Data fusion modeling system for airway epithelial barrier state evaluation

Abstract

The invention discloses a data fusion modeling system for airway epithelial barrier state evaluation, which comprises the following steps of: acquiring time sequence original data in a physiological microenvironment of an airway epithelial cell culture model, and processing to generate a synchronous multi-modal barrier time sequence data set; extracting and calculating multi-scale biophysical features capable of representing the structure and function of the epithelial barrier from the synchronous multi-modal barrier time sequence data set; constructing a multi-criterion weighted fusion model for generating a barrier health index; and comparing the generated barrier health index with a preset barrier health index state threshold, and outputting an evaluation report through a preset logic rule base. The method has the following advantages and effects that through multi-modal data fusion and multi-criterion weighted modeling, comprehensive and quantitative comprehensive evaluation on the structure and function state of the airway epithelial barrier is achieved, and the accuracy and early warning capacity of barrier health state monitoring are remarkably improved.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of biomedical engineering, and particularly relates to a data fusion modeling system for airway epithelial barrier state evaluation. BACKGROUND

[0002] The airway epithelial barrier is a key defense structure of the respiratory tract, and its functional integrity is crucial for preventing pathogens, allergens and harmful substances from invading deep tissues. The impairment of barrier function is closely related to the occurrence and development of various respiratory diseases (such as asthma, chronic obstructive pulmonary disease and pulmonary infection). Therefore, accurate evaluation of the airway epithelial barrier state is of great significance in disease mechanism research, drug toxicity testing and clinical diagnosis. At present, the main methods for evaluating the barrier state include transmembrane resistance measurement, fluorescent marker permeability test and cell morphological observation. However, the most core problem in practice of these methods is that the existing evaluation methods often rely on a single or limited type of biological indicators, which cannot comprehensively and quantitatively reflect the multi-scale structure and functional state of the epithelial barrier. For example, transmembrane resistance can only provide resistance information, but cannot capture the dynamic morphological changes of intercellular junctions; the fluorescent marker permeability test can evaluate the barrier permeability, but is easily interfered by the characteristics of the marker and difficult to quantify the transport kinetics; the cell morphological image can directly show the structural characteristics, but lacks direct correlation with functional parameters. This single-sided or simple combination of evaluation methods leads to one-sided results and insufficient sensitivity, making it difficult to identify subtle damage or overall functional degradation of the barrier at an early stage, thereby limiting its application in real-time monitoring and early warning. Therefore, there is an urgent need for a new method that can integrate multi-modal data and generate comprehensive quantitative indicators to solve the core problem of the existing technology that the evaluation is not comprehensive and accurate. SUMMARY

[0003] The purpose of the present application is to provide a data fusion modeling system for airway epithelial barrier state evaluation to solve the problems presented in the background.

[0004] The above technical purpose of the present application is achieved by the following technical solution: A data fusion modeling system for airway epithelial barrier state evaluation, comprising: A data acquisition and synchronization preprocessing module: acquiring time series raw data in the physiological microenvironment of the airway epithelial cell culture model, the time series raw data including synchronously acquired transmembrane resistance time series data, fluorescent marker permeability time series data and cell morphological time series image data; performing quality inspection and outlier rejection on the time series raw data, and generating a synchronized multi-modal barrier time series data set after aligning the time reference; a multi-criteria weighted fusion model module: constructing a multi-criteria weighted fusion model, taking the impedance spectrum feature set, the permeability kinetic parameter set and the cell morphological metric set as inputs, and generating a comprehensive and quantitative barrier health index; the barrier health index is a normalized scalar value, which is used to intuitively reflect the overall functional state of the epithelial barrier; a multi-criteria weighted fusion model module: constructing a multi-criteria weighted fusion model, taking the impedance spectrum feature set, the permeability kinetic parameter set and the cell morphological metric set as inputs, and generating a comprehensive and quantitative barrier health index; the barrier health index is a normalized scalar value, which is used to intuitively reflect the overall functional state of the epithelial barrier; a multi-criteria weighted fusion model module: constructing a multi-criteria weighted fusion model, taking the impedance spectrum feature set, the permeability kinetic parameter set and the cell morphological metric set as inputs, and generating a comprehensive and quantitative barrier health index; the barrier health index is a normalized scalar value, which is used to intuitively reflect the overall functional state of the epithelial barrier;

[0005] By adopting the above technical solutions, by synchronously collecting transmembrane resistance time series data, fluorescent marker permeability time series data and cell morphology time series image data, and performing strict quality inspection and abnormal value elimination, the reliability and time series consistency of the multi-modal data source are ensured; the impedance spectrum feature set, the permeability kinetic parameter set and the cell morphological metric set are extracted from the data, which can comprehensively characterize the state of the epithelial barrier from multiple scales of electrical characteristics, permeability dynamics and structural morphology; a multi-criteria weighted fusion model is constructed to generate a normalized barrier health index, which provides an intuitive and quantitative comprehensive index to reflect the overall functional state of the barrier; finally, by comparing the index with the preset threshold and combining the trend discrimination features to output the evaluation report, the accurate classification and early risk warning of the barrier state are realized, effectively overcoming the problems of one-sidedness and insufficient sensitivity of the existing technology which relies on a single index.

[0006] Further provided is that the method further comprises the steps of: a model verification and dynamic adjustment module: in a specific verification period, two operations are performed in parallel: one is to obtain and calculate the current barrier health index according to the data acquisition and synchronous preprocessing module, the multi-scale biophysical feature extraction module and the multi-criteria weighted fusion model module, as a model calculation value; the other is to obtain a reference barrier integrity index at the same time point through an offline biological sampling and analysis method; the model calculation value is compared with the reference barrier integrity index, and a model prediction deviation is calculated; based on the model prediction deviation, the multi-criteria weighted fusion model is dynamically adjusted.

[0007] By adopting the technical scheme, the model calculation and offline biology sampling are performed in parallel in a specific verification period to obtain model calculation values and benchmark barrier integrity indexes and calculate model prediction deviation, real-time verification of the multi-criteria weighted fusion model is realized, model parameters are dynamically adjusted based on the prediction deviation to ensure the accuracy and adaptability of the barrier health index, the feedback mechanism can continuously optimize the model performance to improve the robustness and reliability of the model in long-term monitoring, and thus the practical value and stability of the entire evaluation method in actual application are enhanced.

[0008] Further, in the data acquisition and synchronous preprocessing module, the step of acquiring time-series raw data in the physiological microenvironment of the airway epithelial cell culture model includes: Acquiring time-series transmembrane resistance data containing complex impedance information measured by multi-frequency alternating current excitation, acquiring time-series fluorescence marker permeability data obtained by monitoring the fluorescence intensity accumulation curve of the fluorescence intensity on the basal side over time after introducing a fluorescence marker at a specific time point, and acquiring time-series cell morphology image data containing high-resolution cell images under bright field and / or specific fluorescence markers obtained by timed shooting.

