A method and system for assessing spinal health risk based on individual characteristics

Abstract

The application discloses a spine health risk assessment method and system based on individual characteristics, acquires basic characteristic data and spine shape measurement data of an individual, constructs an individual characteristic portrait through professional posture mode matching and body posture deviation calculation; performs risk factor extraction on the individual characteristic portrait to form a candidate risk factor set, establishes a risk factor correlation graph by adopting causal relationship identification, and performs an intervenable assessment to generate an intervenable label; performs risk grading based on the risk factor correlation graph and the intervenable label to generate a risk grading parameter, determines a risk level, and extracts a key risk factor set; classifies the key risk factor set according to spine partitions to obtain partition risk values, identifies a risk concentration area through difference analysis, and obtains a risk distribution type in combination with distribution mode identification; and integrates partition distribution characteristics and the risk level to output a spine health risk assessment result, which can provide a scientific basis for personalized health management and accurate rehabilitation scheme making.

Description

A method and system for assessing spinal health risk based on individual characteristics Technical Field

[0001] This invention relates to the field of health assessment technology, and in particular to a method and system for assessing spinal health risks based on individual characteristics. Background Technology

[0002] Spinal diseases are a common health problem affecting the quality of life of modern people, with complex pathogenesis and numerous influencing factors. Traditional spinal health assessment methods mainly rely on imaging examinations and doctors' experience, but this assessment method often only makes a diagnosis after symptoms have become obvious, making it difficult to achieve early warning.

[0003] Existing spinal health assessment technologies have several limitations. The assessment process typically isolates multi-source data such as occupational information, postural measurements, and flexibility tests, failing to fully explore the correlations between different data sources. Traditional methods lack a deep understanding of the interaction mechanisms between multiple risk factors, making it difficult to identify which factors produce synergistic amplification effects or distinguish between improveable and irreversible factors. Current assessment technologies do not adequately focus on the spatial distribution characteristics of risks across different spinal segments, failing to fully utilize the spatial clustering patterns and continuity of risks to guide the development of precise rehabilitation programs. These shortcomings result in a lack of targeted health management and a lack of prioritization guidance for intervention recommendations. Summary of the Invention

[0004] This invention discloses a method and system for assessing spinal health risks based on individual characteristics. It aims to construct a comprehensive profile by integrating occupational characteristics and postural measurement data, identify causal relationships and synergistic amplification effects among multiple factors, analyze the spatial distribution patterns of risks in spinal segments, and finally generate graded and classified assessment results, providing a scientific basis for personalized health management and precise intervention.

[0005] The first aspect of this invention proposes a method for assessing spinal health risk based on individual characteristics, comprising the following steps:

[0006] Acquire basic feature data and spinal morphology measurement data of an individual, and construct an individual feature profile based on the basic feature data and the spinal morphology measurement data;

[0007] Risk factors are extracted from the individual feature profile to form a candidate risk factor set. Multi-factor coupling analysis is performed on the candidate risk factor set to establish a risk factor association map. An interventionability assessment is performed on the candidate risk factor set to generate interventionable labels.

[0008] Based on the risk factor association map and the interventionability label, risk classification parameters are generated for risk classification. The risk level is determined according to the risk classification parameters. High-weight risk factors are extracted from the risk classification parameters to generate a set of key risk factors.

[0009] The key risk factor set is classified into risk categories according to the spine region to obtain the risk value of each region. The difference analysis of the risk values ​​of each region is performed to identify the risk concentration area. The distribution pattern of the risk concentration area and the risk values ​​of each region is identified to obtain the risk distribution type.

[0010] Based on the risk distribution type and the risk concentration area, a zonal distribution feature is generated by association and integration. The spinal health risk assessment result is then output by combining the zonal distribution feature with the risk level.

[0011] A second aspect of this invention proposes a spinal health risk assessment system based on individual characteristics, comprising:

[0012] The feature acquisition module is used to acquire basic feature data and spinal morphology measurement data of an individual, and to construct an individual feature profile based on the basic feature data and the spinal morphology measurement data.

[0013] The risk analysis module is used to extract risk factors from the individual feature profile to form a candidate risk factor set, perform multi-factor coupling analysis on the candidate risk factor set to establish a risk factor association map, and perform interventionability assessment on the candidate risk factor set to generate interventionability labels.

[0014] The risk grading module is used to generate risk grading parameters based on the risk factor association map and the interventionability label, determine the risk level based on the risk grading parameters, and extract high-weight risk factors from the risk grading parameters to generate a set of key risk factors.

[0015] The zoning assessment module is used to classify the key risk factor set into risk categories according to the spine zoning to obtain the risk value of each zone, perform difference analysis on the risk values ​​of each zone to identify risk concentration areas, and identify the risk distribution type based on the distribution pattern of the risk concentration areas and the risk values ​​of each zone.

[0016] The results output module is used to generate zonal distribution features by associating and integrating the risk distribution type with the risk concentration area, and output the spinal health risk assessment results by combining the zonal distribution features with the risk level.

[0017] The beneficial effects of this invention are reflected in the following points: 1. Based on occupational posture pattern matching technology, occupational type labels are compared with posture pattern databases to obtain typical posture features. Deviation calculations are performed on measured spinal morphology within the standard curvature range, realizing the fusion of occupational factors and postural measurement data at the computational layer. This constructs a comprehensive profile including four evaluation dimensions: occupational posture risk, curvature deviation, center of gravity stability, and flexibility reserve, improving the comprehensiveness and relevance of individual characteristic characterization. 2. Causal relationship identification technology is introduced to distinguish between primary and secondary factors and construct causal association chains. The propagation path and intensity correlation between risk factors are established. The interventionability level classification and improvement difficulty assessment can quantify the intervention feasibility of each factor. Combined with the synergistic factor group identification mechanism, factors with similar interventionability are grouped and introduced into the deterioration rate and urgency assessment, generating a synergistic enhancement coefficient to achieve a quantitative characterization of the nonlinear amplification effect between factors, improving the accuracy of risk classification and the scientific nature of intervention priority. 3. By using the zonal sensitivity coefficient to quantify the differences in the degree of influence of each factor on different spinal segments, and by combining zonal weight adjustment and nonlinear mapping of saturation function to control the rationality of risk accumulation estimation, and by relying on difference analysis to identify risk concentration areas and continuity judgment to identify continuous risk segments, the concentration parameter and continuity characteristics are pattern matched to determine the risk distribution type, effectively distinguishing spatial distribution patterns such as isolated and continuous types, providing spatial positioning basis for the accurate formulation of targeted intervention strategies.

[0018] It should be understood that the above general description and the following detailed description are exemplary and explanatory only, and do not limit this application. Attached Figure Description

[0019] The accompanying drawings illustrate specific examples of the technical solutions described in this invention and, together with the detailed embodiments, form part of the specification, serving to explain the technical solutions, principles, and effects of this invention.

[0020] Unless otherwise specified or defined, the same reference numerals in different figures represent the same or similar technical features, and different reference numerals may be used to represent the same or similar technical features.

[0021] Figure 1 is a flowchart illustrating a spinal health risk assessment method based on individual characteristics according to the present invention.

[0022] Figure 2 is a structural block diagram of a spinal health risk assessment system based on individual characteristics according to the present invention. Detailed Implementation

[0023] In the following description, specific details such as particular system architectures and techniques are set forth for illustrative purposes and not for limitation, in order to provide a thorough understanding of the embodiments of this application. However, those skilled in the art will understand that this application may also be implemented in other embodiments without these specific details. In other instances, detailed descriptions of well-known systems, apparatuses, circuits, and methods have been omitted so as not to obscure the description of this application with unnecessary detail.

[0024] It should be understood that, when used in this application specification and the appended claims, the term "comprising" indicates the presence of the described features, integrals, steps, operations, elements and / or components, but does not exclude the presence or addition of one or more other features, integrals, steps, operations, elements, components and / or a collection thereof.

[0025] References to "one embodiment" or "some embodiments" as described in this specification mean that one or more embodiments of this application include a specific feature, structure, or characteristic described in connection with that embodiment. Therefore, the phrases "in one embodiment," "in some embodiments," "in other embodiments," "in still other embodiments," etc., appearing in different parts of this specification do not necessarily refer to the same embodiment, but rather mean "one or more, but not all, embodiments," unless otherwise specifically emphasized. The terms "comprising," "including," "having," and variations thereof mean "including but not limited to," unless otherwise specifically emphasized.

[0026] The technical solutions of the embodiments of this application will be described below.

[0027] As shown in Figure 1, this embodiment of the invention provides a spinal health risk assessment method based on individual characteristics, including the following steps S110-S150:

[0028] Step S110: Obtain basic feature data and spinal morphology measurement data of the individual, and construct an individual feature profile based on the basic feature data and spinal morphology measurement data.

[0029] Specifically, basic characteristic data and spinal morphology measurement data of individuals are obtained. Basic characteristic data is collected through a standardized questionnaire, gathering information on five dimensions: age, occupation type, years of work experience, body mass index (BMI), and exercise frequency. The questionnaire is completed at a medical institution or health management center. The age item determines the baseline reference for the degree of spinal degeneration and the selection range for flexibility standards. The occupation type item reflects the individual's daily postural load characteristics, and the years of work experience item reflects the cumulative effect of occupational postural load. The BMI item is obtained through height and weight values, reflecting the individual's weight load status. The exercise frequency item records the number of times the individual participates in moderate-intensity exercise per week, with the number of times being an integer value from 0 to 7. Spinal morphology measurement data is acquired using a spinal morphology measurement device in a natural standing position. The measurement requires the individual to maintain a standard measurement posture with feet shoulder-width apart, arms hanging naturally at their sides, and eyes looking forward. The data acquisition device marks the surface projection points of the spinous processes of each vertebra in the cervical, thoracic, and lumbar segments, recording the planar coordinate values ​​of each marked point. These coordinate values ​​include the X-axis (horizontal coordinate) and Z-axis (vertical coordinate). The X-axis coordinate is used for centroid offset analysis, and the Z-axis coordinate is used for spinal curve fitting. The coordinate accuracy reaches the millimeter level, ensuring that the curvature fitting deviation between adjacent vertebral marked points is controlled within an acceptable range. Simultaneously, the device acquires the individual's cervical flexion, extension, and lateral flexion range of motion, thoracic rotation range of motion, and lumbar flexion and extension range of motion. These range of motion data constitute the raw measurements of spinal flexibility parameters, which record the range of motion of each joint segment.

