Method and system for evaluating state of restless patient based on big data analysis

Abstract

The invention discloses a restless patient state evaluation method and system based on big data analysis, and particularly relates to the field of state evaluation. An original multi-source data stream of a patient is collected, feature dimension disassembly and preliminary redundancy marking are carried out, a feature subset with false emotion correlation is identified by applying nonlinear redundancy analysis, a feature dimension combination generating emotion illusion is positioned by adopting a combination effect cross validation technology, and through the feature combination with an emotion illusion effect, the emotion illusion correlation is identified. And obtaining a corrected effective feature data set. And constructing a dysphoria state prediction model based on the effective feature set. By introducing false correlation mode mining of a feature combination layer, the problem of emotion recognition deviation caused by redundant feature interaction in traditional high-dimensional data analysis is solved, the problem of prediction result anti-intuition caused by emotion illusion in a traditional method is solved, and the biological interpretability and clinical prediction accuracy of restless state evaluation are effectively improved.

Description

Technical Field

[0001] The present invention relates to the technical field of state evaluation, and more specifically, the present invention relates to a method and system for evaluating the state of agitated patients based on big data analysis. Background Art

[0002] In the field of monitoring and evaluating the state of patients with mental illnesses, the existing technology generally constructs a big data analysis model through multi-modal data for evaluating the agitation state of patients. However, the traditional method has obvious defects in the integration of high-dimensional heterogeneous data: on the one hand, the non-linear redundant associations between the features of multi-modal data are not effectively decoupled, resulting in the interference of the pseudo-correlations hidden in the feature space on the model learning process; on the other hand, conventional feature screening methods are difficult to identify the false association patterns of cross-dimensional combinations, resulting in the model capturing the pseudo-features of emotional fluctuations that have nothing to do with the real pathological state. These problems produce anti-logical prediction biases in complex scenarios, weakening the medical credibility of the evaluation results of the agitation state, and further affecting the accuracy and timeliness of clinical intervention decisions.

[0003] In order to solve the above problems, a technical solution is provided now. Summary of the Invention

[0004] In order to overcome the above-mentioned defects of the prior art, the embodiments of the present invention provide a method and system for evaluating the state of agitated patients based on big data analysis to solve the problems raised in the above background art.

[0005] To achieve the above object, the present invention provides the following technical solutions: A method for evaluating the state of agitated patients based on big data analysis, comprising the following steps: Collect the original multi-source data stream of the patient within a preset time period, and perform feature dimension decomposition and preliminary redundancy marking on the original data in each data stream to obtain the initial feature dimension set of the original multi-source data stream; For the initial feature dimension set, identify the feature subsets that generate false emotional feature correlations through a non-linear redundancy analysis method to form a redundant feature factor set; Perform a combined effect cross-validation analysis on the redundant feature factor set to determine the feature dimension combinations that exhibit the emotional illusion effect, and generate a set of false emotional association patterns; According to the set of false emotional association patterns, remove the feature combinations with the emotional illusion effect from the initial feature dimension set of the original multi-source data stream to obtain a corrected set of effective feature data; Construct an agitation state evolution prediction model based on the set of effective feature data to evaluate the current agitation state of the patient, and output the recognition result of the patient's agitation state.

[0006] In a preferred embodiment, multi-source data streams of a patient within a preset time period are collected, and the raw data in each data stream is disassembled into feature dimensions and preliminarily marked for redundancy, obtaining an initial feature dimension set of the raw data streams, specifically: Collect the patient's heart rate data stream, skin conductance pulse data stream, patient movement frequency data stream, and spatial displacement trajectory data stream; Disassemble the time domain features and frequency domain features of the patient's heart rate data stream; Disassemble the pulse amplitude features and peak density features of the skin conductance pulse data stream; Disassemble the number of movements per unit time and movement rhythm of the patient movement frequency data stream; Disassemble the trajectory curvature and displacement frequency features of the spatial displacement trajectory data stream; Perform preliminary redundancy marking on each disassembled feature dimension through a data dimension correlation analysis method to obtain an initial feature dimension set.

