A cognitive behavioral testing method combined with medical data analysis

Abstract

The present invention discloses a cognitive behavioral testing method combined with medical data analysis, which relates to the field of data analysis and processing technology. It connects to a medical database, mines cognitive behavioral maps, introduces a penalty coefficient based on specific variables, and supervises the training of a digital testing module. It builds a behavioral cognitive platform, and embeds the digital testing module into the behavioral cognitive platform. It establishes a platform connection based on account information, uploads behavioral source data, makes behavioral cognitive decisions by constructing a bio-behavioral portrait, determines behavioral cognitive results, and generates a platform test pop-up window for interface visualization. It is used to solve the technical problems existing in the prior art, such as the inability to effectively handle complex data correlations and uncertainties, the lack of flexibility and accuracy in cognitive behavioral testing, and the difficulty in providing clear and operational cognitive behavioral assessment results. It can efficiently and specifically analyze behavioral data and identify cognitive disorders.

Description

Technical Field

[0001] The present invention relates to the technical field of data analysis and processing, and in particular to a cognitive behavior testing method combined with medical data analysis. Background Art

[0002] How to conduct effective cognitive behavioral assessments through intelligent means has become a hot topic. Existing cognitive behavioral testing methods are mostly based on static questionnaires and subjective assessments. While these methods can reflect an individual's cognitive state to a certain extent, they still have limitations in terms of accuracy, efficiency, and real-time performance. They struggle to comprehensively consider an individual's performance across multiple behavioral modes, and they lack real-time data updates and feedback.

[0003] In addition, with the large-scale accumulation of medical data, accurate cognitive behavioral analysis can be carried out using these data. However, existing technologies lack intelligent analysis in terms of analytical depth and comprehensiveness. Analyzing multidimensional data accumulated in medical databases, such as physiological data and behavioral data, can not only improve the accuracy of the test, but also better discover individual cognitive differences and potential problems.

[0004] In summary, existing technologies cannot effectively handle complex data correlations and uncertainties, lack flexibility and accuracy in cognitive behavioral testing, and are unable to provide clear and actionable cognitive behavioral assessment results. Summary of the Invention

[0005] The present application provides a cognitive behavioral testing method combined with medical data analysis, which is used to solve the technical problems existing in the existing technology, such as the inability to effectively handle complex data correlations and uncertainties, the lack of flexibility and accuracy in cognitive behavioral testing, and the difficulty in providing clear and actionable cognitive behavioral assessment results.

[0006] In view of the above problems, the present application provides a cognitive behavioral testing method combined with medical data analysis.

[0007] In the first aspect, the present application provides a cognitive behavioral testing method combined with medical data analysis, the method comprising: connecting to a medical database, mining a cognitive behavioral graph, wherein the cognitive behavioral graph is marked with fuzzy association rules, and there is graph node pruning based on mutual information values ​​and symmetric uncertainty coefficients; according to the cognitive behavioral graph, introducing a penalty coefficient based on specific variables, and supervising the training of a digital testing module; building a behavioral cognition platform, and embedding the digital testing module into the behavioral cognition platform, establishing a platform connection based on account information, including a medical side connection and a user side connection; uploading behavioral source data to the behavioral cognition platform, and making behavioral cognitive decisions based on the digital testing module by constructing a biological behavioral portrait, and determining the behavioral cognition results; generating a platform test pop-up window based on the behavioral cognition results, and performing interface visualization.

[0008] In the second aspect, the present application provides a cognitive behavioral testing system combined with medical data analysis, the system comprising: a graph mining unit for connecting to a medical database and mining cognitive behavioral graphs, wherein the cognitive behavioral graphs are marked with fuzzy association rules and there is graph node pruning based on mutual information values ​​and symmetric uncertainty coefficients; a module training unit for introducing a penalty coefficient based on specific variables according to the cognitive behavioral graph to supervise the training of digital testing modules; a connection establishment unit for building a behavioral cognition platform, and embedding the digital testing module into the behavioral cognition platform, establishing a platform connection based on account information, which includes a medical side connection and a user side connection; a behavioral cognition unit for uploading behavioral source data to the behavioral cognition platform, and making behavioral cognition decisions based on the digital test modules by constructing biological behavioral portraits to determine behavioral cognition results; an interface display unit for generating a platform test pop-up window based on the behavioral cognition results for interface visualization.

[0009] One or more technical solutions provided in this application have at least the following technical effects or advantages:

[0010] The embodiment of the present application provides a cognitive behavioral testing method combined with medical data analysis, which connects to a medical database, mines cognitive behavioral maps, introduces a penalty coefficient based on specific variables based on the cognitive behavioral maps, and supervises the training of a digital testing module; builds a behavioral cognitive platform, and embeds the digital testing module into the behavioral cognitive platform, establishes a platform connection based on account information, uploads behavioral source data to the behavioral cognitive platform, and constructs a biological behavioral portrait to make behavioral cognitive decisions based on the digital testing module and determine behavioral cognitive results; generates a platform test pop-up window based on the behavioral cognitive results and performs interface visualization. It is used to solve the technical problems existing in the prior art that it cannot effectively handle complex data correlations and uncertainties, lacks flexibility and accuracy in cognitive behavioral testing, and is difficult to provide clear and operational cognitive behavioral assessment results. It can efficiently and specifically analyze behavioral data and identify cognitive disorders. BRIEF DESCRIPTION OF THE DRAWINGS

[0011] Figure 1 A flowchart of a cognitive behavioral testing method combined with medical data analysis is provided for this application;

[0012] Figure 2 A schematic diagram of the construction process of a digital testing module in a cognitive behavioral testing method combined with medical data analysis is provided for this application;

[0013] Figure 3 A schematic diagram of the structure of a cognitive behavioral testing system combined with medical data analysis is provided for this application.

