Alzheimer's disease auxiliary screening system based on intelligent interaction

Abstract

The invention discloses an Alzheimer's disease auxiliary screening system based on intelligent interaction, particularly relates to the field of computer-aided diagnosis, and is used for solving the problem that in an existing screening method, it is difficult to quantify cognitive performance differences of life knowledge and professional knowledge at the same time. The method comprises the following steps: constructing a task interaction group covering life knowledge and professional knowledge, collecting and coding multi-dimensional interaction operation of a testee in a task process, and generating an interaction vector matrix; further performing jump point identification, vector residual calculation and invalid operation density statistics on the interaction behavior through a preset task logic chain, and quantifying the operation deviation condition of the interaction behavior; respectively generating life and professional cognitive performance indexes based on the analysis result, and constructing a cognitive deviation evolution curve of a difference value of the life and professional cognitive performance indexes along with time change in a set period; and finally, through identifying a continuous rising trend in a cognitive deviation evolution process, auxiliary screening and early warning of asymmetric degradation of cognitive competence are realized.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of computer-aided diagnosis, and more particularly, to an Alzheimer's disease auxiliary screening system based on intelligent interaction. BACKGROUND

[0002] In the early identification process of cognitive impairment, a characteristic cognitive degradation phenomenon can be observed for individuals with strong domain knowledge accumulation or engaged in professional work, that is, the degradation rhythm of life knowledge and professional knowledge is inconsistent. In the early stage of Alzheimer's disease, patients often still maintain relatively complete professional semantic structure and terminology use ability, for example, individuals proficient in organic chemistry can still correctly distinguish the structural differences between alkynes and alkenes, clearly describe the process path of substitution reaction, and even complete abstract tasks such as structure formula judgment and reaction type matching when facing terminology questions and logic analysis tasks. However, the same individuals may exhibit obvious life chain disorder in daily life, such as being unable to correctly complete sequential operations such as boiling water, cooking, and dressing, or frequently making mistakes in placing objects and confusing spatial orientation in familiar environments. Unlike other brain function loss scenarios, such life common sense errors are not caused by simple memory impairment, but are caused by the preferential destruction of the medial temporal lobe and related context construction networks in the early stage of Alzheimer's disease, resulting in the first decline in behavior organization ability driven by actual life experience, while the highly consolidated and repeatedly reinforced professional knowledge system remains relatively stable.

[0003] Therefore, in practice, individuals often perform normally in standard cognitive scale tests but frequently make mistakes in life tasks, causing delays in early identification. The current evaluation mechanism has not established a structural comparison method for the degradation rhythm of different knowledge systems within an individual, and it is urgent to introduce the difference between the cognitive stability of life knowledge and professional knowledge as a basis for judgment to improve the accuracy of early-stage Alzheimer's disease screening. SUMMARY

[0004] In order to overcome the above-mentioned defects of the prior art, embodiments of the present application provide an Alzheimer's disease auxiliary screening system based on intelligent interaction to solve the problems raised in the background art.

[0005] To achieve the above-mentioned purpose, the present application provides the following technical solutions: An Alzheimer's disease auxiliary screening system based on intelligent interaction includes a task construction module, an interaction extraction module, an interaction analysis module, a cognitive performance evaluation module, an evolution curve construction module, and an asymmetric degradation identification module, wherein: The task construction module constructs task interaction groups corresponding to life knowledge and professional knowledge, respectively, and presents them to the testee in turn. The interaction extraction module collects the interactive operations of the test subjects in various tasks, encodes the interactive operations in each round of tasks, and establishes an interaction vector matrix; The interaction analysis module performs jump point statistics and vector residual calculations on the interaction vector matrix according to the preset standard task logic chain, and counts the invalid operation density in the interaction vector matrix; The cognitive performance evaluation module generates life cognitive performance indicators and professional cognitive performance indicators through the calculation and statistical output results of the interactive analysis module; The evolution curve construction module records the difference change sequence between the life cognitive performance index and the professional cognitive performance index during the preset task test cycle, and constructs the life and professional cognitive deviation evolution curve; The asymmetric degradation identification module determines whether the subject's cognitive ability has an asymmetric degradation trend based on the cognitive bias evolution curve.

[0006] In a preferred embodiment, the task construction module constructs task interaction groups corresponding to daily life knowledge and professional knowledge respectively, and presents them to the testee in sequence, specifically including: Based on the test subject's age, educational background, and professional experience, construct interactive tasks related to daily life knowledge for at least a set number of rounds. The tasks include scene recognition, routine operation judgment, and reasoning about the order in which daily items are used. At the same time, we constructed professional knowledge interaction tasks of equal magnitude to the daily life knowledge interaction tasks, including term identification, special process selection, and cause-and-effect judgment within scenarios. All interactive tasks were uniformly coded and integrated into task sequence groups, and presented to the test subjects in sequence in a segmented alternating manner.