[0009] By adopting the technical scheme, time-series transmembrane resistance data containing complex impedance information, time-series fluorescence marker permeability data based on fluorescence intensity accumulation curves, and time-series high-resolution cell morphology image data are acquired, providing rich and multi-dimensional raw information sources; multi-frequency alternating current excitation measurement can capture more detailed impedance spectrum characteristics, fluorescence intensity accumulation curves reflect the permeability dynamics process, and high-resolution images reveal subtle changes in cell morphology, thereby providing a comprehensive and high-quality data basis for subsequent feature extraction and improving the depth and breadth of the overall evaluation.

[0010] Further, in the data acquisition and synchronous preprocessing module, the step of performing quality inspection and outlier rejection on the time-series raw data includes: For the time-series transmembrane resistance data, the signal-to-noise ratio and phase angle are calculated, and data segments with a signal-to-noise ratio below a first preset threshold or a phase angle beyond a reasonable physiological range are rejected; for the time-series fluorescence marker permeability data, the baseline stability is checked and signal saturation is determined, and data segments with baseline drift exceeding a second preset threshold or signal saturation are rejected; for the time-series cell morphology image data, the focusing clarity is evaluated by calculating the image gradient, and the image quality is evaluated by calculating the local contrast, and image frames with a clarity below a third preset threshold or a contrast below a fourth preset threshold are marked and rejected.

[0011] By adopting the technical scheme, through calculating the signal-to-noise ratio and the phase angle of the transmembrane resistance time sequence data and eliminating the data segments with low signal-to-noise ratio or in the super-physiological range, the baseline stability and signal saturation of the fluorescent marker permeability time sequence data are checked, and the data segments with drift or saturation are eliminated, and through evaluating the focusing clarity and local contrast of the cell morphology time sequence image data, the low-quality image frames are eliminated, and a strict data quality control mechanism is implemented. The operation effectively reduces the interference of noise and abnormal values, ensures the accuracy and reliability of the input data of subsequent feature extraction and model construction, and thus improves the consistency and reliability of the entire evaluation method.

[0012] Further, in the multiscale biophysical feature extraction module, the step of extracting and calculating the multiscale biophysical features capable of representing the structure and function of the epithelial barrier from the synchronized multi-modal barrier time sequence data set comprises: extracting the sliding window average of the steady-state value of the transmembrane resistance, the linear regression slope changing over time, and the ratio of the impedance amplitude under high-frequency and low-frequency excitation from the transmembrane resistance time sequence data to form an impedance spectrum feature set; performing nonlinear fitting on the fluorescent intensity accumulation curve in the fluorescent marker permeability time sequence data to extract the time required for the fluorescent intensity accumulation curve to reach half of the plateau, the maximum first derivative value of the curve in the initial stage, and the facilitated diffusion and active transport rate constants obtained by fitting with a two-compartment model to form a permeability kinetic parameter set; performing image segmentation and feature analysis on the cell morphology time sequence image data to quantitatively calculate the cell boundary length in unit area, the continuity index of the tight junction protein fluorescent signal between cells, and the long-width ratio distribution variance of the cell nucleus to form a cell morphology measurement set; standardizing the feature parameters in the impedance spectrum feature set, the permeability kinetic parameter set, and the cell morphology measurement set to eliminate dimensional differences, and collectively constituting the multiscale biophysical features.

[0013] By adopting the technical scheme, the sliding window average of the steady-state value, the linear regression slope, and the impedance amplitude ratio are extracted from the transmembrane resistance time sequence data to form an impedance spectrum feature set reflecting the changes in the electrical behavior of the barrier; the half-time of the plateau, the maximum first derivative value, and the diffusion rate constant are extracted by nonlinear fitting from the fluorescent marker permeability time sequence data to form a permeability kinetic parameter set describing the dynamic process of the barrier permeability; the cell boundary length, the tight junction protein continuity index, and the cell nucleus long-width ratio distribution variance are quantified from the cell morphology time sequence image data to form a cell morphology measurement set representing the structural integrity of the barrier; and after standardization, these feature parameters constitute the multiscale biophysical features, which comprehensively capture the key information of the barrier state from different dimensions and provide rich and consistent inputs for the fusion model.

[0014] Further provided is that the multi-criteria weighted fusion model module specifically comprises the steps of: performing initial weight allocation based on the analytic hierarchy process, to assign an initial weight coefficient to each characteristic parameter in the impedance spectrum feature set, the permeability dynamics parameter set, and the cell morphology metric set; constructing a principal component analysis model to perform dimensionality reduction processing on the normalized multi-scale biophysical features, and extract the first principal components; wherein, the value is determined by the cumulative variance contribution rate exceeding a preset percentage threshold; the variance contribution rate of each principal component is proportionally distributed back to the original characteristic parameters according to the load proportion of the characteristic parameters, to obtain the contribution weight of each characteristic parameter based on data variability; weighting and combining the initial weight coefficient of each characteristic parameter and the contribution weight to calculate the final fusion weight of each characteristic parameter; obtaining the standardized value of each characteristic parameter through the min-max normalization method; based on the standardized values of all characteristic parameters and their corresponding final fusion weights, performing weighted fusion calculation to generate a barrier health index ranging from 0 to 1.

[0015] By adopting the above technical solution, the initial weight of each characteristic parameter is allocated by the analytic hierarchy process, the data variability contribution weight extracted by the principal component analysis dimensionality reduction is combined, and the final fusion weight is calculated, so as to ensure that the weight allocation is integrated with expert judgment and based on data driving; after the characteristic parameters are standardized, weighted fusion is performed to generate a barrier health index ranging from 0 to 1, thereby providing a standardized and comparable comprehensive index; this method makes the fusion model more scientific and objective, improves the accuracy and interpretability of the barrier health index, and effectively integrates the advantages of multi-source information.

[0016] Further provided is that in the multi-criteria weighted fusion model module, the step of performing initial weight allocation based on the analytic hierarchy process to assign an initial weight coefficient to each characteristic parameter in the impedance spectrum feature set, the permeability dynamics parameter set, and the cell morphology metric set comprises: constructing a hierarchical structure model for evaluating the barrier health state, wherein the hierarchical structure model takes the barrier health index as a target layer, takes the impedance spectrum feature set, the permeability dynamics parameter set, and the cell morphology metric set as a criterion layer, and takes the characteristic parameters in each feature set as a scheme layer; Based on the preset judgment scale, the judgment matrix of the target layer and the criterion layer, and the criterion layer and the scheme layer is constructed; the consistency of the judgment matrix is checked, the eigenvector thereof is calculated, and after the eigenvector is normalized, an initial weight coefficient is assigned to each characteristic parameter in the impedance spectrum feature set, the permeability dynamics parameter set and the cell morphology measurement set.

