A microstate analysis method of invasive brain cortex electrical signals

Abstract

This invention discloses an invasive method for microstate analysis of electrical signals from the cerebral cortex, comprising the following analytical steps: Step 1, signal acquisition and preprocessing: A cortical surface electrode array is implanted in a dog to acquire canine ECoG signals, and the acquired canine ECoG signals are used as raw signals for preprocessing; Step 2, frequency band processing: Based on the spectral characteristics of the canine ECoG signals, a digital filter is designed to decompose the signal into multiple frequency bands; Step 3, the signal within each frequency band is segmented using a rolling window method, setting the window length and overlap rate; Step 4, for the signal samples within each sliding window, the differential entropy index is calculated, with the calculation formula employing either a Gaussian distribution assumption or a direct nonparametric method. This invention provides a new analytical tool for basic neuroscience research and can also be applied to disease diagnosis, behavioral research, and intelligent signal processing systems in clinical veterinary medicine, possessing significant scientific research and application value.

Description

Technical Field

[0001] This invention relates to the field of biomedical signal processing technology, and more specifically, to an invasive method for microstate analysis of electrical signals in the cerebral cortex. Background Technology

[0002] With the continuous advancement of neuroscience research and signal processing technology, electrocortical brain surface signals (ECoG), due to their high spatial resolution, low artifact interference, and high signal-to-noise ratio, are gradually replacing traditional electroencephalography (EEG) in clinical neurological function localization, epileptic focus identification, and higher brain function research. Especially in animal models, ECoG data provides valuable information for a deeper understanding of local activity and dynamic microstates in the cerebral cortex. Currently, existing ECoG microstate analysis methods are mostly focused on human signals, with commonly used indicators including power spectral density, phase synchronization, and time-domain waveform characteristics. These methods, to some extent, reveal the stability and dynamic changes of brain states. However, for canine cortical EEG signals, their signal characteristics differ significantly from those of humans, and there is insufficient capture of local dynamic changes in ECoG data, mainly in the following aspects:

[0003] 1. Differences in signal characteristics: Canine EEG signals have their own characteristics in terms of frequency band distribution, signal-to-noise ratio, and dynamic changes. Directly using human analysis methods may ignore the unique physiological information of dogs.

[0004] 2. Insufficient spatiotemporal resolution: Traditional analysis methods generally use fixed time windows for energy statistics, making it difficult to capture short-term, instantaneous state transitions. For the rapid cortical activity in ECoG signals, the micro-state changes and nonlinear characteristics urgently require an analysis method with higher spatiotemporal resolution.

[0005] 3. Limitations in Indicator Selection: Traditional methods primarily focus on energy or power spectral density indicators, which cannot fully reflect the complexity and uncertainty of the signal. Differential entropy, as an indicator of the information content of continuous signals, is more suitable for describing the subtle dynamic changes and complexity in ECoG signals.

[0006] Therefore, there is an urgent need for a novel analytical method that targets the characteristics of canine cortical ECoG signals and can accurately capture microstate changes at high spatiotemporal resolution to address the above-mentioned shortcomings and promote the development of animal neuroscience and clinical veterinary medicine. Summary of the Invention

[0007] In view of the problems existing in the prior art, the purpose of this invention is to provide an invasive method for microstate analysis of electrical signals in the cerebral cortex, so as to solve the problems in the background art.

[0008] To achieve the above objectives, the present invention adopts the following technical solution;

[0009] An invasive method for microstate analysis of electrical signals in the cerebral cortex, comprising the following analytical steps:

[0010] Step 1: Signal Acquisition and Preprocessing. A cortical surface electrode array is implanted into the canine to acquire the canine ECoG signal, and the acquired canine ECoG signal is used as the raw signal for preprocessing.

[0011] Step 2: Frequency banding. Based on the spectral characteristics of canine ECoG signals, design digital filters to decompose the signal into multiple frequency bands.