[0007] In a preferred embodiment, the interaction extraction module collects the interactive operations of the test subject in various tasks, encodes the interactive operations in each round of tasks, and establishes the interaction vector matrix, specifically including: Obtain the interactive operations of the test subjects during each round of tasks, perform structural decomposition on each interactive operation, and extract multi-dimensional interaction data including the semantics of the operation action, the action object, the task code, and the interaction time; Convert multi-dimensional interaction data into interaction operation vector expressions through preset coding mapping rules; After each round of tasks is completed, the interaction operation vectors are spliced ​​into a complete interaction operation vector sequence according to the interaction operation order and marked according to the task code; After completing a set number of rounds of tasks, the interaction operation vector sequence is integrated to establish a time-continuous and task-distinguishable interaction vector matrix, and different weights are set for the columns in the interaction vector matrix.

[0008] In a preferred embodiment, the interactive operation vector expression construction process is to set an enumerable discrete coding space for each dimension of the multi-dimensional interaction data, and splice them into a structure vector in the fixed order of action semantics, action object, task coding and interaction time.

[0009] In a preferred embodiment, the interaction analysis module performs skip point statistics and vector residual calculation on the interaction vector matrix according to the preset standard task logic chain, and the invalid operation density in the interaction vector matrix specifically includes: The preset standard task logic chain defines the ideal vector template of multi-dimensional interaction data in each round of task; Map the interaction vector matrix to the standard task logic chain, identify the position of the task logic jump in the matrix through the task coding sequence in the interaction operation vector, and mark it as a skip point; According to the column weight of the matrix, calculate the weighted vector distance between each row of interaction operation vector in the interaction vector matrix and the ideal vector template, and obtain the vector residual of the interaction operation; Divide the interaction vector matrix into several sub-matrices according to the task round, and perform invalid operation density statistics on the interaction operation vector of the sub-matrix; The invalid operation density is obtained by calculating the proportion of the interaction time corresponding to the non-target operation interaction operation vector in the total duration of the standard task logic chain; The non-target operation is an operation inconsistent with the ideal vector template.

[0010] In a preferred embodiment, the cognitive performance evaluation module generates life cognitive performance indicators and professional cognitive performance indicators through the calculation and statistical output results of the interaction analysis module, specifically including: Divide the skip points, vector residuals and invalid operation densities in the interaction analysis module into task types according to the task coding corresponding to the interaction operation vectors, and integrate and calculate the skip point frequency, vector residual mean and invalid operation density mean of different task types respectively; Based on the skip point frequency, vector residual mean and invalid operation density mean of different task types, generate life cognitive performance indicators and professional cognitive performance indicators through weighted comprehensive scoring.

[0011] In a preferred embodiment, the evolution curve construction module records the difference change sequence between life cognitive performance indicators and professional cognitive performance indicators within a preset task test period, and constructs the life and professional cognitive bias evolution curve, specifically including: The preset task test period includes several times of cognitive performance indicator evaluation; After each cognitive performance index assessment, the difference between the life cognitive performance index and the professional cognitive performance index was calculated and recorded as the cognitive bias assessment value; All cognitive bias assessment values ​​within the preset task test cycle are arranged by time to construct a cognitive bias curve. The vertical axis of the curve represents the difference between life and professional cognitive performance indicators, and the horizontal axis represents the number of cognitive performance indicator assessments.

[0012] In a preferred embodiment, the asymmetric degradation identification module determines whether the cognitive ability of the test subject has an asymmetric degradation trend according to the cognitive bias evolution curve, specifically including: When the cognitive bias evolution curve shows a monotonically increasing trend within a sliding window of a preset length, and the growth rate exceeds the set change rate threshold, it is determined that there is an asymmetric degradation between the subject's life cognitive performance and professional cognitive performance, and a screening result warning is issued.

[0013] The technical effects and advantages of the intelligent interactive Alzheimer's disease auxiliary screening system of the present invention are as follows: The method for identifying asymmetric degeneration of cognitive ability provided by the present invention sets up task interaction groups for daily life knowledge and professional knowledge respectively through a task construction module, effectively covering task response behaviors of different cognitive types; the interaction extraction module performs structured coding on the test subject's operational behavior and constructs a standardized interaction vector matrix to ensure that the behavioral data is comparable and analyzable; the interaction analysis module combines the standard task logic chain to perform jump point identification, residual evaluation and invalid operation density statistics to achieve a multi-dimensional evaluation of the consistency of interactive operation logic and task execution deviation; the cognitive performance evaluation module generates cognitive performance indicators for daily life and professional dimensions based on the analysis results, with clear dimensional division and response explanatory power; the evolution curve construction module performs time series modeling on the difference between the two types of indicators to form a quantifiable cognitive deviation evolution trend; the asymmetric degeneration identification module judges whether there is inconsistency between the daily life and professional dimensions in cognitive ability degeneration based on the changes in the trend curve, which can identify signals of structural cognitive ability decline at an early stage and improve screening accuracy and preclinical recognition rate. BRIEF DESCRIPTION OF THE DRAWINGS

[0014] Figure 1 This is a schematic diagram of an Alzheimer's disease auxiliary screening system based on intelligent interaction in the present invention. DETAILED DESCRIPTION

[0015] The following will provide a clear and complete description of the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. All other embodiments obtained by ordinary technicians in this field based on the embodiments of the present invention without making any creative efforts shall fall within the scope of protection of the present invention.

