A preterm infant language development neuroelectrophysiological signal pattern recognition method and system

Abstract

This invention relates to the field of signal recognition technology, specifically to a method and system for recognizing neurophysiological signal patterns in the language development of premature infants. The method includes acquiring initial neurophysiological response information and auditory stimulation parameters collected during auditory stimulation of the premature infant; extracting language development feature information; determining the neural response state of the premature infant to the current stimulus based on the language development feature information; adjusting the auditory stimulation parameters according to the neural response state to obtain adjusted auditory stimulation parameters; using the adjusted auditory stimulation parameters to provide further auditory stimulation to the premature infant, determining the adjusted neurophysiological response information; and recognizing the neural development of language in the premature infant based on the adjusted neurophysiological response information to obtain the neural recognition result of language development in the premature infant. The purpose of this invention is to solve the problem of inaccurate assessments due to biases in the existing technology for evaluating the language abilities of premature infants.

Description

Technical Field

[0001] This invention relates to the field of signal recognition technology, specifically to a method and system for recognizing neurophysiological signals related to language development in premature infants. Background Technology

[0002] By assessing the potential trajectory of language ability in the early developmental stages of premature infants, potential developmental delays can be identified and intervened in a timely manner. Due to the unique characteristics of premature infant brain development, the high degree of individual variability, and the frequent fluctuations in their physiological state, existing assessment methods rely on pre-set fixed stimulus sequences and assessment criteria based on group averages.

[0003] Furthermore, when applying deep learning models trained on population data to the actual assessment of individual preterm infants, it becomes apparent that preterm infants are not yet fully physiologically mature at birth, and their brain structure and functional development exhibit extremely significant individual differences. The neurophysiological responses of preterm infants to speech stimuli, including waveform morphology, latency, amplitude, and brain region distribution, show enormous variation among individuals, far exceeding the population average for adults or full-term infants. What is a typical neural pattern at the population level shows significant deviations in specific preterm infants, potentially indicating unique developmental pathways. When faced with unique neural responses in preterm infants that deviate from the population average but are still within the normal range, the model may misclassify them as abnormal. This can easily lead to misclassifying normal individuals as abnormal or failing to identify truly at-risk individuals, severely impacting the accuracy and clinical guidance value of early assessments. Consequently, when faced with each preterm infant's unique developmental pathway and rapidly changing physiological condition, inaccurate assessment results and reduced clinical guidance value are highly likely. Summary of the Invention

[0004] The purpose of this invention is to provide a method and system for recognizing neurophysiological signals related to language development in premature infants, in order to solve the problem that the existing technology has biases in assessing the language ability of premature infants, resulting in inaccurate assessments.

[0005] To achieve the above objectives, the present invention adopts the following technical solution: a method for recognizing neurophysiological signals in the language development of premature infants, comprising the following steps: Acquire the initial neurophysiological response information and auditory stimulation parameters collected when auditory stimulation is applied to preterm infants; Extract language development features from the initial neurophysiological response information; Based on the language development characteristics information, determine the neural response state of the premature infant to the current stimulus; Based on the neural response state, the auditory stimulation parameters are adjusted to obtain the adjusted auditory stimulation parameters; Using the adjusted auditory stimulation parameters, premature infants were subjected to further auditory stimulation to determine the neurophysiological adjustment response information. Based on the neurophysiological adjustment response information, the neurodevelopment of language development in premature infants is identified, and the neurodevelopmental identification results of language development in premature infants are obtained.

[0006] Preferably, the step of extracting language development feature information from the initial neurophysiological response information includes: Signal analysis was performed on the initial neurophysiological response information to determine the original characteristics of language development, microphysiological index signals, and background neurophysiological information; Correlation analysis was performed on the microscopic physiological index signals and background neurophysiological information to identify interference characteristic signals; Based on the interference feature signal, the original language development feature information is subjected to feature filtering processing to obtain language development feature information.

[0007] Preferably, the step of performing signal analysis on the initial neurophysiological response information to determine the original language development features, microphysiological index signals, and background neurophysiological information includes: Physiological index analysis was performed on the initial neurophysiological response information to obtain various physiological index information; Information analysis was performed on each physiological indicator to determine the initial developmental information, initial physiological indicator information, and initial neuroelectrophysiological information; Information extraction and processing were performed on the initial developmental information, initial physiological indicator information, and initial neuroelectrophysiological information to obtain the original language development feature information, microscopic physiological indicator signals, and background neuroelectrophysiological information.

[0008] Preferably, the step of performing correlation analysis on the microscopic physiological indicator signals and background neuroelectrophysiological information to determine interference characteristic signals includes: The microscopic physiological index signals and background neurophysiological information are verified separately to obtain the verified microscopic physiological index signals and background neurophysiological information. Based on the pre-set physiological correlation model, correlation analysis is performed on the microscopic physiological indicator signals and background neuroelectrophysiological information after the verification is qualified to determine the interference characteristic signals.

[0009] Preferably, the step of using the adjusted auditory stimulation parameters to re-stimulate the premature infant and determine the neurophysiological adjustment response information includes: Using the adjusted auditory stimulation parameters, the premature infant was subjected to auditory stimulation again to obtain auditory stimulation presentation parameters and each neurophysiological response data point; State analysis was performed on each neurophysiological response data point to obtain the stability coefficient of each physiological state; Weight analysis was performed on each physiological state stability coefficient to determine the weight of each neurophysiological response data point. Based on the weight of each neurophysiological response data point, the auditory stimulus presentation parameters, and each neurophysiological response data point, the neurophysiological adjustment response information is determined.

[0010] Preferably, the step of determining the neurophysiological adjustment response information based on the weight of each neurophysiological response data point, the auditory stimulus presentation parameters, and each neurophysiological response data point includes: Based on the weight of each neurophysiological response data point, the auditory stimulus presentation parameters, and each neurophysiological response data point, the original information on physiological development and physiological adjustment is obtained. Based on the physiological developmental state and the preset developmental state model, determine the confidence level of physiological development; Using the aforementioned physiological development confidence level, the original physiological adjustment information is adjusted to obtain neuroelectrophysiological adjustment response information.

[0011] Preferably, the step of determining the confidence level of physiological development based on the physiological developmental state and the preset developmental state model includes: The physiological developmental status was analyzed to determine the individual developmental stage and type of language developmental abnormality in the premature infant; The developmental model coefficients were determined based on the individual developmental stage and language developmental abnormality type of the premature infant. By using developmental model coefficients, the preset developmental baseline model is adjusted to obtain the preset developmental state model; Based on the physiological developmental state and the preset developmental state model, the confidence level of physiological development is determined.