[0012] Step 3: Divide the signal within each frequency band into segments using the rolling window method, and set the window length and overlap rate;

[0013] Step 4: For each signal sample within the sliding window, calculate the differential entropy index. The calculation formula adopts either the Gaussian distribution assumption or a direct non-parametric method, and is solved by statistically analyzing the signal probability density distribution within the window.

[0014] Step 5: Feature integration and cluster analysis. The differential entropy data of all windows in each frequency band are weighted and fused according to energy proportion to generate a multi-dimensional feature vector sequence.

[0015] Principal component analysis is introduced for dimensionality reduction, retaining more than 90% of the variance and reducing redundant information;

[0016] The fuzzy C-means clustering (FCM) algorithm is used to classify the feature vectors, dividing the continuous differential entropy sequence into several discrete micro-states. The number of clusters is adaptively determined by the gap statistics.

[0017] Step 6: Microstate transition and dynamic analysis. Based on the clustering results, construct a microstate time series diagram of canine EEG signals, analyze the transition patterns between different microstates, and statistically analyze the duration, transition frequency, and transition probability of each microstate. Analyze its correlation with canine behavior, arousal state, and pathological abnormalities.

[0018] Step 7: Results Visualization and Data Mining. Visualization methods such as time series diagrams, heatmaps, and state transition matrices are used to present the microstate analysis results intuitively.

[0019] As a further description of the above technical solution:

[0020] The first step, signal acquisition, includes recording video of canine behavior and electromyography signals. Preprocessing of the raw signals includes bandpass filtering, baseline drift correction, and wavelet denoising.

[0021] As a further description of the above technical solution:

[0022] In step two, the canine ECoG signal is divided into δ (0.5–4Hz), θ (4–8Hz), α (8–13Hz), β (13–30Hz), and γ (30–150Hz).

[0023] As a further description of the above technical solution:

[0024] The rolling window method in step three involves defining a window of fixed length and gradually sliding it across the frequency band to divide a continuous signal into multiple segments. The window length and overlap rate are set according to actual needs.

[0025] As a further description of the above technical solution:

[0026] In step three, the signal within each time window is regarded as a short-time stationary process, serving as the basic unit for differential entropy calculation and feature extraction in step four.

[0027] As a further description of the above technical solution:

[0028] In step four, when the data within the window approximately follows a Gaussian distribution, the formula is used:

[0029]

[0030] Where σ is the standard deviation of the signal within the window. When the data within the window is non-Gaussian, the differential entropy is directly solved by kernel density estimation.

[0031] As a further description of the above technical solution:

[0032] Each vector in step five contains information about a single frequency band and a comprehensive description of the signal complexity across frequency bands.

[0033] Compared with the prior art, the advantages of this invention are:

[0034] (1) This scheme extracts signals at multiple levels: through frequency band processing, the physiological information hidden in each frequency band of the signal can be fully extracted and preserved, which can more comprehensively reflect the characteristics of canine EEG signals.

[0035] (2) This scheme has high spatiotemporal resolution: by using the rolling window method and high sampling rate data, the instantaneous dynamic changes of canine EEG signals are effectively captured, which enhances the spatiotemporal resolution capability of microstate analysis.

[0036] (3) This scheme innovates the quantification of complexity: it adopts differential entropy as the signal complexity index, which can better reflect the nonlinearity, randomness and dynamic change characteristics of the signal compared with the traditional energy or power spectrum index.

[0037] (4) This scheme has highly adaptable data processing: the clustering algorithm adaptively determines the micro-state category, which makes the whole method have good universality and robustness, and is suitable for signal analysis of different dog breeds and various behavioral states.

