A method for assessing physiological age of knee joint muscle function
By constructing a multidimensional phenotypic parameter database and model, the problem of the singularity of traditional assessment methods is solved, enabling accurate assessment and visual representation of individual functional decline, and supporting personalized management of community populations.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- HEBEI UNIV OF TECH
- Filing Date
- 2026-04-28
- Publication Date
- 2026-07-24
AI Technical Summary
Existing technologies lack comprehensive assessment methods based on multidimensional information when evaluating individual functional decline, especially in community populations. It is difficult to achieve standardized comparisons and quantify the degree of individual deviation, and the results of traditional single-indicator assessments are unstable and insufficiently representative.
A physiological age assessment method for knee muscle function was developed. By collecting multidimensional phenotypic parameters such as body composition, knee muscle function, cognitive function, and lifestyle habits, a community sample database was established, significantly relevant parameters were screened, and a generalized linear model with an elastic network penalty term was used for modeling. Individual deviations were represented by multi-domain standardized scores and radar charts.
It enables a multi-dimensional and comprehensive characterization of an individual's functional aging state, improves the accuracy and reliability of the assessment, provides intuitive physiological age assessment results and visualized expressions of individual deviations, and supports early identification and stratified management of community populations.
Smart Images
Figure CN122455338A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the fields of biomedical engineering and health assessment technology, and in particular to a method for assessing and characterizing the physiological age of knee joint muscle function. Background Technology
[0002] Against the backdrop of an accelerating aging population, how to objectively and quantitatively assess the degree of functional decline in individuals has gradually become a key issue in primary public health services and geriatric research. Traditionally, chronological age based on birth year is insufficient to accurately reflect the body's true physiological state, while "physiological age," which comprehensively characterizes the body's functional level, has gained attention because it is closer to actual health conditions. Among numerous functional indicators, the function of knee-related muscles directly relates to lower limb motor ability, postural control, and the ability to perform daily activities. These muscles may show changes such as decreased muscle strength, reduced flexion-extension coordination, and widening functional differences between the left and right sides in the early stages of aging, thus being considered an important window indicator reflecting functional aging.
[0003] Research has found that the aging process in an individual is not a change in a single system, but rather involves the coordinated evolution of multiple physiological and behavioral dimensions. For example, factors such as abnormal body composition (e.g., an imbalance between fat and muscle), cognitive decline, and unhealthy lifestyle habits often occur simultaneously with muscle function decline and interact with each other, jointly driving the process of functional decline. This multi-factor intertwined characteristic indicates that relying on a single indicator is insufficient to fully reveal an individual's true level of aging.
[0004] Existing technologies for assessing physiological age or functional aging can be broadly categorized into two approaches: one is based on a single functional parameter or phenotypic indicator, which offers advantages such as simplicity and ease of data acquisition. However, due to significant individual differences and the complexity of aging manifestations, the assessment results often lack stability and representativeness. The other approach integrates multiple phenotypic information to construct a comprehensive assessment model, characterizing individual aging features from multiple dimensions. This approach can improve the accuracy and reliability of the assessment to some extent and has thus gradually become the main direction of research development. Existing research has explored functional aging from different dimensions. For example, some studies have shown a close relationship between knee extensor strength and individual weakness, serving as a sensitive indicator of early functional decline; others have indicated a correlation between knee muscle strength and emotional state, suggesting that muscle function may reflect a broader range of physical and mental health conditions. Furthermore, research on the relationship between body composition and cognitive function shows a certain correlation between the two, influenced by factors such as age and gender; further research has also found that the coexistence of muscle loss and fat gain may be statistically associated with cognitive impairment.