[0030] In some embodiments, constructing an individual feature profile based on the basic feature data and the spinal morphology measurement data includes: extracting occupational information from the basic feature data to generate occupational type labels; performing occupational posture pattern matching on the occupational type labels to generate typical posture features; performing deviation calculation on the spinal morphology measurement data based on the typical posture features to obtain posture deviation parameters; and performing feature fusion between the posture deviation parameters and the basic feature data to construct an individual feature profile.

[0031] Occupational information is extracted from basic feature data to generate occupational type labels. The occupational name text stored in the occupational type item of the basic feature data is queried and matched against an occupational classification standard library, which contains predefined occupational categories covering a wide range of common occupations. The matching of occupational names in the basic feature data with categories in the standard library uses an edit distance algorithm, which identifies the closest occupational category. Occupational type labels are supplemented with posture risk level and muscle load pattern attributes. The posture risk level of occupational type labels is divided into three levels: low static load, medium static load, and high static load. The classification is based on the statistical characteristics of the duration of typical sitting or standing postures for that occupation. High static load occupations include long-distance drivers who maintain a sitting posture for more than 6 hours continuously per day; medium static load occupations include teachers who alternate between standing while lecturing and walking around; and low static load occupations include construction workers who primarily engage in full-body dynamic activities. The muscle load pattern attribute of the occupation type tag identifies the main muscle load concentration areas for that occupation. These areas are categorized into three types: neck-shoulder concentrated, lower back concentrated, and whole-body distributed. Neck-shoulder concentrated occupations primarily affect the cervical spine and shoulder muscles; prolonged screen time leading to forward neck extension and shoulder shrugging is a typical example of this pattern. Lower back concentrated occupations primarily affect the lumbar spine and back muscles; repetitive bending and lifting causing continuous stress on the lumbar spine and erector spinae muscles is a typical example of this pattern. Whole-body distributed occupations distribute the load relatively evenly across all segments of the body. The occupation type tag stores the posture risk level and muscle load pattern attributes in a structured data format.

[0032] Occupational type tags are matched with occupational posture patterns to generate typical posture features. The posture risk level attribute of the occupational type tag is retrieved from a posture pattern database containing a set of corresponding posture pattern records. Each record describes the spinal morphological characteristics of a specific occupational group. The muscle load pattern attribute of the occupational type tag is compared with the load distribution features of the posture pattern records. The load distribution features are described by vectors representing the load weight distribution ratio of the cervical, thoracic, and lumbar segments. The three components of the vector correspond to the normalized load intensity of the cervical, thoracic, and lumbar segments, respectively. The similarity between the occupational type tag and the posture pattern records is calculated using a cosine similarity algorithm, and the posture pattern with the highest cosine value is selected as the matching result. Typical posture features include standard curvature range information, which includes three intervals: cervical lordosis curvature range, thoracic kyphosis curvature range, and lumbar lordosis curvature range. The standard curvature range of typical postural characteristics is expressed as an upper and lower limit interval. The boundary values ​​of the interval are obtained through statistical analysis of the spinal morphology of the occupational group. The standard curvature range of thoracic kyphosis in high static load occupational groups is usually 20 to 45 degrees Cobb angle. Typical postural characteristics vary among occupational groups with different postural risk levels, and the standard curvature range of high static load occupational groups usually deviates more significantly compared to low static load occupational groups.

[0033] Based on typical posture characteristics, deviation calculations are performed on spinal morphology measurement data to obtain posture deviation parameters. The coordinate points of each segment of the spinal morphology measurement data are fitted into a continuous curve using a cubic spline curve fitting algorithm. The fitting process is performed separately for the cervical, thoracic, and lumbar segments. The maximum curvature point is extracted from the fitted curve in each segment using the first and second derivatives of the spline function as the measured value of the curvature of each segment. The standard curvature range of typical postural characteristics provides reference upper and lower limits for the curvature of each segment. The deviation of each segment is calculated using the formula Di = (V_bound_i - V_actual_i) / V_bound_i, where i is the segment number, i=1 corresponds to the cervical segment, i=2 corresponds to the thoracic segment, and i=3 corresponds to the lumbar segment. Di is the deviation of the i-th segment, V_actual_i is the measured curvature value of the i-th segment, and V_bound_i is the standard curvature boundary value that the i-th segment has exceeded. When V_actual_i is lower than the lower limit of the standard, V_bound_i takes the lower limit value. When V_actual_i exceeds the upper limit of the standard, V_bound_i takes the upper limit value. When V_actual_i is within the range of the upper and lower limits, Di is 0. A positive value of Di indicates that the measured curvature is lower than the standard, and a negative value of Di indicates that the measured curvature exceeds the standard. The deviations of the three segments constitute the deviation vector D=[D_1,D_2,D_3]. This deviation vector, as a core component of the posture deviation parameter, quantifies the degree of posture deviation in each segment. The center of gravity offset is obtained by measuring the distance between the mean X-axis coordinates of all marked points in the spinal morphology measurement data and the body midline. The center of gravity offset and the deviation vector together constitute a complete description of the posture deviation parameter. The posture deviation parameter integrates the two dimensions of the deviation vector and the center of gravity offset. It is stored as a combination of vector and scalar. The vector portion records the curvature deviation information of the three segments, while the scalar portion records the overall balance characteristics of the center of gravity offset.

[0034] Individual profiles are constructed by fusing postural deviation parameters with baseline feature data. Feature fusion employs a weighted combination method to integrate six feature dimensions from the baseline data: age, body mass index (BMI), years of work experience, movement frequency, deviation vector of postural deviation parameters, and center of gravity shift. The individual profile includes four evaluation dimensions: occupational posture risk index, spinal curvature deviation index, center of gravity stability index, and movement compensation index. The occupational posture risk index is calculated by combining years of work experience and the deviation vector of postural deviation parameters from the baseline data; longer work experience and greater deviation result in a higher occupational posture risk index. The spinal curvature deviation index is obtained by weighting the norm of the deviation vector of postural deviation parameters and identifying the high-risk segments with the most significant deviations. The center of gravity stability index is calculated based on BMI and the center of gravity shift from the postural deviation parameters from the baseline data; individuals with high BMI and large center of gravity shifts have low center of gravity stability indices. The motion compensation index of an individual's profile is calculated based on the frequency of movement and age of the basic characteristic data. The higher the frequency of movement and the lower the age, the higher the motion compensation index. The motion compensation index reflects an individual's ability to offset occupational postural load through daily movement. Individuals with a frequency of movement less than twice a week and an age of over 45 years old have a low motion compensation index, indicating a lack of movement buffering for the risks caused by postural deviations.

[0035] Step S120: Extract risk factors from individual feature profiles to form a set of candidate risk factors, perform multi-factor coupling analysis on the set of candidate risk factors to establish a risk factor association map, and perform interventionability assessment on the set of candidate risk factors to generate interventionability labels.

[0036] Specifically, risk factors are extracted from individual profiles to form a candidate risk factor set. When the occupational posture risk index of an individual profile exceeds a preset threshold, occupation-related risk factors are extracted. A higher occupational posture risk index corresponds to three risk factors: prolonged sitting, forward head posture, and excessive lumbar load. The occupational posture risk index comprehensively reflects the level of spinal health risk caused by occupational characteristics. The high-risk segment locations marked by the spinal curvature deviation index in the individual profile determine curvature-related risk factors. For the thoracic spine, high-risk segments correspond to two risk factors: increased thoracic kyphosis and limited thoracic spine mobility. The spinal curvature deviation index quantifies the degree to which each segment deviates from the standard occupational posture. When the center of gravity stability index of an individual profile is below the standard value, balance-related risk factors are extracted. A lower center of gravity stability index corresponds to two risk factors: center of gravity shift and scoliosis tendency. The center of gravity stability index reflects the stability level of the spine's static balance ability. When the motor compensation index in an individual's anatomical profile falls below the standard value, risk factors related to insufficient compensation are identified. Risk factors corresponding to an insufficient motor compensation index include decreased muscle strength and insufficient lumbar spine flexibility. The motor compensation index assesses an individual's ability to cushion occupational postural loads through movement. The candidate risk factor set is compiled from four categories: occupational, curvature-related, balance-related, and flexibility-related, containing nine risk factor items. The factors in the candidate risk factor set are initially ranked using a severity score. The score is obtained by comparing the corresponding index in the individual's anatomical profile with a threshold; a larger difference indicates a greater deviation from the normal level. The factor with the highest score in the candidate risk factor set is ranked first.

[0037] In some embodiments, the step of performing multi-factor coupling analysis on the candidate risk factor set to establish a risk factor association map includes: identifying causal relationships in the candidate risk factor set to obtain primary and secondary risk factors; constructing causal association chains based on the primary and secondary risk factors; performing association strength assessment on the causal association chains to form association path weights; and establishing a risk factor association map based on the association path weights.

