operator and an interpolation operator. Based on this, step S400, according to the constraint implicit representation array and the signal implicit representation array, the noise-added neural signal is de-noised to obtain a noise-cleaning transition signal, which can include:

[0143] Step S410: combining the constraint implicit representation array and the signal implicit representation array into a guidance array based on the residual operator;

[0144] Step S420: first de-noising the noise-added neural signal based on the interpolation operator and the guidance array to obtain a de-noised signal G s-1 , and repeating the de-noising of the de-noised signal G s-1 s times until the s-th de-noised signal G0 obtained by the s-th de-noising is determined as the noise-cleaning transition signal.

[0145] In step S410, the computer system effectively fuses the constraint implicit representation array extracted from the user's visualization requirements and the signal implicit representation array extracted from the signal features to form a guidance array that guides the noise removal process. This fusion process relies on the residual operator in the hidden space mapping branch component, which is good at capturing and emphasizing the difference between the input data, so as to preserve the key information in the fusion process.

[0146] The residual operator can alleviate the gradient vanishing problem in deep network training, and optimize the network performance by learning the difference (residual) between the input and the output. In step S410, the role of the residual operator is slightly different, which is used to effectively fuse the two implicit representation arrays, the constraint implicit representation array and the signal implicit representation array. Specifically, the residual operator can combine the information in the two arrays through element-wise addition or more complex weighted sum to form a guidance array that guides the noise removal. This fusion method can ensure that the guidance array contains both user-defined visualization constraints and signal feature information.

[0147] For example, assume the constraint implicit representation array is a vector C representing a color mapping scheme, where each element corresponds to the color encoding of a different signal intensity interval; the signal implicit representation array is a vector S containing the encoding of the signal's main features, such as signal intensity, frequency, and other statistical properties. The computer system fuses these two vectors using a residual operator to form the guide array G_guide. The fusion process can be similar to C + λS, where λ is a weight factor balancing the importance of constraint information and signal features. By adjusting the value of λ, the system can optimize the fusion effect according to the actual application scenario.

[0148] In step S420, after obtaining the guide array, the computer system enters the core part of noise removal, that is, using the interpolation operator and the guide array to iteratively process the noisy neural signal, gradually removing the noise components in the signal. The role of the interpolation operator in this process is to intelligently interpolate or reconstruct the noisy signal according to the information provided by the guide array, to approximate the state of the original pure signal.

[0149] The interpolation operator can estimate the value of unknown data points from known data points (such as sampling points). In step S420, the interpolation operator intelligently interpolates or reconstructs each point in the noisy neural signal according to the constraint information and signal feature information in the guide array, to reduce the impact of noise. This interpolation process can be local (only considering neighboring points) or global (considering the feature distribution of the entire signal).

[0150] Since the noisy neural signal is obtained by adding s times of noise, in order to recover the pure state of the signal as much as possible, the computer system needs to perform s times of noise removal iterations on the signal. In each iteration, the system will use the interpolation operator and the current guide array to process the signal to generate a noise-removed signal after one iteration. As the number of iterations increases, the noise component gradually decreases, and the signal gradually approaches its pure state.

[0151] For example, assume the noisy neural signal is a one-dimensional array N with a length of L, and each element represents the signal intensity value (possibly containing noise) at a certain time point. In the first iteration, the computer system uses the interpolation operator and the guide array G_guide to process N to generate a noise-removed signal G_1. The processing process may involve local or global interpolation of each element in N according to the color mapping scheme and signal feature information in G_guide to reduce the impact of noise. Subsequently, the system takes G_1 as the new input and repeats the above process to generate G_2, G_3,..., until G_s. After the s-th iteration, G_s (denoted as G0 to avoid confusion) is considered to be a noise-cleared signal, which contains less noise components and retains the main features of the signal and the user's visualization constraints.

[0152] It is worth noting that during the iteration process, the guide array G guide can remain unchanged (i.e., the same guide information is used for each iteration), or it can be dynamically adjusted according to the iteration results (e.g., regenerated according to the characteristics of the current noise removal signal). In addition, the specific implementation of the interpolation operator can also vary depending on the application scenario, such as using linear interpolation, spline interpolation, machine learning-based interpolation methods, etc.

[0153] During the execution of step S400, the computer system can evaluate the effect of noise removal by monitoring the changes in these performance indicators, and adjust the algorithm parameters or optimize the algorithm implementation as needed to improve performance.

[0154] As an implementation, step S420, based on the interpolation operator and the guide array, performs the first noise removal on the noisy neural signal to obtain a noise removal signal G s-1 , including:

[0155] Step S421: Based on the interpolation operator, obtain the center array and the dispersion array of the noisy neural signal, and randomly extract in the center array and the dispersion array of the noisy neural signal to obtain the iteration center array and the iteration dispersion array;

[0156] Step S422: Obtain a random array of noise distribution, adjust the iteration center array and the iteration dispersion array according to the guide array and the random array, and obtain a noise removal signal G s-1 .

[0157] In step S421, the computer system analyzes the noisy neural signal using the interpolation operator to obtain the center array (which can represent the main trend or mean of the signal) and the dispersion array (which describes the fluctuation or variance of the signal relative to the center). Then, the system randomly extracts (or samples, perturbs) on the basis of these arrays to generate the iteration center array and the iteration dispersion array for iteration processing.

[0158] For example, suppose the noisy neural signal has been transformed into a domain suitable for processing by the interpolation operator (e.g., transformed into the frequency domain by Fourier transform). In this domain, the interpolation operator can identify the principal frequency components of the signal (corresponding to the central array) and the fluctuations around these components (corresponding to the dispersion array). The system then generates the iterative central array and the iterative dispersion array by randomly sampling on these components (e.g., slightly adjusting or perturbing certain frequency components in the frequency domain). Assuming the interpolation operator is actually a parametric function or model (e.g., a neural network layer, a Gaussian process, etc.), it generates the central array and the dispersion array by learning the statistical properties of the noisy neural signal. These arrays define a region of probability distribution of the signal in the latent space, and the random sampling is performed within this region to generate the new signal representation (i.e., the iterative central array and the iterative dispersion array).