[0017] By adopting the technical scheme, a hierarchical structure model is constructed with the barrier health index as the target layer, the three feature sets as the criterion layer and the feature parameters as the scheme layer, and based on the preset judgment scale, the judgment matrix is constructed for consistency checking and weight calculation, so that a scientific and reasonable initial weight is assigned to each feature parameter. This structured method ensures the systematicness and logicalness of weight distribution, avoids subjectivity, and enables the initial weight to accurately reflect the relative importance of different features to the barrier health state, thereby laying a solid foundation for the subsequent fusion model.

[0018] Further, a state evaluation and report generation module is provided, which specifically includes the following steps: A preset barrier health index state threshold is called, the barrier health index state threshold includes a health state threshold, a mild damage state threshold and a moderate damage state threshold; the barrier health index is compared with the health state threshold, the mild damage state threshold and the moderate damage state threshold, so as to determine a reference state grade; the reference state grade includes a normal stable state, a slight instability state, a moderate instability state, a severe instability state and a barrier collapse state; A specific trend discrimination feature is selected and a feature abnormal state is marked; The reference state grade and the feature abnormal state are combined and input into the logic rule base for matching query, and according to a preset priority rule, a unique and determined final state classification result is output; According to the final state classification result, a corresponding graded early warning signal is generated, and an evaluation report composed of the final state classification result and the graded early warning signal is output.

[0019] By adopting the technical scheme, the barrier health index is compared with the preset health state threshold to determine the reference state grade, and the abnormal state is marked by combining the selected trend discrimination feature, and the final state classification and the graded early warning signal are output through matching and conflict resolution of the logic rule base; this multi-dimensional evaluation mechanism not only considers the static index value, but also integrates the dynamic trend feature, which can more sensitively identify the early changes and potential risks of the barrier, thereby providing more comprehensive and accurate state evaluation and timely early warning.

[0020] Further, in the state evaluation and report generation module: The specific trend-discriminating features include: extracting a linear regression slope of the transmembrane resistance steady-state value over time from the impedance spectrum features, defined as the transmembrane resistance decline slope; extracting a maximum first derivative value of the fluorescence intensity accumulation curve in the initial stage from the permeability kinetics parameters, defined as the permeability maximum slope; and extracting a tight junction protein fluorescence signal continuity index from the cell morphology metrics, defined as the tight junction continuity index. The labeling operation of the feature abnormal state includes: For the transmembrane resistance decline slope, if the moving average values thereof in consecutive multiple preset time windows are all lower than a preset negative sensitivity threshold, the transmembrane resistance decline slope is labeled as a feature abnormal state with a sustained negative development trend; for the permeability maximum slope, if the moving average values thereof in consecutive multiple preset time windows are all higher than a preset positive sensitivity threshold, the permeability maximum slope is labeled as a feature abnormal state with a sustained positive growth trend; and for the tight junction continuity index, if a single change amount thereof between adjacent two monitoring periods exceeds a preset jump sensitivity threshold, the tight junction continuity index is labeled as a feature abnormal state with a significant negative jump trend.

[0021] By using the above technical solution, the transmembrane resistance decline slope extracted from the impedance spectrum features, the permeability maximum slope extracted from the permeability kinetics parameters, and the tight junction continuity index extracted from the cell morphology metrics are used as the trend-discriminating features, and the features are labeled for the sustained negative development trend, the sustained positive growth trend, and the significant negative jump trend, so that the key dynamic changes of the barrier function can be captured; the trend analysis enhances the sensitivity of the evaluation, and can early identify the subtle damage or deterioration trend of the barrier, thereby achieving more active monitoring and early warning.

[0022] Further, a model verification and dynamic adjustment module is provided, and specifically includes the following steps: In a specific verification period, a first verification operation and a second verification operation are performed in parallel; the first verification operation is to obtain and calculate the current barrier health index as a model calculation value according to the data acquisition and synchronization preprocessing module, the multiscale biophysical feature extraction module, and the multi-criteria weighted fusion model module; and the second verification operation is to obtain a reference barrier integrity index at the same time point by an offline biological sampling and analysis method; The model calculation value and the reference barrier integrity index are compared to calculate a model prediction deviation; the model prediction deviation is an absolute difference or a relative difference between the model calculation value and the reference barrier integrity index; Based on the model prediction deviation, the final fusion weight in the multi-criteria weighted fusion model is dynamically adjusted by using a Bayesian updating algorithm; wherein the Bayesian updating algorithm takes the final fusion weight before adjustment as a prior distribution, takes the benchmark barrier integrity index as an observation value, and obtains the updated final fusion weight by inference.

[0023] By adopting the above technical solution, by performing model calculation and offline biology sampling in parallel in the verification period, the model calculation value and the benchmark barrier integrity index are obtained, the model prediction deviation is calculated, and the final fusion weight is dynamically adjusted by using the Bayesian updating algorithm, thereby realizing self-optimization and continuous improvement of the model; the Bayesian updating combines the prior weight with the observation value to infer the updated weight, so that the model can adapt to data changes and improve prediction accuracy, thereby ensuring the reliability and accuracy of the barrier health index in long-term use.

[0024] In summary, the present application has the following beneficial effects: by multi-modal data fusion and multi-criteria weighted modeling, the present application realizes comprehensive and quantitative comprehensive evaluation of the structure and function state of the airway epithelial barrier, and significantly improves the accuracy and early warning ability of barrier health state monitoring. BRIEF DESCRIPTION OF DRAWINGS

[0025] Figure 1 A main flowchart of the embodiment; Figure 2 A flowchart of the data acquisition and synchronous preprocessing module in the embodiment; Figure 3 A flowchart of the multi-scale biophysical feature extraction module in the embodiment; Figure 4 A flowchart of the multi-criteria weighted fusion model module in the embodiment; Figure 5 A flowchart of the state evaluation and report generation module in the embodiment; Figure 6 A flowchart of the model verification and dynamic adjustment module in the embodiment. DETAILED DESCRIPTION

[0026] The present application will be further described in detail below with reference to the accompanying drawings.

[0027] As shown in the accompanying Figures 1-6 ; The embodiment discloses a data fusion modeling system for airway epithelial barrier state evaluation, comprising the following steps: Data acquisition and synchronization preprocessing module: collect time-series raw data in the physiological microenvironment of the airway epithelial cell culture model, including synchronously collected transmembrane resistance time-series data, fluorescent marker permeability time-series data, and cell morphology time-series image data; perform quality inspection and outlier rejection on the time-series raw data, and generate a synchronized multi-modal barrier time-series data set after aligning the time reference.