[0014] Explanation of the accompanying symbols: graph mining unit 11, module training unit 12, connection establishment unit 13, behavior recognition unit 14, interface display unit 15. DETAILED DESCRIPTION

[0015] This application provides a cognitive behavioral testing method that combines medical data analysis, connects to medical databases, mines cognitive behavioral maps, introduces penalty coefficients based on specific variables, and supervises the training of digital testing modules. It also builds a behavioral cognitive platform and embeds the digital testing module into the behavioral cognitive platform. It establishes a platform connection based on account information, uploads behavioral source data, makes behavioral cognitive decisions by constructing bio-behavioral portraits, determines behavioral cognitive results, and generates a platform test pop-up window for interface visualization. It is used to solve the technical problems existing in the existing technology, such as the inability to effectively handle complex data correlations and uncertainties, the lack of flexibility and accuracy in cognitive behavioral testing, and the difficulty in providing clear and actionable cognitive behavioral assessment results.

[0016] Example 1: Figure 1As shown, the present application provides a cognitive behavioral testing method combined with medical data analysis, the method comprising:

[0017] S1: Connect to the medical database and mine the cognitive behavior map, wherein the cognitive behavior map is marked with fuzzy association rules and there is map node pruning based on mutual information value and symmetric uncertainty coefficient.

[0018] First, the medical database is a local or shared database with access permissions. By connecting to the medical database, it can acquire and integrate multiple patient data sources in real time, such as behavioral data, physiological data, and historical diagnostic information. These data sources provide a rich foundation for subsequent graph construction. By clustering and association analysis of cognitive behaviors, a cognitive behavioral graph is mined. The cognitive behavioral graph analyzes the relationship between patient behavior and environmental factors, revealing key data structures such as behavioral characteristics, cognitive patterns, and potential diseases.

[0019] Specifically, fuzzy association rules refer to relationships that cannot be explicitly expressed but provide potential clues. For example, the relationship between certain behaviors and cognitive abilities may not be direct, but implicit connections can be discovered through data analysis. By mining these fuzzy rules, the system can capture complex patterns in patient behavior and improve the accuracy of assessments.

[0020] Next, the system prunes nodes in the cognitive behavior graph based on mutual information and symmetric uncertainty coefficients. This allows for an assessment of significant correlations between nodes in the graph. If two nodes have low mutual information or symmetric uncertainty, indicating a weak connection, the system removes them from the graph, reducing the presence of insignificant information.

[0021] By effectively screening graph nodes and optimizing the quality of cognitive behavioral graphs, we can focus on the most representative behavioral patterns and cognitive factors, avoid the impact of noise data on the results, and thus improve the accuracy and credibility of behavioral cognitive decisions.

[0022] Furthermore, to mine the cognitive behavior map, step S1 of this application includes:

[0023] The medical database is traversed, clustered by cognitive behavioral classes, and N behavioral classes are determined; cognitive factors are mined based on the N behavioral classes, wherein the cognitive factors meet a frequency threshold and include universal factors and specific factors; the cognitive factors are refined at multiple levels of granularity to construct N types of graphs; the N types of graphs are integrated and graph nodes are pruned to generate the cognitive behavioral graph.

[0024] In this step, we traverse the medical database, comprehensively analyze the various cognitive behavioral data in the database, and classify these data into different cognitive behavioral categories using behavioral classification methods. A cognitive behavioral category is a group of behavioral instances with similar behavioral patterns or cognitive characteristics. These behaviors are usually associated with specific cognitive processes, psychological states, or physiological reactions.

[0025] By traversing a large amount of medical data in the database, the data is clustered based on behavioral patterns and cognitive characteristics, and N behavioral clusters are identified. The N behavioral clusters here represent the diversity of all behavioral patterns in the database. The specific number N depends on the complexity of the data and the set classification criteria.

[0026] Next, cognitive factors are mined based on the determined N behavioral categories. Cognitive factors refer to those factors that play a key role in the process of behavioral cognition and can effectively characterize different dimensions of behavior. In order to ensure that the mined cognitive factors are sufficiently representative and regular, the frequency threshold is set, that is, only when the frequency of a factor exceeds the preset threshold will it be considered a valid cognitive factor. For example, factors that appear frequently in a certain type of behavior and have individual universality are considered universal factors; factors specific to certain individuals, such as allergies, are considered specific factors. That is, universal factors are common to individuals, and specific factors exist in individual cases. By setting the frequency threshold, the system can screen out cognitive factors with sufficient statistical significance and avoid interference from invalid data.

[0027] On this basis, the discovered cognitive factors are refined at multiple levels. Multi-level granularity refinement means a more detailed division and analysis of the different levels and dimensions of cognitive factors, further improving the expressiveness and accuracy of the factors. For example, universal factors may have similar manifestations in multiple cognitive behavioral categories. Therefore, through refinement, they can be divided into smaller sub-factors based on different situations, thus providing more precise cognitive characteristics for each behavioral category. Through this refinement process, the system is able to construct N types of maps, each of which represents the knowledge structure of a specific cognitive behavioral pattern.