[0005] While the aforementioned studies have revealed some characteristics of functional aging from different perspectives, most remain limited to single-dimensional analysis or simple correlational discussions, lacking a comprehensive assessment method that integrates multi-source phenotypic information and conducts systematic modeling with key functional indicators as the core. Particularly in community-based health management scenarios, there is a lack of technical means to standardize and compare individual functional states based on large-scale population norm data, and further quantify the degree of deviation. Therefore, there is an urgent need to construct a technical solution that uses knee joint muscle function as the core, combines multi-dimensional information such as body composition, cognitive function, and lifestyle habits, and relies on a community norm database to achieve physiological age assessment and individual difference characterization, in order to meet the practical application needs of early screening, stratified intervention, and dynamic monitoring. Summary of the Invention
[0006] The technical problem to be solved by the present invention is to overcome the shortcomings of the prior art and provide a method for assessing and characterizing the physiological age of knee joint muscle function.
[0007] This invention is achieved through the following technical solution: A method for assessing and characterizing the physiological age of knee joint muscle function includes the following steps: S1. Collect phenotypic parameters of the subjects and establish a community sample database. The phenotypic parameters include parameters from four categories: body composition, knee joint muscle function, cognitive function, and lifestyle habits. S2. Perform correlation analysis between each of the phenotypic parameters and actual age, and screen candidate phenotypic parameters that are significantly related to age; S3. Based on the candidate phenotypic parameters, establish a physiological age model of knee joint muscle function; S4. Individual deviations are characterized by calculating multi-domain standardized scores and constructing radar charts.
[0008] According to the above technical solution, preferably, in step S1, the community sample database stores the sample size, mean, standard deviation, and preset percentile statistics for each phenotypic parameter. The database fields in the community sample database include stratification identifier, indicator name, sample size, mean, standard deviation, percentile parameter, and data version number.
[0009] According to the above technical solution, preferably, step S2 includes: Missing values and outliers were identified and processed in the data of each phenotypic parameter collected from the subjects, forming a two-dimensional phenotypic parameter matrix of individual and measurement parameter with individuals as rows and measurement parameters as columns; Age-related analyses were performed on each phenotypic parameter, and the correlation coefficient and significance level between each phenotypic parameter and actual age were calculated. Based on the correlation coefficient and significance level results, candidate phenotypic parameters that are significantly related to age are screened.
[0010] According to the above technical solution, preferably, in step S2, the candidate phenotypic parameters include: Height, left upper limb fat-free body mass, left lower limb fat-free body mass, right upper limb fat-free body mass, right lower limb fat-free body mass, waist-to-hip ratio, skeletal muscle mass, fat mass index, fat to muscle mass ratio, body fat percentage, body mass index, percentage of skeletal muscle mass, visceral fat area, relative peak knee extensor strength, relative peak knee flexor strength, hamstring to quadriceps torque ratio, bilateral quadriceps strength difference, bilateral hamstring strength difference, subjective cognitive function decline score, regular exercise habits.
[0011] According to the above technical solution, preferably, in step S3, the expression of the physiological age model of knee joint muscle function is: , Where y represents the subject's physiological age of knee muscle function, x i Let a be the i-th age-related indicator included in the model. i is the regression coefficient corresponding to this indicator, and b is the model intercept.
[0012] According to the above technical solution, preferably, step S3 includes: Based on the candidate phenotypic parameters, a target phenotypic parameter matrix for modeling is constructed, and the target phenotypic parameter matrix is randomly split into a training set and a test set; Using actual age as the training label, a generalized linear modeling method with elastic network penalty terms was used to establish a physiological age model of knee joint muscle function. For each candidate alpha, cross-validation is performed on the training set, and the corresponding optimal regularization parameter lambda is searched. The alpha value with the smallest error and its corresponding optimal lambda value are selected as the final model parameters. Based on the complete training set, the final physiological age model of knee joint muscle function was refitted, and the regression coefficients α corresponding to each input index were obtained. i And the model intercept b.