[0028] The specific implementation process is as follows: collect time-series raw data in the physiological microenvironment of the airway epithelial cell culture model, including synchronously collected transmembrane resistance time-series data, fluorescent marker permeability time-series data, and cell morphology time-series image data. The collection of these data is used to generate a synchronized multi-modal barrier time-series data set subsequently, and serves as the basis for evaluating the state of the epithelial barrier. The collection of transmembrane resistance time-series data is achieved through multi-frequency alternating current excitation measurement, and the specific process is as follows: use an impedance analyzer to apply an alternating current signal with a frequency range of 1 Hz to 100 kHz, measure the complex impedance of the airway epithelial cell layer, record the data sequence of the amplitude and phase angle changing with time, and thus obtain dynamic information reflecting the barrier resistance characteristics. The collection of fluorescent marker permeability time-series data is completed through a fluorescence spectrum monitoring system, and the specific process is as follows: introduce a fluorescent marker with known molecular weight (such as FITC-dextran) into the apical chamber of the airway epithelial cell model, use a microplate reader to detect the fluorescence intensity in the basal chamber at a certain time, and generate a fluorescence intensity cumulative curve, which is used to quantify the permeation rate and cumulative amount of the marker across the barrier. The collection of cell morphology time-series image data utilizes an inverted microscope combined with a high-resolution CCD camera, and the specific process is as follows: take bright-field images and / or fluorescent images under specific fluorescent markers (such as ZO-1 antibody labeled tight junction proteins) at preset time points (such as every 30 minutes), and obtain high-resolution sequence images of cell monolayer morphology, cell boundary structure, and protein distribution.

[0029] The collected time sequence raw data is subjected to quality inspection and abnormal value elimination, and the specific process is as follows: for the transmembrane resistance time sequence data, the signal-to-noise ratio and phase angle thereof are calculated, if the signal-to-noise ratio is lower than a first preset threshold (preferably, the first preset threshold is 20 dB) or the phase angle exceeds a reasonable physiological range (for example, -90° to 90°), it is determined that the data segment has noise interference or measurement error, and is eliminated; for the fluorescent marker permeability time sequence data, the baseline stability thereof is checked and it is determined whether there is signal saturation, if the baseline drift exceeds a second preset threshold (preferably, the first preset threshold is 10% of the initial fluorescence intensity value) or signal saturation (for example, the fluorescence intensity value approaches the maximum range of the detector) occurs, it is determined that the data segment is unreliable, and is eliminated; for the cell morphology time sequence image data, the focusing clarity thereof is evaluated by calculating the image gradient, and the image quality thereof is evaluated by calculating the local contrast, if the clarity is lower than a third preset threshold (preferably, the third preset threshold is 0.1) or the contrast is lower than a fourth preset threshold (preferably, the fourth preset threshold is 50), it is determined that the image frame is blurred or of poor quality, and is marked and eliminated. After completing the quality inspection and abnormal value elimination, the time stamps of the transmembrane resistance time sequence data, the fluorescent marker permeability time sequence data and the cell morphology time sequence image data are unified to the same macro time reference by using a precise time synchronization protocol (such as the PTP protocol); for data streams with different sampling frequencies (for example, the transmembrane resistance data sampling rate is 1 Hz, the fluorescence data is 0.1 Hz, and the image data is 0.033 Hz), a resampling algorithm based on cubic spline interpolation is used to interpolate all the data to a unified high-frequency time axis (for example, 1 kHz), achieving microsecond-level time synchronization; finally, the transmembrane resistance time sequence data, the fluorescent marker permeability time sequence data and the cell morphology time sequence image data after time reference alignment are associated and packaged according to the time stamps to generate a structured synchronous multi-modal barrier time sequence data set, which is stored in the form of a time sequence, each time point containing corresponding impedance values, fluorescence intensity values and image indexes, ensuring that the multi-modal data is strictly aligned in the time dimension, and providing a consistent and reliable data basis for subsequent feature extraction and fusion modeling.

[0030] Multi-scale biophysical feature extraction module: from the synchronous multi-modal barrier time sequence data set, multi-scale biophysical features capable of representing the structure and function of the epithelial barrier are extracted and calculated; the multi-scale biophysical features include: an impedance spectrum feature set extracted from the transmembrane resistance time sequence data, a permeability kinetic parameter set extracted from the fluorescent marker permeability time sequence data, and a cell morphology metric set extracted from the cell morphology time sequence image data.

[0031] The specific implementation process is as follows: the process of extracting the impedance spectrum feature set from the transmembrane resistance time series data is as follows: the transmembrane resistance time series data is a complex impedance sequence obtained by measuring the multi-frequency alternating current excitation, containing impedance amplitude and phase angle information. First, the transmembrane resistance steady-state value is processed by sliding window average, a sliding window with a window size of N (for example, N=10 time points) is used to calculate the arithmetic mean value of the transmembrane resistance steady-state value in each window, to smooth the instantaneous fluctuation and capture the long-term trend, which is expressed as: ; wherein, represents the sliding window average value of the transmembrane resistance steady-state value calculated at time point represents the size of the sliding window; represents the transmembrane resistance steady-state value measured at a specific time point Secondly, the linear regression slope of the transmembrane resistance with time is calculated, the least square linear fitting is performed on the transmembrane resistance time series data, and the fitting equation ; wherein, represents the transmembrane resistance steady-state value at time point represents the intercept constant; represents the linear regression slope, a positive value indicates that the barrier resistance is enhanced, and a negative value indicates that the barrier resistance is weakened. Finally, the ratio of the impedance amplitude under high-frequency and low-frequency excitation is calculated, the impedance amplitudes and at the high-frequency excitation point (for example, frequency ) and the low-frequency excitation point (for example, frequency ) are selected, and the ratio is calculated; wherein, represents the ratio of high-frequency and low-frequency impedance amplitudes, which reflects the capacitive characteristics of the barrier and is related to the tightness of the intercellular connection. The above three characteristic parameters-sliding window average, linear regression slope and impedance amplitude ratio together constitute the impedance spectrum feature set, which is used to represent the electrophysiological characteristics of the barrier.

[0032] The process of extracting the permeability kinetic parameter set from the fluorescent marker permeability time series data is as follows: the fluorescent marker permeability time series data is a fluorescence intensity accumulation curve obtained by monitoring the penetration process of a fluorescent marker (such as FITC-dextran) in the barrier. First, the fluorescence intensity accumulation curve is fitted by a nonlinear function (such as a Logistic function), and the time parameter required for the fluorescence intensity accumulation curve to reach half of the platform is extracted, which represents the time for the marker to penetrate to half of the maximum cumulative amount, reflecting the permeability rate. Secondly, the maximum first derivative value of the curve in the initial stage is calculated, and the first derivative is differentiated by numerical differentiation​​ , represents the fluorescence intensity , represents the first derivative of the fluorescence intensity over time, this value represents the peak of permeation rate, which is used to evaluate the initial permeability of the barrier. Finally, the fluorescence intensity accumulation curve is fitted with a two-compartment model, which contains facilitated diffusion and active transport mechanisms, and its differential equation is ; where, represents the change rate of the marker concentration in the basal compartment over time ; and represent the marker concentration in the apical and basal compartments, respectively, the facilitated diffusion rate constant and the active transport rate constant are estimated by non-linear least squares fitting. The time parameter required for the fluorescence intensity accumulation curve to reach half of the plateau, the maximum first derivative value of the curve at the initial stage , the facilitated diffusion rate constant and the active transport rate constant together constitute the permeability kinetics parameter set, which is used to quantify the dynamic of molecular permeability of the barrier.