[0012] Preferably, the step of determining the neural response state of a premature infant to a current stimulus based on the language development characteristics information includes: The language development feature information is processed by time-frequency conversion to obtain neural activity intensity information at different time points and within different frequency ranges; Identify key feature components in the language development feature information; Based on the information on neural activity intensity and key characteristic components at different time points and within different frequency ranges, neural development characteristics are determined; Using a preset state model, the neural development characteristics are analyzed to obtain the neural response state of the premature infant to the current stimulus.

[0013] Preferably, the steps for obtaining the initial neurophysiological response information and auditory stimulation parameters collected during auditory stimulation of preterm infants include: To obtain raw neurophysiological response information and raw auditory stimulation parameters collected during auditory stimulation of preterm infants; The original neurophysiological response information and the original auditory stimulus parameters are preprocessed to obtain the preprocessed original neurophysiological response information and the preprocessed original auditory stimulus parameters. The preprocessed original neurophysiological response information and the preprocessed original auditory stimulation parameters are verified to obtain the initial neurophysiological response information and auditory stimulation parameters.

[0014] This invention also provides a neuroelectrophysiological signal pattern recognition system for language development in premature infants, the system comprising: The information acquisition module is used to acquire the initial neurophysiological response information and auditory stimulation parameters collected when auditory stimulation is performed on premature infants; The information extraction module is used to extract language development feature information from the initial neurophysiological response information; The state determination module is used to determine the neural response state of the premature infant to the current stimulus based on the language development feature information. The parameter adjustment module is used to adjust the auditory stimulation parameters according to the neural response state to obtain the adjusted auditory stimulation parameters; The information stimulation module is used to provide auditory stimulation to premature infants again using the adjusted auditory stimulation parameters, and to determine the neurophysiological adjustment response information. The result recognition module is used to identify the neural development of language development in premature infants based on the neurophysiological adjustment response information, and to obtain the neural recognition result of language development in premature infants.

[0015] Compared with the prior art, the method and system for pattern recognition of neuroelectrophysiological signals in the language development of premature infants of the present invention have the following advantages: This invention acquires the initial neurophysiological response information and auditory stimulation parameters of premature infants upon auditory stimulation, and extracts language development characteristics from these parameters, enabling in-depth analysis of the neural response state of premature infants to stimulation. Furthermore, it allows for dynamic adjustment of auditory stimulation parameters based on the neural response state, and the adjusted parameters are used for re-stimulation, thereby obtaining more targeted neurophysiological adjustment response information. Ultimately, by identifying the adjusted response information, the neural development of language in premature infants can be accurately assessed. Through dynamic stimulation adjustment and personalized response analysis, the unique developmental pathways and rapidly changing physiological conditions of premature infants can be captured more precisely, significantly improving the accuracy and clinical guidance value of language development assessment in premature infants, and providing a reliable basis for early intervention. Attached Figure Description

[0016] To more clearly illustrate the specific embodiments of the present invention, the accompanying drawings used in the specific embodiments will be briefly described below. In all the drawings, the elements or parts are not necessarily drawn to scale.

[0017] Figure 1 This is a flowchart of a method for recognizing neurophysiological signals related to language development in premature infants according to the present invention.

[0018] Figure 2 This is a structural block diagram of a neurophysiological signal pattern recognition system for language development in premature infants according to the present invention.

[0019] In the diagram: 210, Information Acquisition Module; 220, Information Extraction Module; 230, State Determination Module; 240, Parameter Adjustment Module; 250, Information Stimulation Module; 260, Result Recognition Module.

[0020] The implementation and advantages of the present invention will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0021] The following drawings disclose several embodiments of the present invention. For clarity, many practical details will be described in the following description. However, it should be understood that these practical details are not intended to limit the invention. That is, in some embodiments of the invention, these practical details are not essential. Furthermore, for the sake of simplicity, some conventional structures and components will be shown in the drawings in a simple schematic manner.

[0022] It should be noted that all directional indications (such as up, down, left, right, front, back, etc.) in the embodiments of the present invention are only used to explain the relative positional relationship and movement of each component in a certain specific posture (as shown in the figure). If the specific posture changes, the directional indication will also change accordingly.

[0023] Furthermore, in this invention, the use of terms such as "first" and "second" is for descriptive purposes only and does not specifically refer to any order or sequence, nor is it intended to limit the invention. They are merely used to distinguish components or operations described using the same technical terms, and should not be construed as indicating or implying relative importance or implicitly specifying the number of indicated technical features. Therefore, a feature defined with "first" or "second" may explicitly or implicitly include at least one of those features. Additionally, the technical solutions of various embodiments can be combined with each other, but only if they are feasible for those skilled in the art. If a combination of technical solutions is contradictory or impossible to implement, such a combination should be considered nonexistent and not within the scope of protection claimed by this invention.

[0024] To further understand the content, features, and effects of this invention, the following embodiments are provided, and detailed descriptions are given below in conjunction with the accompanying drawings: Please see Figure 1 This invention provides a method for recognizing neurophysiological signals related to language development in premature infants, comprising the following steps: S100. Acquire the initial neurophysiological response information and auditory stimulation parameters collected during auditory stimulation of the premature infant. The initial neurophysiological response information refers to the raw brain activity data collected using neurophysiological techniques such as electroencephalography (EEG) and event-related potentials (ERPs) during the first auditory stimulation of the premature infant. This reflects the immediate response of the premature infant's brain to auditory stimulation and contains rich neural activity patterns. Auditory stimulation parameters refer to the various attributes of the auditory stimulus used to induce a neural response in the premature infant, such as the frequency, intensity, duration, presentation interval, and stimulus type (e.g., speech and pure tone). The setting of these parameters directly affects the nature and intensity of the neural response. Specifically, a standardized auditory stimulation protocol can be used, with preset auditory stimulation parameters such as pure tones of different frequencies and speech fragments of different phonemes, and EEG equipment can be used to collect the premature infant's EEG signals under stimulation. Alternatively, professionals can manually set the initial auditory stimulation parameters based on the preliminary assessment results of the premature infant and record the premature infant's neurophysiological response.

[0025] S200. Extract language development feature information from the initial neurophysiological response information. This language development feature information is extracted from the initial neurophysiological response information to identify biological markers or patterns directly related to the preterm infant's language processing ability, such as activation patterns in specific brain regions and the latency and amplitude of event-related potentials (ERPs). This reflects the preterm infant's ability to perceive, recognize, and understand speech. This step is achieved through signal processing and analysis of the acquired raw neurophysiological data. For example, the raw EEG signals are preprocessed with filtering and noise reduction, and then specific waveforms or frequency components related to language processing are identified through time-domain analysis (such as ERP analysis) or frequency-domain analysis (such as power spectrum analysis). Alternatively, experienced neurophysiologists can visually examine the EEG waveforms to identify features related to language development, such as the latency and amplitude of P1, N1, and P2 waveforms.