[0016] Example 1, Figure 1 The present invention provides an Alzheimer's disease auxiliary screening system based on intelligent interaction, which includes a task construction module, an interaction extraction module, an interaction analysis module, a cognitive performance evaluation module, an evolution curve construction module, and an asymmetric degradation recognition module, wherein: The task construction module constructs task interaction groups corresponding to daily life knowledge and professional knowledge respectively, and presents them to the test subjects in sequence; The interaction extraction module collects the interactive operations of the test subjects in various tasks, encodes the interactive operations in each round of tasks, and establishes an interaction vector matrix; The interaction analysis module performs jump point statistics and vector residual calculations on the interaction vector matrix according to the preset standard task logic chain, and counts the invalid operation density in the interaction vector matrix; The cognitive performance evaluation module generates life cognitive performance indicators and professional cognitive performance indicators through the calculation and statistical output results of the interactive analysis module; The evolution curve construction module records the difference change sequence between the life cognitive performance index and the professional cognitive performance index during the preset task test cycle, and constructs the life and professional cognitive deviation evolution curve; The asymmetric degradation identification module determines whether the subject's cognitive ability has an asymmetric degradation trend based on the cognitive bias evolution curve.

[0017] The task construction module constructs task interaction groups corresponding to daily life knowledge and professional knowledge respectively, and presents them to the testees in sequence.

[0018] Basic information about the test takers is collected and structured. This information includes, but is not limited to, the test taker's age, years of education, education type (e.g., science, medicine, languages), and professional experience (job title, years of experience, and field of expertise). These basic parameters serve as the basis for individual adaptation of task construction and directly form the foundation for the content design and presentation sequence of interactive tasks. After information collection is completed, the test takers are grouped into task groups. By default, no fewer than six task interaction units are constructed. Each task round consists of one life knowledge task and one professional knowledge task, presented alternately, with a total of no fewer than twelve tasks.

[0019] During the construction of the interactive unit for life-related knowledge tasks, real-life situations were explicitly used as the knowledge foundation, with interactive task content constructed through graphics, text, voice, or embedded user interfaces. Task types include scene recognition tasks, routine operation judgment tasks, and everyday object usage order reasoning tasks. Scene recognition tasks present common environments through images, such as bedrooms, kitchens, bank branches, and subway stations, requiring participants to identify or match multiple environmental options. Routine operation judgment tasks build operational judgment content based on procedural behavioral logic, such as identifying laundry procedures, medicine purchase procedures, and kettle operation procedures. Interactions are conducted through step-by-step completion or operation path selection. Everyday object usage order reasoning tasks present multiple everyday tools or items (such as tableware, condiments, and electrical components) through a combination of graphics and text, requiring users to arrange them in the correct order of use. Each task type is equipped with a set of distractor items to simulate real-life cognitive situations characterized by information redundancy or logical dislocation. The number of distractor items is dynamically adjusted based on the number of task rounds, with a minimum of two examples per round.

[0020] The interactive unit for professional knowledge tasks is constructed based on the professional categories identified in the test-taker's professional background data, and personalized tasks are generated by calling task templates in the professional task library through knowledge extraction rules. Task types include three categories: term identification tasks, dedicated process selection tasks, and in-scenario causal judgment tasks. The term identification task randomly mixes professional terms related to the profession with non-related terms in a unified option list, requiring users to complete term classification and definition matching operations. The dedicated process selection task presents the professional task operation steps through a flowchart module, and presets logical errors or process reordering items. The test-taker must complete error correction or sorting adjustments. The in-scenario causal judgment task sets a combination of key node events in a typical professional workflow, requiring users to judge the logical sequence and causal relationship of each event. The option design of all professional tasks meets the unique logical correctness condition to ensure the stability and discriminant validity of the task scoring.

[0021] All life-related and professional-related tasks are structured and encapsulated according to unified coding rules to form a task metadata structure. This structure includes a task identification code, a type field (life / professional), a task logic label, a preset number of operation nodes, the target task chain length, and a standard answer field. The task construction module then integrates all task structures into a sequence to generate a complete task sequence group. The presentation strategy for the task sequence group adopts a segmented alternating execution mechanism, where each life-related task alternates with its corresponding professional-related task to prevent the clustering of task types from skewing cognitive test results.

[0022] The interaction extraction module collects the interactive operations of the test subjects in various tasks, encodes the interactive operations in each round of tasks, and establishes an interaction vector matrix.