[0028] After constructing N types of maps, they are integrated to form a comprehensive cognitive-behavioral map. This integrated map not only encompasses the characteristics of each cognitive-behavioral class but also incorporates the refined results of all cognitive factors, resulting in a higher resolution and information density for the entire cognitive-behavioral map. However, nodes in the map often contain redundant or irrelevant data. Therefore, after integration, the most representative and influential nodes are pruned to make them more concise, clear, and easier to analyze, providing a high-quality knowledge foundation for subsequent cognitive-behavioral decision-making and testing.

[0029] In summary, by constructing an efficient and accurate cognitive behavioral map, we can reveal the relationship between individual behavior and cognition, and provide strong support for behavioral cognitive testing and decision-making.

[0030] Furthermore, the graph node pruning step S1 of this application includes:

[0031] The calculation formula for obtaining the mutual information value is: f(a,b)=p(a)+p(b)-p(a,b); the calculation formula for obtaining the symmetric uncertainty coefficient is: s(a,b)=2*f(a,b) / [p(a)+p(b)]; wherein, f(a,b) is the mutual information value of the graph nodes a and b with a fuzzy association relationship, p(a) is the information entropy of the graph node a, p(b) is the information entropy of the graph node b, and p(a,b) is the combined information entropy; s(a,b) is the symmetric uncertainty coefficient; set a first threshold based on the mutual information value and a second threshold based on the symmetric uncertainty coefficient, traverse the cognitive behavior graph, and if the first threshold and the second threshold are not met, prune the graph nodes.

[0032] In this step, we first calculate the mutual information value using the formula: f(a,b) = p(a) + p(b) - p(a,b), where f(a,b) represents the mutual information between graph nodes a and b. Mutual information is a metric used to measure the dependency between two random variables, reflecting the amount of information shared between the two nodes.

[0033] Specifically, p(a) is the information entropy of graph node a, representing the individual uncertainty of node a; p(b) is the information entropy of node b, representing the individual uncertainty of node b; and p(a,b) is the combined information entropy of nodes a and b, representing their shared overall uncertainty. By calculating the mutual information value, we can determine the degree of association between nodes a and b, and thus determine whether a fuzzy association exists. If two nodes share a lot of information, they may have some kind of implicit association.

[0034] The symmetric uncertainty coefficient is further calculated using the formula: s(a,b) = 2*f(a,b) / [p(a) + p(b)]. The symmetric uncertainty coefficient s(a,b) is a standardized measure of the mutual information value, representing the relative degree of association between nodes a and b. This coefficient is calculated by multiplying the mutual information value by 2 and then dividing it by the sum of the information entropy of nodes a and b. A larger symmetric uncertainty coefficient indicates a closer relationship between the two nodes; conversely, a smaller coefficient indicates a weaker association between the nodes. This coefficient can more accurately determine whether there is a strong association between graph nodes and further provide a basis for graph pruning.

[0035] Then, a first threshold based on the mutual information value and a second threshold based on the symmetric uncertainty coefficient are set. The setting of the thresholds depends mainly on the actual application requirements and data characteristics. The first threshold is based on the mutual information value. When the mutual information value is lower than this threshold, the correlation between the nodes is considered to be weak and may not have significant cognitive behavioral characteristics. The second threshold is based on the symmetric uncertainty coefficient. If this coefficient is lower than the set second threshold, it indicates that the correlation between the nodes is weak or the correlation is unclear.

[0036] The cognitive behavioral map is traversed, and each node is judged based on a preset first and second thresholds. If the mutual information value and symmetric uncertainty coefficient of the map node do not reach the corresponding threshold, the correlation between the nodes is considered low, and pruning can be performed. This simplifies the cognitive behavioral map, making it clearer and more concise, and more efficient for subsequent behavioral cognitive testing and decision analysis.

[0037] S2: According to the cognitive behavioral map, a penalty coefficient based on specific variables is introduced to supervise the training of the digital test module.

[0038] In this step, a penalty coefficient based on specific variables is first introduced according to the cognitive behavioral map. Specific variables refer to unique variables in the cognitive behavioral map that can significantly affect behavioral patterns or test results. They usually correspond to specific individual behavioral characteristics or environmental factors. For example, under the test results of universal factors, some individuals may have specific variables, such as allergic factors, which will cause the test results based on universal factors to be inapplicable to the individual. By introducing a penalty coefficient, the fit of individual tests can be improved.

[0039] By introducing a penalty coefficient based on the specific variables, we can meet the testing requirements and accuracy for both universal individuals and individuals with specific factors. The specific variables refer to the different factor states based on the specific factors. By constructing the digital testing module through sample supervised learning, we provide a more stable and accurate model support for subsequent behavioral cognitive decision-making.

[0040] Further, such as Figure 2 As shown, the supervised training digital test module, step S2 of this application includes:

[0041] Determine the time series based on the behavioral cycle, perform periodic transformation on the cognitive behavioral map, and determine the time series map, which includes multiple periodic behavioral stages; retrieve training samples based on the medical database, wherein the training samples include a first sample and a second sample; traverse the time series map, for the universal factor part, use the inference analysis based on the probability matrix as a benchmark, perform sample-driven training based on the first sample, and determine the first training block; for the specific factor part, determine the penalized linear relationship under the factor trend change, perform probability matrix migration training based on the second sample on the first training block, and determine the second training block; construct the digital test module based on the first training block and the second training block.