[0013] According to the above technical solution, preferably, step S4 includes: Based on the gender and age information of the individual to be evaluated, the data is matched to the corresponding age-gender stratified norm unit in the community sample database; For the j-th indicator, based on the original observed value x of the indicator... j Based on the distribution information of the index and its corresponding norm unit, calculate the relative quantile position r corresponding to the j-th index. jAnd convert it into a standardized score S j ; Calculate the comprehensive index D for each field k Construct a multi-domain radar map of the individual to be evaluated.
[0014] The beneficial effects of this invention are: The proposed method for assessing physiological age of knee joint muscle function, by constructing a community sample database and incorporating multidimensional phenotypic parameters such as body composition, knee joint muscle function, cognitive function, and lifestyle habits, achieves a comprehensive characterization of an individual's functional aging state, overcoming the problem that traditional single-indicator assessment methods cannot fully reflect the body's true functional level. By conducting correlation analysis between various phenotypic parameters and actual age and screening candidate indicators, the method improves the specificity and effectiveness of modeling variables while ensuring model interpretability, avoiding interference from redundant information in the assessment results. During the modeling process, a generalized linear model with an elastic network penalty term is introduced, and cross-validation is used to optimize model parameters. This not only improves the model's stability and generalization ability but also effectively alleviates the problem of collinearity among multiple indicators, thereby improving the accuracy and reliability of physiological age estimation.
[0015] Meanwhile, this invention performs hierarchical matching and standardization of individual indicators based on a community norm database, and visualizes individual deviations through multi-domain standardized scores and radar charts. This allows the assessment results to not only output an intuitive physiological age but also reveal the strengths and weaknesses of different functional dimensions, providing a more refined and operable technical means for early identification, hierarchical management, and individualized intervention in community populations. Overall, this application demonstrates significant improvements in the integration of assessment dimensions, the scientific rigor of model construction, and the intuitiveness of result presentation, possessing considerable application and promotion value. Attached Figure Description
[0016] Figure 1 This is a schematic diagram of the technical route of the present invention.
[0017] Figure 2 This is a schematic diagram of the correlation analysis between various phenotypic parameters and actual age in this invention. The significance level of the correlation is marked by an asterisk (*p<0.05, **p<0.01, ***p<0.001) and indicated in the corresponding cell. The correlation range is from negative to positive.
[0018] Figure 3 This is a schematic diagram of the construction and evaluation of the physiological age model of knee joint muscle function in this invention, where R is the Spearman correlation coefficient and MAE is the mean absolute error.
[0019] Figure 4 This is the individual deviation characterization radar chart in this invention. Detailed Implementation
[0020] To enable those skilled in the art to better understand the technical solutions of the present invention, the present invention will be further described in detail below with reference to the accompanying drawings and preferred embodiments. All other embodiments obtained by those skilled in the art based on the embodiments of the invention without inventive effort are within the scope of protection of the invention.
[0021] As shown in the figure, the present invention includes the following steps: Step S1. Collect phenotypic parameters of the subjects and establish a community sample database. The phenotypic parameters include parameters from four categories: body composition, knee joint muscle function, cognitive function, and lifestyle habits.
[0022] First, basic information of potential volunteers was collected based on the pre-set inclusion and exclusion criteria, and eligible subjects were selected for inclusion in the research cohort, providing a basic sample for subsequent collection of multi-domain phenotypic parameters and model construction.
[0023] The inclusion criteria included: (1) being 18 years of age or older; (2) having no significant intellectual or cognitive impairment; and (3) voluntarily participating in the study and signing an informed consent form. The exclusion criteria included: (1) abnormal body fluid status, such as edema or dehydration, or the presence of metal implants or external devices that may interfere with bioelectrical impedance analysis; (2) a history of musculoskeletal injury or surgery within the past 6 months; (3) joint pain during muscle strength measurement, or recent muscle strength training that has not yet been completed, making it impossible to complete the full muscle strength test; (4) the presence of a serious chronic disease that may affect musculoskeletal function; and (5) poor compliance, making it impossible to complete the questionnaire.