[0033] The process of extracting the set of cell morphological metrics from cell morphological time-lapse image data is as follows: cell morphological time-lapse image data includes high-resolution sequence images under bright field and specific fluorescent markers (such as ZO-1 antibody labeling). First, the image is preprocessed and segmented, and a deep learning model based on U-Net is used to segment the cell boundary and nucleus, calculate the cell boundary length per unit area, and obtain the cell boundary length per unit area ; where, represents the cell boundary length; represents the total number of cell boundary pixels in the image; represents the total pixel area of the image; the cell boundary length reflects the complexity of cell shape and the degree of cell-cell contact. Second, the continuity index of the intercellular tight junction protein fluorescence signal is quantified, the tight junction signal is extracted from the ZO-1 fluorescence image, and the continuity index ; where, represents the continuity index; represents the local signal variance, represents the square of the global signal mean, the higher the continuity index value, the more continuous the connection. Finally, the distribution variance of the aspect ratio of the nucleus is analyzed, the aspect ratio of each nucleus is calculated ; where, represents the aspect ratio of the nucleus; represents the length of the long axis of the fitted ellipse of the nucleus; represents the length of the short axis of the fitted ellipse of the nucleus. Then, the variance of the aspect ratio of all nuclei is calculated , which indicates the strong heterogeneity of the nuclear morphology and may indicate the cell stress or differentiation state. The cell boundary length per unit area, the tight junction continuity index, and the variance of the nuclear aspect ratio distribution together constitute the set of cell morphological metrics for evaluating the structural integrity of the barrier.

[0034] Finally, all the feature parameters in the impedance spectrum feature set, the permeability kinetics parameter set, and the cell morphological metric set are standardized to eliminate the dimensional differences. The minimum-maximum normalization method is used to map each feature parameter value to the interval [0, 1], and the standardized feature parameters together constitute the multi-scale biophysical feature vector, which provides consistent and comparable input for the subsequent multi-criteria weighted fusion model.

[0035] Multi-criteria weighted fusion model module: a multi-criteria weighted fusion model is constructed, which takes the impedance spectrum feature set, the permeability kinetics parameter set, and the cell morphological metric set as inputs to generate a comprehensive and quantitative barrier health index; the barrier health index is a normalized scalar value that intuitively reflects the overall functional state of the epithelial barrier.

[0036] The specific implementation process is as follows: first, the initial weight distribution based on the analytic hierarchy process is performed, and an initial weight coefficient is assigned to each feature parameter in the impedance spectrum feature set, the permeability dynamics parameter set and the cell morphology measurement set. The analytic hierarchy process can systematically integrate expert knowledge by constructing a hierarchical structure model and a judgment matrix, and ensure the rationality and consistency of the weight distribution. In the specific implementation, a hierarchical structure model for evaluating the barrier health status is constructed, which takes the barrier health index as the target layer (the highest layer), takes the impedance spectrum feature set, the permeability dynamics parameter set and the cell morphology measurement set as the criterion layer (the middle layer), and takes the specific feature parameters in each feature set as the scheme layer (the bottom layer). For example, the impedance spectrum feature set includes the sliding window average value of the transmembrane resistance steady-state value, the linear regression slope of the transmembrane resistance change over time, and the ratio of the impedance amplitude under high-frequency and low-frequency excitation; the permeability dynamics parameter set includes the time required for the fluorescence intensity accumulation curve to reach half of the plateau, the maximum first derivative value of the curve in the initial stage, the facilitated diffusion rate constant and the active transport rate constant; the cell morphology measurement set includes the cell boundary length per unit area, the continuity index of the intercellular tight junction protein fluorescence signal, and the long-width ratio distribution variance of the cell nucleus. Based on the preset judgment scale (such as the 1-9 scale method), the judgment matrix of the criterion layer to the target layer and the judgment matrix of the scheme layer to the criterion layer are constructed. For example, the criterion layer judgment matrix is used to compare the relative importance of the impedance spectrum feature set, the permeability dynamics parameter set and the cell morphology measurement set to the barrier health index; the scheme layer judgment matrix is used to compare the relative importance of each feature parameter under the same criterion layer. The consistency of each judgment matrix is checked, the consistency ratio is calculated, and if the consistency ratio is less than 0.1, the consistency check is passed, indicating that the judgment matrix is reasonable; then the characteristic vector of the judgment matrix is calculated and normalized to obtain the initial weight coefficient of each feature parameter. For example, through calculation, the sliding window average value of the transmembrane resistance steady-state value may be assigned an initial weight of 0.15, the linear regression slope of the transmembrane resistance change over time may be assigned an initial weight of 0.10, and the ratio of the impedance amplitude under high-frequency and low-frequency excitation may be assigned an initial weight of 0.05, and so on, to ensure that the sum of the initial weight coefficients of all feature parameters is 1.

[0037] Secondly, a principal component analysis model is constructed to reduce the dimensionality of the standardized multi-scale biophysical features. Specifically, the standardized feature parameters are combined into a feature matrix, the covariance matrix is calculated, and the eigenvalues and eigenvectors are solved. According to the eigenvalues from large to small, the first ​The value is determined by the cumulative variance contribution rate exceeding a preset percentage threshold (e.g., 85%). The variance contribution rate of each principal component represents its ability to explain the variability of the original data. Subsequently, the variance contribution rate of each principal component is allocated back to the original feature parameters according to the proportion of the feature parameter's loading on that principal component, resulting in the contribution weight of each feature parameter based on data variability. For example, if the variance contribution rate of the first principal component is 50%, and the sliding window average value of the transmembrane resistance steady-state value has a loading of 0.3 on that principal component, then the contribution weight obtained by this feature parameter from the first principal component is 50% × 0.3 = 0.15; this process is repeated for all principal components and feature parameters, and normalized to obtain the final contribution weight of each feature parameter. Next, the initial weight coefficient of each feature parameter is weighted and combined with the contribution weight to calculate the final fusion weight of each feature parameter. Specifically, a linear weighting method is used, and a combination coefficient (e.g., 0.5 to balance subjective and objective weights) is set, expressed by the formula: ;in, Indicates the first The final fusion weights of the feature parameters; Represents the combination coefficients; Representing characteristic parameters The initial weighting coefficients; Representing characteristic parameters The contribution weights are determined by this step, which incorporates expert experience and takes into account the variability of the data itself, making the weight allocation more scientific and adaptable.