[0038] Causal relationship identification was performed on the candidate risk factor set to obtain primary and secondary risk factors. The causality of the factor items in the candidate risk factor set was determined using a causal relationship knowledge base. This knowledge base contains causal relationship rules in the field of spinal health, which describe the causal relationships and propagation paths between different risk factors in a condition-conclusion format. The factor item "prolonged sitting" in the candidate risk factor set was matched with an antecedent attribute in the knowledge base. This factor, as a primary risk factor, does not depend on the existence of other factors; its occurrence is autonomous or directly determined by the external work environment. The factor item "excessive lumbar load" in the candidate risk factor set was matched with a consequence attribute in the knowledge base. This factor, as a secondary risk factor, is triggered by the "prolonged sitting" factor. The causal determination of the candidate risk factor set used a rule matching algorithm. The algorithm retrieved rule entries in the knowledge base; if a match was successful, a causal relationship was established between the factors. The primary risk factor set contains the factor items in the candidate risk factor set that were determined to be antecedent attributes. This set includes three factor items: "prolonged sitting," "forward neck posture," and "decreased muscle strength." The set of secondary risk factors includes factors identified as consequences from the candidate risk factor set. This set includes six factors, such as "excessive lumbar load," "increased thoracic kyphosis," and "center of gravity shift." The distinction between primary and secondary risk factors is represented by a directed graph structure, where directed edges from primary risk factor nodes to secondary risk factor nodes indicate the direction of the causal relationship. The number of nodes for primary risk factors is less than the number of nodes for secondary risk factors; one primary risk factor may trigger multiple secondary risk factors.

[0039] For example, constructing a causal chain based on the primary risk factor and the secondary risk factor includes: identifying the causal direction of the primary risk factor and the secondary risk factor to obtain the factor-inducing direction; distinguishing between dominant factors and subordinate factors based on the factor-inducing direction; performing a correlation strength assessment on the dominant factor and the subordinate factor to form a causal path strength parameter; and constructing a causal chain based on the causal path strength parameter.

[0040] Causal direction identification was performed on primary and secondary risk factors to determine their triggering direction. The temporal relationship between the primary risk factor "prolonged sitting" and the secondary risk factor "excessive lumbar load" was determined using medical literature evidence to establish the triggering direction attribute. Sitting predates the accumulation of lumbar load, and this temporal sequence is a fundamental characteristic of causality. The triggering direction between the primary and secondary risk factors was quantified using rule confidence levels in a causal relationship knowledge base. Confidence levels, ranging from 0 to 1, indicate the reliability of the causal relationship, and are determined based on clinical research data and expert consensus. The rule confidence level for the primary risk factor "forward neck posture" triggering the secondary risk factor "increased thoracic kyphosis" was 0.78. Forward neck posture reduces cervical lordosis, and to maintain visual level and head balance, the thoracic spine compensates by increasing the kyphosis angle. This causal relationship has been repeatedly verified in clinical observations. The causal paths of risk propagation are represented by directed edges. Each edge originates from a primary risk factor node, terminates at a secondary risk factor node, and has a weight equal to its confidence level. These directed edges form the pathways for risk propagation. The set of causal paths includes all directed edges between primary and secondary risk factors, and the number of edges in the set equals the number of identified causal relationships. A confidence threshold of 0.6 is set for each causal path. Paths with confidence levels below this threshold are filtered out, while high-confidence causal relationships form the core framework of the causal relationship network. Weak associations with low confidence levels are excluded to avoid interfering with the identification of the main risk paths.

[0041] Factors are categorized into dominant and subordinate factors based on their initiation direction. The out-degree statistics of the initiation direction show that the "prolonged sitting" factor has an out-degree of 5, triggering 5 different secondary risk factors. A high out-degree indicates that this factor is the common starting point of multiple risk chains. Factors with an out-degree reaching a preset threshold in their initiation direction are marked as dominant factors. Dominant factors include three primary risk factors: "prolonged sitting," "forward neck posture," and "decreased muscle strength." These factors control the main propagation channels of the risk network. Factors with an in-degree reaching a preset threshold in their initiation direction are marked as subordinate factors. Subordinate factors include secondary risk factors such as "center of gravity shift" and "scoliosis tendency," which are jointly triggered by multiple factors. These factors are influenced by the convergence of multiple risk paths. Dominant factors occupy an upstream position in the causal network; changes in their severity affect multiple downstream subordinate factors through causal chains. The higher the out-degree of a dominant factor, the wider its influence within the network. Subordinate factors occupy a downstream position in the causal network, and their occurrence is influenced by multiple upstream dominant factors. Controlling a single dominant factor is insufficient to completely eliminate the risk of subordinate factors. The distinction between dominant and subordinate factors determines the priority of risk intervention; prioritizing the control of dominant factors can simultaneously block multiple causal propagation paths.

[0042] A causal path strength parameter is generated by assessing the correlation strength between the dominant and subordinate factors. The correlation strength between the dominant factor 'prolonged sitting' and the subordinate factor 'center of gravity shift' is obtained by multiplying the path confidence scores. The path strength S is equal to the product of the confidence scores of all edges on the path, S = c_1 × c_2 × ... × c_n, where c_i is the confidence score of the i-th edge on the path, i ranges from 1 to n, and n is the number of edges in the path. The product reflects the propagation reliability of the entire path. When there are multiple causal paths between the dominant and subordinate factors, the strongest correlation channel is represented by the maximum value of all path strengths. The maximum value represents the most reliable propagation path between the two factors. The path strength from the dominant factor "prolonged sitting" to the subordinate factor "center of gravity shift" via the intermediate factors "excessive lumbar load" and "scoliosis tendency" is the product of the confidence scores of the three sides. If the confidence scores of the three sides are 0.78, 0.72, and 0.65 respectively, then the path strength S = 0.78 × 0.72 × 0.65 = 0.365. The longer the path, the greater the confidence decay. The direct path strength between the dominant and subordinate factors is higher than the indirect path strength. The direct path contains fewer intermediate nodes, resulting in less confidence decay, and direct causal relationships are more reliable than indirect relationships. The causal path strength parameters are organized in matrix form, with rows corresponding to dominant factors, columns corresponding to subordinate factors, and matrix elements representing path strength values. The matrix form facilitates batch calculation and analysis. In the matrix of causal path strength parameters, some matrix elements corresponding to the absence of a causal path between some dominant and subordinate factors are 0. Non-zero elements are concentrated in specific rows and columns, reflecting the selectivity of the association between factors. High-intensity matrix elements of the causal path strength parameters mark the key propagation channels from the dominant factor to the subordinate factor.

[0043] A causal relationship chain is constructed based on the causal path strength parameter. The row and column indices of the matrix elements with a causal path strength parameter greater than a preset threshold determine the start and end points of the causal relationship chain. Threshold filtering retains strongly correlated causal paths, while weakly correlated paths are excluded to simplify the network structure. The topology of the causal relationship chain is represented by a directed acyclic graph. The causal relationship knowledge base itself is constructed based on the principle of temporal sequence, eliminating loop structures where factor A causes factor B and factor B also causes factor A. The network after threshold filtering by the causal path strength parameter inherits this acyclic characteristic. The numerical value of the causal path strength parameter serves as the weight attribute of the edges in the causal relationship chain. The weight quantifies the strength level of the causal relationship; a higher weight indicates a more reliable causal relationship. The nodes included in the causal relationship chain cover the main factors and some mediating factors in the candidate risk factor set, and the set of nodes constitutes the complete participants in risk propagation. The longest path depth of the causal relationship chain represents the number of intermediate propagation levels from the primary risk factor to the final secondary risk factor; the path depth reflects the complexity of risk propagation. In the causal chain, the dominant factor "prolonged sitting" has the most outgoing edges, making this node the core hub of the network and controlling multiple risk propagation paths. The causal chain is stored as a combination of node sets and edge sets. Nodes carry factor names and severity rating attributes, while edges carry causal direction and path strength attributes.

[0044] For causal chains, a correlation strength assessment is performed to form correlation path weights. The path strength value carried by each edge in the causal chain is converted into a correlation path weight through local node normalization. The normalization method uses the sum of the outgoing edge strengths of each node as the denominator, ensuring that the sum of the correlation path weights of the outgoing edges of each node is 1. In the causal chain, the normalized correlation path weights of the multiple outgoing edge strengths of the "long-term inactivity" node correspond to different propagation directions, and the correlation path weight allocation reflects the proportion of risk distribution along each path. The correlation path weights satisfy the constraint that the sum of the outgoing edge weights of each node is 1. This constraint gives the correlation path weights a probabilistic interpretation meaning; that is, the correlation path weight can be understood as the probability distribution of risk propagation from this node to each downstream node. The larger the edge weight value of the causal chain, the higher the relative importance of that path among all outgoing edges. The correlation path weight allocation reflects the primary and secondary directions of risk propagation. The correlation path weights are stored through the edge attribute fields, which include the starting node, ending node, and normalized weight, comprehensively recording the correlation information of the path. The allocation of path weights distinguishes between the main path and branch paths in a causal chain. The path weight of the main path is significantly higher than that of the branch paths, indicating that the main path is the primary channel for risk propagation. When multiple paths in a causal chain connect the same pair of start and end points, the cumulative value of the path weights represents the overall correlation strength of that factor pair. The allocation of path weights shows that a few high-weight edges carry most of the causal propagation traffic, while most low-weight edges only undertake secondary, auxiliary propagation.

[0045] A risk factor correlation graph is constructed based on the weights of associated paths. The topology of the risk factor correlation graph is constructed by combining the set of edges with the set of nodes in the causal chain. The graph contains factor nodes and weighted directed edges. Each node in the graph is labeled with its factor name and severity score; this score helps identify high-risk nodes. The graph reveals a network pattern where "prolonged sitting" acts as a central hub, radiating outwards to influence multiple secondary risk factors. Outgoing edges from this node point to downstream factors such as "excessive lumbar load," "limited thoracic spine mobility," and "insufficient lumbar spine flexibility." The associated path weights of these edges reflect the proportion of risk distribution along each path. The graph structure supports path query operations, which can retrieve all intermediate paths and the total path strength between any two factors. The total path strength equals the sum of the associated path weights of all paths connecting the factor pair. Edges with high associated path weights in the graph form the backbone network of risk propagation, connecting the core causal channels between dominant factors and key subordinate factors. The topological structure of the risk factor association map shows a star-shaped distribution pattern with a few dominant factors as the core and spreading outwards, with the "long-term sitting" node located at the center of the star structure.