[0159] For example, suppose the noisy neural signal has been mapped into a high-dimensional latent space and forms a distribution centered at a certain point with a certain dispersion in this space. The interpolation operator generates the central array and the dispersion array by analyzing the geometric properties of this distribution in the latent space (e.g., mean vector, covariance matrix, etc.). The system then performs random sampling within this distribution region (e.g., by sampling in a Gaussian distribution) to generate the iterative central array and the iterative dispersion array. These arrays will then be used to generate the noise-removed signal G s-1 .

[0160] After obtaining the iterative central array and the iterative dispersion array, step S422 further adjusts these iterative arrays in combination with a random array of the noise distribution and a guidance array extracted from the user's visualization needs to generate the noise-removed signal G s-1 .

[0161] The random array of the noise distribution is an array containing random noise samples that reflect the statistical properties of the noise components in the noisy neural signal. This array can be generated in various ways, such as sampling from a pre-defined noise model (e.g., Gaussian noise, salt and pepper noise, etc.). In practical applications, the random array of the noise distribution can be closely related to the generation process of the noisy neural signal, i.e., it can be generated based on the noise model and parameters used when adding noise in step S200.

[0162] The guidance array is an implicit representation array extracted from user-defined visualization constraint parameters. It contains the visualization feature information desired by the user (such as color mapping schemes, dynamic effect settings, etc.). During noise removal, the guidance array serves as a global reference framework, guiding the adjustment direction of the iteration center array and the iteration distribution array. By continuously aligning the iteration array with the guidance array and optimizing the matching degree between them (possibly measured by some form of loss function), the system can progressively generate a noise removal signal that meets the user's expectations.

[0163] In step S422, the computer system acquires a random array and a guidance array of the noise distribution. Then, it combines these arrays with the iteration center array and the iteration dispersion array for comprehensive adjustment. This adjustment process may involve various operations such as weighted summation, nonlinear transformation, and optimization algorithm iteration, depending on the algorithm design and application scenario requirements.

[0164] One approach involves initially adjusting the iterative center array based on a guiding array to ensure it roughly matches the user's desired visualization characteristics; then, fine-tuning the iterative dispersion array based on a random array of noise distribution to introduce some randomness and uncertainty; finally, combining the adjusted iterative center array and iterative dispersion array to generate the noise-removed signal G. s-1 .

[0165] In step S422, the computer system first adjusts the iteration center array according to the guide array. For example, if the value of an element in the guide array is larger than the corresponding value in the iteration center array, the system can increase the value of that element in the iteration center array to make it closer to the value in the guide array.

[0166] The system then fine-tunes the iteratively distributed array using a random array of noise distributions. This fine-tuning process may involve combining elements from the random array in a certain way (such as through weighted summation) with elements from the iteratively distributed array to introduce randomness. However, since it is desirable to preserve the naturalness and diversity of the signal while removing noise, this fine-tuning should be moderate and should not destroy the main characteristics of the signal.

[0167] Finally, the system combines the adjusted iterative center array and iterative dispersion array to generate the noise-removed signal Gs-1. This signal should be closer to the user's desired position represented by the guidance array in the latent space than the original noisy neural signal, while also retaining a certain degree of randomness and naturalness to avoid over-smoothing or distortion.

[0168] As an implementation, the noise-cleaning transition signal is generated in a hidden space mapping branch component in the target machine learning algorithm, which includes a feature representation operator and an interpolation operator. Based on this, in step S400, the noise-cleaning transition signal is restored into a visual neural signal conforming to the target visual feature, including:

[0169] Step S430: feature extraction of the noise-cleaning transition signal based on the feature representation operator, to obtain a center array and a dispersion array of the noise-cleaning transition signal, random extraction in the center array to obtain a hidden space center array, and random extraction in the dispersion array to obtain a hidden space dispersion array;

[0170] Step S440: generating a visual neural signal conforming to the target visual feature based on the interpolation operator, the hidden space center array, and the hidden space dispersion array.

[0171] In step S430, the computer system uses the feature representation operator to deeply analyze the noise-cleaning transition signal to extract its key features and map these features into the hidden space. Through feature extraction and hidden space parameter extraction, the system can more accurately control the generation process of the signal, thereby ensuring that the finally generated visual neural signal not only conforms to the target visual feature, but also retains important information of the original signal.

[0172] In step S430, the feature representation operator extracts a center array and a dispersion array from the noise-cleaning transition signal. These two arrays respectively represent the main trend (or mean) of the signal and the fluctuation (or variance) around the main trend, which together describe the distribution characteristics of the signal in the hidden space.

[0173] After extracting the center array and the dispersion array, the system further randomly extracts (or samples, perturbs) in these arrays to generate a hidden space center array and a hidden space dispersion array. This process increases the diversity and randomness of the signal, which helps to avoid overfitting and improve the generalization ability of the model.

[0174] For example, assume that the noise-cleaning transition signal has been mapped into a high-dimensional hidden space, and in this space, a distribution centered at a certain point with a certain dispersion is formed. The feature representation operator generates a center array and a dispersion array by analyzing the geometric characteristics (such as mean, variance, shape, etc.) of this distribution in the hidden space. These arrays represent the position and directionality of the signal in the hidden space in the form of numerical vectors.

[0175] To generate the visualized neural signal that conforms to the target visualization feature, random sampling is performed in these arrays. Specifically, the system can fine-tune the elements of the center array within a small range (e.g., add a small random perturbation) to explore the subtle changes of the signal feature in the latent space; meanwhile, the system can scale or rotate the elements of the scatter array to adjust the dispersion or direction of the signal in the latent space. These operations collectively constitute the sampling process of the latent space parameters.