[0024] Then, body composition testing, knee muscle function testing, cognitive function testing, and lifestyle information were collected from the subjects to obtain multi-domain phenotypic parameters, as shown in Table 1.
[0025] Table 1 ; Specifically, in body composition testing, bioelectrical impedance analysis is used to measure the subject's body composition. During measurement, the subject is required to wear light clothing and stand barefoot on the electrode plate, with both feet completely covering the electrodes and shoulder-width apart, while holding the electrodes with both hands in natural contact. The subject is required to remain still during the measurement, which lasts for several tens of seconds to obtain the corresponding body composition parameters.
[0026] In the knee joint muscle function test, an isokinetic testing system was used to assess the strength of the knee flexor and extensor muscles in the subjects' lower limbs. During the test, the subjects sat on the equipment seat, adjusting the seat to align the knee joint axis with the equipment axis; the thigh, lower leg, and torso were secured with straps, while the non-tested limbs were placed in a stable position. The test preferably used an isokinetic concentric mode, with an angular velocity set at 30° / s, repeated 3 times per set, with a 10-second rest between sets. During the test, the subjects exerted their best effort to complete the movement under resistance, and the system simultaneously recorded peak torque and joint angle data.
[0027] In the cognitive function test, the 9-item Subjective Cognitive Decline Rapid Screening Scale (SCD-9) was used to assess the cognitive function of the subjects. The scale scores ranged from 0 to 9, with higher scores indicating poorer cognitive function.
[0028] In the lifestyle habit data collection, information on participants' drinking, smoking, and exercise habits was collected through face-to-face questionnaires and defined as binary variables. Among them: Drinking is defined as follows, according to WHO standards: less than 140g of pure alcohol per week for men and less than 70g for women is defined as low-risk drinking, and all others are defined as high-risk drinking. Smoking is defined as: those who currently engage in smoking behavior, including daily smokers and occasional smokers, are categorized as smokers; those who do not currently engage in smoking behavior are categorized as non-smokers. Regular exercise is defined as: engaging in at least 150 minutes of moderate-intensity or higher-intensity physical activity per week.
[0029] Subsequently, a community sample database is established based on the collected multi-domain phenotypic parameters of the subjects. This database serves as the foundation for subsequent age-related parameter screening, the establishment of a physiological age model for knee muscle function, and the calculation of individual standardized scores. The database stores the sample size, mean, standard deviation, and preset percentile statistics for each indicator. Preferably, the preset percentile statistics include P25, P50, and P75; the database fields include at least the stratification identifier, indicator name, sample size, mean, standard deviation, percentile parameters, and data version number.
[0030] In terms of assessment dimensions, this application is no longer limited to a single physiological or functional indicator. Instead, it collects multiple types of phenotypic parameters, such as body composition, knee muscle function, cognitive function, and lifestyle habits, and constructs a community sample database to achieve a multi-dimensional comprehensive characterization of an individual's functional aging state. This multi-source information fusion approach can more comprehensively reflect the changes in the body at different system levels, effectively overcoming the one-sidedness caused by the single indicator in traditional methods, and making the assessment results closer to the individual's true physiological state.
[0031] Step S2. Perform correlation analysis between each phenotypic parameter and actual age, and screen candidate phenotypic parameters that are significantly related to age.
[0032] Specifically, missing and outlier values were identified and processed in the data of each phenotypic parameter collected from the subjects, forming a two-dimensional phenotypic parameter matrix with individuals as rows and measurement parameters as columns. In this matrix, each row corresponds to a subject, and each column corresponds to a phenotypic parameter.
[0033] Age-related analysis was performed on each phenotypic parameter. In this example, Spearman correlation analysis was preferred. The correlation coefficient and significance level between each phenotypic parameter and the actual age were calculated, and the parameters were distinguished as positively correlated with age and negatively correlated with age based on the sign of the correlation coefficient.