[0038] Finally, based on the standardized values ​​of all feature parameters and their corresponding final fusion weights, a weighted fusion calculation is performed to generate the barrier health index. First, the value of each feature parameter is mapped to the [0,1] interval using a min-max normalization method. Then, the barrier health index is calculated using a weighted summation formula: ;in, Indicates the barrier health index; Indicates the total number of characteristic parameters; Representing characteristic parameters The standardized value is the value of the barrier health index. Since the sum of all weights is 1 and the standardized value is in the range [0,1], the barrier health index will also fall between 0 and 1, with a higher value indicating a better barrier health status. For example, if all feature parameters are in an ideal state, the barrier health index is close to 1; if most feature parameters are abnormal, the index tends to be close to 0. As a comprehensive indicator, this index can effectively integrate multimodal data, intuitively reflect the overall functional status of the epithelial barrier, and provide a reliable basis for subsequent status assessment and early warning.

[0039] State assessment and report generation module: compare the generated barrier health index with the preset barrier health index state threshold, and combine the selected specific trend discriminant features in the impedance spectrum feature set, permeability dynamics parameter set and cell morphology metric set, output the assessment report through the preset logic rule base.

[0040] The specific implementation process is as follows: first, the system calls the preset barrier health index state threshold, which is based on a large amount of historical experimental data, clinical verification and expert consensus. The system compares the calculated barrier health index with the preset threshold level by level to determine a baseline state level. The classification of the baseline state level provides a preliminary quantitative evaluation of the overall function of the barrier, laying the foundation for subsequent fine discrimination combined with specific trend discriminant features. The baseline state level includes normal stable state, slight instability state, moderate instability state, severe instability state and barrier collapse state. The correspondence between the barrier health index state threshold and the baseline state level is shown in Table 1: Table 1 Correspondence between barrier health index state threshold and baseline state level

[0041] Next, the system selects specific trend discriminant features and performs feature abnormal state marking operations. The specific trend discriminant features include: extracting the linear regression slope of the transmembrane resistance steady-state value over time from the impedance spectrum feature set, defined as the transmembrane resistance decline slope; extracting the maximum first derivative value of the fluorescence intensity cumulative curve in the initial stage from the permeability dynamics parameter set, defined as the permeability maximum slope; extracting the continuity index of the intercellular tight junction protein fluorescence signal from the cell morphology metric set, defined as the tight junction continuity index. For the transmembrane resistance decline slope, the system calculates its moving average value in consecutive multiple preset time windows (for example, every 3 time points as a window, the time window length can be dynamically adjusted according to the experimental period), if the moving average value is lower than the preset negative sensitive threshold (for example, -0.05 Ω / hour), it is marked as a feature abnormal state with a sustained negative development trend, indicating that the barrier resistance is continuously deteriorating. For the permeability maximum slope, the system calculates its moving average value in consecutive multiple preset time windows, if the moving average value is higher than the preset positive sensitive threshold (for example, 0.1 RFU / min), it is marked as a feature abnormal state with a sustained positive growth trend, reflecting abnormal increase of barrier permeability. For the tight junction continuity index, the system compares the single change amount in adjacent two monitoring periods (for example, every 6 hours as a period), if the change amount exceeds the preset jump sensitive threshold (for example, a decrease of more than 0.15), it is marked as a feature abnormal state with a significant negative jump trend, indicating that the tight junction structure may be acutely damaged.

[0042] Then, the system combines the benchmark state level with the feature abnormal state, inputs into the logic rule base for matching query. The logic rule base is a rule set predefined based on domain expert knowledge (such as pathophysiological mechanism, preclinical model validation data) and statistical analysis of a large number of historical experimental data, which uses the production rule of "IF-THEN" to map multi-source information (benchmark state level and dynamic trend feature) to the final state classification. For example, a barrier with a benchmark state level of "mild instability state", if its "transmembrane electrical resistance descending slope", "permeability maximum slope" and "tight junction continuity index" three trend discriminant features are all marked as abnormal, it indicates that although the current comprehensive index is acceptable, multiple key aspects of the barrier are continuously deteriorating, and the system will accordingly upgrade the final state classification to "moderate instability state" to achieve early risk warning. When the input combination condition meets multiple rules at the same time, the system will start the conflict resolution mechanism, and its preset priority rules are: 1. Feature abnormality quantity priority: under the same benchmark state level, the more the number of triggered feature abnormalities, the higher the priority of the corresponding rule; 2. State severity priority: if the triggered rules point to different final states, the rule pointing to the more serious state has a higher priority. Through this mechanism, the system can ensure that for any input combination, a unique and determined final state classification result is output.

[0043] Finally, the system generates corresponding hierarchical warning signals according to the final state classification result, and outputs a structured comprehensive evaluation report. The report aims to provide users with an overview of the state and decision support at a glance. The hierarchical warning signal adopts the internationally recognized five-level color coding and text identification: green represents a normal stable state, blue represents a mild instability state and suggests attention, yellow represents a moderate instability state and issues a warning, orange represents a serious instability state and triggers an alarm, and red represents a barrier collapse state and issues a serious alarm. The output evaluation report is a comprehensive document, and the core content includes the final state classification conclusion, the prominent warning signal color identification, and the systematic integration of key data and in-depth analysis. The report will clearly list the current barrier health index value, the specific values of each trend discriminant feature and its abnormal mark, forming a key parameter abstract; at the same time, it provides time series change charts of transmembrane electrical resistance, fluorescence permeability curve and cell morphology image to visually present the data dynamics. In addition, the report will automatically generate a brief mechanism analysis based on the final state and specific abnormal features, such as indicating that "the current state is driven by continuous electrical resistance decline and significant permeability increase, indicating that the barrier structure and function are damaged"; and it will also provide preliminary action recommendations according to the state level, from "green" state without operation, to "yellow" state with increased monitoring frequency, to "red" state requiring immediate manual intervention confirmation.

[0044] Model verification and dynamic adjustment module: in a specific verification period, two operations are performed in parallel: one is to obtain and calculate the current barrier health index according to the data acquisition and synchronous preprocessing module, the multiscale biophysical feature extraction module and the multi-criteria weighted fusion model module, as the model calculation value; the second is to obtain the reference barrier integrity index at the same time point through offline biological sampling and analysis method; the model calculation value is compared with the reference barrier integrity index, and the model prediction deviation is calculated; based on the model prediction deviation, the multi-criteria weighted fusion model is dynamically adjusted.