[0176] After obtaining the latent space center array and the latent space scatter array, step S440 utilizes an interpolation operator to generate the visualized neural signal that conforms to the target visualization feature based on these parameters. The interpolation operator maps the parameters in the latent space back to the original signal space (or the user-defined visualization space), thereby generating a visually presentable signal representation.

[0177] Under the action of the interpolation operator, the latent space center array and the latent space scatter array are mapped back to the original signal space (or the visualization space), generating a preliminary representation of the visualized neural signal. However, this preliminary representation may still need to go through a series of post-processing steps (such as smoothing, normalization, color mapping, etc.) before it can be finally presented to the user. These post-processing steps aim to further optimize the visual effect of the signal and improve its interpretability.

[0178] To conform to the target visualization feature, the system fully considers the user's definition and expectations in the process of generating the visualized neural signal. For example, if the user defines a specific color mapping scheme to distinguish different intensity regions of the signal, the system will apply this mapping scheme when generating the signal to assign corresponding color values to each time point or spatial point. Similarly, if the user expects to see dynamic change effects (such as flickering, gradient, etc.) in the visualization, the system will also introduce corresponding dynamic effect parameters when generating the signal to achieve this expectation.

[0179] For example, continuing with the previous example, suppose the latent space center array and the latent space scatter array have been obtained. Now the system utilizes an interpolation operator to map these parameters back to the original signal space (or the visualization space) to generate the visualized neural signal. Specifically, the interpolation operator can determine the main position of the signal on the time axis or spatial axis according to each element value in the latent space center array; at the same time, it can adjust the fluctuation range or shape change of the signal at these positions according to the element values in the latent space scatter array.

[0180] If the user wants to highlight areas with high signal strength in the visualization, the system can apply an enhancement filter to amplify the signal values ​​in these areas; if the user defines a specific color mapping scheme, the system will assign corresponding color values ​​to each time point or spatial point according to the magnitude of the signal value to form an intuitive color-coded map; if the user wants to see the dynamic changes of the signal in the visualization, the system can introduce time-dimension variation parameters (such as periodic changes, random jitter, etc.) when generating the signal to achieve the display of dynamic effects.

[0181] Finally, through the optimization of the mapping of interpolation operators and post-processing steps, a visual neural signal that conforms to the target visualization characteristics was successfully generated. This signal not only removes most of the noise components in the original noisy neural signal, but also retains the main features of the signal and user-defined visualization constraints, providing medical professionals with intuitive and clear results of neural activity monitoring.

[0182] This application also provides a training process for the target machine learning algorithm, which may specifically include the following steps:

[0183] Step S10: Obtain the learning sample library; the learning sample library includes original live neural signal samples, visualized neural signal samples, and visualized neural signal parameters; the visualized neural signal parameters are used to describe the visualization features of the visualization results; the original live neural signal samples and the visualized neural signal samples have the same number of sampling points.

[0184] The learning sample library is a collection of samples used to train the machine learning algorithm. The original live neural signal samples are directly derived from real-time neural signal data acquired by implanted medical chips. Each sample is a series of voltage values ​​that change over time, representing the activity of neurons within a specific time period. For example, a raw live neural signal sample might be a one-dimensional array containing 1000 sampling points at a sampling rate of 1000Hz, covering one second of neural activity. Visualized neural signal samples are the results of some visualization processing of the raw signals. They have been transformed according to the user's visualization needs (such as color mapping, dynamic effects, etc.). Each visualized neural signal sample not only contains the processed signal values ​​but may also contain visualization-related metadata (such as color encoding schemes, view parameters, etc.). Importantly, these samples strictly correspond to the original signal samples in time, meaning they have the same number of sampling points, ensuring data consistency and comparability.

[0185] Visualization neural signal parameters are used to describe specific characteristics of the visualization display results, such as color mapping rules, dynamic effect parameters, view scaling ratios, etc. They exist in the form of structured data (such as JSON objects, XML files, etc.), which is convenient for computer systems to parse and process. For example, a visualization neural signal parameter may specify a color mapping rule of "displaying red when the signal intensity is greater than a threshold value, and displaying blue when the signal intensity is less than the threshold value".

[0186] The computer system obtains these learning samples through various ways. On the one hand, it can extract historical data from existing databases as samples; on the other hand, it can also cooperate with medical professionals to generate new samples through actual collection and labeling. During the acquisition process, the system will strictly screen and preprocess the samples to ensure that their quality meets the training requirements.

[0187] Before training the target machine learning algorithm, it is crucial to ensure the diversity and representativeness of the learning sample library. This means that the sample library should contain data from different patients, different time periods, and different physiological states, so that the algorithm can learn more extensive and general feature representations and mapping rules. At the same time, accurate labeling and classification of samples are also essential steps, which will directly affect the training effect and generalization ability of the algorithm.

[0188] Step S20: Abstracting signal features from the original living neural signal samples to obtain abstracted signal samples, and loading the original living neural signal samples and the abstracted signal samples into the initial machine learning algorithm.

[0189] In step S20, the computer system deeply analyzes the original living neural signal samples in the learning sample library to extract and abstract the key features in the signal. These features may include signal waveform features, frequency components, energy distributions, etc., which collectively describe the inherent laws and characteristics of neural activity.

[0190] The computer system can first preprocess the original living neural signal samples to remove noise, standardize signal amplitude, adjust sampling rate, etc. This step helps to reduce interference factors in the subsequent feature extraction process, improving the accuracy and stability of the features. Then, the system uses feature extraction algorithms (such as time-frequency analysis, wavelet transform, principal component analysis, etc.) to extract key features from the preprocessed signals. These features can be statistical quantities (such as mean, variance, skewness, kurtosis, etc.), waveform parameters (such as peak value, half-width, rise time, etc.), frequency domain features (such as spectral energy distribution, dominant frequency, etc.), or time-frequency features (such as instantaneous frequency, instantaneous energy, etc.).

[0191] After extracting a large number of original features, the system needs to further abstract these features (refer to the process of step S100 described above).