[0007] In a preferred embodiment, perform preliminary redundancy marking on each disassembled feature dimension through a data dimension correlation analysis method to obtain an initial feature dimension set, specifically: Calculate the Pearson correlation coefficient matrix between each disassembled feature dimension to determine the degree of linear correlation between feature dimensions; Compare the correlation coefficients between any two feature dimensions in the Pearson correlation coefficient matrix. When the correlation coefficient is greater than the linear correlation threshold, mark the feature dimension with a lower diagnostic contribution degree among the two feature dimensions as a preliminary redundant dimension; Classify and mark all feature dimensions marked as preliminary redundant dimensions to form a preliminary redundant feature marking set; Map and compare the preliminary redundant feature marking set with each disassembled feature dimension set to obtain an initial feature dimension set.

[0008] In a preferred embodiment, for the initial feature dimension set, identify the feature subsets that generate false emotion feature correlations through a non-linear redundancy analysis method to form a redundant feature factor set, specifically: Apply the kernel function mapping method to the initial feature dimension set to project the feature dimensions into a high-dimensional non-linear feature space; Evaluate the non-linear correlation degree between the projected feature dimensions based on the mutual information metric; When the mutual information metric between the projected feature dimensions is greater than the non-linear redundancy threshold, determine the corresponding feature dimension combination as a feature subset with false emotion feature correlations; After integrating all identified feature subsets, form a redundant feature factor set.

[0009] In a preferred embodiment, a combined effect cross-validation analysis is performed on the redundant feature factor set to determine the feature dimension combinations that exhibit the emotional illusion effect, and a set of false emotional association patterns is generated. Specifically: Each feature subset in the redundant feature factor set is randomly combined to form multiple different combination patterns; For each combination pattern, the data of the combination pattern is input into a preset emotional state classifier using the cross-validation method, and the interference degree of each combination pattern on the classification result of the restlessness state is calculated; When the interference degree of the combination pattern on the classification of the restlessness state exceeds a predetermined interference threshold, it is determined that the combination pattern is a feature dimension combination that exhibits the emotional illusion effect; All the feature dimension combinations that exhibit the emotional illusion effect are integrated to generate a set of false emotional association patterns.

[0010] Each feature dimension combination in the set of false emotional association patterns is sequentially removed from the initial feature dimension set; A correlation review analysis of the remaining features is performed on the feature dimension set after removal; The remaining feature dimension set after the correlation review analysis is re-labeled to obtain a corrected set of effective feature data.

[0011] In a preferred embodiment, a restlessness state evolution prediction model is constructed based on the effective feature data set to evaluate the current restlessness state of the patient and output the recognition result of the patient's restlessness state. Specifically: The effective feature data set is used as an input variable to construct a restlessness state evolution prediction model based on time series; The historical effective feature data set is used to train the parameters of the restlessness state evolution prediction model; The effective feature data of the patient at the current time node is input into the restlessness state evolution prediction model to calculate the probability of the patient's current restlessness state; The current restlessness state level of the patient is determined according to the restlessness state probability.

[0012] In a preferred embodiment, the current restlessness state level of the patient is determined according to the restlessness state probability. Specifically: The restlessness state probability is compared with a preset probability threshold interval; When the restlessness state probability is within the first probability threshold interval, it is determined that the current restlessness state level of the patient is mild restlessness; When the restlessness state probability is within the second probability threshold interval, it is determined that the current restlessness state level of the patient is moderate restlessness; When the restlessness state probability is within the third probability threshold interval, it is determined that the current restlessness state level of the patient is severe restlessness.

[0013] On the other hand, the present invention provides a restlessness patient state assessment system based on big data analysis, including a data processing module, a redundancy analysis module, a cross-validation analysis module, a feature correction module, and a state assessment module; Data processing module: Collect the original multi-source data stream of the patient within a preset time period, and perform feature dimension decomposition and preliminary redundancy marking on the original data in each data stream to obtain the initial feature dimension set of the original multi-source data stream; Redundancy analysis module: For the initial feature dimension set, identify the feature subset that generates the correlation of false emotion features through a non-linear redundancy analysis method, and form a redundant feature factor set; Cross-validation analysis module: Perform combined effect cross-validation analysis on the redundant feature factor set, determine the feature dimension combination that shows the emotion false appearance effect, and generate a set of false emotion association patterns; Feature correction module: According to the set of false emotion association patterns, remove the feature combinations with emotion false appearance effects from the initial feature dimension set of the original multi-source data stream to obtain a corrected set of effective feature data; State assessment module: Build an evolution prediction model of the restlessness state based on the set of effective feature data, evaluate the current restlessness state of the patient, and output the recognition result of the patient's restlessness state.