[0026] S300. Based on the language development characteristics information, determine the premature infant's neural response state to the current stimulus. The neural response state refers to the comprehensive performance of the premature infant's brain in response to the current auditory stimulus, including its activity level, processing efficiency, and sensitivity to the stimulus. It reflects the premature infant's current neurophysiological and cognitive state. This step is achieved by comparing the extracted language development characteristics information with preset neurodevelopmental standards. For example, a database containing language development characteristics of premature infants at different developmental stages can be established. The current premature infant's characteristics information can be matched with the data in the database to determine whether their neural response state is normal, delayed, or advanced. Alternatively, expert evaluation can be used, where clinicians or neuropsychologists, based on the language development characteristics information and their professional knowledge and experience, make a qualitative or semi-quantitative judgment on the premature infant's neural response state.

[0027] S400. Based on the neural response state, adjust the auditory stimulation parameters to obtain the adjusted auditory stimulation parameters. If the premature infant's neural response state to the current stimulus indicates insensitivity to low-frequency sounds, the auditory stimulation parameters can be manually adjusted to increase the intensity or duration of the low-frequency stimulus. Alternatively, according to preset adjustment rules, the auditory stimulation parameters can be automatically adjusted when a certain neural response state is detected. For example, if the premature infant's response to a certain speech stimulus is not obvious, it can be adjusted to a simpler and more repetitive speech stimulus to reduce its cognitive load.

[0028] S500. Using the adjusted auditory stimulation parameters, the premature infant is subjected to auditory stimulation again to determine the neurophysiological adjustment response information. The neurophysiological adjustment response information refers to the neurophysiological data collected when auditory stimulation is performed again after adjusting the auditory stimulation parameters according to the neural response state. This data is used to assess the impact of the adjusted stimulation on the preterm infant's language development. Specifically, the implementation method for the first auditory stimulation is similar, but the adjusted auditory stimulation parameters are used. For example, an operator can manually set the auditory stimulation device according to the adjusted parameters and collect the preterm infant's neurophysiological response data again. Alternatively, an automated system can automatically generate and present auditory stimulation based on the adjusted parameters and simultaneously collect neurophysiological data.

[0029] S600. Based on the neurophysiological adjustment response information, the neurodevelopmental status of language development in premature infants is identified, resulting in a neurodevelopmental recognition result for language development in premature infants. This recognition result is a comprehensive judgment and assessment of the neurodevelopmental level, presence of abnormalities, and types of abnormalities in premature infants' language development, based on the neurophysiological adjustment response information. Specifically, this is achieved through in-depth analysis of the adjusted neurophysiological response information. For example, pattern recognition can be performed on the adjusted response information, and machine learning algorithms can be used to classify and cluster the data to identify specific patterns in the premature infant's language development. These patterns can then be compared with normal developmental patterns to determine if any developmental abnormalities exist. Alternatively, expert consultation can be used, where multiple clinical experts jointly analyze the adjusted neurophysiological response information, combining it with the premature infant's clinical manifestations and other assessment results to comprehensively judge the neurodevelopmental status of their language development.

[0030] This invention acquires initial neurophysiological response information in real time and extracts language development features from it to determine the neural response state of preterm infants to current stimuli. This allows the assessment to fully consider the immediate physiological and cognitive state of the preterm infant. Based on the preterm infant's neural response state, auditory stimulation parameters are dynamically adjusted, and re-stimulation is performed to obtain neurophysiological adjustment response information. This allows the stimulation protocol to be optimized according to the actual response of the preterm infant, thereby more effectively inducing and capturing neural activity related to language development. For example, if the preterm infant does not respond well to a certain stimulus, the intensity, frequency, or type of the stimulus can be automatically adjusted to find the most suitable stimulation pattern for the preterm infant, thereby obtaining more representative neurophysiological data. Finally, through in-depth analysis of the neurophysiological adjustment response information, precise identification of the neurodevelopment of language development in preterm infants can be achieved. This not only identifies potential developmental delays but also provides clinicians with more detailed and personalized diagnostic evidence to guide the formulation of early intervention measures. Thus, it provides a more scientific and reliable tool for the assessment and intervention of language development in preterm infants.

[0031] In some embodiments of this application described above, the step of extracting language development feature information from the initial neurophysiological response information includes: Signal analysis is performed on the initial neurophysiological response information to determine the original language development features, microphysiological indicators, and background neurophysiological information. Specifically, various signal processing techniques, such as time-domain analysis, frequency-domain analysis, time-frequency analysis, independent component analysis, or principal component analysis, are used to decompose and identify the initial neurophysiological response information collected from premature infants. Through this analysis, original language development features directly related to language development, microphysiological indicators related to the physiological state of premature infants (e.g., heart rate, respiration, and electromyography), and background neurophysiological information related to the environment or equipment (e.g., power line noise and motion artifacts) can be separated from the complex initial neurophysiological response information. The aim is to decompose the raw, mixed neurophysiological signals into components from different sources for subsequent targeted processing.

[0032] Correlation analysis is performed on the microscopic physiological indicator signals and background neuroelectrophysiological information to identify interfering characteristic signals. Specifically, statistical methods or machine learning algorithms, such as correlation analysis, regression analysis, or pattern recognition, are used to evaluate the relationship between the microscopic physiological indicator signals and background neuroelectrophysiological information. The correlation analysis aims to identify and quantify the influence patterns of non-language development-related signals on the initial neuroelectrophysiological response information, thereby identifying interfering characteristic signals. These interfering characteristic signals are unrelated to language development but are mixed in with the initial neuroelectrophysiological response information and may negatively impact the extraction of language development features. The purpose is to accurately identify and characterize the interfering components that need to be suppressed or removed.

[0033] Based on the aforementioned interference signals, the original language development feature information is subjected to feature filtering processing to obtain language development feature information. Specifically, various signal filtering techniques or denoising algorithms, such as adaptive filtering, notch filtering, wavelet denoising, independent component separation, or blind source separation, are used to remove the identified interference signals from the original language development feature information. Through filtering, the contamination of the original language development feature information by non-language development-related factors can be effectively eliminated or significantly reduced, thereby obtaining purer and more accurate language development feature information. The aim is to improve the signal-to-noise ratio and effectiveness of the language development feature information, providing high-quality input for subsequent determination of neural response states.