[0023] Capture the test subject's full interaction data stream during each round of task execution. Interaction data should include every action the test subject performs during the interactive task. Records must include basic information such as timestamp, action content, target response object, and task identifier. All interactive actions must be captured and archived immediately upon execution. The collection frequency should meet the response time requirements of the task granularity. The recommended interval is within 50ms to ensure high-precision behavior capture.

[0024] The interactive behavior data of the test subjects in each round of the task is first subjected to a structured decomposition operation. Specifically, each interaction behavior record is broken down into four basic dimensions: action semantics, action object identifier, task coding identifier, and interaction time parameter. Action semantics are descriptive categories that reflect the nature of the behavior, such as "click," "slide," and "long press." These can be standardized using a pre-set behavior type dictionary within the task control logic. Action object identifiers are the labels of the specific interactive elements targeted by the corresponding action, such as images, options, and text boxes. These identifiers are extracted by combining the identifier mapping of the task interface elements. Task coding identifiers are uniquely assigned based on the order in which the tasks are presented, such as 0101 for the first round of life tasks and 0203 for the third round of professional tasks. The interaction time parameter is the triggering time of the action, recorded as a relative timestamp from the start of the task. After the basic dimensions are extracted, the interaction operation vector is constructed. To ensure the discreteness and comparability of the vector representation, a discrete coding space is defined for each of these dimensions. The construction of the encoding space is based on preset enumeration rules. To enhance semantic relevance and the effectiveness of subsequent vector distance calculations, semantically similar actions and objects are assigned adjacent or similar encoding identifiers to preserve their semantic relevance. For example, in the action semantics dimension, "click" is assigned an encoding value of 01, "slide" is assigned 02, and "long press" is assigned 03. In the action object identifier dimension, a discrete mapping relationship is established based on the fixed numbering of task elements in the interface layout. The task encoding dimension directly inherits the sequential numbering defined during task setup. The interaction time dimension is interval-coded according to time periods, and so on. After encoding each dimension, the fixed sequence of "action semantics - action object - task encoding - time" is concatenated to generate a structural vector, which serves as the vector representation of each interaction operation.

[0025] Here's an example: A daily life interaction, "clicking the wall switch to turn on the light," is encoded as the structure vector [01, 12, 0103, 1], where 01 represents the "click" action, 12 represents the "wall switch" object, 0103 is the current task code, and 1 indicates the time interval within which the action occurred, coded 1. Another professional knowledge interaction, "dragging a reagent bottle to the lab bench," is encoded as [05, 21, 0205, 2], corresponding to the following encoding: "drag," "reagent bottle," "task code 1105," and "time interval code 2." These structure vectors serve as the basic units of the interaction vector matrix and are directly used in subsequent vector distance calculations and jump point identification operations, ensuring a consistent representation and engineering parsability.

[0026] After each round of tasks, the interaction vectors generated by all interactions in the current round are concatenated according to the actual operation order to form a complete interaction vector sequence. This sequence has temporal characteristics and is used to analyze the logical sequence of operation behaviors and the consistency of task progress. At the same time, each vector in the sequence is appended with the current task code field as a tag.

[0027] After all tasks for the specified number of rounds are completed, the interaction vector sequences for each round are integrated into a unified interaction vector matrix. This matrix is ​​arranged in temporal order, with rows representing the order of interaction operations and columns corresponding to the four dimensions and their associated encoding information. Different weights are assigned to the columns of the interaction vector matrix based on the varying importance of different dimensions in the evaluation process. For example, in subsequent use for skip point identification and task consistency comparison, the "task encoding" and "action semantics" dimensions have a higher discriminant contribution and are assigned default weights of 0.35 and 0.30, respectively. The "interaction time" and "action object" dimensions are less important, with weights of 0.20 and 0.15, respectively. The weights are normalized to sum to 1 and are adjusted and optimized based on the correlation of each dimension with cognitive degradation indicators in the training data.

[0028] The interaction analysis module performs jump point statistics and vector residual calculation on the interaction vector matrix according to a preset standard task logic chain, and counts the invalid operation density in the interaction vector matrix.

[0029] After constructing the interaction vector matrix and assigning column weights, the interaction analysis phase begins. First, the task structure is standardized and modeled, creating a standard task logic chain. This logic chain combines the expected interaction actions for each task round in an ideal sequence, using a predefined task execution sequence, the standard operational steps for each task round, and the interaction time set. This chain forms a standard task path, indicating that all task options meet the unique logical correctness criteria. Each standard task path is indexed by the corresponding task round and corresponds to the task code in the interaction data, ensuring that each task round can be independently mapped to its corresponding standard path. To facilitate quantitative comparison, each operation in the standard task path is converted into an ideal vector template, which is structurally identical to the interaction operation vector. The ideal vector template consists of four dimensions: the expected operation action, the corresponding object, the standard task code, and the recommended operation time period. It is populated with the ideal operation data for each step in the standard task execution process, forming a multi-row standard structure vector set.