[0046] In some embodiments, the step of performing an interventionability assessment on the candidate risk factor set to generate interventionability labels includes: identifying factor types in the candidate risk factor set to obtain factor classification results; dividing the interventionability levels based on the factor classification results to form an interventionability grading table; performing an improvement difficulty assessment on the interventionability grading table to generate an improvement difficulty coefficient; and marking the candidate risk factor set with interventionability based on the improvement difficulty coefficient to generate interventionability labels.

[0047] Factor type identification was performed on the candidate risk factor set to obtain factor classification results. Factor items in the candidate risk factor set were classified into three main types using a factor classification rule base. The rule base defines three main types: physiological factors, behavioral factors, and genetic factors. The classification is based on the essential attributes and formation mechanisms of the factors. The factor item "decreased muscle strength" in the candidate risk factor set matched the physiological factor type in the rule base. This type of factor involves the body's tissue structure or functional state, and improvement requires physiological intervention. The factor item "prolonged sitting" in the candidate risk factor set matched the behavioral factor type. This type of factor is related to individual lifestyle habits or work patterns, and improvement mainly relies on behavioral adjustments. The factor items in the candidate risk factor set did not include genetic factor types; genetic factors such as family hereditary predisposition were not extracted in this assessment. The factor classification results divided the factor items in the candidate risk factor set into a physiological factor subset and a behavioral factor subset. The physiological factor subset included four factors: "decreased muscle strength," "insufficient lumbar spine flexibility," "limited thoracic spine mobility," and "scoliosis tendency." These factors require rehabilitation training or medical intervention. The behavioral factor subset of the factor classification results includes five items: "prolonged sitting", "forward neck posture", "excessive lumbar load", "increased thoracic kyphosis", and "center of gravity shift". The number of elements in the two subsets of the factor classification results reflects the dominance of behavioral factors in the individual; a high proportion of behavioral factors suggests a better prospect for intervention.

[0048] An intervention grading table is created based on the factor classification results to determine the level of interventionizability. Behavioral factors, as identified in the factor classification results, correspond to a high intervention level in the intervention grading standard. These behavioral factors can be improved through habit changes and environmental adjustments, with low cost and quick results. Physiological factors, as identified in the factor classification results, correspond to a medium intervention level in the intervention grading standard. These physiological factors require rehabilitation training or medical intervention for improvement, with a long improvement cycle and the need for professional guidance. Genetic factors, as identified in the factor classification results, correspond to a low intervention level in the intervention grading standard. Genetic factors are the most difficult to change or cannot be changed; their impact can only be mitigated by controlling other factors. The intervention grading table marks behavioral factors in the factor classification results as high intervention levels, indicating the highest intervention feasibility; these factors are the priority targets for intervention. The intervention grading table also marks physiological factors in the factor classification results as medium intervention levels, indicating moderate intervention feasibility; these factors require moderate-intensity intervention investment. The statistical distribution of the interventionability grading scale shows 5 items at the high interventionability level, 4 items at the medium interventionability level, and 0 items at the low interventionability level. This distribution suggests that the individual's risk factors have good intervention prospects, and most factors can be improved. The data in the interventionability grading scale are sorted from highest to lowest interventionability level, with high-level factors listed first.

[0049] An improvement difficulty coefficient was generated by performing an improvement difficulty assessment on the interventionizability grading scale. The improvement difficulty coefficient for the high-interventionibility factor "prolonged sitting" on the interventionizability grading scale was obtained using an improvement difficulty assessment model. The model is calculated using the formula H = 0.3 × T + 0.2 × R + 0.5 × C, where H is the improvement difficulty coefficient, T is the time dimension score indicating the duration required for improvement, R is the resource input dimension score indicating the economic and equipment input required for improvement, and C is the compliance dimension score indicating the degree of individual self-discipline required for improvement. The scores for all three dimensions range from 1 to 5. The time dimension score T for the "prolonged sitting" factor on the interventionizability grading scale is 2, indicating that improvement requires a behavioral adjustment cycle of 2 to 3 months, with a moderate time cost. The resource input dimension score R for the "prolonged sitting" factor on the interventionizability grading scale is 1, indicating that improvement only requires low-cost environmental adjustments or equipment assistance, with a low economic burden. The compliance dimension score C for the "prolonged sitting" factor on the interventionizability grading scale is 3, indicating that improvement requires high individual self-discipline and habit adherence; compliance is a key challenge in behavior change. The difficulty coefficient for improving the "prolonged sitting" factor is H = 0.3 × 2 + 0.2 × 1 + 0.5 × 3 = 2.3. For the "decreased muscle strength" factor, the time dimension score T is 4, requiring more than 6 months of systematic rehabilitation training; the resource input score R is 3, requiring professional rehabilitation equipment and therapist guidance; and the compliance score C is 3, requiring long-term adherence to the training plan. The difficulty coefficient for improving these factors is H = 0.3 × 4 + 0.2 × 3 + 0.5 × 3 = 3.3. The interventionability grading table assigns difficulty coefficients to each factor, ranging from lowest to highest. Physiological factors generally have higher difficulty coefficients than behavioral factors, reflecting the difference in difficulty between the two types of factors.

[0050] Interventional labels are generated by labeling the candidate risk factor set based on the difficulty coefficient of improvement. Factors with an improvement difficulty coefficient below 2.0 are labeled with an "easy to improve" interventional label. Factors with an improvement difficulty coefficient below this threshold face little resistance to improvement, and behavioral adjustments can achieve the improvement goal. Factors with an improvement difficulty coefficient between 2.0 and 3.0 are labeled with a "moderate difficulty" interventional label. Factors in this range require appropriate intervention support and professional guidance. Factors with an improvement difficulty coefficient above 3.0 are labeled with a "difficult to improve" interventional label. Factors above this threshold are difficult to improve, typically involving the restoration of physiological structure or functional reconstruction. The three categories of interventional labels correspond to different intervention strategy recommendations: "easy to improve" suggests self-adjustment, "moderate difficulty" suggests professional guidance, and "difficult to improve" suggests medical intervention. Each factor in the candidate risk factor set is associated with an interventional label, establishing a mapping relationship between factors and intervention strategies. The distribution of interventionability labels reflects the overall difficulty of intervention for the individual. Behavioral factors are concentrated in the "easy to improve" and "moderate difficulty" label ranges, while physiological factors are concentrated in the "difficult to improve" label range. The distribution characteristics suggest that the individual needs a comprehensive, stratified intervention program.

[0051] Step S130: Based on the risk factor association map and the intervention capability label, risk classification parameters are generated for risk classification. The risk level is determined according to the risk classification parameters. High-weight risk factors are extracted from the risk classification parameters to generate a set of key risk factors.

[0052] In some embodiments, the step of generating risk classification parameters based on the risk factor association map and the interventionlability label includes: extracting high-weight factors from the risk factor association map to obtain a core factor set; identifying synergistic relationships in the core factor set based on the interventionlability label to generate a synergistic factor group; performing a synergistic enhancement assessment on the synergistic factor group to form a synergistic enhancement coefficient; and adjusting the core factor set using the synergistic enhancement coefficient to generate risk classification parameters.

[0053] The core factor set is obtained by extracting high-weight factors from the risk factor association graph. Nodes in the risk factor association graph are sorted by the weighted sum of their out-degree and in-degree, calculated using the formula W = α × out-degree + β × in-degree, where α is the out-degree weight (0.6) and β is the in-degree weight (0.4). The out-degree weight is higher than the in-degree weight because factors with higher out-degrees have stronger risk propagation capabilities; controlling this factor can simultaneously affect multiple downstream factors. The node "prolonged sitting" in the risk factor association graph has an out-degree of 5 and an in-degree of 0, ranking high in weighted sum among all nodes. The node "excessive lumbar load," although a secondary risk factor, also has high out-degree and in-degree, acting as a propagation relay in the risk factor association graph, and its weighted sum is also high. Factors whose node weighted sums exceed a preset threshold are marked as high-weight factors. These high-weight factors include the nodes "prolonged sitting," "forward neck posture," "decreased muscle strength," and "excessive lumbar load." High-weight factors in the risk factor association graph occupy key positions in the network; their removal leads to a significant decrease in network connectivity. The core factor set summarizes the four high-weight factors identified in the risk factor association graph. While the core factor set accounts for 44% of the total number of nodes in the risk factor association graph, it also accounts for 65% of the number of associated edges. The four factors in the core factor set form a tightly connected core subgraph in the risk factor association graph, with a direct causal edge between "prolonged sitting" and "excessive lumbar load." The core factor set is extracted from the risk factor association graph as a set of nodes, with each element retaining the factor name and severity score of the node.

[0054] Synergistic factor groups were generated by identifying synergistic relationships within the core factor set based on interventionability labels. The four factors in the core factor set were grouped according to the category attributes of the interventionability labels. The grouping objective was to identify factor combinations with similar interventionability and synergistically implementable intervention strategies. The interventionability labels showed that "prolonged sitting" and "neck extension posture" in the core factor set were both categorized as "moderate difficulty," with improvement difficulty coefficients of 2.3 and 2.5, respectively. Factors with the same interventionability labels in the core factor set were grouped into candidate synergistic factor groups; the candidate synergistic factor group corresponding to the "moderate difficulty" category contained two behavioral factors. The interventionability labels showed that "decreased muscle strength" was categorized as "difficult to improve," and "excessive lumbar load" was categorized as "moderate difficulty"; these two factors, with different interventionability label categories, did not constitute a synergistic factor group. The identification rules for synergistic factor groups require that factors within the group have the same modifiable label category and a difference in improvement difficulty coefficient of less than 0.5. The difference in improvement difficulty coefficient between "prolonged sitting" and "neck extension posture" is 0.2, meeting this condition. Although "excessive lumbar load" shares the same modifiable label category as "prolonged sitting," it is a subordinate rather than a dominant factor in the risk factor association map, and because the implementing entities for synergistic interventions are different, it is not included in the same synergistic factor group. The two factors in the "moderate difficulty" synergistic factor group have an indirect causal path in the risk factor association map, connected by the intermediate factor "excessive lumbar load." Synergistic intervention of the two factors can simultaneously control the risk input of this intermediate factor. The synergistic grouping results of the modifiable labels identified one effective synergistic factor group, which includes factors from the two core factor sets: "prolonged sitting" and "neck extension posture." Synergistic factor groups are stored in a nested list structure, with the outer list containing multiple synergistic groups and the inner list containing the factor items within each group.