[0042] Among them, the behavioral cycle refers to the time span of a complete cognitive behavior. During this period, the individual's behavioral pattern will show a certain regularity, that is, the behavior in different cycle stages is different, and the corresponding cognitive results are also different. Through the definition of the behavioral cycle, the cognitive behavioral map can be converted into a time series map, wherein the time series map not only contains the time information of the behavioral cycle, but also can subdivide the multiple behavioral stages in the cycle. These behavioral stages reflect the changes in the individual's cognitive state within a specific time period. For example, the behavioral stage may correspond to the peak or trough of cognitive ability, or some special behavioral patterns.

[0043] Next, training samples are retrieved based on the medical database, and the training samples include a first sample and a second sample. The first sample usually contains a wide range of universal data, representing the common cognitive behavioral characteristics of most individuals; while the second sample contains the basic entropy of the universal data, as well as data on specific factors. The existence of specific factors may cause changes in test results. By obtaining these sample data, the system can fully consider individual differences and distinguish between universal factors and specific factors in the subsequent training process. Universal factors refer to behavioral characteristics that can be commonly seen in most individuals, while specific factors correspond to the unique behavioral characteristics of certain specific individuals or groups.

[0044] On this basis, the system traverses the time series graph and, targeting the universal factors, employs a probability matrix-based inference analysis method for sample-driven training. A probability matrix is ​​a mathematical tool that can represent the relationships between multiple behavioral factors. It helps reveal the connection between universal factors and behavior. Using the first sample as training data, the system derives the probability distribution between each behavioral stage based on the probability matrix and determines the first training block. This first training block contains the training results for the universal factors, with the goal of ensuring that the model can effectively predict common cognitive-behavioral patterns in the majority of individuals.

[0045] For the specific factors, the system further determines a penalized linear relationship under factor trends to constrain the impact of the specific factors on the test results. The penalized linear relationship characterizes the behavioral impact of the specific factors under different states. Based on the data of the second sample, probability matrix transfer training is performed with the probabilistic impact based on the penalized linear relationship, thereby adjusting the parameters in the first training block to form the second training block.

[0046] By combining the results of the first and second training blocks, the digital testing module was ultimately constructed. After training, this module was able to accurately identify the differences between universal and specific behaviors in different individuals and conduct behavioral cognitive tests based on these differences.

[0047] In summary, this step successfully constructed a digital test module that can adapt to individual differences by dividing universal factors and specific factors in detail, combining the time series graph of periodic behavior, and using a combination of probability matrix derivation and transfer training, providing scientific and precise support for behavioral cognitive decision-making.

[0048] S3: Build a behavior recognition platform, and embed the digital test module into the behavior recognition platform, and establish a platform connection based on the account information, including a medical side connection and a user side connection.

[0049] In this implementation step, the first step is to build a behavioral cognition platform. The behavioral cognition platform is an integrated software platform designed to provide users with behavioral cognitive analysis, data processing, decision support and other services. By integrating multiple technical modules, such as data acquisition modules, data analysis modules, cognitive assessment modules, and decision support modules, the platform can effectively process and analyze individual behavioral data and provide personalized cognitive intervention plans. Through this platform, users can identify behavioral patterns, assess cognitive abilities, and provide a scientific basis for individual health management, education and training, etc.

[0050] Next, the digital testing module is integrated into the behavioral cognition platform. This module is a functional module embedded within the platform for behavioral cognition assessment and testing. This module uses digital tools and algorithmic models to analyze and evaluate individual behavioral characteristics, helping to determine their cognitive status, behavioral patterns, and more. With this module integrated, the platform can perform automated behavioral cognition testing and generate corresponding assessment reports.

[0051] Among them, the account information refers to the basic data used to identify and manage user identity, including user name, password, permission settings, etc. The medical side connection refers to the connection between the platform and the medical institution or system. It can access and share medical data, such as individual health records, medical history information, etc., to help the platform conduct behavioral cognitive analysis and health management. Through the medical side connection, the platform can obtain individual medical data in real time, combine it with behavioral data for comprehensive evaluation, and thus provide users with personalized health intervention recommendations. The user side connection refers to the connection between the platform and individual users, which includes the user's personal information, behavior records, test results and other data. Through the user side connection, the platform can provide users with real-time cognitive analysis, test feedback and behavior improvement suggestions to help users better manage their behavioral health.

[0052] S4: uploading the behavioral source data to the behavioral cognition platform, constructing a biological behavioral profile, making behavioral cognition decisions based on the digital test module, and determining behavioral cognition results;

[0053] S5: Generate a platform test pop-up window based on the behavioral recognition results and perform interface visualization.

[0054] In this step, behavioral source data, i.e., collected behavioral performance data of individuals or groups in specific situations, is first uploaded to the behavioral recognition platform. This data reflects the cognitive and behavioral characteristics of individuals in different situations.

[0055] Next, the platform constructs a biobehavioral profile based on the uploaded behavioral source data. A biobehavioral profile is a multidimensional user image encompassing physiological and psychological characteristics, generated through a comprehensive analysis of individual or group behavioral data. This process involves extracting behavioral features, pattern recognition, and mining feature associations, aiming to capture the behavioral patterns of individuals in different situations. For example, an individual's response patterns in stressful situations and performance in cognitive tasks will be reflected in the biobehavioral profile. Through biobehavioral profiles, the platform can conduct a comprehensive, multi-faceted assessment of an individual's cognitive behavioral characteristics, providing an important basis for subsequent cognitive decision-making.