[0014] Technical effects and advantages of a restlessness patient state assessment method and system based on big data analysis according to the present invention: By constructing a dynamic screening and verification mechanism for feature dimensions, the accuracy and reliability of the emotion state assessment of restlessness patients are improved. By introducing non-linear redundancy analysis and combined effect cross-validation methods, the problem of misjudgment of emotion features caused by feature redundancy in traditional assessments is effectively solved. It can accurately identify and remove false association patterns between multi-modal data, and avoid the interference of emotion false appearance effects on the assessment results. By building an evolution prediction model of the restlessness state based on the corrected set of effective feature data, in-depth analysis and real-time tracking of the patient's restlessness state are realized, and the interpretability of clinical decisions is enhanced. It breaks through the limitations of traditional single-dimensional assessment methods, provides new ideas for the accurate identification of complex mental symptoms, has the intelligent analysis advantage of adapting to individual differences, and provides a more reliable objective basis for clinical intervention. Brief Description of the Drawings

[0015] Figure 1 It is a schematic diagram of a restlessness patient state assessment method based on big data analysis according to the present invention; Figure 2 It is a schematic diagram of the structure of a restlessness patient state assessment system based on big data analysis according to the present invention. Detailed Description of the Invention

[0016] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0017] Embodiment 1

[0018] Figure 1 A method for evaluating the state of agitated patients based on big data analysis according to the present invention is provided, which includes the following steps: Collect the original multi-source data stream of the patient within a preset time period, and perform feature dimension decomposition and preliminary redundancy marking on the original data in each data stream to obtain the initial feature dimension set of the original multi-source data stream; For the initial feature dimension set, identify the feature subset that generates the false emotion feature correlation through the non-linear redundancy analysis method to form the redundant feature factor set; Perform combined effect cross-validation analysis on the redundant feature factor set to determine the feature dimension combination that exhibits the emotion illusion effect, and generate the false emotion association pattern set; According to the false emotion association pattern set, remove the feature combinations with emotion illusion effects from the initial feature dimension set of the original multi-source data stream to obtain the corrected effective feature data set; Based on the effective feature data set, construct an agitated state evolution prediction model to evaluate the current agitated state of the patient and output the recognition result of the patient's agitated state.

[0019] Specifically, collecting the multi-source data stream of the patient within a preset time period, and performing feature dimension decomposition and preliminary redundancy marking on the original data in each data stream to obtain the initial feature dimension set of the original data stream, includes: Collect the patient's heart rate data stream, skin conductance pulse data stream, patient's action frequency data stream, and spatial displacement trajectory data stream; Specifically, within the preset time period, collect the patient's heart rate data through a biosensor, collect the skin conductance pulse data using a skin electrical sensor, and at the same time collect the patient's action frequency data and spatial displacement trajectory data through a camera and a motion sensor.

[0020] Each data stream reflects information in different dimensions. For example, the heart rate data stream is used to describe the patient's autonomic nerve regulation; the skin conductance data stream reflects the patient's physiological stress response; the action frequency and spatial displacement data streams reflect the patient's behavioral activity.

[0021] Perform dimensional decomposition on the time-domain features and frequency-domain features of the patient's heart rate data stream; Specifically, the time-domain features include calculating the mean, standard deviation, maximum value, minimum value, etc. of the heart rate data; the frequency-domain features include calculating the low-frequency power, high-frequency power, and the ratio of low-frequency to high-frequency using the fast Fourier transform.

[0022] Perform dimensional decomposition on the skin conductance pulse data stream into pulse amplitude features and peak density features; Specifically, the pulse amplitude feature refers to the amplitude of each pulse in the skin conductance signal, and the peak density feature refers to the frequency of the appearance of peaks within a unit time.

[0023] For example, by detecting the amplitude difference between the signal peaks and valleys, calculate the amplitude of each pulse, and count the number of pulses per minute as the peak density.

[0024] Perform dimensional decomposition on the patient's action frequency data stream into the number of actions per unit time and action rhythm; Specifically, the number of actions per unit time can be obtained by counting the number of actions of the patient within a specified time; the action rhythm reflects the distribution law of the time intervals between actions.