[0034] Specifically, when premature infants receive auditory stimulation, their electroencephalogram (EEG) data is collected as initial neurophysiological response information. First, signal analysis is performed on the collected EEG data. This includes performing Fourier transform on the EEG signals to analyze their frequency domain components, or using independent component analysis (ICA) to separate different signal sources. Analysis can identify event-related potential (ERP) components associated with auditory stimulation as primitive features of language development, while artifacts generated by the premature infant's heartbeat (ECG artifacts) are separated as microphysiological indicators, and electromyographic artifacts caused by 50Hz or 60Hz noise from the power line and the premature infant's body movement are separated as background neurophysiological information. Second, correlation analysis is performed on the separated ECG and electromyographic artifacts (as microphysiological indicators and background neurophysiological information). For example, by determining the correlation between artifacts and specific frequency components or time points in the EEG signal, the specific interference patterns and intensities affecting primitive features of language development can be identified, thus determining the interference characteristic signals. Finally, based on the identified interference signals, the original language development feature information is filtered. For example, an adaptive filtering algorithm can be used to subtract interference components from the original language development feature information in real time based on the characteristics of electrocardiogram and electromyography artifacts. After filtering, the obtained language development feature information will be purer and can more accurately reflect the neural response of premature infants to auditory stimuli, providing high-quality data for subsequent language development assessment.

[0035] This embodiment employs refined signal analysis of the initial neurophysiological response information, decomposing the complex raw signal into three main components: raw language development features, microscopic physiological indicators, and background neurophysiological information. Subsequently, by performing correlation analysis on the microscopic physiological indicators and background neurophysiological information, non-language development-related interference signals can be accurately identified and quantified. Finally, using the identified interference signals, targeted feature filtering is applied to the raw language development features, effectively removing noise and artifacts from the raw signal. This results in purer and more accurate language development features extracted from the initial neurophysiological response information of premature infants, significantly reducing the interference of non-language developmental factors on the recognition results.

[0036] In some embodiments of this application described above, the step of performing signal analysis on the initial neurophysiological response information to determine the original language development feature information, microphysiological index signals, and background neurophysiological information includes: Physiological index analysis was performed on the initial neurophysiological response information to obtain various physiological index information. Specifically, through professional physiological signal processing techniques, multiple physiological indicators related to language development in preterm infants were identified and quantified from the raw neurophysiological data. These physiological index information may include, but is not limited to, specific frequency band activity of electroencephalograms (EEG), latency and amplitude of event-related potentials (ERPs), and heart rate variability (HRV), aiming to reflect the neurophysiological state of preterm infants from multiple dimensions.

[0037] Information analysis is performed on each physiological indicator to determine initial developmental information, initial physiological indicator information, and initial neuroelectrophysiological information. This step involves in-depth analysis of each physiological indicator using specific algorithms and models to distinguish and extract initial developmental information directly related to language development, initial physiological indicator information reflecting the individual's physiological state, and background neuroelectrophysiological activity information. For example, spectral analysis of EEG signals can be performed to obtain power spectral density at different frequency bands as initial developmental information; waveform analysis of ERP components can be performed to obtain latency and amplitude as initial physiological indicator information; and non-specific background noise can be separated as initial neuroelectrophysiological information.

[0038] Information extraction and processing are performed on initial developmental information, initial physiological indicator information, and initial neuroelectrophysiological information to obtain original language development feature information, microscopic physiological indicator signals, and background neuroelectrophysiological information. This step involves using specialized data processing modules or algorithms to refine and classify the various types of initial information obtained from the analysis. For example, methods such as Independent Component Analysis (ICA) or Principal Component Analysis (PCA) can be used to separate pure original language development feature information from complex physiological signals, extract microscopic physiological indicator signals from physiological indicators, and extract background neuroelectrophysiological information from background activities. The purpose is to provide a clear and independent data foundation for subsequent determination of interference feature signals and feature filtering.

[0039] This embodiment employs a multi-level and multi-dimensional detailed analysis of initial neurophysiological response information. First, it identifies various physiological indicators at a macroscopic level. Then, it analyzes each indicator in depth, decomposing it into initial developmental information directly related to language development, initial physiological indicators reflecting the individual's physiological state, and initial background neurophysiological information. This allows for the subsequent extraction of these information separately, resulting in a more accurate and comprehensive acquisition of original language development features, microscopic physiological indicator signals, and background neurophysiological information. This lays a solid foundation for subsequent identification of interference features and feature filtering.

[0040] In some embodiments of this application described above, the step of performing correlation analysis on the microscopic physiological indicator signals and background neuroelectrophysiological information to determine interference characteristic signals includes: The microphysiological indicator signals and background neurophysiological information are verified separately to obtain verified microphysiological indicator signals and background neurophysiological information. Specifically, information verification refers to the quality check and validity assessment of the acquired microphysiological indicator signals and background neurophysiological information. For example, the integrity of the signals, noise levels, and the presence of artifacts or abnormal fluctuations can be checked. Its purpose is to ensure that the data used for subsequent correlation analysis is reliable and physiologically meaningful. Through information verification, erroneous or inaccurate analysis results caused by data quality issues can be effectively eliminated, thereby obtaining verified microphysiological indicator signals and verified background neurophysiological information.

[0041] Based on a pre-defined physiological correlation model, correlation analysis is performed on verified microscopic physiological indicator signals and background neuroelectrophysiological information to identify interfering characteristic signals. The pre-defined physiological correlation model is a mathematical model or algorithm pre-established to describe the relationship between microscopic physiological indicator signals and background neuroelectrophysiological information. This model is constructed and trained based on extensive clinical data, physiological knowledge, or machine learning methods. For example, the model can be a statistical regression model, a neural network model, or a causal model based on physiological principles. Its purpose is to accurately capture the intrinsic connections between different physiological signals, thereby effectively identifying interfering characteristic signals related to language development features in the correlation analysis.

[0042] This embodiment first verifies the microscopic physiological indicator signals and background neuroelectrophysiological information to ensure the quality and reliability of the data used for correlation analysis. This allows subsequent correlation analysis to be performed on a clean and accurate data basis. Based on this, a pre-defined physiological correlation model is used to perform correlation analysis on the verified microscopic physiological indicator signals and background neuroelectrophysiological information, enabling more precise identification of the physiological correlation patterns between the two, thereby accurately determining the interference feature signals. This effectively improves the accuracy and robustness of interference feature signal identification.

[0043] In some embodiments of this application described above, the step of using the adjusted auditory stimulation parameters to re-stimulate the premature infant and determine the neurophysiological adjustment response information includes: Using the adjusted auditory stimulation parameters, the premature infant is subjected to further auditory stimulation to obtain auditory stimulus presentation parameters and each neurophysiological response data point. This step refers to the specific attributes of the auditory stimulus actually applied to the premature infant during this stimulation process, such as the frequency, intensity, duration, interval, and presentation method of the stimulus. Each neurophysiological response data point refers to the neural activity data collected in real time at a specific time point or time window during the re-auditory stimulation using neurophysiological equipment (such as electroencephalography (EEG) and event-related potentials (ERPs)). This reflects the immediate physiological response of the premature infant's brain to auditory stimulation.