[0056] The digital test module is further used to analyze the behavioral data of individuals or groups, and use probability matrices, cognitive factors, feature derivation, etc. to help evaluate the behavioral cognitive status.

[0057] Next, based on the behavioral recognition results, the platform generates a test pop-up window and visualizes the interface. A test pop-up window is an interface window that automatically pops up after the platform generates behavioral recognition results, displaying the test results and related analysis. This pop-up window may include various methods such as charts, text, and color coding to enable users to intuitively understand the behavioral recognition test results.

[0058] For example, a behavioral test might display results through color changes (e.g., red for abnormal, green for normal), or chart an individual's performance on a cognitive task. Furthermore, the platform's visual interface is designed to enhance the user experience, allowing users to easily understand complex behavioral cognitive data and make timely decisions.

[0059] In summary, by uploading behavioral source data to the platform, combining it with the construction of biobehavioral profiles, and utilizing digital testing modules to make behavioral cognitive decisions, behavioral cognitive results are ultimately generated and visualized through the platform's test pop-up window. This not only ensures accurate analysis and evaluation of behavioral cognitive data, but also enables users to quickly and intuitively access key cognitive information through visualization, thereby supporting decision-making and intervention in fields such as healthcare, education, and psychology.

[0060] Furthermore, by constructing a biological behavior profile, a behavior cognition decision is made based on the digital test module to determine the behavior cognition result. Step S4 of this application includes:

[0061] If it is a medical side connection, the user detection data stream is called according to the interactive network of medical data; the behavior recognition platform receives the user detection data stream as the behavior source data and constructs a first biological behavior portrait; according to the digital test module, a cognitive test is performed on the first biological behavior portrait to determine the behavior recognition result.

[0062] In the embodiments of this application, the medical connection refers to the data interaction interface between the behavioral recognition platform and the medical system or institution. This connection can obtain medical data, including but not limited to individual health records, test reports, medical history, physical examination data, etc. Through the interactive network of medical data, the platform can access the user's test data stream in real time, that is, various data streams related to the user's health status stored in the medical system, which helps to understand the user's physiological condition and health trends.

[0063] After invoking the user's detection data stream, the behavior recognition platform receives this data and uses it as the behavior source data. Behavior source data refers to the basic data used to construct a biological behavior profile in the behavior recognition platform.

[0064] Among them, the biobehavioral portrait is a comprehensive description that reflects the individual's physiological characteristics, health status and its relationship with behavior. The purpose of constructing a biobehavioral portrait is to comprehensively portray the individual through multi-dimensional data (such as physiological, psychological, behavioral, etc.) to help better understand their health behavior patterns. Specifically, the first biobehavioral portrait includes the individual's basic physiological information (such as age, gender, weight, health status, etc.) and health indicators obtained through test data (such as blood sugar, blood lipids, exercise status, etc.). These portraits provide an important reference for subsequent behavioral cognitive assessment and decision-making.

[0065] Preferably, the positions where different physiological features occur are marked based on the biological body, that is, the contour model of the user, which can improve the intuitiveness of the portrait.

[0066] Next, based on the digital test module, an automated cognitive analysis is performed on the first bio-behavioral portrait to evaluate the individual's cognitive performance and behavioral responses in different health states, and determine the behavioral cognitive results. The behavioral cognitive results refer to the comprehensive evaluation results based on the bio-behavioral portrait and the digital test module, which can reflect the individual's cognitive state, behavioral patterns, and possible health risks. The behavioral cognitive results provide users with personalized health management recommendations, help them understand their current health status and behavioral patterns, and guide them to make necessary health interventions.

[0067] Through this step, the platform can conduct a comprehensive cognitive assessment of users and develop personalized behavioral intervention plans based on the results, thereby improving the efficiency of individual health management while helping to enhance behavioral cognition.

[0068] Furthermore, by constructing a biological behavior profile, a behavior cognition decision is made based on the digital test module to determine the behavior cognition result. Step S4 of this application includes:

[0069] If it is a user-side connection, the user uploads behavioral data to the behavioral recognition platform, wherein the behavioral data includes scanned text and user subjective description; the behavioral data is used as the behavioral source data to construct a second biological behavioral portrait, wherein there is natural language processing based on the user's subjective description; according to the digital test module, a cognitive test is performed on the second biological behavioral portrait to determine the behavioral recognition result.

[0070] Among them, the user-side connection refers to the data interaction interface between the behavior recognition platform and the end user. Users upload their daily behavior data to the platform through smart devices or other data collection methods. The sources of behavioral data can be diverse, including various types of data recorded by users through mobile devices, wearable devices or other data collection tools, such as exercise data, sleep data, diet records, etc. If it is text data, such as medical records, etc., it can be scanned and uploaded through optical character recognition (OCR) technology. In addition, behavioral data can also include subjective descriptions entered by users through text, reflecting their personal feelings, emotional state or daily living habits.

[0071] Using the aforementioned behavioral data as the behavioral source data, the behavioral recognition platform constructs a second bio-behavioral profile based on this data. This second bio-behavioral profile is an individualized health and behavioral profile formed by comprehensively analyzing the user's physiological state, behavioral habits, and psychological feelings based on the behavioral data uploaded by the user. This bio-behavioral profile encompasses not only physiological data but also multi-dimensional data such as the user's emotional state, subjective experience, and behavioral motivations. By analyzing the scanned text and subjective descriptions uploaded by users, the platform can more comprehensively characterize the user's behavioral patterns and form a precise second bio-behavioral profile.