[0030] Subsequently, the interaction vector matrix is ​​mapped to the standard task logic chain, and the operation sequence of each round of tasks is compared vertically with the standard path sequence. The specific operation is to divide the matrix into rounds according to the task coding field in the interaction operation vector, and extract the operation sequence in the corresponding task segment in chronological order. Then compare it with the expected operation sequence in the standard task logic chain. In the case of skipped steps, omissions, misordered or redundant operations, the task execution sequence in the interaction vector will have breakpoints that are inconsistent with the standard chain. Such breakpoints are defined as "jump points". The identification of jump points uses the task coding sequence analysis algorithm to detect whether there is any deviation from the expected sequence in the continuous vector sequence, and once found, the jump point is marked. The following is a specific example: The standard task logic chain involves sequentially executing subtasks related to life knowledge, coded as 0201 → 0202 → 0203 → 0204 → 0205. This represents the normal execution path from "kitchen appliance identification" to "bathroom supplies order determination." However, in a test subject's actual interaction vector sequence, the extracted task code sequence was: 0201 → 0202 → 0204 → 0203 → 0205. In this execution sequence, the jump from 0202 to 0204 violates the linear progression requirement of the standard logic chain, and 0204 is identified as a premature jump point. Furthermore, the reverse sequence from 0204 back to 0203 also constitutes a jump point, and 0203 is considered a retrograde jump point.

[0031] After identifying jump points, the accuracy of interactive operations is evaluated by calculating the vector distance between each row of interactive operation vectors and their corresponding ideal vector templates. Because each dimension contributes differently to cognitive performance, a weighted Euclidean distance is used as the weighted vector distance calculation method. The calculated weights are based on a combination of pre-set weights in the interaction extraction module.

[0032] The interaction vector matrix is ​​divided into multiple segments based on task rounds. The invalid operation density is calculated for each submatrix within each task segment. Within each submatrix, all interaction operation vectors are traversed to determine whether they are non-target operations. The criterion for this determination is: if any dimension of an interaction operation vector is completely inconsistent with the ideal vector template (i.e., not included in the encoding, representing completely unrelated operation semantics or unrelated operation objects), it is considered a non-target operation. The total duration of non-target operations is calculated (accumulated by the difference in operation timestamps), and its ratio to the total duration of the standard task logic chain is used as the invalid operation density for that submatrix. For example, if the standard execution time for a task is 60 seconds, and the actual detected non-target operation duration is 24 seconds, the invalid operation density is 0.4. This density metric reflects the subject's disruptive behaviors and the degree of deviation during task execution, indicating whether there are signs of "repetitive groping" or "ineffective attempts" during task execution. A significantly high behavior density (far above the mean) indicates cognitive hesitation or comprehension difficulties at that stage.

[0033] The cognitive performance evaluation module generates life cognitive performance indicators and professional cognitive performance indicators through the calculation and statistical output results of the interactive analysis module.

[0034] After completing the interaction vector matrix's jump point identification, vector residual calculation, and invalid operation density statistics, the cognitive performance assessment phase begins. This phase aims to quantify the subject's cognitive level in tasks related to everyday knowledge and professional knowledge based on behavioral analysis results across different task types, and to generate highly comparable and clearly interpretable cognitive performance indicators. During implementation, the three indicators generated in the interaction analysis module—jump points, vector residuals, and invalid operation density—are first divided into task type dimensions based on the task encoding information embedded in the interaction operation vector. Specifically, based on the task code prefix or identification bit, the interaction operation vector is divided into two sets: everyday knowledge tasks and professional knowledge tasks. The jump point frequency, vector residual values, and invalid operation density values ​​corresponding to each task type are then statistically analyzed and aggregated.

[0035] The skipping frequency metric is calculated by counting the number of interaction vector entries with skipping within each task type and dividing the result by the total number of operation vectors for that task type. For example, if a certain type of professional task contains 100 interaction operation vectors, 28 of which experience skipping, then the skipping frequency for that task type is 0.28. This value reflects the frequency of logical structure disruption within the operation sequence and indirectly measures task sequence comprehension and cognitive process stability.

[0036] The mean vector residual is calculated by averaging the residual values ​​of all interaction action vectors within a task type. The result represents the average deviation of the action behavior from the ideal path for that task type. Because the residual values ​​are normalized using the weighted vector distance calculation method described above, the mean is comparable across task types. This metric can be considered an overall reflection of operational accuracy and understanding.

[0037] The mean invalid operation density is calculated by first dividing the matrix into submatrices by task round, obtaining the invalid operation density value for each submatrix, and then averaging the density values ​​of all submatrices by task type to obtain the average invalid operation density index for daily life tasks and professional tasks. This index reflects the comprehensive degree of concentration, goal orientation, and understanding of the task purpose during task execution.