[0025] For example, if 30 actions are recorded within one minute, the number of actions per unit time is 30; if the average interval between actions is 2 seconds, it can be used as a reference index for the action rhythm.

[0026] Perform dimensional decomposition on the spatial displacement trajectory data stream into trajectory curvature and displacement frequency features; Specifically, the trajectory curvature reflects the degree of bending of the patient's walking trajectory, which can be obtained by calculating the curvature of the curve formed by consecutive position points; the displacement frequency feature is the number of position changes per unit time.

[0027] Perform preliminary redundancy marking on each decomposed feature dimension through the data dimension correlation analysis method to obtain the initial feature dimension set.

[0028] Specifically, perform preliminary redundancy marking on each decomposed feature dimension through the data dimension correlation analysis method to obtain the initial feature dimension set, including: Calculate the Pearson correlation coefficient matrix between each decomposed feature dimension to determine the degree of linear correlation between feature dimensions; Specifically, arrange all the decomposed feature dimension data in matrix form, calculate the Pearson correlation coefficient for any two columns (representing two feature dimensions) in the matrix to form a symmetric matrix.

[0029] Compare the correlation coefficients between any two feature dimensions in the Pearson correlation coefficient matrix. When the correlation coefficient is greater than the linear correlation threshold, mark the feature dimension with a lower diagnostic contribution among the two feature dimensions as a preliminary redundant dimension; Specifically, when the Pearson correlation coefficient between two feature dimensions is greater than a predetermined linear correlation threshold (e.g., 0.8), it is considered that there is linear redundancy between them.

[0030] For example, if the correlation coefficient between feature A and feature B is 0.85 and the predetermined threshold is 0.8, then it is determined that there is redundancy between feature A and feature B.

[0031] By comparing the diagnostic contributions of each feature in the disease state assessment (which can be verified through historical data), select the feature dimensions with lower diagnostic contributions for marking.

[0032] For example, if both feature A and feature B are correlated, and through statistical analysis, it is found that feature B has a small contribution to the prediction of the restlessness state, then feature B is marked as a preliminary redundant dimension.

[0033] Classify and mark all the feature dimensions marked as preliminary redundant dimensions to form a preliminary redundant feature marking set; Map and compare the preliminary redundant feature marking set with each disassembled feature dimension set to obtain the initial feature dimension set; Specifically, compare the classified redundant feature marking set with all the disassembled feature dimension sets, remove the feature dimensions marked as preliminary redundant dimensions, and the remaining feature dimensions constitute the initial feature dimension set.

[0034] For example, if there are 10 feature dimensions after disassembly, and 3 redundant ones are removed after comparison, then the initial feature dimension set contains 7 non-redundant feature dimensions.

[0035] Specifically, for the initial feature dimension set, through a non-linear redundancy analysis method, identify the feature subsets that generate false emotion feature correlations to form a redundant feature factor set, including: Adopt a kernel function mapping method for the initial feature dimension set to project the feature dimensions into a high-dimensional non-linear feature space; Specifically, use a kernel function (such as a Gaussian kernel function) to map the initial feature dimensions so that the non-linear relationships that are difficult to distinguish in the low-dimensional space can be reflected in the high-dimensional space.

[0036] Evaluate the non-linear correlation degree between the projected feature dimensions based on the mutual information index; Specifically, for random variables X and Y, first consider all possible combinations of values; for each pair of values (x, y), calculate the probability of their simultaneous occurrence, that is, the joint probability; then calculate the product of the probabilities of their independent occurrences, that is, the product of their respective marginal probabilities; then, take the logarithm of the ratio of the joint probability to the product of the marginal probabilities, which represents the difference in the amount of information between the actual joint occurrence and the independent occurrence of this pair of values; finally, multiply the logarithmic value by the corresponding joint probability and sum over all possible combinations of values, and the resulting sum is the mutual information index.

[0037] When the mutual information index between the projected feature dimensions is greater than the non-linear redundancy threshold, the corresponding feature dimension combination is determined as the feature subset with false emotion feature correlation. Specifically, set the non-linear redundancy threshold. When the mutual information index of any pair or multiple dimensions of features is greater than the non-linear redundancy threshold, it is judged that there is non-linear redundancy in these features in the high-dimensional space, that is, there is false emotion feature correlation.