[0072] In the process of constructing the second bio-behavioral portrait, there is natural language processing based on the user's subjective description. Natural language processing (NLP) refers to the process of analyzing and understanding human language through computer technology. In this embodiment, the platform analyzes the user's subjective description through natural language processing technology, which may be more colloquial, converts text information into structured data, and extracts emotional information, behavioral intentions or other potential behavioral characteristics. For example, by analyzing the user's description of daily activities, the platform can identify the user's emotional fluctuations, living habits and other information, thereby more accurately constructing the user's cognitive behavioral portrait.

[0073] Next, based on the second bio-behavioral profile, the behavioral cognition platform uses a digital testing module to conduct cognitive testing and determine behavioral cognition results. The behavioral cognition results are a comprehensive assessment of the user's cognitive state and behavioral patterns, revealing individual behavioral characteristics, emotional changes, and cognitive abilities.

[0074] For example, the platform might use cognitive test results to determine whether a user has a high level of emotional regulation or potential cognitive biases. Behavioral cognitive results provide users with personalized health management recommendations to help them improve their behavior, adjust their emotional state, or implement other health interventions.

[0075] Furthermore, step S4 of this application includes:

[0076] Identify the behavioral source data, standardize the data format and perform redundancy removal, and determine multiple data sequences, wherein the data sequence includes data location and data characteristics, and the data location is the location where the behavior occurs; based on the multiple data sequences, perform distribution integration based on the data location, perform distribution identification based on the data characteristics, and generate a biological behavior portrait.

[0077] First, identify the behavioral source data, extract valuable behavioral data from the raw data stream received by the behavioral recognition platform, and standardize the data format and remove redundancy for the behavioral source data. Standardizing the data format refers to converting the behavioral source data into a unified format so that it can be easily processed and analyzed. Data normalization usually involves operations such as timestamps, standardization of data types, and unit unification. De-redundancy is the removal of duplicate or redundant data items during the processing process to ensure the uniqueness and validity of each piece of data. This is helpful in improving data processing efficiency, reducing computational burden, and avoiding data analysis deviations caused by redundant data.

[0078] After standardizing the data format and removing redundancies, the platform divides behavioral data into multiple data sequences. A data sequence is a collection of behavioral data that is continuous in time and relatively fixed in location. Each data sequence typically represents the trajectory of a specific behavior or physiological state, such as heart rate changes, cadence, or mood swings. Each data sequence contains two key pieces of information: data location and data characteristics.

[0079] Data location refers to the location where the behavior occurred, such as an abnormal heart position, or the scene location, such as a specific room, work area, or outdoor environment. Data features describe the content or nature of the behavioral data itself, such as heart rate, step count, sound intensity, etc. The combination of these two types of information effectively helps the platform identify and analyze behavioral patterns in different environments and situations.

[0080] Based on the multiple data sequences, distributed integration is performed. Distributed integration refers to unifying and aligning multiple data sequences in time and space, analyzing the relationships between different data sequences, and obtaining more comprehensive and accurate behavioral data.

[0081] Next, the platform distributes and labels data based on its characteristics. For example, for an abnormal heart rate, the data location is the heart position, and the abnormal frequency is the data characteristic. Through integration, this effectively improves intuitiveness and generates a bio-behavioral profile. A bio-behavioral profile is a comprehensive portrayal of user behavior, integrating multi-dimensional data such as physiological data, behavioral habits, psychological state, and environmental information to accurately depict individual behavioral characteristics.

[0082] Furthermore, the cognitive test is performed, and step S4 of this application includes:

[0083] Identify the biological behavior profile and determine whether a specific factor exists; if so, perform a cognitive test based on the second training block; if not, perform a cognitive test based on the first training block.

[0084] First, identify the biobehavioral profile. Determine whether specific factors exist within the biobehavioral profile. Specific factors refer to specific variables or characteristics within an individual's behavior or physiological state that significantly influence cognition. This determination is made by matching the profile with the cognitive-behavioral profile. If present, this indicates that the individual's cognitive level may be influenced by specific factors, necessitating further cognitive testing through relevant training blocks.

[0085] If no specific factor is found, cognitive testing is performed based on the first training block. The first training block is targeted at individuals without a specific factor, or individuals whose cognitive performance is not significantly affected by a specific factor. It typically includes conventional cognitive ability assessment tasks.

[0086] If the presence of a specific factor is identified, the system conducts a cognitive test based on the first training block. The first training block refers to a cognitive test module for individuals without specific factors. Such training blocks usually include some behavioral tasks or cognitive assessments for specific factors, such as emotion recognition, stress testing, and attention training. These tasks can effectively help the system evaluate changes in cognitive performance caused by specific factors, and further determine the extent of the impact of the factor on individual behavioral cognition. Through this cognitive test, the system can obtain more accurate behavioral cognitive results and provide data support for subsequent interventions. Improve the pertinence of the test and user fit.

[0087] This application provides a cognitive behavioral testing method combined with medical data analysis, which has the following technical effects:

[0088] 1. By analyzing individual behavioral data, a comprehensive biobehavioral profile is constructed. This includes information such as individual physiological data, behavioral patterns, and emotional fluctuations, comprehensively presenting the individual's behavior in different environments. This accurately identifies individual behavioral characteristics and provides a reliable data foundation for subsequent cognitive testing.