[0038] After completing the task type classification and statistics for the three indicators above, a weighted scoring strategy can be used to comprehensively evaluate cognitive performance in different task types. To ensure controllability and adaptability of the scoring, the default scoring weights for the three indicators are: 0.4 for jump frequency, 0.35 for vector residual mean, and 0.25 for invalid operation density mean. Using this configuration, the normalized values ​​of the three indicators are substituted into a comprehensive scoring function (e.g., weighted summation) to calculate a comprehensive score for each task type. This comprehensive score serves as the final output for both the everyday cognitive performance indicator and the professional cognitive performance indicator. Lower scores indicate closer adherence to the standard task path, more accurate operation, and lower deviation. These indicators are standardized floating-point values ​​in the [0, 1] range and can be used for cross-temporal comparisons and population screening assessments. In practical applications, if a subject's everyday cognitive performance indicator is significantly lower than their professional cognitive performance indicator, it may indicate a risk of deterioration in their ability to handle routine tasks; conversely, a lower score may reflect a decline in their mastery of professional knowledge.

[0039] The evolution curve construction module records the difference change sequence between the life cognition performance index and the professional cognition performance index during the preset task test period, and constructs the life and professional cognition deviation evolution curve.

[0040] After completing the generation of life cognitive performance indicators and professional cognitive performance indicators, in order to further monitor their changing trends within the set time period and thus identify their potential deviation relationships and evolutionary characteristics, it is necessary to construct a cognitive bias evolution curve. This step takes the task test cycle as the basic unit, maps the cognitive performance differences in each round of testing into time series data, and realizes the quantitative tracking of the dynamic differences between life and professional cognitive performance. First, in the test plan setting stage, the basic structure of the task test cycle is preset. The cycle consists of several rounds of interactive tasks to ensure that no less than the set number of cognitive performance indicator evaluation operations can be performed within the cycle. In the example setting, a task test cycle can be defined as including 10 rounds of task evaluation, each round of evaluation consists of a set of interactive tasks, and corresponds to a set of calculation outputs of life cognitive performance indicators and professional cognitive performance indicators.

[0041] After each round of assessment is completed and the corresponding daily life and professional cognitive performance indicators are obtained, the difference between the two is immediately calculated, defining this as the "cognitive bias assessment value" for that round of assessment. This difference is expressed as an absolute value, or, depending on the indicator definition, retains a positive or negative sign to reflect the direction of the deviation. For example, if the assessment results for a round are: daily life cognitive performance index is 0.32 and professional cognitive performance index is 0.24, the cognitive bias assessment value is +0.08; if the opposite is true, it is −0.08. The sign of this difference can be used to subsequently determine which cognitive ability is more significantly impaired.

[0042] After calculating the cognitive bias assessment values ​​for each round, the system chronologically numbers all assessment values ​​according to the order of the task test cycle, forming an ordered data sequence. Each item in this sequence corresponds to the degree of cognitive bias at a task evaluation point. This sequence is then plotted as a graph to construct a cognitive bias evolution curve. The horizontal axis of this curve represents the number of assessments within the task test cycle (e.g., Round 1, Round 2, ..., Round 10), and the vertical axis represents the cognitive bias assessment value corresponding to each assessment. This curve visually depicts the changing trend of the differences between everyday and professional cognitive performance throughout the test cycle. The curve's trajectory reveals the volatility, growth, and phases of the bias, as well as potential inflection points. To ensure the curve's engineering applicability, the time span of the horizontal axis should be evenly spaced and the assessment values ​​should be sourced consistently to avoid outliers due to uneven task numbers or fluctuations in the evaluation method. To enhance readability and analysis, key points (such as maximum deviation points and deviation reversal points) can be annotated on the curve, and local fluctuations can be smoothed using a sliding window to identify trend changes rather than isolated fluctuations.

[0043] As a fundamental data structure for subsequent asymmetric cognitive decline identification, the cognitive bias evolution curve offers the following technical value: First, it can be used to assess whether cognitive ability differences widen as task testing progresses; second, it can provide a dynamic basis for determining the direction of cognitive deviation; and third, it can provide interpretable, quantitative clues for developing personalized intervention strategies or screening for early decline. In particular, in population-based screening or continuous follow-up, if the curve exhibits a stable upward trend or dramatic fluctuations, further analysis is necessary to determine whether it represents asymmetric cognitive decline.

[0044] The asymmetric degradation identification module determines whether the cognitive ability of the test subject has an asymmetric degradation trend according to the cognitive bias evolution curve.

[0045] After the cognitive bias evolution curve is constructed, a fixed number of consecutive evaluations, starting from the first evaluation value, are selected to form a sliding window sequence. The sliding window length should be determined based on the total number of evaluations within the task test cycle and the required trend sensitivity. The recommended value is 5 to 8 consecutive evaluation results within a cycle. For example, if the value is set to 6, the cognitive bias evaluation values ​​of the most recent 6 rounds will be used as a subsequence for each calculation to serve as the judgment basis.