[0192] Step S30: Perform feature embedding on the original living body neural signal sample to obtain a first signal feature representation sample, and perform feature embedding on the abstracted signal sample to obtain a second signal feature representation sample; the number of sampling points of the original living body neural signal sample is greater than the number of sampling points of the first signal feature representation sample and the second signal feature representation sample.

[0193] In step S30, the computer system converts the original living body neural signal sample and the abstracted signal sample into lower-dimensional feature representations, i.e., the first signal feature representation sample and the second signal feature representation sample, using feature embedding technology. This process can refer to the description of step S100 described above.

[0194] For the original living body neural signal sample, the computer system first inputs it into a pre-trained feature embedding model, which is, for example, a deep neural network (such as an autoencoder, a convolutional neural network, etc.), to map the high-dimensional original signal into a low-dimensional feature space by learning the internal structure of the signal and the correlation between features. Specifically, each layer in the model transforms the input signal to extract features at different levels, and finally outputs a low-dimensional feature vector as the first signal feature representation sample.

[0195] For the abstracted signal sample, since it is already an abstraction and simplification of the original signal, the feature embedding process can be relatively simple. The computer system can directly input the abstracted signal sample into the same feature embedding model (or a model specially designed for abstracted signals) for further feature extraction and dimension reduction. Since the abstracted signal sample already contains the main feature information of the original signal, this step is mainly to further refine and compress these features.

[0196] For the abstracted signal sample (assuming it is a 10-dimensional feature vector), the computer system can choose to use a lighter feature embedding model (such as a single-layer fully connected network) for further processing. This model will fine-tune the abstracted signal sample to generate a lower-dimensional second signal feature representation sample (e.g., 5-dimensional). This feature representation not only contains the main feature information of the abstracted signal, but also further reduces the redundancy of the data and the computational complexity.

[0197] Through the feature embedding process, the number of sampling points of the original live neural signal sample is reduced from the original 1000 to 10 (or less) of the first signal feature representation sample, and the dimension of the abstracted signal sample is also reduced from the original possibly higher dimension to 5 dimensions (or less) of the second signal feature representation sample. The reduction of the number of sampling points not only reduces the storage and transmission cost of the data, but also improves the efficiency of subsequent processing and analysis.

[0198] Step S40: adding noise to the second signal feature representation sample to obtain a noisy neural signal sample.

[0199] Step S50: mapping the visualized neural signal parameter to a constraint implicit representation array, mapping the second signal feature representation sample to a signal implicit representation array, removing noise from the noisy neural signal sample according to the constraint implicit representation array and the signal implicit representation array, obtaining a noise-cleaning transitional signal sample, and restoring the noise-cleaning transitional signal sample to a predicted visualized neural signal conforming to the target visualized feature.

[0200] The processes of steps S40 and S50 can refer to the descriptions of the aforementioned steps S200-S400, and the principles are consistent, which will not be described here.

[0201] Step S60: determining an algorithm debugging error based on the predicted visualized neural signal and the learning sample library, and optimizing the algorithm parameters of the initial machine learning algorithm based on the algorithm debugging error until the initial machine learning algorithm meets the preset debugging stop requirement, obtaining a target machine learning algorithm; the target machine learning algorithm is used to map the original live neural signal to a visualized neural signal conforming to the target visualized feature.

[0202] In step S60, the computer system first evaluates the similarity between the predicted visualized neural signal and the real visualized neural signal in the learning sample library, and quantifies this difference by calculating the algorithm debugging error. Then, based on this error signal, the system adjusts the parameters of the initial machine learning algorithm in order to reduce the error in subsequent iterations and improve the prediction accuracy of the algorithm. This process will continue until the algorithm performance meets the preset debugging stop requirement, thereby obtaining the final target machine learning algorithm.

[0203] The algorithm debugging error is an important indicator to measure the difference between the algorithm prediction result and the real result in the learning sample library. The error may involve multiple dimensions, including but not limited to the similarity of signal shape, the accuracy of color mapping, the coincidence of dynamic effect, etc. In order to comprehensively evaluate these differences, the computer system can use a composite loss function that combines multiple error measurement standards.

[0204] For example, assume that the Mean Squared Error (MSE) is used to measure the prediction accuracy of signal strength, while the Cross-Entropy Loss is used to evaluate the correctness of color mapping. The composite loss function can be represented as:

[0205] L = a · MSE(y pred ,y true ) + b · Cross Entropy(p pred ,p true );

[0206] where y pred and y true represent the predicted signal strength and the true signal strength, respectively, p pred and p true represent the predicted color distribution and the true color distribution, respectively, and a and b are hyperparameters that balance the weights of different loss terms.

[0207] For example, assume that the learning example library contains a patient's original live neural signal and its corresponding visualized neural signal. During the training process, the computer system generates a predicted visualized neural signal and calculates the MSE and Cross Entropy loss between the predicted and true visualized neural signals. For example, the MSE may calculate the average deviation of signal strength, while the Cross Entropy evaluates the accuracy of color mapping (assuming that the visualized signal uses color coding to represent signal strength).

[0208] After calculating the algorithm debugging error, the computer system optimizes the parameters of the initial machine learning algorithm based on this error signal. This process can be achieved through optimization algorithms such as gradient descent (or its variants, such as Adam, RMS prop, etc.), which can automatically adjust algorithm parameters according to error gradients to minimize the error function.

[0209] During the optimization process, the system first calculates the gradient of the error function with respect to the algorithm parameters, i.e., the partial derivative of the error function with respect to each parameter. Then, the system updates the parameter values according to these gradient information, which can be achieved by subtracting a step proportional to the gradient from the current parameter value. This step (learning rate) is an important hyperparameter that needs to be carefully adjusted to balance the convergence speed and stability.

[0210] For example, assume that the system uses the Adam optimization algorithm to adjust the weights and biases in the neural network. In each iteration, the system first forward propagates the predicted visualized neural signal, then calculates the MSE and Cross Entropy loss, and backward propagates these losses to calculate the gradient of each parameter in the network. Then, the system updates the network parameters, including weights and biases, according to the rules of the Adam algorithm.