[0044] State analysis was performed on each neurophysiological response data point to obtain a physiological state stability coefficient. This coefficient measures the stability of the physiological state reflected by each data point at the time of acquisition. For example, stability was assessed by analyzing indicators such as noise level, baseline drift, and artifacts (e.g., eye movement and electromyography). Higher stability indicates a greater likelihood that the data point reflects a genuine neural response. The aim is to select high-quality neurophysiological response data to provide a reliable basis for subsequent analysis.

[0045] Weighted analysis is performed on each physiological state stability coefficient to determine the weight of each neurophysiological response data point. The weight of each neurophysiological response data point refers to the importance or influence assigned to that data point in the subsequent information determination process based on its corresponding physiological state stability coefficient. For example, physiological state stability coefficients with higher stability are assigned higher weights, and vice versa. Weighted analysis employs various algorithms, such as those based on statistical methods, machine learning models, or expert rules. Its purpose is to ensure that more reliable and stable data points play a greater role in determining neurophysiological adjustment response information, thereby improving the accuracy of the final results.

[0046] Based on the weight of each neurophysiological response data point, the auditory stimulus presentation parameters, and each neurophysiological response data point, neurophysiological adjustment response information is determined. Specifically, this information is derived by comprehensively considering the auditory stimulus presentation parameters, each neurophysiological response data point, and its corresponding weight. For example, the weighted neurophysiological response data points can be correlated with the auditory stimulus presentation parameters to construct comprehensive information that fully reflects the neural response pattern of preterm infants to adjusted auditory stimuli. The aim is to provide optimized and refined neural response data for subsequent identification of the neurodevelopment of language development in preterm infants.

[0047] This embodiment effectively improves the quality and reliability of neurophysiological adjustment response information by performing detailed state analysis and weight allocation on the raw neurophysiological response data points collected during re-auditory stimulation. Specifically, firstly, auditory stimulus presentation parameters and each neurophysiological response data point are acquired, providing basic data for subsequent analysis. Subsequently, by performing state analysis on each neurophysiological response data point, the physiological stability of each data point can be identified and quantified, thereby obtaining the stability coefficient of each physiological state. This effectively eliminates or reduces the impact of physiological artifacts and noise on data quality. Weight analysis is performed on each physiological state stability coefficient, so that more stable and reliable neurophysiological response data points occupy a larger proportion in the final determination of adjustment response information, thereby reducing the interference of low-quality data on the overall results. Finally, by comprehensively considering the weighted neurophysiological response data points and the actual auditory stimulus presentation parameters, the neural response pattern of preterm infants to adjusted auditory stimuli, i.e., neurophysiological adjustment response information, can be constructed more accurately and comprehensively.

[0048] In some embodiments of this application described above, the step of determining neurophysiological adjustment response information based on the weight of each neurophysiological response data point, the auditory stimulus presentation parameters, and each neurophysiological response data point includes: Based on the weight of each neurophysiological response data point, the auditory stimulus presentation parameters, and each neurophysiological response data point, the physiological developmental state and original physiological adjustment information are obtained. The physiological developmental state refers to the physiological and neurodevelopmental stage of the premature infant at the current time point, such as gestational age, corrected gestational age, birth weight, and the presence of complications. This information can be obtained from medical records, clinical observation, or through preliminary analysis of the neurophysiological response data points. The original physiological adjustment information refers to the neural activity data directly reflecting the auditory stimulus from the neurophysiological response data points, before considering individual developmental differences.

[0049] Based on the stated physiological developmental state and the pre-defined developmental state model, a confidence level for physiological development is determined. Specifically, the pre-defined developmental state model is a reference model established based on extensive clinical data and developmental patterns of preterm infants, used to assess the expected patterns and range of variation in neurophysiological responses under specific physiological developmental states. By comparing and analyzing the current physiological developmental state of the preterm infant with the pre-defined developmental state model, the reliability of its neural responses and the degree of deviation from normal developmental patterns can be quantified, thereby obtaining the confidence level for physiological development. This confidence level can be a value between 0 and 1, with a higher value indicating that the neural response data under the current physiological developmental state is more reliable or more in line with the expected developmental pattern.

[0050] Using the aforementioned physiological development confidence level, the original physiological adjustment information is adjusted to obtain neuroelectrophysiological adjustment response information. This adjustment can involve weighting, modifying, or filtering the original information. The aim is to eliminate or reduce interference caused by individual developmental differences and physiological fluctuations, ensuring that the adjusted neuroelectrophysiological adjustment response information more accurately reflects the true language development neurological state of premature infants.

[0051] Specifically, when performing auditory stimulation on preterm infants with a corrected gestational age of 32 weeks, after obtaining the weights of each neurophysiological response data point, auditory stimulation presentation parameters, and each neurophysiological response data point, the current physiological developmental status (e.g., in the early prelingual developmental stage with a slight risk of delayed auditory processing) and original physiological adjustment information were first determined based on the information and the infant's clinical data (such as gestational age at birth and the presence of brain injury). Subsequently, the preterm infant's physiological developmental status was compared with a pre-defined developmental status model. The pre-defined developmental status model included baselines and ranges of variation in neurophysiological responses for preterm infants with different corrected gestational ages and different complication conditions. Through comparison, the physiological developmental confidence level of the preterm infant's current neurophysiological response data was determined to be 0.85, indicating that the data has high reliability at the current developmental stage, but some degree of correction is still needed to eliminate individual differences. Finally, using the physiological development confidence score of 0.85, the raw information on physiological adjustment was weighted. For example, the signal intensity of specific frequency bands or time windows related to auditory processing delay in the raw information was appropriately adjusted to obtain more accurate neurophysiological adjustment response information that better reflects the individual case of the preterm infant. The adjusted information will be used for subsequent language development neurorecognition to ensure the accuracy of the assessment results.

[0052] This embodiment refines the original physiological adjustment information by introducing physiological developmental states and a preset developmental state model, and determining the physiological development confidence level based on these models. Since the language developmental neurological state of premature infants is influenced by their individual physiological developmental stage and potential developmental abnormalities, information extracted directly from neurophysiological response data points may not fully reflect their true situation. By first acquiring the physiological developmental state and original physiological adjustment information of premature infants, individual differences can be incorporated into the reference. Subsequently, by comparing the physiological developmental state with the preset developmental state model, the reliability of the current neural response data and its conformity with the expected developmental pattern can be assessed, thereby generating a physiological development confidence level. Because this confidence level can quantify the impact of individual developmental differences on neural response data, subsequent adjustments to the original physiological adjustment information become possible, effectively filtering out or correcting noise and biases caused by individual differences, ensuring that the final neurophysiological adjustment response information is more targeted and accurate.