[0045] The specific implementation process is as follows: in a specific verification period, the system performs two verification operations in parallel to evaluate the accuracy and reliability of the multi-criteria weighted fusion model and dynamically adjusts it accordingly. The first verification operation is to obtain and calculate the current barrier health index in real time according to the data acquisition and synchronous preprocessing module, the multiscale biophysical feature extraction module and the multi-criteria weighted fusion model module, as the model calculation value; specifically, the system synchronously acquires transmembrane resistance time series data, fluorescent marker permeability time series data and cell morphology time series image data, generates synchronous multi-modal barrier time series data set after quality inspection and time alignment, then extracts multiscale biophysical features (including impedance spectrum feature set, permeability dynamics parameter set and cell morphology measurement set), and calculates the barrier health index through the multi-criteria weighted fusion model. The index is a normalized scalar value that quantifies the overall functional state of the epithelial barrier. The second verification operation is to obtain the reference barrier integrity index at the same time point through offline biological sampling and analysis method; the offline method includes using a transmembrane resistance meter to detect the end-point resistance value, calculating the apparent permeability coefficient through a fluorescent marker permeability experiment, or evaluating the continuity of tight junction proteins by immunofluorescence staining combined with microscope observation. These offline indicators serve as the gold standard for verifying the accuracy of the model calculation value.

[0046] The model calculation value is compared with the reference barrier integrity index, and the model prediction deviation is calculated; the model prediction deviation is defined as the absolute difference or relative difference between the model calculation value and the reference barrier integrity index, and the specific formula is: absolute difference or relative difference ; wherein, represents the model calculation value, represents the reference barrier integrity index. The system presets a deviation tolerance threshold (for example, the absolute difference threshold is set to 0.05, and the relative difference threshold is set to 10%), if the calculated model prediction deviation exceeds the threshold, the weight adjustment mechanism is triggered.

[0047] Based on the model prediction deviation, the system adopts a Bayesian updating algorithm to dynamically adjust the final fusion weight in the multi-criteria weighted fusion model. In this algorithm, the system regards the final fusion weight vector before adjustment as the prior distribution, which is set as a multivariate Gaussian distribution. The mean vector of the prior distribution is the final fusion weight vector currently used, and the covariance matrix is initialized and set according to historical data or expert experience. At the same time, the system takes the benchmark barrier integrity index obtained by offline measurement as the observation value, and establishes an observation model: the observation value is considered to be a Gaussian distribution with the barrier health index calculated by the model as the mean value and the preset observation noise variance as the variance, where the size of the observation noise variance can be determined according to the repeatability data of multiple experiments. Through the Bayesian theorem, the system combines the prior distribution with the observation likelihood function to derive the posterior distribution about the final fusion weight. The specific updating process is as follows: first, the system obtains the numerical value of all standardized feature parameters at the current time to form a feature vector. Then, the mean vector of the posterior distribution is calculated using the Bayesian updating formula, which integrates the prior weight vector, the prior covariance matrix, the current feature vector, the observation noise variance, and the difference between the model calculation value and the benchmark observation value. This calculation is essentially a correction of the prior weight, and the correction amplitude depends on the prior uncertainty, the observation noise, and the performance of the current feature parameter. The posterior mean vector calculated is used as the updated final fusion weight vector. Subsequently, the system normalizes the vector to ensure that the sum of all weight components is 1, thereby obtaining the adjusted final fusion weight that can be directly used for the next barrier health index calculation. This complete dynamic adjustment process enables the model to adaptively calibrate the weight distribution according to the offline verification data, continuously improve the accuracy and robustness of barrier health index prediction, and ultimately ensure that the evaluation results are consistent with the true biological state.

[0048] The specific embodiments are only an explanation of the present application, which is not a limitation of the present application. Those skilled in the art can make modifications to the embodiments without creative contribution after reading the specification, as long as the modifications are within the scope of the claims of the present application.

Claims

1. A data fusion modeling system for assessing the state of the airway epithelial barrier, characterized in that, include: Data acquisition and synchronous preprocessing module: Acquires time-series raw data in the physiological microenvironment of the airway epithelial cell culture model. The time-series raw data includes synchronously acquired transmembrane resistance time-series data, fluorescent label permeability time-series data, and cell morphology time-series image data. The original time-series data is subjected to quality checks and outlier removal, and after being aligned with the time reference, a synchronous multimodal barrier time-series dataset is generated. Multi-scale biophysical feature extraction module: Extracts and calculates multi-scale biophysical features that characterize the structure and function of the epithelial barrier from the synchronous multimodal barrier time-series dataset; The multi-scale biophysical features include: an impedance spectrum feature set extracted from transmembrane resistance time series data, a permeability kinetic parameter set extracted from fluorescent marker permeability time series data, and a cell morphology metric set extracted from cell morphology time series image data. Multi-criteria weighted fusion model module: Constructs a multi-criteria weighted fusion model, taking the impedance spectrum feature set, permeability kinetic parameter set and cell morphology measurement set as inputs to generate a comprehensive and quantitative barrier health index; the barrier health index is a normalized scalar value used to intuitively reflect the overall functional status of the epithelial barrier; Status assessment and report generation module: The generated barrier health index is compared with the preset barrier health index status threshold, and the specific trend discrimination features selected in the impedance spectrum feature set, permeability dynamic parameter set and cell morphology measurement set are combined with the preset logic rule base to output an assessment report.

2. The data fusion modeling system for assessing the state of the airway epithelial barrier according to claim 1, characterized in that, It also includes the following steps: Model Validation and Dynamic Adjustment Module: During a specific validation cycle, two operations are performed in parallel: First, the current barrier health index is acquired and calculated using the data acquisition and synchronous preprocessing module, the multi-scale biophysical feature extraction module, and the multi-criteria weighted fusion model module, and used as the model calculation value; second, a baseline barrier integrity index at the same time point is acquired through offline biological sampling and analysis methods; the model calculation value is compared with the baseline barrier integrity index to calculate the model prediction deviation; and based on the model prediction deviation, the multi-criteria weighted fusion model is dynamically adjusted.

3. The data fusion modeling system for assessing the state of the airway epithelial barrier according to claim 1, characterized in that, In the data acquisition and synchronous preprocessing module, the steps for acquiring time-series raw data in the physiological microenvironment of the airway epithelial cell culture model include: The system acquires transmembrane resistance time-series data, which includes complex impedance information obtained by multi-frequency AC excitation; it acquires fluorescent label permeability time-series data, which is obtained by monitoring the cumulative fluorescence intensity curve of the basal side fluorescence intensity over time after the introduction of the fluorescent label at a specific time point; and it acquires cell morphology time-series image data, which includes high-resolution cell images under bright field and / or specific fluorescent labels obtained by timed imaging.

4. The data fusion modeling system for assessing the state of the airway epithelial barrier according to claim 1, characterized in that, In the data acquisition and synchronization preprocessing module, the steps for quality inspection and outlier removal of the raw time-series data include: For the transmembrane resistance time-series data, its signal-to-noise ratio and phase angle are calculated, and data segments with a signal-to-noise ratio lower than a first preset threshold or a phase angle exceeding a reasonable physiological range are discarded; for the fluorescent label permeability time-series data, its baseline stability is checked and signal saturation is determined, and data segments with baseline drift exceeding a second preset threshold or signal saturation are discarded; for the cell morphology time-series image data, its focus sharpness is evaluated by calculating the image gradient, and its image quality is evaluated by calculating the local contrast, and image frames with sharpness lower than a third preset threshold or contrast lower than a fourth preset threshold are marked and discarded.