[0211] The optimization process will continue until preset debugging stopping requirements are met. These requirements can include reaching a maximum number of iterations, reducing error below a certain threshold, no longer significant performance improvement on the validation set, etc. In the context of visualization of monitoring information based on implantable medical chips, the debugging stopping requirements can take into account the particularity of medical applications, such as ensuring the stability and reliability of the algorithm, avoiding overfitting to improve generalization ability, etc.

[0212] For example, during the training of the system, a maximum number of iterations (e.g. 100 epochs) and an early stopping criterion on the validation set (e.g. stop training if the validation loss does not decrease for 10 consecutive epochs) can be set. When the training reaches these stopping requirements, the system considers that the algorithm has converged to a sufficiently good solution, and saves the algorithm parameters at this time as the final target machine learning algorithm model.

[0213] As an implementation, in step S50, the visual neural signal parameter is mapped to a constraint implicit representation array, the second signal feature representation sample is mapped to a signal implicit representation array, and the noise removal is performed on the noisy neural signal sample according to the constraint implicit representation array and the signal implicit representation array to obtain a noise-cleared transitional signal sample. The noise-cleared transitional signal sample is restored to a predicted visual neural signal conforming to the target visual feature, including:

[0214] In step S51, the visual neural signal parameter in the learning sample library is divided into u visual labels, and v visual difference parameters are determined according to the corresponding uniform visual elements of the u visual labels, 1≤u≤v.

[0215] In step S52, the v visual difference parameters are mapped to v constraint implicit representation arrays, the second signal feature representation sample is mapped to a signal implicit representation array, and the noise removal is performed on the noisy neural signal sample according to the v constraint implicit representation arrays and the signal implicit representation array to obtain v diverse neural signals. The parameter mean value of the v diverse neural signals is calculated to obtain a noise-cleared transitional signal sample, and the noise-cleared transitional signal sample is restored to a predicted visual neural signal conforming to the target visual feature.

[0216] In step S51, the computer system finely divides the visual neural signal parameter in the learning sample library, and determines the visual difference parameter based on the division, so as to be mapped to the constraint implicit representation array subsequently.

[0217] Visualizing neural signal parameters can include various user expectations for the visualization results, such as color mapping schemes, dynamic effect settings, view angle selections, etc. These parameters can exist in various forms, such as color coding tables, dynamic effect parameter tables, view angle conversion matrices, etc. The computer system first needs to uniformly process these parameters and divide them into multiple visualization tags. Visualization tags are high-level abstractions of visualized neural signal parameters, and each tag represents a class of visualization characteristics. For example, a color mapping scheme can be a visualization tag, which includes the mapping relationship between signal intensity and color; dynamic effect settings can be another visualization tag, which describes the display effect of the signal over time (such as flashing, gradient, etc.).

[0218] After dividing the visualization tags, the computer system further analyzes the specific visual elements represented by these tags and their differences. Unified visual elements refer to common or similar parts in different visualization tags, such as basic colors in a color wheel, time periods in dynamic effects, etc. Through the analysis of these unified visual elements, the system can identify the differences between different visualization tags, i.e., the visualization difference parameters.

[0219] Visualization difference parameters are used to quantify the specific differences between different visualization tags, which are crucial for subsequent generation of diversified visualized signals. For example, in a color mapping scheme, the visualization difference parameters may include the difference in RGB values corresponding to different color intervals; in dynamic effect settings, the visualization difference parameters may include the variation range of parameters such as flashing frequency and gradient speed.

[0220] For example, suppose the learning example library contains two color mapping schemes as visualized neural signal parameters. The first scheme (labeled as Tag1) maps the signal intensity in the range of 0-50 to blue and the range of 50-100 to red; the second scheme (labeled as Tag2) maps the range of 0-50 to green and the range of 50-100 to yellow. In this example, "color mapping" is a unified visual element, and "blue and red" and "green and yellow" are difference parameters between the two color mapping schemes.

[0221] In step S52, the computer system maps the visualization difference parameters determined in step S51 to the constraint implicit representation array and maps the second signal feature representation example to the signal implicit representation array. Then, based on these implicit representation arrays, the noise removal processing is performed on the noisy neural signal example, a diversified noise-cleared transitional signal example is generated, and finally the predicted visualized neural signal conforming to the target visualized characteristics is restored.

[0222] Constraint implicit representation arrays are numerical representations of the visualization difference parameters, which exist in the form of vectors, matrices, or other data structures, and are used to pass the user's visualization requirements within the algorithm. The computer system generates the corresponding constraint implicit representation arrays according to the specific values of each visualization difference parameter. These arrays may contain information about color coding, dynamic effect parameters, view transformation matrices, and other elements.

[0223] Signal implicit representation arrays are another numerical representation of the second signal feature representation examples, which retain the main feature information of the original signal but reduce the dimensionality and complexity of the data. This array can be extracted from the original live neural signal examples through feature embedding, dimensionality reduction processing, and other steps.

[0224] After obtaining the constraint implicit representation arrays and the signal implicit representation arrays, the computer system uses these arrays as guidance information to perform noise removal processing on the noisy neural signal examples. This processing process may involve parallel running or alternating optimization strategies of multiple noise removal models to ensure that the generated signal both removes noise and retains diversified features.

[0225] Specifically, the system may use generative models such as conditional generative adversarial networks (cGAN) or variational autoencoders (VAE) to generate diversified noise-cleaning transitional signal examples according to the constraint implicit representation arrays. These models can introduce visualization differences specified by the constraint implicit representation arrays while maintaining the main features of the signal, thereby achieving the diversified generation of the signal.

[0226] For the generated multiple noise-cleaning transitional signal examples, the system may use parameter mean calculation and other methods to integrate their common features and reduce individual differences. This process helps to obtain a more stable and user-desired predicted visualized neural signal.