[0038] (5) This method has broad application prospects: It not only provides new analytical tools for basic neuroscience research, but can also be applied to disease diagnosis (such as epilepsy monitoring), behavioral research and intelligent signal processing systems in clinical veterinary medicine, and has significant scientific research and application value. Attached Figure Description

[0039] Figure 1 This is a schematic diagram illustrating the principle of the present invention;

[0040] Figure 2 This is a schematic diagram of the process of the present invention. Detailed Implementation

[0041] The technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention;

[0042] Please see Figure 1 and Figure 2 The present invention provides the following embodiments:

[0043] An invasive method for microstate analysis of electrical signals in the cerebral cortex, comprising the following analytical steps:

[0044] Step 1: Signal Acquisition and Preprocessing. A cortical surface electrode array is implanted into the dog to acquire the canine ECoG signal. During the acquisition process, a high sampling rate device (e.g., sampling rate ≥ 500 Hz) is used to ensure high spatiotemporal resolution data. The acquired canine ECoG signal is then used as the raw signal for preprocessing.

[0045] Step 2: Frequency band processing. Based on the spectral characteristics of canine ECoG signals, a digital filter is designed to decompose the signal into multiple frequency bands. In Step 2, the canine ECoG signal is divided into δ (0.5–4Hz), θ (4–8Hz), α (8–13Hz), β (13–30Hz), and γ (30–150Hz). In some experiments, the γ band can be further subdivided to capture high-frequency activity. By adjusting the filter parameters, it is ensured that each frequency band has a high signal-to-noise ratio and fully preserves cortical activity information, providing reliable data for subsequent feature calculations.

[0046] Step 3: Segment the signal within each frequency band using a rolling window method, setting the window length and overlap rate. For example, the window length can be 1 to 3 seconds, and the overlap rate can be set to 50% or higher to ensure the local stability and continuity of the signal within the window.

[0047] Step 4: For each signal sample within a sliding window, calculate the differential entropy index. As a measure of the information entropy of a continuous signal, differential entropy can reflect the complexity and uncertainty of the signal. The calculation formula adopts either the Gaussian distribution assumption or a direct non-parametric method. It is solved by statistically analyzing the probability density distribution of the signal within the window. This step will yield a series of numerical characteristics representing the signal complexity of each time window.

[0048] Step 5: Feature integration and cluster analysis. The differential entropy data of all windows in each frequency band are weighted and fused according to energy proportion to generate a multi-dimensional feature vector sequence.

[0049] Principal component analysis is introduced for dimensionality reduction, retaining more than 90% of the variance, reducing redundant information and computational load;

[0050] The fuzzy C-means clustering (FCM) algorithm is used to classify the feature vectors, dividing the continuous differential entropy sequence into several discrete micro-states. The number of clusters is adaptively determined by the gap statistics, avoiding subjective setting.

[0051] Step Six: Microstate transitions and dynamic analysis. Based on the clustering results, construct a microstate time series diagram of canine EEG signals, analyze the transition patterns between different microstates, and statistically analyze the duration, transition frequency, and transition probability of each microstate. Analyze the correlation between microstates and canine behavior, arousal state, and pathological abnormalities (such as epileptic seizures) to provide a basis for subsequent clinical applications or behavior prediction.

[0052] Step 7: Results Visualization and Data Mining. Using visualization methods such as time series diagrams, heatmaps, and state transition matrices, the microstate analysis results are presented intuitively. Machine learning algorithms can be combined to further perform pattern recognition and classification prediction on the spatiotemporal distribution of different microstates, so as to achieve real-time monitoring and early warning of canine neurological states.

[0053] In this invention, the first step of signal acquisition includes recording canine behavior videos and electromyographic signals. The preprocessing of the raw signals includes bandpass filtering, baseline drift correction, and wavelet noise reduction to remove environmental interference and motion artifacts in order to obtain high-quality signal data.

[0054] The rolling window method in step three involves defining a window of fixed length and gradually sliding it across the frequency band to divide a continuous signal into multiple segments. The window length and overlap rate are set according to actual needs.

[0055] The rolling window method is a commonly used segmentation technique in time series signal processing, especially suitable for the analysis of ECoG (electrocortical imaging) signals. It can be applied in frequency band processing, event detection, and real-time signal processing, and can balance the requirements of time resolution and frequency resolution.

[0056] In step three, the signal within each time window is regarded as a short-time stationary process, serving as the basic unit for differential entropy calculation and feature extraction in step four.