[0046] Within each sliding window, the cognitive bias assessment values ​​within that window are sequentially compared to see if they strictly adhere to a monotonically increasing relationship—that is, for any subsequence, the subsequent value is greater than the preceding value. If so, the overall growth rate of the cognitive bias values ​​within that window is assessed to see if it exceeds a preset rate-of-change threshold. The rate of growth is defined as the difference between the final and initial assessment values ​​within the sliding window divided by the absolute value of the initial value. If the initial value is close to 0, a lower limit (e.g., 0.05) is introduced to mitigate extreme amplification. The threshold is typically set between 15% and 30%, based on the test baseline for a population with normal cognition. For example, if the rate-of-change threshold is set to 0.2, and the growth rate of the difference between the initial and final values ​​in the current window is 0.27, then the trend significance criteria are met. If the dual criteria of a sustained monotonically increasing trend and a rate of change exceeding the threshold are met within the sliding window, the subject is determined to exhibit an asymmetric deterioration trend between their everyday and professional cognitive performance. A screening result warning should be generated at this time, and the abnormal label should be attached to the current task test cycle for reference in subsequent report generation, medical intervention recommendations, or further cognitive testing sessions.

[0047] The above formulas are all dimensionless and numerical calculations. The formulas are obtained by collecting a large amount of data and performing software simulation to obtain the most recent real situation. The preset parameters and thresholds in the formulas are set by technicians in this field according to actual conditions.

[0048] The above embodiments can be implemented in whole or in part via software, hardware, firmware, or any other combination. 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 comprises one or more computer instructions or computer programs. When loaded or executed on a computer, the processes or functions described in the embodiments of this application are fully or partially performed. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. 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 via wired means (e.g., infrared, wireless, microwave, etc.). The computer-readable storage medium can be any available medium accessible by a computer or a data storage device such as a server or data center that contains a collection of one or more available media. The available medium can be magnetic media (e.g., floppy disks, hard disks, tapes), optical media (e.g., DVDs), or semiconductor media. The semiconductor media can be a solid-state drive.

[0049] Those skilled in the art will appreciate that the modules and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0050] Those skilled in the art will 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 aforementioned method embodiments and will not be repeated here.

[0051] In the several embodiments provided in this 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 schematic. For example, the division of the modules is only a logical function division. In actual implementation, there may be other division methods, such as 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 mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of devices or modules, which can be electrical, mechanical or other forms.

[0052] The modules described as separate components may or may not be physically separate, and the components shown as modules may or may not be physical modules, and may be located in one place or distributed across multiple network modules. Some or all of the modules may be selected to achieve the purpose of this embodiment according to actual needs.

[0053] In addition, each functional module in each embodiment of the present application may be integrated into one processing module, or each module may exist physically separately, or two or more modules may be integrated into one module.

[0054] 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, or the part that contributes to the prior art, or the part of the technical solution, can be embodied in the form of a software product. The computer software product is stored in a storage medium and includes several instructions for enabling a computer device (which can be a personal computer, server, or network device, etc.) to execute all or part of the steps of the method described in each embodiment of the present application. The aforementioned storage medium includes various media that can store program codes, such as a USB flash drive, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk, or an optical disk.

[0055] The above description is merely a specific embodiment of the present application, but the scope of protection of the present application is not limited thereto. Any changes or substitutions that can be easily conceived by a person skilled in the art within the technical scope disclosed in this application should be included in the scope of protection of this application. Therefore, the scope of protection of this application should be based on the scope of protection of the claims.

[0056] Finally: The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. An Alzheimer's disease auxiliary screening system based on intelligent interaction, characterized in that: It includes a task construction module, an interaction extraction module, an interaction analysis module, a cognitive performance evaluation module, an evolution curve construction module, and an asymmetric degradation identification module, among which: The task construction module constructs task interaction groups corresponding to daily life knowledge and professional knowledge respectively, and presents them to the test subjects in sequence; The interaction extraction module collects the interactive operations of the test subjects in various tasks, encodes the interactive operations in each round of tasks, and establishes an interaction vector matrix; The interaction analysis module performs jump point statistics and vector residual calculations on the interaction vector matrix according to the preset standard task logic chain, and counts the invalid operation density in the interaction vector matrix; The cognitive performance evaluation module generates life cognitive performance indicators and professional cognitive performance indicators through the calculation and statistical output results of the interactive analysis module; The evolution curve construction module records the difference change sequence between the life cognitive performance index and the professional cognitive performance index during the preset task test cycle, and constructs the life and professional cognitive deviation evolution curve; The asymmetric degradation identification module determines whether the subject's cognitive ability has an asymmetric degradation trend based on the cognitive bias evolution curve.