[0227] Finally, the system restores the integrated noise-cleaning transitional signal examples to the final signal representation that meets the target visualization features according to the visualization difference parameters in the constraint implicit representation arrays. This process may involve multiple steps such as color mapping, dynamic effect rendering, view transformation, etc., depending on the specific content of the visualization requirements.

[0228] For example, continue the previous color mapping scheme example. Suppose the system has generated two noise-cleaning transition signal samples based on different color mapping schemes (corresponding to the color mapping of Tag1 and Tag2, respectively). During the calculation of the parameter mean value, the system may find that the color difference between the two signal samples is large in the low signal intensity part (0-50 range) but small in the high signal intensity part (50-100 range). Therefore, the system may choose to perform a weighted average processing on the color of the low signal intensity part to reduce the color difference; and for the high signal intensity part, directly use one of the color mapping schemes (such as the red color of Tag1) as the final representation.

[0229] After such processing, the system obtains a predicted visualized neural signal that not only removes noise but also conforms to the user-defined color mapping scheme. This signal can be further used for visualizing neural activity, assisting in disease diagnosis and analysis, and other fields.

[0230] Step S50 successfully combines the user-defined visualization requirements with the signal characteristics processed by the algorithm through a series of complex operations such as detailed division of the visualized neural signal parameter, determination of the visualization difference parameter, mapping of the constraint implicit representation array and the signal implicit representation array, and noise removal and signal restoration, to generate a predicted visualized neural signal that conforms to the target visualization characteristics. This process not only relies on advanced machine learning algorithms and techniques, but also requires a deep understanding of neural signal processing, visualization theory, and user requirements.

[0231] As an implementation, the learning sample library further includes a parameter representation label; the parameter representation label is semantic information associated with the visualized neural signal parameter; and the initial machine learning algorithm includes an initial constraint embedding branch component and an initial noise adding branch component. Based on this, in step S60, the algorithm debugging error is determined based on the predicted visualized neural signal and the learning sample library, and the algorithm parameters of the initial machine learning algorithm are optimized based on the algorithm debugging error until the initial machine learning algorithm meets the preset debugging stop requirement, to obtain the target machine learning algorithm, which can include: mapping the constraint implicit representation array to a to-be-executed semantic feature, determining the algorithm debugging error of the initial noise adding branch component and the initial constraint embedding branch component based on the predicted visualized neural signal and the action potential difference value of the visualized neural signal sample, and the semantic difference value between the parameter representation label and the to-be-executed semantic feature, optimizing the algorithm parameters in the initial noise adding branch component and the initial constraint embedding branch component based on the algorithm debugging error and the constraint condition, until the initial machine learning algorithm meets the constraint condition, to obtain the target machine learning algorithm.

[0232] In step S60, the computer system maps the constraint implicit representation arrays (which are generated by step S50 to guide the noise removal and signal restoration process) into to-be-executed semantic features. To-be-executed semantic features are the abstraction and interpretation of constraint implicit representation arrays, which not only contain numerical constraint information, but also contain the semantic meaning behind these constraints. For example, suppose a color mapping scheme is defined in the visualization of the neural signal parameter, which divides the signal intensity into three levels and uses red, yellow and green to represent low, medium and high intensity respectively. The corresponding constraint implicit representation array may contain a set of color coding values. In the mapping process, the computer system interprets these color coding values as the to-be-executed semantic features of “low intensity signals should be displayed in red, medium intensity signals should be displayed in yellow, and high intensity signals should be displayed in green”.

[0233] Next, the computer system calculates the algorithm debugging error to evaluate the quality of the current predicted visualized neural signal. This process not only considers the difference in action potential between the predicted signal and the real visualized neural signal sample (i.e. direct comparison of signal intensity), but also takes into account the semantic difference between the parameter representation label and the to-be-executed semantic feature. The calculation of semantic difference is an important indicator to measure the degree of understanding of the algorithm for the user's visualization needs.

[0234] For example, when comparing the predicted signal with the real signal, the computer system calculates the difference in action potential at each time point, resulting in a series of error values. Then it further analyzes whether these error values are contrary to the semantic information in the parameter representation label. For example, if the real signal should be displayed in green in a high intensity area, but the predicted signal is incorrectly displayed in yellow, this will result in a larger semantic difference.

[0235] The calculation of algorithm debugging error can be a composite loss function that combines the quantitative error of action potential difference and the qualitative evaluation of semantic difference. A simplified loss function example is as follows:

[0236] ;

[0237] where N is the number of time points of the signal, and are the action potential values of the predicted signal and the real signal at the i-th time point, and are the color mapping schemes (or other visualization features) of the predicted signal and the real signal, λ1 and λ2 are weight coefficients to balance different loss terms, and SD (Semantic Difference) function is used to calculate the semantic difference.

[0238] After obtaining the algorithm debugging error, the computer system will optimize the parameters of the initial machine learning algorithm based on this error signal. The goal of optimization is to reduce the error, so that the predicted visualized neural signal is closer to the true signal, and better meets the user's visualization needs.

[0239] Since the initial machine learning algorithm includes an initial constraint embedding branch component and an initial noise adding branch component, the optimization process will be carried out for these two components respectively. For the constraint embedding branch component, optimization may involve adjusting the weights of the feature embedding operator, changing the parameters of the hidden space mapping, etc.; for the noise adding branch component, optimization may focus on adjusting the parameters of the noise distribution, optimizing the noise adding strategy, etc.

[0240] For example, during the optimization process, the computer system may use gradient descent (or its variants such as Adam optimizer) to adjust the algorithm parameters. For the constraint embedding branch component, it can fine-tune the weight distribution in the feature embedding process according to the feedback of the semantic difference, to ensure that the semantic features to be executed can be accurately mapped to the hidden space; for the noise adding branch component, it may adjust the noise intensity and noise type according to the feedback of the action potential difference, to reduce the impact of noise on signal quality.