[0089] 2. Different cognitive testing blocks are used depending on the presence of specific factors. The first training block is designed for general cognitive testing, with tasks assessing the impact of specific factors such as emotion and stress. The second training block is used for testing with specific factors, assessing an individual's basic cognitive abilities. Technically, this customized testing approach can improve assessment accuracy and ensure more representative and practical test results.

[0090] In summary, the accuracy of cognitive tests can be improved at different levels, ensuring that individual cognitive assessment and intervention are more effective and targeted.

[0091] Example 2: Based on the same inventive concept as the cognitive behavioral testing method combined with medical data analysis in the above embodiment, Figure 3 As shown, the present application provides a cognitive behavioral testing system combined with medical data analysis, the system comprising:

[0092] A graph mining unit 11 is used to connect to a medical database and mine cognitive behavioral graphs, wherein the cognitive behavioral graphs are marked with fuzzy association rules and have graph node pruning based on mutual information values ​​and symmetric uncertainty coefficients;

[0093] A module training unit 12 is configured to introduce a penalty coefficient based on specific variables according to the cognitive behavior map to supervise the training of the digital test module;

[0094] The connection establishment unit 13 is used to build a behavior recognition platform, embed the digital test module into the behavior recognition platform, and establish a platform connection based on the account information, including a medical side connection and a user side connection;

[0095] The behavior recognition unit 14 is used to upload the behavior source data to the behavior recognition platform, build a biological behavior profile, make behavior recognition decisions based on the digital test module, and determine the behavior recognition results;

[0096] The interface display unit 15 is used to generate a platform test pop-up window based on the behavior recognition result to perform interface visualization.

[0097] Among them, the module training unit 12 is also used to perform the following steps: traverse the medical database, cluster with cognitive behavior classes, and determine N behavior classes; based on the N behavior classes, mine cognitive factors, wherein the cognitive factors meet the frequency threshold and include universal factors and specific factors; perform multi-layer granularity refinement on the cognitive factors to construct N types of graphs; integrate the N types of graphs and perform graph node pruning to generate the cognitive behavior graph.

[0098] In which, the module training unit 12 is also used to perform the following steps: obtain the mutual information value calculation formula: f(a,b)=p(a)+p(b)-p(a,b); obtain the symmetric uncertainty coefficient calculation formula: s(a,b)=2*f(a,b) / [p(a)+p(b)]; wherein, f(a,b) is the mutual information value of graph nodes a and b with a fuzzy association relationship, p(a) is the information entropy of graph node a, p(b) is the information entropy of graph node b, and p(a,b) is the combined information entropy; s(a,b) is the symmetric uncertainty coefficient; set a first threshold based on the mutual information value and a second threshold based on the symmetric uncertainty coefficient, traverse the cognitive behavior graph, and if the first threshold and the second threshold are not met, prune the graph nodes.

[0099] Among them, the module training unit 12 is also used to perform the following steps: determine the time series based on the behavioral cycle, perform periodic transformation on the cognitive behavioral map, and determine the time series map, which includes multiple periodic behavioral stages; retrieve training samples based on the medical database, wherein the training samples include a first sample and a second sample; traverse the time series map, for the universal factor part, based on the inference analysis based on the probability matrix, perform sample-driven training based on the first sample to determine the first training block; for the specific factor part, determine the penalized linear relationship under the factor trend change, perform probability matrix migration training based on the second sample on the first training block, and determine the second training block; construct the digital test module based on the first training block and the second training block.

[0100] Among them, the behavioral recognition unit 14 is also used to perform the following steps: if it is a medical side connection, the user detection data stream is called according to the interactive network of medical data; the behavioral recognition platform receives the user detection data stream as the behavioral source data to construct a first biological behavioral portrait; according to the digital test module, the first biological behavioral portrait is cognitively tested to determine the behavioral recognition result.

[0101] Among them, the behavioral recognition unit 14 is also used to perform the following steps: if it is a user-side connection, the user end uploads behavioral data to the behavioral recognition platform, wherein the behavioral data includes scanned text and user subjective description; the behavioral data is used as the behavioral source data to construct a second biological behavioral portrait, wherein there is natural language processing based on the user's subjective description; according to the digital test module, the second biological behavioral portrait is cognitively tested to determine the behavioral recognition result.

[0102] Among them, the behavior recognition unit 14 is also used to perform the following steps: identifying the behavior source data, standardizing the data format and removing redundancy, determining multiple data sequences, wherein the data sequence includes data position and data characteristics, and the data position is the location where the behavior occurs; according to the multiple data sequences, distribution integration is performed based on the data position, and distribution identification is performed based on the data characteristics to generate a biological behavior portrait.

[0103] The behavioral recognition unit 14 is further configured to perform the following steps: identifying the biological behavioral profile and determining whether a specific factor exists; if so, performing a cognitive test based on the second training block; if not, performing a cognitive test based on the first training block.

[0104] Through the above detailed description of a cognitive behavioral testing method combined with medical data analysis in this specification, those skilled in the art can clearly understand a cognitive behavioral testing method combined with medical data analysis in this embodiment. For the device disclosed in the embodiment, since it corresponds to the method disclosed in the embodiment, the description is relatively simple, and the relevant parts can be referred to the method part.