[0053] In some embodiments of this application described above, the step of determining the confidence level of physiological development based on the physiological developmental state and a preset developmental state model includes: The physiological developmental status is analyzed to determine the individual developmental stage and type of language developmental abnormality in premature infants. Specifically, the physiological developmental status is analyzed to gain a deeper understanding of the current developmental characteristics of premature infants. This includes identifying their individual developmental stage, such as classifying them into a specific developmental period based on factors like corrected gestational age, birth weight, and complications. Simultaneously, it is also necessary to identify the type of language developmental abnormality, such as the presence of language comprehension difficulties, expression difficulties, or mixed disorders.

[0054] Developmental model coefficients are determined based on the individual developmental stage and language developmental abnormality type of the preterm infant. These developmental model coefficients are parameters used to personalize and adjust the general developmental model to more accurately reflect the actual developmental trajectory of a specific preterm infant. For example, the neurophysiological response patterns of preterm infants at different developmental stages or with different abnormality types may differ significantly; by introducing specific developmental model coefficients, these individual differences can be captured.

[0055] By adjusting the developmental model coefficients, a pre-defined developmental baseline model is obtained, resulting in a pre-defined developmental state model. This pre-defined developmental baseline model is a general model representing the neurophysiological response patterns of language development in typical children or preterm infants. Through adjustments to the developmental model coefficients, this baseline model is personalized to a specific pre-defined developmental state model applicable to the current preterm infant, enabling more accurate prediction and assessment of the infant's language development.

[0056] Based on the stated physiological developmental state and the preset developmental state model, the physiological development confidence level is determined. Specifically, the physiological development confidence level is an indicator that measures the degree of matching between the preterm infant's current physiological developmental state and the personalized preset developmental state model; it quantifies the reliability of the recognition results for the preterm infant's language development and neurodevelopment.

[0057] This embodiment analyzes the physiological development of premature infants in detail, identifying their individual developmental stages and types of language development abnormalities, thereby enabling the targeted determination of developmental model coefficients. Due to these personalized developmental model coefficients, the preset basic developmental model can be precisely adjusted to better reflect the actual situation of the premature infant. Therefore, when determining the confidence level of physiological development, it is no longer a simple comparison between the premature infant's physiological development and a general model, but rather a comparison with a personalized model that fully considers individual differences and specificities. This allows for a more accurate reflection of the premature infant's true developmental level, thereby improving the accuracy and reliability of the confidence level of physiological development.

[0058] In some embodiments of this application described above, the step of determining the neural response state of a premature infant to a current stimulus based on the language development characteristic information includes: The language development feature information is processed using time-frequency transformation to obtain neural activity intensity information at different time points and within different frequency ranges. This step involves transforming the original language development feature information from a single dimension of the time or frequency domain to a two-dimensional time-frequency plane, simultaneously revealing the frequency components and intensity changes of the signal at different time points. The aim is to obtain richer and more comprehensive dynamic information on neural activity. For example, wavelet transform, short-time Fourier transform, or Hilbert-Huang transform methods can be used to obtain neural activity intensity information at different time points and within different frequency ranges.

[0059] This step involves identifying key feature components within the language development characteristics information. Specifically, it involves extracting specific patterns, waveforms, or frequency bands of activity that significantly indicate the neural response status of preterm infants' language development from the time-frequency converted neural activity intensity information or from the original language development characteristics information. For example, it can identify the latency, amplitude, or topological distribution of specific peaks in event-related potentials (ERPs), or identify power spectral density changes in specific frequency bands. The aim is to focus on the most important neurophysiological indicators for language development assessment.

[0060] Based on the information on neural activity intensity and key feature components at different time points and frequency ranges, neurodevelopmental characteristics are determined. This step involves integrating the dynamic information revealed by time-frequency analysis with the identified static or dynamic key feature components to form a comprehensive neurodevelopmental feature vector or pattern. The aim is to construct a feature set that comprehensively reflects the neural response state of preterm infants' language development.

[0061] Using a pre-defined state model, the neurodevelopmental characteristics are analyzed to obtain the neural response state of the premature infant to the current stimulus. The pre-defined state model is a machine learning model, such as a support vector machine (SVM), neural network, decision tree, or hidden Markov model (HMM), which is established by training on the neurodevelopmental characteristics of a large number of premature infants with known developmental states. Its purpose is to classify or regress the current neurodevelopmental characteristics of the premature infant based on existing knowledge and patterns, thereby objectively and accurately determining its neural response state to auditory stimuli, such as normal development, developmental delay, or the presence of a specific abnormal pattern.

[0062] Specifically, the study obtained initial neurophysiological response information of a preterm infant to a specific auditory stimulus (e.g., the 'ba' syllable) and extracted language development features from it. To more accurately determine the preterm infant's neural response to the stimulus, wavelet transform processing was first applied to the language development features to obtain the energy distribution of neural activity at different time scales (corresponding to different frequency ranges), i.e., the intensity of neural activity at different time points and within different frequency ranges. For example, a significant increase in the activity intensity of the theta band (4-8 Hz) was observed within 200-300 milliseconds after stimulation, which may be related to auditory attention and early semantic processing. Simultaneously, key feature components were identified from the language development features, such as the amplitude and latency of the N1 wave (approximately 100 milliseconds latency) and the P2 wave (approximately 200 milliseconds latency) in the event-related potential (ERP). The amplitude of the N1 wave reflects the early processing capacity of the auditory cortex, while the P2 wave is associated with higher-level cognitive processing. Subsequently, the time-frequency intensity information obtained from wavelet transform (such as the power change of the theta band) is integrated with the identified key feature components such as the amplitude and latency of the N1 and P2 waves to form a multi-dimensional neurodevelopmental feature vector. For example, this vector may include the N1 amplitude, P2 latency, and the average power of the theta band 200-300 ms after stimulation. Finally, a pre-trained support vector machine (SVM) model is used as the preset state model to analyze this neurodevelopmental feature vector. This SVM model has been trained on a large amount of neuroelectrophysiological data from preterm infants known to have normal development and preterm infants known to have language developmental delays, learning feature patterns under different developmental states. Through the classification of this model, the neural response state of the preterm infant to the current stimulus can be obtained, for example, judging whether the language development neural state of the preterm infant is normal development or mild developmental delay, thus providing a precise basis for subsequent adjustment of auditory stimulation parameters.