[0034] Based on the correlation coefficient and significance level results, only phenotypic parameters with a significance level less than 0.05 were retained as candidate phenotypic parameters significantly related to age. The candidate phenotypic parameters significantly related to age were selected as follows: age-related characteristics were selected from height, left upper limb fat-free body mass (LA-FFM), left lower limb fat-free body mass (RA-FFM), right upper limb fat-free body mass (LL-FFM), right lower limb fat-free body mass (RL-FFM), waist-to-hip ratio, skeletal muscle mass (SMM), fat mass index (FMI), fat to muscle mass ratio (FMR), body fat percentage (Fat%), body mass index (BMI), skeletal muscle mass percentage (SMM%), visceral fat area (VFA), relative knee extensor peak strength (KES), relative knee flexor peak strength (KFS), hamstring to quadriceps torque ratio (H / Q ratio), bilateral quadriceps strength difference (BHD), bilateral hamstring strength difference (BQD), subjective cognitive decline score (SCD-9), and regular exercise habits. Among them, waist-to-hip ratio, lean body mass of the left upper limb, lean body mass of the right upper limb, FMI, FMR, Fat%, BMI, VFA, H / Q ratio, and SCD-9 were significantly positively correlated with age (P < 0.05); height, SMM, lean body mass of the left lower limb, lean body mass of the right lower limb, SMM%, KES, KFS, BHD, BQD, and regular exercise were significantly negatively correlated with age (P < 0.05). Since the lean body mass of the corresponding segments on the left and right sides belong to the same type of indicator in the same anatomical region and have similar representational meanings, in order to avoid the influence of redundant information caused by the simultaneous inclusion of indicators from both sides in the model, and to maintain consistency with the calculation method of the knee joint muscle strength index which uses the dominant side, the lean body mass of the dominant side was selected as the representative indicator for inclusion in the model in this case. The dominant side was defined as the side with the larger lean body mass value between the two sides.
[0035] Regarding the feature selection mechanism, this application conducts correlation analysis between each phenotypic parameter and actual age, and selects candidate indicators based on significance levels, ensuring from the outset that the variables entering the model have clear age-related relationships and biological significance. Compared to directly selecting empirical indicators or modeling with all variables, this method can reduce the introduction of redundant information and noise interference, thereby improving the simplicity and interpretability of the model structure.
[0036] Step S3. Based on the candidate phenotypic parameters, establish a physiological age model of knee joint muscle function.
[0037] Specifically, a target phenotypic parameter matrix for modeling is constructed based on the candidate phenotypic parameters. The target phenotypic parameter matrix is then randomly split into a training set and a test set in a 1:1 ratio. The training set is used for model training and parameter optimization, while the test set is used for model performance verification.
[0038] Using actual age as the training label, a generalized linear modeling method with elastic network penalty terms is used to establish a physiological age model of knee joint muscle function. In this example, the model can reduce the coefficients of some indicators with high collinearity or low contribution to 0 by the penalty term constraint, thereby improving the stability of the model.
[0039] The candidate set of the preset mixing parameter alpha is 0.1, 0.2, 0.3, 0.4, 0.5, 0.6, 0.7, 0.8, and 0.9. For each candidate alpha, cross-validation is performed on the training set, and the corresponding optimal regularization parameter lambda is searched. Preferably, the minimum cross-validation error is used as the criterion for model parameter selection. After comparing the cross-validation errors of the models corresponding to different alpha values, the alpha value with the smallest error and its corresponding optimal lambda value are selected as the final model parameters. In this example, standardization, cross-validation, and elastic network modeling are preferably implemented using scikit-learn (version 1.7.0).
[0040] After determining the optimal alpha and lambda, the final physiological age model of knee muscle function is refitted based on the complete training set, thereby obtaining the regression coefficients ai and model intercepts b corresponding to each input index. The physiological age y of the subject's knee muscle function is calculated using the following formula:
[0041] Where y represents the subject's physiological age of knee muscle function, x i Let a be the i-th age-related indicator included in the model. i is the regression coefficient corresponding to this indicator, and b is the model intercept.