[0055] For example, the step of performing a synergy enhancement assessment on the synergy factor group to form a synergy enhancement coefficient includes: assessing the deterioration rate of irreversible factors in the synergy factor group to obtain a deterioration process coefficient; classifying urgency based on the deterioration process coefficient to form an urgency level; performing a risk acceleration assessment on the urgency level and the synergy factor group to generate a risk acceleration coefficient; and determining the synergy enhancement coefficient based on the risk acceleration coefficient.

[0056] The rate of deterioration of irreversible factors in the synergistic factor group was assessed to obtain the deterioration process coefficient. Irreversible factors in the synergistic factor group were identified by analyzing common secondary factors induced by factors within the group. Secondary factors jointly induced by "prolonged sitting" and "forward neck posture" in the synergistic factor group constituted the candidate range of irreversible factors. The irreversibility attributes of candidate secondary factors in the synergistic factor group were identified using reversibility determination rules, which determined reversibility based on factor type and physiological degenerative characteristics. "Prolonged sitting" in the synergistic factor group was a behavioral factor with complete reversibility; this factor could be completely eliminated through behavioral changes. The secondary factor "straightening of cervical curvature" induced by "forward neck posture" in the synergistic factor group was partially reversible in the early stages, but after long-term persistence, it transformed into an irreversible structural change. Irreversible factors were identified in the common secondary factors of the synergistic factor group; among the three common secondary factors, "increased thoracic kyphosis" was determined to be an irreversible factor. The rate of deterioration of the irreversible factor "increased thoracic kyphosis" was assessed using a longitudinal data model. The model predicted the deterioration trend based on three variables: age, disease duration, and intervention response. The model output was the current deterioration rate, R_actual. The deterioration rate assessment showed that the annual deterioration rate of "increased thoracic kyphosis" without intervention was an increase of 1.5 degrees of Cobb angle per year. In normal adults, the thoracic kyphosis angle naturally increases by approximately 0.5 degrees per year with age. A deterioration rate of 1.5 degrees is three times the normal rate of deterioration. The deterioration progression coefficient was defined as D_prog = R_actual / R_normal, where R_actual is the current deterioration rate and R_normal is the normal rate of deterioration, and D_prog = 1.5 / 0.5 = 3.0. A deterioration progression coefficient greater than 1 indicates that the deterioration rate exceeds the normal rate of deterioration; the coefficient value reflects the multiple relationship of the deterioration rate. The deterioration progression coefficient is recorded as a floating-point number in the attributes of the irreversible factor.

[0057] Urgency levels are established based on a deterioration process coefficient. The deterioration process coefficient is mapped to an urgency level using a grading standard, which defines a coefficient less than 1.5 as low urgency, 1.5 to 2.9 as medium urgency, and 3.0 or higher as high urgency. Different urgency levels correspond to different intervals in the grading standard, with the interval boundary values ​​serving as thresholds for level division. The urgency grading standard considers the interaction between the deterioration process coefficient and individual age; older individuals have higher urgency levels for the same deterioration process coefficient. The urgency level adjustment rule lowers the threshold value of the deterioration process coefficient for individuals in specific age groups, reflecting the amplifying effect of age on risk urgency. The high urgency level corresponds to a time window of 6 to 12 months requiring intervention. Failure to intervene within this time window may lead to irreversible spinal morphological changes due to increased thoracic kyphosis exceeding structural compensation limits. Urgency levels are stored as an enumeration type, with three options: "low urgency," "medium urgency," and "high urgency." The urgency level is assigned to the record of irreversible factors. A high urgency level triggers an increase in the priority of the synergistic factor group, ensuring that the synergistic factor group receives priority in intervention resource allocation.

[0058] Risk acceleration assessment is performed on urgency levels and synergy factor groups to generate risk acceleration coefficients. The synergy baseline score of the synergy factor group is obtained by averaging the severity scores of factors within the group. The synergy baseline score quantifies the risk level of the synergy factor group before considering urgency. The high urgency indicator of the urgency level is coupled with the synergy baseline score of the synergy factor group for analysis. This coupling analysis assesses the amplification effect of urgency factors on synergistic risk. The risk acceleration assessment model converts urgency levels into acceleration factors: high urgency corresponds to an acceleration factor of 1.5, medium urgency to 1.2, and low urgency to 1.0. The acceleration factor reflects the impact of urgency on the rate of risk accumulation. Under high urgency, the rate of deterioration of irreversible factors far exceeds the normal rate of degradation, requiring a higher acceleration factor to reflect the urgency of the risk. The synergy baseline score of the synergy factor group is multiplied by the acceleration factor corresponding to the urgency level, and the result is the accelerated risk intensity. The risk acceleration factor is defined as R_accel = S_accelerated - S_base, where S_accelerated is the accelerated risk intensity and S_base is the synergy baseline score. The risk acceleration factor quantifies the additional risk increment brought about by urgency factors, and the increment value reflects the degree of impact of urgency. Risk acceleration assessment identifies the non-linear amplification of overall risk caused by the existence of irreversible factors in the synergy factor group. The higher the urgency level, the larger the risk acceleration factor. In low urgency cases, an acceleration factor of 1.0 corresponds to a risk acceleration factor of 0, indicating that urgency has not generated any additional risk increment. The risk acceleration factor is stored as a floating-point number in the assessment results of the synergy factor group.

[0059] The synergy enhancement coefficient is determined based on the risk acceleration coefficient. The synergy enhancement coefficient is calculated using the formula E = S_base + R_accel, where E is the synergy enhancement coefficient, S_base is the synergy base score of the synergy factor group, and R_accel is the risk acceleration coefficient. This additive approach combines the additional risk increment from urgency into the synergy base score. The synergy enhancement coefficient reflects the true risk level of the synergy factor group after considering the urgency of irreversible factors. The value is significantly higher than the synergy base score, with the magnitude of the increase depending on the acceleration factor size corresponding to the urgency level. The final value of the synergy enhancement coefficient is used to replace the initial severity scores of factors in the synergy factor group. This replacement operation updates the score fields of each factor in the synergy factor group. After replacement, factors within the synergy factor group participate in ranking and classification with a unified synergy enhancement coefficient value in subsequent risk grading. The synergy enhancement coefficient ranks highest among the severity scores of factors in all core factor sets, exceeding any single score of non-synergistic factors in the core factor set. The results of the determination of the synergy enhancement coefficient are appended to the data record of the synergy factor group in numerical form. The record also retains two intermediate quantities, the synergy base score and the risk acceleration coefficient, to support score traceability.

[0060] Risk grading parameters were generated by adjusting the core factor set using a synergy enhancement coefficient. The synergy enhancement coefficient replaced the original severity scores of two factors in the synergistic factor group within the core factor set, updating the scores for "prolonged sitting" and "neck extension posture" to a synergy enhancement coefficient of 12.23. The remaining two non-synergistic factors in the core factor set retained their original severity scores: "decreased muscle strength" (7.8 points) and "excessive lumbar load" (6.5 points). The score distribution of the core factor set after adjustment with the synergy enhancement coefficient showed significant differentiation, with the scores of the synergistic factor group being significantly higher than the highest scores of the non-synergistic factors. The risk grading parameters used the adjusted core factor set scores and included the severity values ​​of all nine factors in the candidate risk factor set. Factors within the synergistic factor group used the scores replaced by the synergy enhancement coefficient, non-synergistic factors in the core factor set used their original scores, and the five factors outside the core factor set used the initial severity scores from the candidate risk factor set. The nine factors in the risk grading parameters were sorted from highest to lowest score, and the sorting result determined the risk priority sequence. The risk grading parameters are stored in a list data structure, which contains each factor item and its corresponding score.

[0061] Risk levels are determined based on risk grading parameters. The comprehensive risk score of the risk grading parameters is divided into three risk levels—high risk, medium risk, and low risk—using a quantile segmentation method, with quantile thresholds set at the 75th and 25th percentiles, respectively. Factors with risk grading parameter scores greater than the 75th percentile threshold are classified as high-risk. The high-risk level includes three factors: "prolonged sitting," "forward neck posture," and "decreased muscle strength." These three factors are ranked among the top three in terms of scores influenced by the synergistic enhancement coefficient or the individual's high severity score. Factors with risk grading parameter scores between the 25th and 75th percentile thresholds are classified as medium-risk. The medium-risk level includes three factors: "excessive lumbar load," "increased thoracic kyphosis," and "insufficient lumbar flexibility." Factors with risk grading parameter scores below the 25th percentile threshold are classified as low-risk. The low-risk level includes three factors: "limited thoracic spine mobility," "center of gravity shift," and "scoliosis tendency." The risk level classification results are appended as labels to the factor records of the risk grading parameters, with label values ​​of three enumeration types: "high risk," "medium risk," and "low risk." The risk level distribution statistics show that high-risk (3 items), medium-risk (3 items), and low-risk (3 items) each account for 33%. The risk level determination results are used to guide the priority allocation of intervention resources, with high-risk factors receiving priority intervention resources and intensive monitoring frequency. The three-level classification results of the risk levels are stored in the risk grading parameter data table, where row records correspond to each risk factor, and column fields contain scores and risk level labels.