[0063] This embodiment, through time-frequency conversion processing of language development feature information, can meticulously characterize the dynamic changes of neurophysiological signals in both time and frequency dimensions, thereby revealing neural activity patterns that are difficult to capture with existing single time-domain or frequency-domain analyses. Simultaneously, by identifying key feature components, it can focus on neurophysiological indicators that have specific indicative roles in language development. Therefore, combining time-frequency dynamic information with key feature component information allows for the construction of more comprehensive and refined neurodevelopmental features, effectively avoiding the one-sidedness that may result from single-feature analysis. Finally, by analyzing these comprehensive features using a pre-defined state model, the judgment of the neural response state of preterm infants is no longer a simple threshold comparison, but rather based on complex pattern recognition, thus enabling a more accurate and objective assessment of the language development neural state of preterm infants.

[0064] In some embodiments of this application, the steps of obtaining the initial neurophysiological response information and auditory stimulation parameters collected during auditory stimulation of premature infants include: This study aims to acquire raw neurophysiological response information and raw auditory stimulation parameters of premature infants during auditory stimulation. Specifically, using specialized neurophysiological acquisition equipment (such as electroencephalography (EEG) and event-related potential (ERP) devices), the electrical activity data of the cerebral cortex of premature infants is recorded in real time while they receive auditory stimulation. Simultaneously, the specific parameters of the applied auditory stimulation, such as stimulation type, frequency, intensity, and duration, are also recorded. The raw neurophysiological response information can be understood as the unprocessed raw electroencephalogram (EEG) signals, while the raw auditory stimulation parameters refer to the initial, unadjusted stimulation settings.

[0065] The raw neurophysiological response information and raw auditory stimulus parameters are preprocessed to obtain preprocessed raw neurophysiological response information and preprocessed raw auditory stimulus parameters. Preprocessing includes, but is not limited to, filtering the raw neurophysiological response information (e.g., bandpass filtering to remove power line interference and baseline drift), denoising (e.g., independent component analysis (ICA) to remove electrooculography and electromyography artifacts), baseline correction, and segmentation to improve the signal-to-noise ratio and usability. For the raw auditory stimulus parameters, preprocessing may involve standardizing the parameter format, filling in missing values, or preliminary screening for outliers to ensure the parameters' standardization and completeness.

[0066] The preprocessed raw neurophysiological response information and preprocessed raw auditory stimulation parameters are validated to obtain initial neurophysiological response information and auditory stimulation parameters. Specifically, the purpose of validation is to further confirm the validity and reliability of the data. For example, the quality of the preprocessed neurophysiological signals can be assessed to check for residual artifacts or channels with poor signal quality, and these can be removed or corrected according to preset thresholds. For auditory stimulation parameters, validation may include comparison with a preset stimulation scheme to ensure that the actual applied stimulation is consistent with the plan, or to check whether the parameter range is reasonable. Through validation, low-quality or unacceptable data can be effectively eliminated, thereby obtaining high-quality initial neurophysiological response information and auditory stimulation parameters, laying a solid foundation for subsequent language development feature extraction.

[0067] Specifically, during auditory stimulation of a premature infant, multiple electrodes attached to the infant's scalp are used to collect real-time electroencephalogram (EEG) signals, obtaining raw neurophysiological response information. Simultaneously, raw auditory stimulation parameters such as frequency, intensity, and duration of the specific speech stimulus (e.g., phonemes ba or da) are recorded. Next, the collected raw neurophysiological response information is preprocessed. Specifically, a 0.5-30Hz bandpass filter is used to remove low-frequency drift and high-frequency electromyographic interference, and independent component analysis (ICA) is used to identify and remove artifacts caused by eye movements or blinking. The raw auditory stimulation parameters are standardized to a standard format, and any outliers outside the reasonable range are checked. Finally, the preprocessed neurophysiological signals are validated. For example, signal quality indicators (such as signal-to-noise ratio) for each channel can be determined, and channel data with signal quality below a preset threshold (e.g., signal-to-noise ratio less than 5 dB) are discarded. Simultaneously, the auditory stimulus parameters are verified to ensure a complete match with the preset stimulus scheme. For example, the error between the actual played stimulus volume and the target volume is within an acceptable range. This ultimately yields high-quality and highly reliable initial neurophysiological response information and auditory stimulus parameters, which are used for subsequent language development feature extraction and pattern recognition.

[0068] This embodiment establishes the foundation for the entire recognition process by acquiring raw neurophysiological response information and raw auditory stimulus parameters, ensuring the source of the data. Secondly, the preprocessing step systematically removes interference components from the raw signal, improving signal purity and making the language development feature information extracted from the signal more accurate. For example, filtering and denoising can effectively separate neural activities related to language development, preventing them from being masked by irrelevant physiological activities or environmental noise. Finally, a verification step acts as a quality control checkpoint, performing a final validity evaluation of the preprocessed data to ensure that only high-quality and highly reliable data is used for subsequent analysis. Therefore, this invention guarantees the quality of input data from the source, providing a solid data foundation for subsequent accurate pattern recognition.

[0069] Based on any of the above embodiments, a method for recognizing neurophysiological signals related to language development in premature infants is provided. Figure 2 The present invention also provides a neurophysiological signal pattern recognition system for language development in premature infants, the system comprising an information acquisition module 210, an information extraction module 220, a state determination module 230, a parameter adjustment module 240, an information stimulation module 250, and a result recognition module 260.

[0070] The information acquisition module 210 is used to acquire the initial neurophysiological response information and auditory stimulation parameters collected when auditory stimulation is performed on premature infants.

[0071] The information extraction module 220 is used to extract language development feature information from the initial neurophysiological response information.

[0072] The state determination module 230 is used to determine the neural response state of the premature infant to the current stimulus based on the language development feature information.

[0073] The parameter adjustment module 240 is used to adjust the auditory stimulation parameters according to the neural response state to obtain the adjusted auditory stimulation parameters.

[0074] The information stimulation module 250 is used to provide auditory stimulation to the premature infant again using the adjusted auditory stimulation parameters to determine the neurophysiological adjustment response information.

[0075] The result recognition module 260 is used to recognize the neural development of language development in premature infants based on the neurophysiological adjustment response information, and obtain the neural recognition result of language development in premature infants.

[0076] In this embodiment, raw data is collected by the information acquisition module 210, and the information extraction module 220 performs preliminary processing on the data to obtain key features. The state determination module 230 determines the neural response state of the premature infant based on these features. The parameter adjustment module 240 dynamically optimizes the auditory stimulation parameters according to this state, and the information stimulation module 250 uses the adjusted parameters for further stimulation to obtain more accurate response data. Finally, the result recognition module 260 performs in-depth analysis on these data to provide personalized neural recognition results for language development. This invention can effectively address the challenges of large individual differences and frequent fluctuations in the physiological state of premature infants, improve the accuracy and real-time nature of assessment, and thus provide a more reliable basis for subsequent clinical intervention.