[0038] For example, if the mutual information index between the heart rate frequency domain feature and the skin conductance peak density feature after mapping is greater than the non-linear redundancy threshold, then these two features are grouped into a false emotion feature correlation combination.

[0039] After integrating all the identified feature subsets, a redundant feature factor set is formed.

[0040] Specifically, perform combined effect cross-validation analysis on the redundant feature factor set to determine the feature dimension combination that exhibits the emotion illusion effect, and generate a set of false emotion association patterns, including: Randomly combine each feature subset in the redundant feature factor set to form multiple different combination patterns. Specifically, for each feature subset in the redundant feature factor set, generate several combination patterns through random permutation and combination. When generating combination patterns, the minimum and maximum numbers of combinations can be set to ensure that the combination patterns are representative.

[0041] For each combination pattern, use the cross-validation method to input the data of the combination pattern into a preset emotion state classifier respectively, and calculate the degree of interference of each combination pattern on the classification result of the restlessness state. Specifically, the cross-validation method can adopt K-fold validation, in which the data is divided into K parts, and alternately use K - 1 parts as training data and 1 part as test data.

[0042] For each combination pattern, calculate the classification accuracy or error rate after the input of this combination pattern, and define the degree of interference as: Degree of interference = max(0, Standard classification accuracy - Current combined mode classification accuracy). That is, the interference effect is only calculated when the classification accuracy under the current combined mode is lower than the standard classification accuracy; if the classification accuracy under the current combined mode is higher than the standard classification accuracy, it is determined that there is no interference effect.

[0043] Among them, the standard classification accuracy is the classification accuracy when no redundant features are added.

[0044] When the degree of interference of the combined mode on the classification of the restlessness state exceeds the predetermined interference threshold, determine that this combined mode is the feature dimension combination with the emotional illusion effect; Integrate all the feature dimension combinations with the emotional illusion effect to generate a set of false emotion association patterns.

[0045] Specifically, according to the set of false emotion association patterns, remove the feature combinations with the emotional illusion effect from the initial feature dimension set of the original multi-source data stream to obtain a corrected set of effective feature data, including: Remove each feature dimension combination in the set of false emotion association patterns from the initial feature dimension set in turn; Specifically, for each feature dimension combination included in the set of false emotion association patterns, locate the corresponding feature dimension in the initial feature dimension set in turn and delete it to remove the redundant features that have a false emotion impact on the assessment of the restlessness state.

[0046] Conduct a correlation review analysis of the remaining features in the removed feature dimension set; Specifically, for the remaining feature dimension set after the removal process, use the correlation analysis method (such as the Pearson correlation coefficient) again to verify the correlation between the features to ensure that no new redundancy is introduced during the removal process.

[0047] Relabel the remaining feature dimension set after the correlation review analysis to obtain a corrected set of effective feature data; Specifically, after the correlation review, relabel and organize the remaining feature dimensions according to their contributions to the assessment of the patient's restlessness state to obtain a corrected set of effective feature data.

[0048] Specifically, based on the set of effective feature data, construct an evolution prediction model for the restlessness state, evaluate the patient's current restlessness state, and output the recognition result of the patient's restlessness state, including: Use the set of effective feature data as the input variable to construct an evolution prediction model for the restlessness state based on time series; Specifically, using the corrected effective feature data set, the data is sorted according to the time series, and the feature vectors at each time node are used as the input of the restlessness state evolution prediction model. When constructing the restlessness state evolution prediction model, a prediction model suitable for time series data is adopted to ensure that the model can capture the evolution law of the patient's characteristics over time.

[0049] Use the historical effective feature data set to train the parameters of the restlessness state evolution prediction model; Specifically, use the historically collected effective feature data and the corresponding labeled restlessness states to train the parameters of the restlessness state evolution prediction model, so that the restlessness state evolution prediction model can accurately reflect the change law of the patient's historical restlessness state. During the training process, the least mean square error can be used as the loss function.

[0050] Input the effective feature data of the patient at the current time node into the restlessness state evolution prediction model to calculate the current restlessness state probability of the patient; Specifically, after the restlessness state evolution prediction model is trained, input the effective feature data of the patient at the current time node into the restlessness state evolution prediction model, and the restlessness state evolution prediction model outputs the current restlessness state probability of the patient according to the training parameters.