[0062] High-weighted risk factors were extracted from the risk grading parameters to generate a key risk factor set. Three factors with a "high risk" risk level in the risk grading parameters were directly included in the key risk factor set; the comprehensive risk scores of these factors all exceeded the high-risk threshold. Factors with a "medium risk" risk level in the risk grading parameters but an out-degree greater than or equal to 3 in the risk factor association graph were also included in the key risk factor set. Factors with high out-degrees have strong risk propagation capabilities, and simply classifying them based on scores might miss such factors. The extraction rules of the risk grading parameters identified the factor "excessive lumbar load," which, although of medium risk, has an out-degree of 4. This factor was added to the key risk factor set due to its strong propagation capabilities. The key risk factor set summarizes high-risk factors and factors with high propagation capabilities, resulting in a set containing 4 factors, accounting for 44% of the original 9 factors. The 4 factors in the key risk factor set are "prolonged sitting," "decreased muscle strength," "forward neck posture," and "excessive lumbar load," covering both behavioral and physiological factors. The key risk factor set retains three key attributes from the risk grading parameters: comprehensive risk score, risk level, and interveneability label. The key risk factor set is stored in a list data structure, with list elements encapsulating factor names and related attributes as factor objects. The data size of the key risk factor set is reduced to 44% of the original candidate risk factor set, focusing on four key risk points.

[0063] Step S140: Classify the key risk factors set by spine partition to obtain the risk value of each partition, perform difference analysis on the risk values ​​of each partition to identify the risk concentration area, and identify the risk distribution type based on the distribution pattern of the risk concentration area and the risk values ​​of each partition.

[0064] In some embodiments, the step of classifying the set of key risk factors by spine partition to obtain the risk value of each partition includes: performing a partition impact degree analysis on the set of key risk factors to obtain a partition sensitivity coefficient; adjusting the partition weights of the set of key risk factors based on the partition sensitivity coefficients to form adjusted partition weights; accumulating the risk values ​​of each spine partition according to the adjusted partition weights to generate the original cumulative value of partition risk; and determining the risk saturation of the original cumulative value of partition risk to determine the risk value of each partition.

[0065] A regional sensitivity coefficient was obtained by performing a regional impact analysis on the set of key risk factors. The sensitivity of each of the four factors in the set of key risk factors to the spinal region was assessed using an anatomical impact model. The model calculates the stress impact of each factor on the regional structure based on biomechanical principles. The sensitivity of the "prolonged sitting" factor in the set of key risk factors to the lumbar region was obtained through intervertebral disc pressure analysis. The analysis showed that sitting posture increases lumbar intervertebral disc pressure by 40%, with the lumbar region showing the highest sensitivity to this factor. The sensitivity of the "forward neck posture" factor in the set of key risk factors to the cervical region was assessed through cervical torque analysis. Torque analysis showed that forward neck posture increases the torque on the cervical spine by 2.5 times, with the cervical region showing the highest sensitivity to this factor. The regional sensitivity coefficient quantifies the influence of each factor in the set of key risk factors on each region as a value between 0 and 1. The closer the value is to 1, the higher the sensitivity of the factor to that region. The sensitivity coefficients for the "prolonged sitting" factor in the key risk factor set are 0.85 for the lumbar spine, 0.45 for the thoracic spine, and 0.20 for the cervical spine. The coefficient distribution reflects the biomechanical characteristics of this factor; the lumbar spine bears the most direct load from sitting, the thoracic spine is indirectly affected through postural compensation, and the cervical spine is least affected. The calculated sensitivity coefficients are organized in matrix form, with rows corresponding to factor items in the key risk factor set, columns corresponding to spinal regions, and matrix elements representing the sensitivity coefficient values. The sensitivity of each of the four factor items in the key risk factor set to three regions is calculated, resulting in a 4x3 matrix. The matrix elements exhibit sparsity; some factors have a sensitivity close to 0 to specific regions, indicating a weak impact.

[0066] The weights of key risk factors are adjusted based on the zonal sensitivity coefficients to form the adjusted weights. Each element of the zonal sensitivity coefficient matrix is ​​multiplied by the comprehensive risk score of the corresponding factor in the key risk factor set. The result represents the weighted risk contribution of that factor to that zonal. This adjustment process combines the overall risk level of a factor with its sensitivity to a specific zonal, reflecting the differentiated impact paths of different factors on different segments. The comprehensive risk score of the "prolonged sitting" factor in the key risk factor set is 12.23. This score is multiplied by the lumbar spine zonal sensitivity coefficient of 0.85 to obtain the adjusted weight for the lumbar spine zonal region. The multiplication operation for adjusting the zonal weights is performed on each element of the zonal sensitivity coefficient matrix, and the adjusted weight matrix maintains a 4x3 dimensional structure. The adjusted weights of the factors in the key risk factor set differ significantly across zonal regions, reflecting the uneven risk contribution of factors to different zonal regions. For example, prolonged sitting contributes much more to the lumbar spine than to the cervical spine, while forward head posture has a major impact on the cervical spine zonal region. This zonal difference conforms to the transmission law of biomechanical load. Key risk factors with high overall risk scores generally contribute significantly to all partitions in the adjusted weight matrix, but their relative contribution ratio within each partition is still modulated by the partition sensitivity coefficient. The weight distribution after partition adjustment shows a trend towards concentration in specific partitions. This concentration trend is determined by the differences in partition sensitivity coefficients. For office workers, the weight distribution typically concentrates towards the lumbar and cervical spine, while for manual laborers, the weight distribution is more concentrated in the lumbar spine single partition. The adjusted weights are stored in a matrix data structure.

[0067] After zoning adjustments, the weights are applied to each spinal region, and the risk values ​​are accumulated to generate the original cumulative risk value for each region. The column vectors of the zoning-adjusted weight matrix represent the weight contributions received by each region from all risk factors. The summation of the elements in the column vectors calculates the original cumulative risk value for that region. This summation process quantifies the cumulative effect of multiple factors on a single region. For the lumbar spine region, the risk value accumulation process sums the adjusted weight values ​​of the four key risk factors. This summation operation integrates the cumulative influence of multiple factors on the region. When multiple high-weight factors act on the same region simultaneously, the accumulated value increases significantly, indicating that the region bears multiple risk loads. The first column of the zoning-adjusted weight matrix corresponds to the lumbar spine region; the summation of the four elements in this column yields the original cumulative risk value for that region. The calculation of the original cumulative risk value for each region is performed separately for the cervical, thoracic, and lumbar spine regions, and the results are the cumulative risk values ​​for each of the three regions. The lumbar spine region had the highest cumulative risk value, followed by the cervical spine region, and the thoracic spine region had the lowest. This ranking of cumulative values ​​is consistent with the weighted distribution of the region sensitivity coefficients. This ranking pattern is common among sedentary professionals, reflecting the vulnerability of the lumbar spine under static load. The magnitude of the cumulative risk value is influenced by both the size of the set of key risk factors and the overall risk score; a larger size or higher score results in a larger cumulative value. When an individual has multiple high-score risk factors, the cumulative value of a specific region may significantly exceed the level of a single factor, forming a risk clustering effect. The cumulative risk values ​​of the three regions are stored in vector form, with the vector elements arranged according to the anatomical location of the cervical, thoracic, and lumbar spine.

[0068] Risk saturation is used to determine the risk value of each zone based on the original cumulative risk values ​​of the zones. The original cumulative risk values ​​of the zones are non-linearly mapped using a saturation function, mathematically expressed as S(x) = x / (1 + x / x_sat), where x is the original cumulative risk value of the zone and x_sat is the saturation threshold parameter, which is set based on the full score of a clinical risk scoring scale. The physical significance of risk saturation determination lies in simulating the upper limit of risk accumulation that a spinal zone can bear; beyond this limit, the marginal effect of additional risk diminishes. The zone values ​​of the original cumulative risk values ​​are mapped using the saturation function, and the mapping result is the final risk value for each zone. A uniform saturation threshold parameter is used for the saturation mapping of risk values ​​across all zones, ensuring the comparability of risk values ​​across different zones. Zones with higher original cumulative risk values ​​exhibit a more pronounced compression effect after saturation function mapping; high scores close to the saturation threshold result in mapping values ​​lower than the original cumulative value. Zones with lower original cumulative risk values ​​have less impact on their risk values ​​after saturation mapping, with values ​​before and after mapping being similar. The upper limit of the risk value of each partition after saturation determination is determined by the asymptotic properties of the saturation function. When the original cumulative value of partition risk approaches infinity, the risk value of each partition approaches the saturation threshold x_sat. The correspondence between the original cumulative value of partition risk and the risk value of each partition is established through the saturation function. The nonlinearity of the function introduces the robustness of risk assessment and avoids the partition risk value from losing its discriminative power due to extremely high scores for individual factors.

[0069] A variance analysis was performed on the risk values ​​of each zone to identify areas of concentrated risk. The calculated risk values ​​for each zone showed that the lumbar spine zone had the highest risk value, followed by the cervical spine zone, and the thoracic spine zone had the lowest, exhibiting a differentiated distribution across the zones. The standard deviation statistic of the risk values ​​for each zone was used to assess the dispersion of risk levels across the zones; the standard deviation formula is as follows: , where x_i is the risk value of each partition, μ is the average value, and n is the number of partitions. Range analysis of the risk values ​​of each partition identifies the numerical difference between the highest and lowest risk partitions. A range exceeding a set threshold indicates a significant risk difference between the partitions. A dual-threshold determination method is used to identify risk concentration areas. The determination rule requires that the risk value of each partition simultaneously meet two conditions: exceeding an absolute threshold and exceeding a multiple of the average value. Partitions that meet the dual-threshold conditions are marked as risk concentration areas. The identification results show that the lumbar spine partition is the only high-risk concentration area. The spatial location of the risk concentration area is recorded as the lower lumbar spine region. Isolated distribution means that this high-risk area is isolated from other partitions by low-risk areas, and the pathological mechanisms of different partitions are relatively independent. The location markers of the risk concentration areas are used for subsequent distribution pattern identification.

[0070] In some embodiments, the step of identifying the risk distribution type based on the distribution pattern of the risk concentration area and the risk values ​​of each partition includes: extracting the number of concentration areas based on the risk concentration area to obtain a regional concentration parameter; determining the continuity of the risk values ​​of each partition to identify continuous risk segments; generating a distribution pattern identifier by performing pattern matching between the regional concentration parameter and the continuous risk segments; and determining the risk distribution type based on the distribution pattern identifier.