[0042] In this embodiment, independent models were built for the male and female groups respectively. The optimal alpha value for the male group was 0.9, corresponding to an optimal regularization parameter lambda of 0.037; the optimal alpha value for the female group was 0.9, corresponding to an optimal regularization parameter lambda of 0.012. Subsequently, based on their respective optimal parameters, the final physiological age models of knee joint muscle function for the male and female groups were reconstructed on the complete training set. The model parameter results are shown in Table 2, which lists the regression coefficient α corresponding to each input index. i And the model intercept b.
[0043] Table 2 ; In terms of model construction methods, this application adopts a generalized linear modeling method with an elastic network penalty term, and optimizes the selection of regularization parameters through cross-validation. This allows the model to balance sparsity and stability in the process of variable selection and parameter estimation. On the one hand, it can effectively deal with the possible collinearity problem among multiple indicators, and on the other hand, it improves the model's generalization ability on different sample data, thereby significantly improving the accuracy and robustness of physiological age prediction results.
[0044] The final physiological age model of knee joint muscle function was constructed and applied to the test set to validate the model's predictive performance. Figure 3 In this example, the mean absolute error was 6.322 for the male group and 5.643 for the female group. The correlation coefficient (Spearman rank correlation) between predicted age and actual age was 0.841 for the male group and 0.863 for the female group (P<0.001), both of which were statistically significant (P<0.001), indicating that the established model has high predictive accuracy and good application value.
[0045] Step S4. Individual deviations are characterized by calculating multi-domain standardized scores and constructing radar charts.
[0046] Taking a 45-year-old woman as an example, her multi-domain indicator information was input, and individual deviations were characterized based on the "women 40-49 years old" age-gender stratified norm unit in the community norm database. The multi-domain indicators preferably include body composition domain indicators, knee joint muscle function domain indicators, and cognitive function domain indicators. Height, as a physical characteristic correction variable, is used to correct relevant functional indicators and is not directly used as a scoring indicator with superiority or inferiority in the calculation of domain standardized scores.
[0047] The specific calculation process includes the following steps: (1) Based on the gender and age information of the individual to be evaluated, the individual is matched to the corresponding age-gender stratified norm unit in the community norm database. In this embodiment, the "female 40-49 years old" norm unit is called. For the j-th indicator, the distribution information of the indicator is extracted from the norm unit to determine the relative quantile position r of the indicator in the norm unit. j .
[0048] For the j-th indicator, record its original observed value as x. j According to the preset scoring rules, the indicators are divided into monotonically positive indicators, monotonically negative indicators, and interval-optimal indicators.
[0049] Among them: monotonic positive indicators are those whose larger values represent better physiological condition, preferably including muscle mass and muscle strength; monotonic negative indicators are those whose smaller values represent better condition, preferably including the Subjective Cognitive Decline Scale (SCD-9); and interval-optimal indicators are those whose values are within a certain optimal range when the condition is optimal, and whose deviations in any direction indicate a decline in condition, preferably including BMI.
[0050] (2) Calculate the relative quantile position r j : Based on the original observed value x of the indicator j Based on the distribution information of the index and its corresponding norm unit, calculate the relative quantile position r corresponding to the j-th index. j , where r j ∈[0,1].
[0051] Where, r j The closer the value is to 1, the closer it is to the high quantile within its corresponding norm unit; r j The closer the value is to 0, the closer it is to the lower quantile in its corresponding norm unit; r j =0.5 indicates that the index is close to the median level of the corresponding norm unit.
[0052] (3) Calculate the standardized score S j : To eliminate the dimensional differences between different indicators and to unify the direction of their merits and demerits, all indicators are uniformly converted into standardized scores S ranging from -100% to +100%. j .