[0051] Determine the current restlessness state level of the patient according to the restlessness state probability.

[0052] Specifically, determining the current restlessness state level of the patient according to the restlessness state probability includes: Compare the restlessness state probability with the preset probability threshold interval; When the restlessness state probability is within the first probability threshold interval, determine that the current restlessness state level of the patient is mild restlessness; When the restlessness state probability is within the second probability threshold interval, determine that the current restlessness state level of the patient is moderate restlessness; When the restlessness state probability is within the third probability threshold interval, determine that the current restlessness state level of the patient is severe restlessness; Specifically, the preset probability threshold interval is set based on the probability distribution of the effective feature data collected historically and the corresponding labeled restlessness states in the output of the restlessness state evolution prediction model. Combining with the medical experts' behavioral judgment criteria for different restlessness levels (in medical behavior science, there are usually empirical thresholds for grading restlessness. Mapping these behavioral labels to the model output probabilities, representing the true clinical grading criteria with probability intervals, and forming a matching relationship between probability and medical labels), the probability values output by the restlessness state evolution prediction model are divided into three grade intervals. Specifically, they include: the first probability threshold interval is [0, 0.33), the second probability threshold interval is [0.33, 0.66), and the third probability threshold interval is [0.66, 1]. Ensure that the restlessness levels output by the restlessness state evolution prediction model have clear and verifiable probability boundaries.

[0053] Embodiment 2 The difference between Embodiment 2 and Embodiment 1 of the present invention is that this embodiment introduces a restlessness patient state assessment system based on big data analysis.

[0054] Figure 2 The structural schematic diagram of a restlessness patient state assessment system based on big data analysis according to the present invention is given. A restlessness patient state assessment system based on big data analysis includes a data processing module, a redundancy analysis module, a cross-validation analysis module, a feature correction module, and a state assessment module; Data processing module: Collect the original multi-source data streams of the patient within a preset time period, and perform feature dimension decomposition and preliminary redundancy marking on the original data in each data stream to obtain the initial feature dimension set of the original multi-source data streams; Redundancy analysis module: For the initial feature dimension set, identify the feature subsets that generate false emotional feature correlations through a non-linear redundancy analysis method to form a redundant feature factor set; Cross-validation analysis module: Perform a combined effect cross-validation analysis on the redundant feature factor set to determine the feature dimension combinations that exhibit emotional illusion effects, and generate a set of false emotional association patterns; Feature correction module: According to the set of false emotional association patterns, remove the feature combinations with emotional illusion effects from the initial feature dimension set of the original multi-source data streams to obtain a corrected set of effective feature data; State assessment module: Based on the set of effective feature data, construct a restlessness state evolution prediction model to evaluate the current restlessness state of the patient and output the recognition result of the patient's restlessness state.

[0055] The above formulas are all dimensionless and take their numerical values for calculation. The formulas are obtained by collecting a large amount of data and performing software simulations to get a formula that is closest to the actual situation. The preset parameters and threshold selection in the formulas are set by those skilled in the art according to the actual situation.

[0056] The above embodiments can be implemented in whole or in part by software, hardware, firmware, or any combination thereof. When implemented using software, the above embodiments can be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions or computer programs. When the computer instructions or computer programs are loaded or executed on a computer, the processes or functions described in the embodiments of the present application are generated in whole or in part. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable devices. The computer instructions can be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, the computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center by wired (such as infrared, wireless, microwave, etc.) means. The computer-readable storage medium can be any available medium that the computer can access or a data storage device such as a server or data center that contains one or more collections of available media. The available medium can be a magnetic medium (such as a floppy disk, hard disk, magnetic tape), an optical medium (such as a DVD), or a semiconductor medium. The semiconductor medium can be a solid-state drive.

[0057] Those of ordinary skill in the art can realize that the modules and algorithm steps of the examples described in conjunction with the embodiments disclosed herein can be implemented by electronic hardware or a combination of computer software and electronic hardware. Whether these functions are executed in a hardware or software manner depends on the specific application and design constraints of the technical solution. Professional technicians can use different methods to implement the described functions for each specific application, but such implementation should not be considered to exceed the scope of this application.

[0058] Those skilled in the art can clearly understand that for the convenience and brevity of description, the specific working processes of the systems, devices, and modules described above can refer to the corresponding processes in the foregoing method embodiments and will not be elaborated herein.