[0071] The concentration parameter is obtained by extracting the number of concentrated risk areas. The identification results of concentrated risk areas are obtained by counting the number of concentrated areas, specifically the number of partitions that meet the risk threshold. The spatial distribution characteristics of concentrated risk areas vary among individuals. Individuals who work at desks for long periods often show a single concentrated risk area in the cervical spine. This is because prolonged looking down at screens leads to accumulated forward flexion load on the cervical spine, with the risk primarily focused on the cervical segment. Manual laborers often show a single concentrated risk area in the lumbar spine, as repetitive movements such as lifting and bending subject the lumbar spine to repeated loads. Individuals with poor posture may show a dual concentrated risk area in both the cervical and lumbar spine, exhibiting both forward neck extension during upper limb work and lumbar compression from prolonged sitting. The concentration parameter is defined as the ratio of the number of concentrated areas to the total number of partitions. This ratio reflects the spatial diffusion range of risk; a smaller value indicates fewer concentrated risk areas, while a larger value indicates the risk has spread to more partitions. The lowest regional concentration parameter value is when the number of risk-concentrated areas is a single partition. A low regional concentration parameter indicates that the risk is mainly focused on a local area rather than spreading throughout the spine. In this case, intervention can concentrate resources on intensive training for specific segments. The regional concentration parameter ranges from 0 to 1, where 0 indicates no risk-concentrated areas and 1 indicates that all partitions are risk-concentrated areas. The regional concentration parameter is divided into three concentration levels based on its value range: less than 0.35 is the single-point concentration level, 0.35 to 0.65 is the multi-point dispersion level, and greater than 0.65 is the whole-spine diffusion level. The regional concentration parameter, together with the spatial location information of the risk-concentrated areas, describes the distribution characteristics of the risk. The regional concentration parameter quantifies the diffusion range, and the location information locates specific areas. The single-partition concentration pattern of risk-concentrated areas allows for targeted treatment strategies in clinical intervention, and rehabilitation programs can be designed around a single segment for targeted movement training and posture correction.

[0072] Continuous risk segments were identified by determining the continuity of risk values ​​for each spinal region. The risk values ​​for each region were arranged in anatomical order according to the cervical, thoracic, and lumbar vertebrae, with the elements in the sequence arranged from top to bottom corresponding to the natural arrangement of the spine. The determination of continuous risk segments checked whether the risk values ​​of adjacent regions in each region's risk value sequence all exceeded a continuity threshold. The continuity threshold was set based on the boundary score between medium and high risk levels in the risk grading parameters; adjacent regions exceeding this threshold constituted continuous risk segments. The existence of continuous risk segments indicated that the risk was extending longitudinally along the spine, commonly seen in the gradual spread of degenerative diseases or a chain reaction caused by postural compensation. Adjacent pairs of cervical and thoracic vertebrae in each region's risk value sequence did not meet the continuity condition because the thoracic vertebra risk value was below the continuity threshold. Adjacent pairs of thoracic and lumbar vertebrae in each region's risk value sequence also did not meet the continuity condition, as the thoracic vertebra risk value did not reach the continuity threshold requirement. The results of continuous risk segment identification showed that there were no continuous risk segments with a length greater than or equal to 2 in each region's risk value sequence, and all high-risk regions exhibited isolated distribution characteristics. The isolated distribution of risk values ​​across different zones indicates that high-risk areas are isolated by low-risk zones. This isolation reduces the likelihood of chain transmission of risk, suggesting that individual spinal problems have not yet formed a global pathological continuum. The statistical results for the length of continuous risk segments show 0 segments of length 2 and 0 segments of length 3, indicating that no valid continuous risk segments exist. The determination results for continuous risk segments are stored in a list format, including the start and end zones and segment length. An empty list indicates the absence of continuous risk segments.

[0073] Distribution pattern identifiers are generated by matching the regional concentration parameter with the execution mode of continuous risk segments. The regional concentration parameter and continuous risk segment features are combined for pattern matching in a distribution pattern knowledge base containing typical spine risk distribution patterns. Each pattern type in the knowledge base corresponds to a specific distribution pattern identifier. The distribution pattern knowledge base defines four categories of pattern types: isolated, concentrated, continuous, and skipping. The category division is based on two dimensions: regional concentration parameter and continuous risk segments. Each of the four categories corresponds to one of four candidate distribution pattern identifiers. Pattern matching uses a rule-based elimination method to determine the final distribution pattern identifier category. First, it checks whether the regional concentration parameter exceeds the threshold for concentrated patterns; if the regional concentration parameter is below this threshold, concentrated patterns are excluded. Then, it checks whether there are valid continuous risk segments; if the continuous risk segment length is 0, continuous patterns are excluded. The definition characteristics of an isolated pattern in the distribution pattern knowledge base are a low regional concentration parameter, no continuous risk segments, and a risk concentration area of ​​1. The isolated pattern perfectly matches the current regional concentration parameter and continuous risk segment features. The defining characteristic of a skip pattern is that the number of risk concentration areas is greater than or equal to 2, and there are low-risk partitions between each risk concentration area. In the current case, only a single risk concentration area does not meet the skip pattern condition. The distribution pattern identifier is stored as a string representing the name of the matched pattern type. The current distribution pattern identifier is "isolated". The name of the pattern type after a successful match is used as the final value of the distribution pattern identifier.

[0074] The risk distribution type is determined based on the distribution pattern identifier. The "isolated" distribution pattern identifier corresponds to a specific risk management strategy in the risk distribution type classification system, which maps the distribution pattern identifier to intervention program types. The determined risk distribution type is "isolated high-risk," indicating that the risk is mainly concentrated in a single area and is discontinuous with other areas. This distribution pattern allows rehabilitation resources to be concentrated on intensive interventions in a single segment. The clinical significance of the "isolated high-risk" risk distribution type lies in indicating the need for targeted interventions for specific areas. Intervention measures can be designed around specific training movements for that area without considering the interconnected effects of other segments. The mapping relationship between the distribution pattern identifier and the risk distribution type is stored in the mapping table of the classification system, which contains three related fields: distribution pattern identifier, risk distribution type, and intervention recommendation. The process of determining the risk distribution type combines the absolute value of the risk value of each area and the spatial location of the risk concentration area to comprehensively determine the type classification result. In the current case, the risk distribution type "isolated high-risk type" specifically refers to a single point of high-risk distribution in the lumbar spine region. The distribution pattern identifier "isolated type" and the risk distribution type "isolated high-risk type" describe the distribution characteristics from the perspectives of spatial distribution pattern and risk severity, respectively.

[0075] Step S150: Based on the risk distribution type and the risk concentration area, the partition distribution characteristics are generated by association and integration. The spinal health risk assessment results are output by combining the partition distribution characteristics and the risk level.

[0076] Specifically, zonal distribution features are generated by associating and integrating risk distribution types and risk concentration areas. The type identifier of the risk distribution type "isolated high-risk type" is associated and combined with the lumbar spine location information of the risk concentration area. This combination operation integrates distribution pattern features and spatial location features into a unified descriptive structure. The type identifier of the risk distribution type provides information on the risk distribution pattern, while the lumbar spine location information of the risk concentration area provides specific location information of the risk. Zonal distribution features integrate information from both the risk distribution type and the risk concentration area, resulting in a composite feature description of "isolated high-risk lumbar spine distribution." The risk distribution type "isolated high-risk type" may correspond to different risk concentration area locations in different individuals. Lumbar spine concentration is common among office workers, cervical spine concentration is common among teachers, and lumbar spine concentration is common among drivers. Different locations of risk concentration areas correspond to different rehabilitation training focuses and daily protection points. The composite description format of zonal distribution features is a combination of "location + distribution type". The quantity information of risk concentration areas is incorporated into the description of zonal distribution features through concentration levels. The regional concentration parameter of the current case corresponds to the 'single-point concentration' level. The zonal distribution characteristics are stored in the form of structured text, which contains three descriptive elements: the location identifier of the risk concentration area, the distribution pattern of the risk distribution type, and the concentration level. These three elements fully depict the distribution characteristics of the risk in the spine space.

[0077] The spinal health risk assessment results are output by combining zonal distribution characteristics and risk levels. The zonal distribution characteristic "isolated high-risk distribution of the lumbar spine" is combined with the risk level "high risk" to form a comprehensive assessment conclusion, which includes information on both the degree of risk and the spatial distribution of risk. The generation of spinal health risk assessment results uses a template-filling method. The template defines a standardized assessment report format and content structure, and includes five core sections: risk level, risk concentration area, distribution type, main risk factors, and intervention recommendations. The risk level section of the spinal health risk assessment results is filled with the risk level label "high risk," which is marked as the highest warning level. The risk levels "medium risk" and "low risk" are marked as secondary and normal levels, respectively. The risk concentration area section of the spinal health risk assessment results is filled with the location label "lumbar spine," which is highlighted in a color block in the segment area corresponding to the risk concentration area using a spinal anatomy diagram. The zonal distribution characteristics are converted into a readable description in the distribution type section. The main risk factors section of the spinal health risk assessment results lists the key risk factor set, with the four key risk factor items arranged from high to low risk level. Each factor item is accompanied by a corresponding intervention tagged category. The intervention recommendations section generates targeted recommendations based on risk level and regional distribution characteristics. When the risk level is "high risk," a time window based on the urgency level is used to set the priority initiation period. When the risk concentration area is the lumbar spine, rehabilitation training focuses on reducing lumbar spine load and strengthening the lumbar core muscles. The spinal health risk assessment results are output in a structured document format, containing both text descriptions and data charts. The data charts use bar charts to display the risk values ​​of each region in the cervical, thoracic, and lumbar spine.

[0078] To implement the individual-characteristic-based spinal health risk assessment method corresponding to the above-described method embodiments, and to achieve the corresponding functional and technical effects, refer to Figure 2. Figure 2 shows a structural block diagram of an individual-characteristic-based spinal health risk assessment system 200 provided in this application embodiment. For ease of explanation, only the parts relevant to this embodiment are shown. The individual-characteristic-based spinal health risk assessment system 200 provided in this application embodiment includes:

[0079] The feature acquisition module 201 is used to acquire basic feature data and spinal morphology measurement data of an individual, and to construct an individual feature profile based on the basic feature data and the spinal morphology measurement data.