[0053] For monotonically positive indicators, the standardized score S of the j-th indicator is... j Defined as: S j = (2r j - 1) × 100% For a monotonically negative indicator, the standardized score S of the j-th indicator is...j Defined as: S j = (1 - 2r j ) × 100% For interval-optimal indicators, let the quantile range corresponding to the optimal interval be [0.25, 0.75], then the standardized score S of the j-th indicator is... j Determine as follows: When 0.25≤r j When ≤0.75, the definition is: S j = +100% When r j When <0.25, the definition is: S j = [1 - 2(0.25 - r j [0.25] × 100% When rj > 0.75, the definition is: S j = [1 - 2(r j [-0.75) / (1 - 0.75)] × 100% In this embodiment, for BMI, the original value range corresponding to the preferred interval is [18.5, 23.9]. The program preferably first determines the relative quantile position r of the BMI within its corresponding norm unit based on the original BMI observation value. j Then, calculate its standardized score S according to the mapping rules of the above interval optimal index. j .
[0054] After the above transformation, continuous indicators of different directions and dimensions are all uniformly mapped to a standardized score range of -100% to +100%, where: S j A value close to +100% indicates that the indicator is in a relatively good state within its corresponding normative unit; S j =0 indicates that the index is close to the median level of its corresponding norm unit; S j A value close to -100% indicates that the indicator is in a poor state within its corresponding normative unit.
[0055] (4) Computing Domain Comprehensive Index D k : For multiple indicators within the same domain, the program calculates the domain composite index according to the principle of equal weighting. Let the k-th domain contain m... k One indicator, the comprehensive index D of this field k Defined as:
[0056] Therefore, the corresponding domain comprehensive index D can be calculated for the body composition domain, knee joint muscle functional domain, and cognitive functional domain. k Due to the standardized scores S of each individual indicator j All fall between -100% and +100%, therefore the comprehensive index D for each field is... k The values also fall within the range of -100% to +100%, thus ensuring a uniform numerical range and comparability across different fields. Individual indicator results and field-specific comprehensive indices can be found in Table 3.
[0057] Table 3 ; Based on the comprehensive standardized scores of the body composition domain, knee joint muscle functional domain, and cognitive functional domain, 18.5%, 19.0%, and 30.2% were used as radar values, respectively. Figure 3 The values of each coordinate axis are used to construct a three-domain radar chart for the individual being evaluated. The value range for each coordinate axis is set from -100% to +100%, where 0 represents the median level based on gender and age group groups, the closer to +100% indicates a better state in the corresponding domain, and the closer to -100% indicates a worse state in the corresponding domain. For example... Figure 4 As shown, the radar chart can intuitively represent the strengths and weaknesses of the individual being evaluated in the body composition domain, knee joint muscle function domain, and cognitive function domain, and can serve as a graphical output of the individual deviation report.
[0058] In summary, this application constructs a community norm database and integrates multidimensional phenotypic information such as body composition, knee muscle function, cognitive function, and lifestyle habits. By introducing correlation analysis and a modeling method with elastic network penalties during feature selection and modeling, it achieves accurate estimation of an individual's physiological age and a comprehensive representation of their multidimensional functional states. Furthermore, through hierarchical standardization and visualization, the comparability and intuitiveness of the evaluation results are effectively improved. Thus, while ensuring model stability and generalization ability, this application provides a more scientific, comprehensive, and feasible technical solution for early population identification, hierarchical management, and individualized intervention.
[0059] The above description is only a preferred embodiment of the present invention. It should be noted that for those skilled in the art, several improvements and modifications can be made without departing from the principle of the present invention, and these improvements and modifications should also be considered within the scope of protection of the present invention.