[0241] Finally, the computer system regularly checks whether the preset debugging stop conditions are met. These conditions may include reaching the maximum number of iterations, the algorithm debugging error falling below the predetermined threshold, the performance on the validation set no longer significantly improving, etc. Only when all conditions are met, the training process will be terminated, and the algorithm parameters at this time are considered as the optimal solution, and are used to generate the target machine learning algorithm.

[0242] Through the implementation of the above step S60, the computer system considers the quantitative error between the predicted signal and the true signal and the semantic understanding degree of the user's visualization needs, and finely optimizes the parameters of the initial machine learning algorithm. This process not only improves the performance and stability of the algorithm, but also ensures that the final generated target machine learning algorithm can accurately map the original living neural signal to the visualized neural signal that meets the user's expectations

[0243] Figure 2 A hardware entity schematic diagram of a computer system provided by an embodiment of the present application is shown in FIG. 1, which includes a processor 1001 and a memory 1002, wherein the memory 1002 stores a computer program executable on the processor 1001, and the processor 1001 implements the steps in the method of any of the above embodiments when executing the program. Figure 2

[0244] ​The above merely provides the implementation of the present application, but the protection scope of the present application is not limited thereto, any person skilled in the art can easily think of the change or replacement within the technical range disclosed by the present application, which should be covered in the protection scope of the present application.

Claims

1. A method for visualizing regulatory monitoring information based on an implantable medical chip, characterized in that, include: The process involves acquiring raw live neural signals, abstracting the signal features of the raw live neural signals to obtain abstract signals, embedding features into the raw live neural signals to obtain a first signal feature representation, and embedding features into the abstract signals to obtain a second signal feature representation; the number of sampling points of the raw live neural signals is greater than the number of sampling points of the first signal feature representation and the second signal feature representation. Noise is added to the first signal feature representation to obtain a noisy neural signal; Obtain the visualization constraint parameters, map the visualization constraint parameters to an array of implicit constraint representations, and map the second signal feature representation to an array of implicit signal representations. The visualization constraint parameters are used to describe the visualization features of the visualization display results. Based on the constraint implicit representation array and the signal implicit representation array, noise removal is performed on the noisy neural signal to obtain a noise-cleaned transition signal. The noise-cleaned transition signal is then restored to a visual neural signal that conforms to the target visualization characteristics. The original live neural signal and the visual neural signal have the same number of sampling points. The step of embedding features from the original live neural signal to obtain a first signal feature representation includes loading the original live neural signal into a target machine learning algorithm; the target machine learning algorithm includes a latent space mapping branch component, which includes a feature representation operator; performing feature extraction on the original live neural signal based on the feature representation operator to obtain a center array and a dispersion array of the original live neural signal; randomly extracting from the center array and dispersion array of the original live neural signal to obtain a latent space center array and a latent space dispersion array; and generating a first signal feature representation based on the latent space center array and the latent space dispersion array; the first signal feature representation is generated in the latent space mapping branch component of the target machine learning algorithm, which includes a normalization operator and a pooling operator; The step of adding noise to the first signal feature representation to obtain a noisy neural signal includes acquiring a noise distribution and an iteration number s; performing a standardization operation on the waveform feature values ​​of the action potential in the first signal feature representation based on the standardization operator and the noise distribution; s≥1; and adding noise s times to the standardized first signal feature representation based on the pooling operator and the iteration number s to obtain a noisy neural signal, wherein the noise distribution represents the confidence distribution information of the added noise, and the iteration number s represents the degree of noise added to the first signal feature representation according to the noise distribution.

2. The method according to claim 1, characterized in that, The process of acquiring raw in vivo neural signals and abstracting the signal features of the raw in vivo neural signals to obtain abstracted signals includes: Obtain the waveform feature values ​​corresponding to x action potentials in the original live neural signal, and perform a smoothing operation on the original live neural signal based on the waveform feature values ​​to obtain a smoothed signal; x≥1; The time axis change rate and amplitude axis change rate corresponding to x action potentials in the smoothed signal are obtained respectively. The intensity and angle corresponding to x action potentials in the smoothed signal are determined according to the time axis change rate and amplitude axis change rate. Local extremum detection is performed on the smoothed signal based on the intensity and the angle to obtain the extremum neural signal. An abstract signal is generated based on the time axis change rate and amplitude axis change rate corresponding to the x action potentials in the extreme neural signal.

3. The method according to claim 2, characterized in that, The process of generating an abstract signal based on the time axis rate of change and amplitude axis rate of change corresponding to x action potentials in the extreme neural signal includes: Obtain the first reference value and the second reference value, and determine the rate of change corresponding to each of the x action potentials in the extreme neural signal based on the rate of change of the time axis and the rate of change of the amplitude axis corresponding to the x action potentials in the extreme neural signal respectively; Among the x action potentials of the extreme neural signal, a significant action potential is identified, wherein the rate of change of the significant action potential is greater than or equal to the first reference value; A fuzzy action potential is determined from x action potentials of the extreme neural signal, wherein the rate of change of the fuzzy action potential is greater than or equal to the second reference value and less than the first reference value; The fuzzy action potentials adjacent to the significant action potential are identified as candidate action potentials, and an abstract signal is generated based on the significant action potential and the candidate action potentials.