[0080] Risk analysis module 202 is used to extract risk factors from the individual feature profile to form a candidate risk factor set, perform multi-factor coupling analysis on the candidate risk factor set to establish a risk factor association map, and perform interventionability assessment on the candidate risk factor set to generate interventionability labels.

[0081] Risk classification module 203 is used to generate risk classification parameters based on the risk factor association map and the interventionability label, determine the risk level according to the risk classification parameters, and extract high-weight risk factors from the risk classification parameters to generate a set of key risk factors.

[0082] The zoning assessment module 204 is used to classify the key risk factor set into risk categories according to the spine zoning to obtain the risk value of each zone, perform difference analysis on the risk values ​​of each zone to identify risk concentration areas, and identify the risk distribution type based on the distribution pattern of the risk concentration areas and the risk values ​​of each zone.

[0083] The result output module 205 is used to generate zonal distribution features by associating and integrating the risk distribution type with the risk concentration area, and output the spinal health risk assessment result by combining the zonal distribution features with the risk level.

[0084] The aforementioned individual-characteristic-based spinal health risk assessment system 200 can implement one of the individual-characteristic-based spinal health risk assessment methods described in the above-described method embodiments. The options in the above method embodiments are also applicable to this embodiment and will not be detailed here. The remaining content of this application's embodiments can be referred to the content of the above method embodiments, and will not be repeated in this embodiment.

[0085] The purpose of the above embodiments is to reproduce and derive the technical solution of the present invention by way of example, and to fully describe the technical solution, purpose and effect of the present invention. The purpose is to enable the public to have a more thorough and comprehensive understanding of the disclosure of the present invention, and not to limit the scope of protection of the present invention.

[0086] The above embodiments are not an exhaustive list based on the present invention, and there may be many other embodiments not listed. Any substitutions and improvements made without departing from the concept of the present invention are within the protection scope of the present invention.

Claims

1. A method for assessing spinal health risk based on individual characteristics, characterized in that, include: Acquire basic characteristic data and spinal morphology measurement data of an individual. The basic characteristic data includes the individual's age, occupation type, years of work experience, body mass index (BMI), and exercise frequency. Construct an individual characteristic profile based on the basic characteristic data and the spinal morphology measurement data, including: extracting occupational information from the basic characteristic data to generate occupational type labels; performing occupational posture pattern matching on the occupational type labels to generate typical posture features; calculating the deviation of the spinal morphology measurement data based on the typical posture features to obtain posture deviation parameters; and fusing the posture deviation parameters with the basic characteristic data to construct the individual characteristic profile. The individual characteristic profile includes four evaluation dimensions: occupational posture risk index, spinal curvature deviation index, center of gravity stability index, and movement compensation index. The occupational posture risk index is calculated by combining the years of work experience from the basic characteristic data and the deviation vector of the posture deviation parameters. The spinal curvature deviation index is obtained through the weighted norm of the deviation vector of the posture deviation parameters. The center of gravity stability index is calculated based on the body mass index (BMI) from the basic characteristic data and the center of gravity offset of the posture deviation parameters. The movement compensation index is calculated based on the exercise frequency and age from the basic characteristic data. The process involves: extracting risk factors to form a candidate risk factor set; performing multi-factor coupling analysis on the candidate risk factor set to establish a risk factor association map; identifying causal relationships in the candidate risk factor set to obtain primary and secondary risk factors; constructing causal association chains based on the primary and secondary risk factors; evaluating the association strength of the causal association chains to form association path weights; establishing a risk factor association map based on the association path weights; performing an interventionizability assessment on the candidate risk factor set to generate interventionizability labels; and conducting risk factor association analysis based on the risk factor association map and the interventionizability labels. Risk grading parameters are generated based on risk grading parameters. Risk levels are determined based on these parameters. High-weight risk factors are extracted from the risk grading parameters to generate a set of key risk factors. The set of key risk factors is then categorized by spinal region to obtain risk values ​​for each region. Difference analysis is performed on the risk values ​​of each region to identify risk concentration areas. Distribution patterns are identified based on the risk concentration areas and the risk values ​​of each region to obtain risk distribution types. Based on the correlation and integration of the risk distribution types and the risk concentration areas, regional distribution features are generated. Finally, the spinal health risk assessment results are output by combining the regional distribution features with the risk levels.

2. The method according to claim 1, characterized in that, The step of performing an interventionability assessment on the candidate risk factor set to generate interventionability labels includes: identifying factor types in the candidate risk factor set to obtain factor classification results; dividing the interventionability levels based on the factor classification results to form an interventionability grading table; performing an improvement difficulty assessment on the interventionability grading table to generate an improvement difficulty coefficient; and marking the candidate risk factor set with interventionability labels according to the improvement difficulty coefficient to generate interventionability labels.

3. The method according to claim 1, characterized in that, The step of generating risk classification parameters based on the risk factor association map and the interventionlability label includes: extracting high-weight factors from the risk factor association map to obtain a core factor set; identifying synergistic relationships in the core factor set based on the interventionlability label to generate a synergistic factor group; performing a synergistic enhancement assessment on the synergistic factor group to form a synergistic enhancement coefficient; and adjusting the core factor set using the synergistic enhancement coefficient to generate risk classification parameters.

4. The method according to claim 1, characterized in that, The step of classifying the set of key risk factors by spine partitions to obtain risk values ​​for each partition includes: performing a partition impact degree analysis on the set of key risk factors to obtain a partition sensitivity coefficient; adjusting the partition weights of the set of key risk factors based on the partition sensitivity coefficients to form adjusted partition weights; accumulating the risk values ​​of each spine partition according to the adjusted partition weights to generate an original cumulative value of partition risk; and determining the risk saturation of the original cumulative value of partition risk to determine the risk value of each partition.

5. The method according to claim 1, characterized in that, The step of identifying the risk distribution type based on the distribution pattern of the risk concentration area and the risk values ​​of each partition includes: extracting the number of concentration areas based on the risk concentration area to obtain a regional concentration parameter; determining the continuity of the risk values ​​of each partition to identify continuous risk segments; generating a distribution pattern identifier by performing pattern matching between the regional concentration parameter and the continuous risk segments; and determining the risk distribution type based on the distribution pattern identifier.

6. The method according to claim 1, characterized in that, The step of constructing a causal chain based on the primary risk factor and the secondary risk factor includes: identifying the causal direction of the primary risk factor and the secondary risk factor to obtain the factor-initiating direction; distinguishing between dominant factors and subordinate factors based on the factor-initiating direction; performing a correlation strength assessment on the dominant factor and the subordinate factor to form a causal path strength parameter; and constructing a causal chain based on the causal path strength parameter.

7. The method according to claim 3, characterized in that, The step of performing a synergistic enhancement assessment on the synergistic factor group to form a synergistic enhancement coefficient includes: assessing the deterioration rate of irreversible factors in the synergistic factor group to obtain a deterioration process coefficient, wherein the irreversible factors refer to factors with irreversible attributes identified by analyzing secondary factors jointly caused by various factors in the synergistic factor group and determining reversibility based on factor type and physiological degeneration characteristics; classifying urgency based on the deterioration process coefficient to form an urgency level; performing a risk acceleration assessment on the urgency level and the synergistic factor group to generate a risk acceleration coefficient; and determining the synergistic enhancement coefficient based on the risk acceleration coefficient.

8. A spinal health risk assessment system based on individual characteristics, characterized in that, include: The feature acquisition module is used to acquire basic feature data and spinal morphology measurement data of an individual. The basic feature data includes the individual's age, occupation type, years of work experience, body mass index (BMI), and exercise frequency. Based on the basic feature data and the spinal morphology measurement data, an individual feature profile is constructed, including: extracting occupational information from the basic feature data to generate occupational type labels; performing occupational posture pattern matching on the occupational type labels to generate typical posture features; calculating the deviation of the spinal morphology measurement data based on the typical posture features to obtain posture deviation parameters; and fusing the posture deviation parameters with the basic feature data to construct the individual feature profile. The individual profile includes four evaluation dimensions: occupational posture risk index, spinal curvature deviation index, center of gravity stability index, and movement compensation index. The occupational posture risk index is calculated by combining the years of work experience in the basic characteristic data and the deviation vector of the posture deviation parameters. The spinal curvature deviation index is obtained by weighting the norm of the posture deviation parameter deviation vector. The center of gravity stability index is calculated based on the body mass index (BMI) in the basic characteristic data and the center of gravity offset of the posture deviation parameters. The movement compensation index is calculated based on the exercise frequency and age in the basic characteristic data. A risk analysis module is used to perform [analysis] on the individual profile. Risk factors are extracted to form a candidate risk factor set. Multi-factor coupling analysis is then performed on the candidate risk factor set to establish a risk factor association map. This includes: identifying causal relationships in the candidate risk factor set to obtain primary and secondary risk factors; constructing causal association chains based on the primary and secondary risk factors; performing association strength assessment on the causal association chains to form association path weights; establishing a risk factor association map based on the association path weights; performing an interventionizability assessment on the candidate risk factor set to generate interventionizability labels; and a risk grading module for performing risk grading based on the risk factor association map and the interventionizability labels to generate risk levels. The system comprises a risk grading parameter module, which determines the risk level based on the risk grading parameter and extracts high-weight risk factors from the risk grading parameter to generate a key risk factor set; a zonal assessment module, which classifies the key risk factor set according to spinal zonal regions to obtain risk values ​​for each region, performs differential analysis on the risk values ​​of each region to identify risk concentration areas, and identifies the risk distribution type by recognizing the distribution pattern between the risk concentration areas and the risk values ​​of each region; and a result output module, which integrates the risk distribution type with the risk concentration areas to generate zonal distribution features, and outputs the spinal health risk assessment result by combining the zonal distribution features with the risk level.

Images