[0059] In several embodiments provided in the present application, it should be understood that the disclosed systems, devices, and methods can be implemented in other ways. For example, the device embodiments described above are merely illustrative. For example, the division of the modules is only a logical function division. In actual implementation, there may be other division methods. For example, multiple modules or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the displayed or discussed coupling, direct coupling, or communication connection to each other can be through some interfaces. The indirect coupling or communication connection of the devices or modules can be in electrical, mechanical, or other forms.

[0060] The modules described as separate components may or may not be physically separated. The components shown as modules may or may not be physical modules. They can be located in one place or distributed to multiple network modules. Some or all of the modules can be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0061] In addition, in each embodiment of the present application, the functional modules can be integrated into one processing module, or each module can exist physically alone, or two or more modules can be integrated into one module.

[0062] If the functions are implemented in the form of software function modules and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present application, in essence, or the part that contributes to the prior art, or this part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions for causing a computer device (which can be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods described in each embodiment of the present application. The foregoing storage medium includes: various media such as USB flash drives, mobile hard disks, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical discs that can store program codes.

[0063] As described above, the above are only specific implementation manners of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art within the technical scope disclosed in the present application can easily think of changes or substitutions, which should all be covered by the protection scope of the present application. Therefore, the protection scope of the present application should be subject to the protection scope of the claims.

[0064] Finally, the above are only the preferred embodiments of the present invention and are not intended to limit the present invention. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principle of the present invention shall be included within the protection scope of the present invention.

Claims

1. A method for assessing the state of agitated patients based on big data analysis, characterized in that: The steps include: Collecting the original multi-source data streams of the patient within a preset time period, and performing feature dimension decomposition and preliminary redundancy marking on the original data in each data stream to obtain an initial feature dimension set of the original multi-source data stream; For the initial feature dimension set, a nonlinear redundancy analysis method is used to identify the feature subsets that produce false emotional feature correlations to form a redundant feature factor set; Performing combined effect cross-validation analysis on the redundant feature factor set to determine the feature dimension combination that produces the emotion illusion effect and generate a set of false emotion association patterns; According to the false emotion association pattern set, the feature combination with the emotion illusion effect is removed from the initial feature dimension set of the original multi-source data stream to obtain the corrected effective feature data set; Based on the effective feature data set, a prediction model for the evolution of the agitation state is constructed to evaluate the patient's current agitation state and output the recognition result of the patient's agitation state.

2. A method for assessing the state of agitated patients based on big data analysis according to claim 1, characterized in that: Collect multi-source data streams of patients within a preset time period, and perform feature dimension decomposition and preliminary redundancy marking on the original data in each data stream to obtain the initial feature dimension set of the original data stream, specifically: Collect patient heart rate data stream, skin conductance pulse data stream, patient motion frequency data stream and spatial displacement trajectory data stream; Decompose the patient's heart rate data stream into time domain features and frequency domain features; Decompose the skin conductance pulse data stream into dimensions of pulse amplitude features and peak density features; Decompose the patient's action frequency data stream into the dimensions of the number of actions per unit time and action rhythm; Decompose the spatial displacement trajectory data stream into dimensions of trajectory curvature and displacement frequency characteristics; The data dimension correlation analysis method is used to perform preliminary redundancy marking on each feature dimension after decomposition to obtain the initial feature dimension set.

3. A method for assessing the state of agitated patients based on big data analysis according to claim 2, characterized in that: The data dimension correlation analysis method is used to perform preliminary redundancy marking on each feature dimension after disassembly, and the initial feature dimension set is obtained, which is specifically: Calculate the Pearson correlation coefficient matrix between each feature dimension after decomposition to determine the degree of linear correlation between feature dimensions; Compare the correlation coefficients between any two feature dimensions in the Pearson correlation coefficient matrix. When the correlation coefficient is greater than the linear correlation threshold, mark the feature dimension with low diagnostic contribution in the two feature dimensions as a preliminary redundant dimension. Classify and mark all feature dimensions marked as preliminary redundant dimensions to form a preliminary redundant feature mark set; The preliminary redundant feature tag set is mapped and compared with each feature dimension set after disassembly to obtain the initial feature dimension set.