[0105] The above description of the disclosed embodiments is intended to enable one skilled in the art to implement or use the present application. Various modifications to these embodiments will be readily apparent to one skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present application. Therefore, the present application is not limited to the embodiments shown herein, but is intended to conform to the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A cognitive behavioral testing method combined with medical data analysis, characterized in that: The method comprises: Connecting to a medical database to mine cognitive behavioral graphs, wherein the cognitive behavioral graphs are identified by fuzzy association rules and have graph node pruning based on mutual information values ​​and symmetric uncertainty coefficients; According to the cognitive behavioral map, a penalty coefficient based on specific variables is introduced to supervise the training of the digital test module; Building a behavior recognition platform, and embedding the digital testing module into the behavior recognition platform, and establishing a platform connection based on account information, including a medical side connection and a user side connection; Uploading behavioral source data to the behavioral cognition platform, constructing a biological behavioral profile, making behavioral cognition decisions based on the digital test module, and determining behavioral cognition results; Generate a platform test pop-up window based on the behavioral recognition results and perform interface visualization; The supervised training digital test module includes: Determine a time series based on a behavioral cycle, perform a periodic transformation on the cognitive behavioral map, and determine a time series map, which includes a plurality of periodic behavioral stages; Retrieving training samples according to the medical database, wherein the training samples include a first sample and a second sample; Traversing the time series graph, for the universal factor portion, based on the inference analysis based on the probability matrix, performing sample-driven training according to the first sample to determine a first training block; For the specific factor part, determine the penalty linear relationship under the factor trend change, perform probability matrix migration training based on the second sample on the first training block, and determine the second training block; The digital test module is constructed according to the first training block and the second training block.

2. A cognitive behavioral testing method combined with medical data analysis as claimed in claim 1, characterized in that: Mining cognitive behavioral maps, including: Traversing the medical database, clustering by cognitive behavioral categories, and determining N behavioral categories; Mining cognitive factors based on the N behavioral categories, wherein the cognitive factors meet a frequency threshold and include universal factors and specific factors; Perform multi-layer granularity refinement on the cognitive factors to construct N types of maps; The N types of graphs are integrated and graph nodes are pruned to generate the cognitive behavior graph.

3. A cognitive behavioral testing method combined with medical data analysis as claimed in claim 2, characterized in that: The graph node pruning includes: The formula for calculating the mutual information value is: f(a,b)=p(a)+p(b)-p(a,b); The formula for calculating the symmetric uncertainty coefficient is: s(a,b)=2 f(a,b) / [p(a)+p(b)]; Among them, f(a,b) is the mutual information value of graph nodes a and b with fuzzy association relationship, p(a) is the information entropy of graph node a, p(b) is the information entropy of graph node b, and p(a,b) is the combined information entropy; s(a,b) is the symmetric uncertainty coefficient; A first threshold based on the mutual information value and a second threshold based on the symmetric uncertainty coefficient are set, and the cognitive behavior graph is traversed. If the first threshold and the second threshold are not met, the graph nodes are pruned.

4. The cognitive behavioral testing method combined with medical data analysis according to claim 1, characterized in that: By constructing a biological behavior profile, a behavioral cognitive decision is made based on the digital test module to determine the behavioral cognitive results, including: If it is a medical connection, the user detection data flow is called according to the interactive network of medical data; The behavior recognition platform receives the user detection data stream as the behavior source data and constructs a first biological behavior profile; According to the digital testing module, a cognitive test is performed on the first biological behavior portrait to determine a behavioral cognitive result.

5. A cognitive behavioral testing method combined with medical data analysis as claimed in claim 4, characterized in that: By constructing a biological behavior profile, a behavioral cognitive decision is made based on the digital test module to determine the behavioral cognitive results, including: If it is a user-side connection, the user uploads behavior data to the behavior recognition platform, wherein the behavior data includes scanned text and user subjective description; Using the behavior data as the behavior source data, constructing a second bio-behavioral profile, wherein natural language processing based on the user's subjective description is performed; According to the digital test module, a cognitive test is performed on the second biological behavior portrait to determine a behavioral cognitive result.

6. A cognitive behavioral testing method combined with medical data analysis as claimed in claim 5, characterized in that: include: Identify the behavior source data, standardize the data format and perform redundancy removal, and determine multiple data sequences, wherein the data sequences include data locations and data features, and the data locations are locations where the behavior occurs; According to the multiple data sequences, distribution integration is performed based on data location, and distribution identification is performed based on data features to generate a biological behavior portrait.

7. A cognitive behavioral testing method combined with medical data analysis as claimed in claim 6, characterized in that: The cognitive tests performed include: Identify the biological behavior profile and determine whether specific factors exist; If so, conduct cognitive testing based on the first training block; If not present, cognitive testing is performed based on the second training block.

8. A cognitive behavioral testing system combined with medical data analysis, characterized in that: A system for performing a cognitive behavioral testing method combined with medical data analysis according to any one of claims 1 to 7, the system comprising: A graph mining unit, configured to connect to a medical database and mine cognitive behavioral graphs, wherein the cognitive behavioral graphs are identified by fuzzy association rules and have graph node pruning based on mutual information values ​​and symmetric uncertainty coefficients; A module training unit, configured to introduce a penalty coefficient based on specific variables according to the cognitive behavior map, and supervise the training of a digital test module; A connection establishment unit is used to build a behavior recognition platform, embed the digital test module into the behavior recognition platform, and establish a platform connection based on account information, including a medical side connection and a user side connection; A behavior recognition unit is used to upload behavior source data to the behavior recognition platform, construct a biological behavior profile, make behavior recognition decisions based on the digital test module, and determine the behavior recognition results; The interface display unit is used to generate a platform test pop-up window based on the behavioral recognition results to perform interface visualization.

Images