5. A data fusion modeling system for assessing the state of the airway epithelial barrier according to claim 3, characterized in that, In the multi-scale biophysical feature extraction module, the steps of extracting and calculating multi-scale biophysical features that characterize the structure and function of the epithelial barrier from the synchronous multimodal barrier time-series dataset include: From the transmembrane resistance time-series data, the sliding window average value of the steady-state transmembrane resistance, the linear regression slope changing with time, and the ratio of impedance amplitude under high-frequency and low-frequency excitation are extracted to form an impedance spectrum feature set. Nonlinear fitting was performed on the fluorescence intensity accumulation curve in the permeability time series data of the fluorescent label to extract the time required for the fluorescence intensity accumulation curve to reach half of the plateau period, the maximum first derivative value of the curve in the initial stage, and the facilitated diffusion and active transport rate constants obtained by fitting with a two-compartment model, forming a permeability kinetic parameter set. The cell morphology time-series image data is segmented and feature analyzed to quantify and calculate the cell boundary length per unit area, the continuity index of fluorescence signals of tight junction proteins between cells, and the aspect ratio distribution variance of the cell nucleus, thus forming a cell morphology metric set. The characteristic parameters in the resulting impedance spectrum feature set, permeability kinetic parameter set, and cell morphology measurement set are standardized to eliminate dimensional differences and together constitute the multi-scale biophysical features.

6. The data fusion modeling system for assessing the state of the airway epithelial barrier according to claim 5, characterized in that, The multi-criteria weighted fusion model module includes the following steps: Perform an initial weight allocation based on the analytic hierarchy process (AHP) to assign an initial weight coefficient to each feature parameter in the impedance spectrum feature set, the permeability kinetic parameter set, and the cell morphology metric set. A principal component analysis model was constructed to perform dimensionality reduction on the standardized multi-scale biophysical features, and to extract the unprocessed features. One principal component; among which... The value is determined by the cumulative variance contribution rate exceeding a preset percentage threshold; the variance contribution rate of each principal component is allocated back to the original feature parameters according to the feature parameter loading ratio to obtain the contribution weight of each feature parameter based on data variability; The initial weight coefficient of each feature parameter is weighted and combined with the contribution weight to calculate the final fusion weight of each feature parameter; The value of each feature parameter is normalized by the min-max normalization method. Based on the normalized values ​​of all feature parameters and their corresponding final fusion weights, a weighted fusion calculation is performed to generate a barrier health index ranging from 0 to 1.

7. The data fusion modeling system for assessing the state of the airway epithelial barrier according to claim 6, characterized in that, In the multi-criteria weighted fusion model module, the step of performing initial weight allocation based on the analytic hierarchy process (AHP) to assign an initial weight coefficient to each feature parameter in the impedance spectrum feature set, the permeability kinetic parameter set, and the cell morphology metric set includes: A hierarchical model for assessing barrier health status is constructed. The hierarchical model uses the barrier health index as the target layer, the impedance spectrum feature set, the permeability dynamic parameter set and the cell morphology measurement set as the criterion layer, and the feature parameters in each feature set as the scheme layer. Based on a preset judgment scale, a judgment matrix is ​​constructed between the criterion layer and the target layer, and between the scheme layer and the criterion layer. The judgment matrix is ​​subjected to a consistency check, and its feature vector is calculated after passing the check. After normalizing the feature vector, an initial weighting coefficient is assigned to each feature parameter in the impedance spectrum feature set, the permeability dynamic parameter set, and the cell morphology metric set.

8. The data fusion modeling system for assessing the state of the airway epithelial barrier according to claim 1, characterized in that, The status assessment and report generation module includes the following steps: A preset barrier health index state threshold is invoked, which includes a healthy state threshold, a slightly damaged state threshold, and a moderately damaged state threshold. The barrier health index is compared with the healthy state threshold, the slightly damaged state threshold, and the moderately damaged state threshold to determine a baseline state level. The baseline state level includes a normal stable state, a slightly unstable state, a moderately unstable state, a severely unstable state, and a barrier collapse state. Select specific trend discrimination features and mark abnormal feature states; The baseline state level is combined with the characteristic abnormal state and input into the logical rule base for matching query. Conflicts are resolved according to preset priority rules, and a unique and definite final state classification result is output. Based on the final state classification results, corresponding graded early warning signals are generated and an evaluation report consisting of the final state classification results and the graded early warning signals is output.

9. A data fusion modeling system for assessing the state of the airway epithelial barrier according to claim 8, characterized in that, In the status assessment and report generation module: The specific trend discrimination features include: extracting the linear regression slope of the steady-state value of transmembrane resistance over time from the impedance spectrum feature set, defined as the transmembrane resistance decrease slope; extracting the maximum first derivative value of the fluorescence intensity accumulation curve in the initial stage from the permeability kinetic parameter set, defined as the maximum permeability slope; and extracting the continuity index of the fluorescence signal of intercellular tight junction proteins from the cell morphology measurement set, defined as the tight junction continuity index. The marking of characteristic anomalous states includes: For the slope of the transmembrane resistance decrease, if its moving average value within multiple consecutive preset time windows is lower than a preset negative sensitivity threshold, it is marked as a characteristic abnormal state with a continuous negative development trend; for the maximum permeability slope, if its moving average value within multiple consecutive preset time windows is higher than a preset positive sensitivity threshold, it is marked as a characteristic abnormal state with a continuous positive growth trend; for the tight connection continuity index, if its single change in two adjacent monitoring periods exceeds a preset jump sensitivity threshold, it is marked as a characteristic abnormal state with a significant negative jump trend.

10. A data fusion modeling system for assessing the state of the airway epithelial barrier according to claim 6, characterized in that, The model validation and dynamic adjustment module includes the following steps: During a specific verification cycle, the first verification operation and the second verification operation are executed in parallel. The first verification operation involves acquiring and calculating the current barrier health index using the data acquisition and synchronous preprocessing module, the multi-scale biophysical feature extraction module, and the multi-criteria weighted fusion model module, and using it as the model calculation value. The second verification operation involves acquiring the baseline barrier integrity index at the same time point using offline biological sampling and analysis methods. The model prediction deviation is calculated by comparing the calculated value with the benchmark barrier integrity index; wherein the model prediction deviation is the absolute or relative difference between the calculated value and the benchmark barrier integrity index. Based on the model prediction bias, a Bayesian update algorithm is used to dynamically adjust the final fusion weights in the multi-criteria weighted fusion model. The Bayesian update algorithm uses the final fusion weights before adjustment as the prior distribution and the benchmark barrier integrity index as the observed value to infer the updated final fusion weights.

Images