[0057] In step four, when the data within the window approximately follows a Gaussian distribution, the formula is used:

[0058]

[0059] Where σ is the standard deviation of the signal within the window. When the data within the window is non-Gaussian, the differential entropy is directly solved by kernel density estimation.

[0060] Each vector in step five contains information about a single frequency band and a comprehensive description of the signal complexity across frequency bands.

[0061] We designed corresponding data processing and frequency banding methods based on the unique physiological characteristics of canine ECoG signals, which facilitates the extraction of representative frequency band features. This allows us to apply microstate analysis to real-time monitoring of dogs, providing support for canine medical care. We capture microstate changes in canine EEG signals at high spatiotemporal resolution. In particular, we utilize the rolling window method combined with differential entropy calculation to quantify signal complexity and uncertainty, improving monitoring accuracy. We also employ data mining techniques such as clustering to transform continuous signal indicators into discrete microstates, and further explore their relationship with canine behavior or pathological states.

[0062] The above description is merely a preferred embodiment of the present invention; however, the scope of protection of the present invention is not limited thereto. Any equivalent substitutions or modifications made by those skilled in the art within the scope of the technology disclosed in the present invention, based on the technical solution and its improved concepts, should be covered within the scope of protection of the present invention.

Claims

1. A microstate analysis method of a brain cortical electrical signal, characterized by, The analysis includes the following steps: Step 1: Signal preprocessing. The collected canine ECoG signals are used as raw signals for preprocessing. Step 2: Frequency banding. Based on the spectral characteristics of canine ECoG signals, design digital filters to decompose the signal into multiple frequency bands. Step 3: Segmentation using a rolling window method. Segment the signal within each frequency band using a rolling window method, setting the window length and overlap rate. Step 4: Differential entropy calculation. For each signal sample within the sliding window, the differential entropy index is used for calculation. The calculation formula adopts either the Gaussian distribution assumption or a direct non-parametric method, and is solved by statistically analyzing the probability density distribution of the signal within the window. Step 5: Feature integration and cluster analysis. The differential entropy data of all windows in each frequency band are weighted and fused according to energy proportion to generate a multi-dimensional feature vector sequence. Principal component analysis is introduced for dimensionality reduction, retaining more than 90% of the variance and reducing redundant information; The fuzzy C-means clustering algorithm is used to classify the feature vectors, dividing the continuous differential entropy sequence into several discrete micro-states. The number of clusters is adaptively determined by the gap statistic. Step 6: Microstate transition and dynamic analysis. Based on the clustering results, construct a microstate time series diagram of canine EEG signals, analyze the transition patterns between different microstates, and statistically analyze the duration, transition frequency, and transition probability of each microstate. Analyze its correlation with canine behavior, arousal state, and pathological abnormalities. Step 7: Results Visualization and Data Mining. Visualization methods such as time series diagrams, heat maps, and state transition matrices are used to present the microstate analysis results intuitively. The signals in step one include recorded videos of canine behavior and electromyography signals. Preprocessing of the raw signals includes bandpass filtering, baseline drift correction and wavelet noise reduction. In step two, the canine ECoG signal is divided into δ: 0.5–4Hz, θ: 4–8Hz, α: 8–13Hz, β: 13–30Hz and γ: 30–150Hz; The rolling window method in step three involves defining a window of fixed length and gradually sliding it across the frequency band to divide a continuous signal into multiple segments. The window length and overlap rate are set according to actual needs. In step three, the signal within each time window is regarded as a short-time stationary process, serving as the basic unit for differential entropy calculation and feature extraction in step four.

2. The method of claim 1, wherein the method is characterized by: In step four, when the data within the window approximately follows a Gaussian distribution, the formula is used: where σ is the standard deviation of the signal within the window. When the data within the window is not Gaussian, the differential entropy is solved directly using kernel density estimation.

3. The method of claim 1, wherein the method is characterized by: Each vector in step five contains information about a single frequency band and a comprehensive description of the signal complexity across frequency bands.

Images