4. The method according to claim 1, characterized in that, The step of mapping the visualized constraint parameters to an array of implicit constraint representations includes: The visualized constraint parameters are loaded into the target machine learning algorithm, wherein the target machine learning algorithm includes a constraint embedding branch component; the constraint embedding branch component includes a feature embedding operator, a weight allocation operator, and a fusion operator; Based on the feature embedding operator, the visualized constraint parameters are embedded to obtain y constraint array representations; y≥1; Based on the weight allocation operator, generate the focus weights corresponding to the y constraint arrays respectively, and adjust the focus of the constraint arrays according to the focus weights to obtain y focus arrays to be used; The fusion operator is used to perform weight fusion on the y focus arrays to be used, thereby obtaining a constrained implicit representation array; The noisy neural signal is obtained by adding noise s times to the first signal feature representation, which is generated in the latent space mapping branch component of the target machine learning algorithm. The latent space mapping branch component includes a residual operator and an interpolation operator. The step of removing noise from the noisy neural signal based on the constraint implicit representation array and the signal implicit representation array to obtain a noise-cleaned transition signal includes: Based on the residual operator, the constraint implicit representation array and the signal implicit representation array are combined into a guidance array; Based on the interpolation operator and the guidance array, the noisy neural signal is subjected to initial noise removal to obtain the noise-removed signal G. s-1 Repeat the noise removal signal G s-1 Noise removal is performed until s noise removals are performed. The noise removal signal G0 obtained from the sth noise removal is determined as the noise cleanup transition signal.

5. The method according to claim 4, characterized in that, The noise removal signal G is obtained by performing the first noise removal on the noisy neural signal based on the interpolation operator and the guidance array. s-1 ,include: The center array and dispersion array of the noisy neural signal are obtained based on the interpolation operator. Randomly extracting from the center array and dispersion array of the noisy neural signal, an iterative center array and an iterative dispersion array are obtained. Obtain a random array of noise distributions, and adjust the iteration center array and the iteration dispersion array according to the guide array and the random array to obtain the noise removal signal G. s-1 .

6. The method according to claim 1, characterized in that, The noise-sweeping transition signal is generated in the latent space mapping branch component of the target machine learning algorithm, the latent space mapping branch component including feature representation operators and interpolation operators; the step of restoring the noise-sweeping transition signal into a visual neural signal that conforms to the target visualization features includes: Based on the feature representation operator, feature extraction is performed on the noise-sweeping transition signal to obtain the center array and the dispersion array of the noise-sweeping transition signal. Randomly extracting from the center array, a latent space center array is obtained, and randomly extracting from the dispersion array, a latent space dispersion array is obtained. Based on the interpolation operator, the latent space center array, and the latent space dispersion array, a visual neural signal that conforms to the target visualization characteristics is generated.

7. The method according to any one of claims 5 to 6, characterized in that, The method further includes: Obtain a learning sample library; the learning sample library includes original live neural signal samples, visualized neural signal samples, and visualized neural signal parameters; the visualized neural signal parameters are used to describe the visualization features of the visualization results; the original live neural signal samples and the visualized neural signal samples have the same number of sampling points; The original live neural signal sample is abstracted to obtain an abstract signal sample, and the original live neural signal sample and the abstract signal sample are loaded into the initial machine learning algorithm. The original live neural signal sample is subjected to feature embedding to obtain a first signal feature representation sample, and the abstracted signal sample is subjected to feature embedding to obtain a second signal feature representation sample; the number of sampling points of the original live neural signal sample is greater than the number of sampling points of the first signal feature representation sample and the second signal feature representation sample; Noise is added to the second signal feature representation sample to obtain a noisy neural signal sample; The visualized neural signal parameters are mapped to a constrained implicit representation array, and the second signal feature representation example is mapped to a signal implicit representation array. Based on the constrained implicit representation array and the signal implicit representation array, noise removal is performed on the noisy neural signal example to obtain a noise-cleaned transition signal example. The noise-cleaned transition signal example is then restored to a predicted visualized neural signal that conforms to the target visualized features. The algorithm debugging error is determined based on the predicted visualized neural signals and the learning sample library. The algorithm parameters of the initial machine learning algorithm are optimized based on the algorithm debugging error until the initial machine learning algorithm meets the preset debugging stop requirements, and the target machine learning algorithm is obtained. The target machine learning algorithm is used to map the original live neural signals into visualized neural signals that conform to the target visualization features.

8. The method according to claim 7, characterized in that, The process of mapping the visualized neural signal parameters to a constrained implicit representation array, mapping the second signal feature representation example to a signal implicit representation array, removing noise from the noisy neural signal example based on the constrained implicit representation array and the signal implicit representation array to obtain a noise-cleaned transition signal example, and restoring the noise-cleaned transition signal example to a predicted visualized neural signal that conforms to the target visualized features includes: The visual neural signal parameters in the learning example library are divided into u visual labels, and v visual difference parameters are determined according to the unified visual elements corresponding to the u visual labels, where 1≤u≤v; The v visualization difference parameters are mapped to v constraint implicit representation arrays, and the second signal feature representation example is mapped to a signal implicit representation array. Based on the v constraint implicit representation arrays and the signal implicit representation arrays, noise removal is performed on the noisy neural signal example to obtain v diverse neural signals. The parameter mean of the v diverse neural signals is calculated to obtain a noise-cleaned transition signal example. The noise-cleaned transition signal example is restored to a predicted visual neural signal that conforms to the target visualization features. The learning example library also includes parameter representation labels; the parameter representation labels are semantic information associated with the parameters of the visualized neural signal; the initial machine learning algorithm includes an initial constraint embedding branch component and an initial noise-adding branch component; the step of determining the algorithm debugging error based on the predicted visualized neural signal and the learning example library, optimizing the algorithm parameters of the initial machine learning algorithm based on the algorithm debugging error, until the initial machine learning algorithm meets the preset debugging stop requirements, to obtain the target machine learning algorithm, includes: The implicit constraint representation array is mapped to the semantic features to be executed. Based on the predicted visualized neural signal and the action potential difference of the visualized neural signal sample, as well as the semantic difference between the parameter representation label and the semantic features to be executed, the algorithm debugging error of the initial noisy branch component and the initial constraint embedding branch component is determined. Based on the algorithm debugging error and the constraint conditions, the parameters in the initial noisy branch component and the initial constraint embedding branch component are optimized until the initial machine learning algorithm meets the constraint conditions, and the target machine learning algorithm is obtained.

9. A computer system comprising a memory and a processor, the memory storing a computer program executable on the processor, characterized in that, When the processor executes the program, it implements the steps of the method according to any one of claims 1 to 8.

Images