4. The method for assessing the state of agitated patients based on big data analysis according to claim 3, characterized in that: For the initial feature dimension set, a nonlinear redundancy analysis method is used to identify the feature subset that produces false emotional feature correlations and form a redundant feature factor set, specifically: The kernel function mapping method is used for the initial feature dimension set to project the feature dimensions into a high-dimensional nonlinear feature space; The nonlinear correlation between the projected feature dimensions is evaluated based on the mutual information index; When the mutual information index between the projected feature dimensions is greater than the nonlinear redundancy threshold, the corresponding feature dimension combination is determined as a feature subset with false emotion feature correlation; All identified feature subsets are integrated to form a redundant feature factor set.

5. The method for assessing the state of agitated patients based on big data analysis according to claim 4, characterized in that: Perform combined effect cross-validation analysis on the redundant feature factor set to determine the feature dimension combination that produces the emotion illusion effect and generate a set of false emotion association patterns, specifically: Each feature subset in the redundant feature factor set is randomly combined to form multiple different combination patterns; For each combination pattern, the data of the combination pattern is input into the preset emotional state classifier by using the cross-validation method, and the interference degree of each combination pattern on the classification result of the agitated state is calculated; When the interference degree of the combination pattern on the classification of the agitated state exceeds a predetermined interference threshold, the combination pattern is determined to be a characteristic dimension combination that causes an emotion illusion effect; All characteristic dimension combinations that appear to have false emotion effects are integrated to generate a set of false emotion association patterns.

6. A method for assessing the state of agitated patients based on big data analysis according to claim 5, characterized in that: According to the set of false emotion association patterns, the feature combinations with emotion illusion effects are removed from the initial feature dimension set of the original multi-source data stream to obtain the corrected effective feature data set, which is specifically: Each feature dimension combination in the false emotion association pattern set is removed from the initial feature dimension set in turn; Conduct relevance review and analysis on the remaining features of the removed feature dimension set; The remaining feature dimension set after the correlation review analysis is re-labeled to obtain a corrected valid feature data set.

7. A method for assessing the state of agitated patients based on big data analysis according to claim 6, characterized in that: Based on the effective feature data set, a prediction model for the evolution of agitation state is constructed to evaluate the patient's current agitation state and output the recognition result of the patient's agitation state, specifically: The effective feature data set is used as input variables to build a time series-based prediction model for the evolution of agitated states. The historical effective feature data set is used to train the parameters of the agitation state evolution prediction model; Input the effective feature data of the patient at the current time node into the agitation state evolution prediction model to calculate the patient's current agitation state probability; The patient's current agitation level is determined based on the probability of the agitation state.

8. The method for assessing the state of agitated patients based on big data analysis according to claim 7, characterized in that: The patient's current agitation level is determined based on the probability of agitation, specifically: Comparing the probability of the agitated state with a preset probability threshold interval; When the probability of the agitation state is within the first probability threshold interval, the patient's current agitation state level is determined to be mild agitation; When the probability of the agitation state is within the second probability threshold interval, the patient's current agitation state level is determined to be moderate agitation; When the probability of the agitation state is within the third probability threshold interval, it is determined that the patient's current agitation state level is severe agitation.

9. A system for assessing the state of an agitated patient based on big data analysis, used to implement a method for assessing the state of an agitated patient based on big data analysis as described in any one of claims 1 to 8, characterized in that: It includes a data processing module, a redundancy analysis module, a cross-validation analysis module, a feature correction module and a state evaluation module; Data processing module: collects the original multi-source data streams of patients within a preset time period, and performs feature dimension decomposition and preliminary redundancy marking on the original data in each data stream to obtain the initial feature dimension set of the original multi-source data stream; Redundancy analysis module: for the initial feature dimension set, a nonlinear redundancy analysis method is used to identify the feature subsets that produce false emotional feature correlations and form a redundant feature factor set; Cross-validation analysis module: performs combined effect cross-validation analysis on redundant feature factor sets, determines the feature dimension combination that produces emotion illusion effects, and generates a set of false emotion association patterns; Feature correction module: based on the false emotion association pattern set, removes the feature combination with emotion illusion effect from the initial feature dimension set of the original multi-source data stream to obtain the corrected valid feature data set; State evaluation module: constructs an agitation state evolution prediction model based on a valid feature data set, evaluates the patient's current agitation state, and outputs the recognition result of the patient's agitation state.

Images