[0077] The above embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit it. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some or all of the technical features therein. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention, and they should all be covered within the scope of the present invention specification.

Claims

1. A method for recognizing neuroelectrophysiological signals in the language development of premature infants, characterized in that, Includes the following steps: Acquire the initial neurophysiological response information and auditory stimulation parameters collected when auditory stimulation is applied to preterm infants; Extract language development features from the initial neurophysiological response information; Based on the language development characteristics information, determine the neural response state of the premature infant to the current stimulus; Based on the neural response state, the auditory stimulation parameters are adjusted to obtain the adjusted auditory stimulation parameters; Using the adjusted auditory stimulation parameters, premature infants were subjected to further auditory stimulation to determine the neurophysiological adjustment response information. Based on the neurophysiological adjustment response information, the neurodevelopment of language development in premature infants is identified, and the neurodevelopmental identification results of language development in premature infants are obtained.

2. The method for pattern recognition of neuroelectrophysiological signals in the language development of premature infants according to claim 1, characterized in that, The steps for extracting language development feature information from the initial neurophysiological response information include: Signal analysis was performed on the initial neurophysiological response information to determine the original characteristics of language development, microphysiological index signals, and background neurophysiological information; Correlation analysis was performed on the microscopic physiological index signals and background neurophysiological information to identify interference characteristic signals; Based on the interference feature signal, the original language development feature information is subjected to feature filtering processing to obtain language development feature information.

3. The method for recognizing neurophysiological signals related to language development in premature infants according to claim 2, characterized in that, The steps of performing signal analysis on the initial neurophysiological response information to determine the original language development features, microphysiological index signals, and background neurophysiological information include: Physiological index analysis was performed on the initial neurophysiological response information to obtain various physiological index information; Information analysis was performed on each physiological indicator to determine the initial developmental information, initial physiological indicator information, and initial neuroelectrophysiological information; Information extraction and processing were performed on the initial developmental information, initial physiological indicator information, and initial neuroelectrophysiological information to obtain the original language development feature information, microscopic physiological indicator signals, and background neuroelectrophysiological information.

4. The method for pattern recognition of neuroelectrophysiological signals in the language development of premature infants according to claim 2, characterized in that, The steps for performing correlation analysis between the microscopic physiological indicator signals and background neuroelectrophysiological information to determine interference characteristic signals include: The microscopic physiological index signals and background neurophysiological information are verified separately to obtain the verified microscopic physiological index signals and background neurophysiological information. Based on the pre-set physiological correlation model, correlation analysis is performed on the microscopic physiological indicator signals and background neuroelectrophysiological information after the verification is qualified to determine the interference characteristic signals.

5. The method for pattern recognition of neuroelectrophysiological signals in language development of premature infants according to claim 1, characterized in that, The steps for determining the neurophysiological adjustment response information by re-stimulating the premature infant with the adjusted auditory stimulation parameters include: Using the adjusted auditory stimulation parameters, the premature infant was subjected to auditory stimulation again to obtain auditory stimulation presentation parameters and each neurophysiological response data point; State analysis was performed on each neurophysiological response data point to obtain the stability coefficient of each physiological state; Weight analysis was performed on each physiological state stability coefficient to determine the weight of each neurophysiological response data point. Based on the weight of each neurophysiological response data point, the auditory stimulus presentation parameters, and each neurophysiological response data point, the neurophysiological adjustment response information is determined.

6. The method for pattern recognition of neuroelectrophysiological signals in the language development of premature infants according to claim 5, characterized in that, The steps for determining the neurophysiological adjustment response information based on the weight of each neurophysiological response data point, the auditory stimulus presentation parameters, and each neurophysiological response data point include: Based on the weight of each neurophysiological response data point, the auditory stimulus presentation parameters, and each neurophysiological response data point, the original information on physiological development and physiological adjustment is obtained. Based on the physiological developmental state and the preset developmental state model, determine the confidence level of physiological development; Using the aforementioned physiological development confidence level, the original physiological adjustment information is adjusted to obtain neuroelectrophysiological adjustment response information.

7. The method for pattern recognition of neuroelectrophysiological signals in language development of premature infants according to claim 6, characterized in that, The steps for determining the confidence level of physiological development based on the physiological developmental state and the preset developmental state model include: The physiological developmental status was analyzed to determine the individual developmental stage and type of language developmental abnormality in the premature infant; The developmental model coefficients were determined based on the individual developmental stage and language developmental abnormality type of the premature infant. By using developmental model coefficients, the preset developmental baseline model is adjusted to obtain the preset developmental state model; Based on the physiological developmental state and the preset developmental state model, the confidence level of physiological development is determined.

8. The method for pattern recognition of neuroelectrophysiological signals in language development of premature infants according to claim 1, characterized in that, The steps for determining the neural response state of a premature infant to a current stimulus based on the aforementioned language development characteristics include: The language development feature information is processed by time-frequency conversion to obtain neural activity intensity information at different time points and within different frequency ranges; Identify key feature components in the language development feature information; Based on the information on neural activity intensity and key characteristic components at different time points and within different frequency ranges, neural development characteristics are determined; Using a preset state model, the neural development characteristics are analyzed to obtain the neural response state of the premature infant to the current stimulus.

9. The method for pattern recognition of neuroelectrophysiological signals in the language development of premature infants according to claim 1, characterized in that, The steps for obtaining initial neurophysiological response information and auditory stimulation parameters during auditory stimulation of preterm infants include: To obtain raw neurophysiological response information and raw auditory stimulation parameters collected during auditory stimulation of preterm infants; The original neurophysiological response information and the original auditory stimulus parameters are preprocessed to obtain the preprocessed original neurophysiological response information and the preprocessed original auditory stimulus parameters. The preprocessed original neurophysiological response information and the preprocessed original auditory stimulation parameters are verified to obtain the initial neurophysiological response information and auditory stimulation parameters.

10. A neuroelectrophysiological signal pattern recognition system for language development in premature infants, characterized in that, The system includes: The information acquisition module is used to acquire the initial neurophysiological response information and auditory stimulation parameters collected when auditory stimulation is performed on premature infants; The information extraction module is used to extract language development feature information from the initial neurophysiological response information; The state determination module is used to determine the neural response state of the premature infant to the current stimulus based on the language development feature information. The parameter adjustment module is used to adjust the auditory stimulation parameters according to the neural response state to obtain the adjusted auditory stimulation parameters; The information stimulation module is used to provide auditory stimulation to premature infants again using the adjusted auditory stimulation parameters, and to determine the neurophysiological adjustment response information. The result recognition module is used to identify the neural development of language development in premature infants based on the neurophysiological adjustment response information, and to obtain the neural recognition result of language development in premature infants.

Images