2. The Alzheimer's disease auxiliary screening system based on intelligent interaction according to claim 1 is characterized in that: The task construction module constructs task interaction groups corresponding to daily life knowledge and professional knowledge respectively, and presents them to the testee in sequence, including: Based on the test subject's age, educational background, and professional experience, construct interactive tasks related to daily life knowledge for at least a set number of rounds. The tasks include scene recognition, routine operation judgment, and reasoning about the order in which daily items are used. At the same time, we constructed professional knowledge interaction tasks of equal magnitude to the daily life knowledge interaction tasks, including term identification, special process selection, and cause-and-effect judgment within scenarios. All interactive tasks were uniformly coded and integrated into task sequence groups, and presented to the test subjects in sequence in a segmented alternating manner.

3. The Alzheimer's disease auxiliary screening system based on intelligent interaction according to claim 1 is characterized in that: The interaction extraction module collects the interactive operations of the test subjects in various tasks, encodes the interactive operations in each round of tasks, and establishes an interaction vector matrix, which specifically includes: Obtain the interactive operations of the test subjects during each round of tasks, perform structural decomposition on each interactive operation, and extract multi-dimensional interaction data including the semantics of the operation action, the action object, the task code, and the interaction time; Convert multi-dimensional interaction data into interaction operation vector expressions through preset coding mapping rules; After each round of tasks is completed, the interaction operation vectors are spliced ​​into a complete interaction operation vector sequence according to the interaction operation order and marked according to the task code; After completing a set number of rounds of tasks, the interaction operation vector sequence is integrated to establish a time-continuous and task-distinguishable interaction vector matrix, and different weights are set for the columns in the interaction vector matrix.

4. The Alzheimer's disease auxiliary screening system based on intelligent interaction according to claim 3 is characterized in that: The interactive operation vector expression construction process is to set an enumerable discrete coding space for each dimension in the multi-dimensional interactive data, and splice it into a structural vector in a fixed order of action semantics, action object, task coding and interaction time.

5. The Alzheimer's disease auxiliary screening system based on intelligent interaction according to claim 1 is characterized in that: The interaction analysis module performs jump point statistics and vector residual calculation on the interaction vector matrix according to the preset standard task logic chain, and counts the invalid operation density in the interaction vector matrix, specifically including: Preset standard task logic chain to define the ideal vector template of multi-dimensional interaction data in each round of tasks; Map the interaction vector matrix to the standard task logic chain, identify the position of task logic jump vertically in the matrix through the task encoding order in the interaction operation vector, and mark it as a jump point; According to the column weights of the matrix, the weighted vector distance between each row of the interaction operation vector in the interaction vector matrix and the ideal vector template is calculated to obtain the vector residual of the interaction operation; The interaction vector matrix is ​​divided into several sub-matrices according to the task rounds, and the invalid operation density statistics of the interaction operation vectors of the sub-matrices are performed; The invalid operation density is obtained by calculating the ratio of the interaction time corresponding to the interaction operation vector of the non-target operation to the total duration of the standard task logic chain; The non-target operation is an operation that is inconsistent with any interactive operation corresponding to the ideal vector template.

6. The Alzheimer's disease auxiliary screening system based on intelligent interaction according to claim 1, characterized in that: The cognitive performance evaluation module generates life cognitive performance indicators and professional cognitive performance indicators through the calculation and statistical output results of the interactive analysis module, specifically including: The jump points, vector residuals, and invalid operation density in the interaction analysis module are divided into task types according to the task codes corresponding to the interaction operation vectors. The jump point frequency, vector residual mean, and invalid operation density mean of different task types are integrated and calculated respectively. Based on the jump point frequency, vector residual mean and invalid operation density mean of different task types, life cognitive performance indicators and professional cognitive performance indicators are generated by empowering comprehensive scoring.

7. The Alzheimer's disease auxiliary screening system based on intelligent interaction according to claim 1 is characterized in that: The evolution curve construction module records the difference change sequence between the life cognition performance index and the professional cognition performance index during the preset task test period, and constructs the life and professional cognition deviation evolution curve specifically including: Preset task test cycle, one task test cycle includes several cognitive performance indicator assessments; After each cognitive performance index assessment, the difference between the life cognitive performance index and the professional cognitive performance index was calculated and recorded as the cognitive bias assessment value; All cognitive bias assessment values ​​within the preset task test cycle are arranged by time to construct a cognitive bias curve. The vertical axis of the curve represents the difference between life and professional cognitive performance indicators, and the horizontal axis represents the number of cognitive performance indicator assessments.

8. The Alzheimer's disease auxiliary screening system based on intelligent interaction according to claim 1 is characterized in that: The asymmetric degradation identification module determines whether the cognitive ability of the test subject has an asymmetric degradation trend according to the cognitive bias evolution curve, specifically including: When the cognitive bias evolution curve shows a monotonically increasing trend within a sliding window of a preset length, and the growth rate exceeds the set change rate threshold, it is determined that there is an asymmetric degradation between the subject's life cognitive performance and professional cognitive performance, and a screening result warning is issued.

Images