Claims
1. A method for assessing and characterizing the physiological age of knee joint muscle function, characterized in that, Includes the following steps: S1. Collect phenotypic parameters of the subjects and establish a community sample database. The phenotypic parameters include parameters from four categories: body composition, knee joint muscle function, cognitive function, and lifestyle habits. S2. Perform correlation analysis between each of the phenotypic parameters and actual age, and screen candidate phenotypic parameters that are significantly related to age; S3. Based on the candidate phenotypic parameters, establish a physiological age model of knee joint muscle function.
2. The method for assessing and characterizing the physiological age of knee joint muscle function according to claim 1, characterized in that, In step S1, the community sample database stores the sample size, mean, standard deviation, and preset percentile statistics for each phenotypic parameter. The database fields in the community sample database include stratification identifier, indicator name, sample size, mean, standard deviation, percentile parameter, and data version number.
3. The method for assessing and characterizing the physiological age of knee joint muscle function according to claim 1, characterized in that, Step S2 includes: Missing values and outliers were identified and processed in the data of each phenotypic parameter collected from the subjects, forming a two-dimensional phenotypic parameter matrix of individual and measurement parameter with individuals as rows and measurement parameters as columns; Age-related analyses were performed on each phenotypic parameter, and the correlation coefficient and significance level between each phenotypic parameter and actual age were calculated. Based on the correlation coefficient and significance level results, candidate phenotypic parameters that are significantly related to age are screened.
4. The method for assessing and characterizing the physiological age of knee joint muscle function according to claim 3, characterized in that, In step S2, the candidate phenotypic parameters include: Height, left upper limb fat-free body mass, left lower limb fat-free body mass, right upper limb fat-free body mass, right lower limb fat-free body mass, waist-to-hip ratio, skeletal muscle mass, fat mass index, fat to muscle mass ratio, body fat percentage, body mass index, percentage of skeletal muscle mass, visceral fat area, relative peak knee extensor strength, relative peak knee flexor strength, hamstring to quadriceps torque ratio, bilateral quadriceps strength difference, bilateral hamstring strength difference, subjective cognitive function decline score, regular exercise habits.
5. The method for assessing and characterizing the physiological age of knee joint muscle function according to any one of claims 1-4, characterized in that, In step S3, the expression for the physiological age model of knee joint muscle function is: ， Where y represents the subject's physiological age of knee muscle function, x i Let a be the i-th age-related indicator included in the model. i is the regression coefficient corresponding to this indicator, and b is the model intercept.
6. The method for assessing and characterizing the physiological age of knee joint muscle function according to claim 5, characterized in that, Step S3 includes: Based on the candidate phenotypic parameters, a target phenotypic parameter matrix for modeling is constructed, and the target phenotypic parameter matrix is randomly split into a training set and a test set; Using actual age as the training label, a generalized linear modeling method with elastic network penalty terms was used to establish a physiological age model of knee joint muscle function. For each candidate alpha, cross-validation is performed on the training set, and the corresponding optimal regularization parameter lambda is searched. The alpha value with the smallest error and its corresponding optimal lambda value are selected as the final model parameters. Based on the complete training set, the final physiological age model of knee joint muscle function was refitted, and the regression coefficients α corresponding to each input index were obtained. i And the model intercept b.
7. The method for assessing and characterizing the physiological age of knee joint muscle function according to claim 1, characterized in that, After establishing the physiological age model of knee joint muscle function in step S3, the method further includes: S4. Individual deviations are characterized by calculating multi-domain standardized scores and constructing radar charts.
8. The method for assessing and characterizing the physiological age of knee joint muscle function according to claim 7, characterized in that, Step S4 includes: Based on the gender and age information of the individual to be evaluated, the data is matched to the corresponding age-gender stratified norm unit in the community sample database; For the j-th indicator, based on the original observed value x of the indicator j Based on the distribution information of the index and its corresponding norm unit, calculate the relative quantile position r corresponding to the j-th index. j And convert it into a standardized score S j ; Calculate the comprehensive index D for each field k Construct a multi-domain radar map of the individual to be evaluated.