Tongue image physical health assessment method and assessment system, and storage medium

Abstract

The application discloses a tongue image constitution health assessment method and system and a storage medium, and comprises the following steps: analyzing a tongue image to obtain first matching degree data; normalizing questionnaire screening data to obtain second matching degree data; fusing and calculating the first matching degree data and the second matching degree data to obtain fusion discrimination degree data corresponding to at least one candidate constitution; calculating and comprehensively obtaining disease risk degree data corresponding to a health risk dimension according to each fusion discrimination degree data in a constitution discrimination degree set and a risk reference value in a corresponding candidate constitution; matching a constitution risk correction set corresponding to different disease risks according to user personal data, calculating and comprehensively obtaining correction risk degree data according to each fusion discrimination degree data and a risk correction value; and comprehensively obtaining constitution health assessment data from the disease risk degree data and the correction risk degree data. The application can flexibly correct multi-dimensional data and improve assessment accuracy.

Description

Technical Field

[0001] This invention relates to the field of health risk assessment technology, and in particular to a tongue image-based physical health assessment method, assessment system, and storage medium. Background Technology

[0002] Traditional Chinese medicine (TCM) constitution identification can predict potential health risks and guide personalized health maintenance and disease prevention by identifying a user's constitution type. It is widely used in health checkups, chronic disease intervention, and home health management. Traditional TCM constitution identification mainly relies on the manual diagnostic model of "observation, auscultation, inquiry, and palpation" by physicians, heavily depending on the physician's personal clinical experience and professional accumulation, and thus possesses a high degree of subjectivity. With the development of artificial intelligence, machine vision, and big data technologies, intelligent TCM constitution assessment technology is gradually replacing traditional manual identification, becoming the mainstream development trend in the industry.

[0003] However, single identification technologies such as tongue image, constitution, and facial features cannot fully and dynamically reflect the user's true constitution status and combined constitution characteristics. They are prone to misjudgment and omission of constitution, and the overall identification accuracy and generalization ability are limited. Moreover, they can only output basic constitution type results and cannot specifically quantify disease and health risks in different dimensions. The assessment dimensions are relatively simple.

[0004] Therefore, current TCM constitution intelligent assessment technology generally suffers from technical drawbacks such as single assessment dimensions, insufficient data fusion accuracy, lack of health risk stratification assessment, lack of personalized risk correction mechanism, and poor accuracy and adaptability of assessment results. Summary of the Invention

[0005] This invention aims to at least solve one of the technical problems existing in the prior art. To this end, this invention proposes a tongue image-based physical health assessment method and system, a storage medium, and flexible correction of multi-dimensional data to improve the accuracy of the assessment.

[0006] A tongue image-based physical health assessment method according to a first aspect of the present invention includes: acquiring a tongue image; analyzing the tongue image to obtain a tongue image recognition result; mapping at least one candidate physical constitution and first matching degree data corresponding to the candidate physical constitution based on the tongue image recognition result; acquiring questionnaire screening data; normalizing the questionnaire screening data to obtain second matching degree data corresponding to at least one candidate physical constitution; fusing the first matching degree data and the second matching degree data to obtain fusion discrimination degree data corresponding to at least one candidate physical constitution, and constructing a physical constitution discrimination degree set from the fusion discrimination degree data corresponding to various candidate physical constitutions; and calculating and synthesizing disease risk degree data corresponding to the corresponding health risk dimension based on each fusion discrimination degree data in the physical constitution discrimination degree set and the risk benchmark value in the corresponding candidate physical constitution under different health risk dimensions. In this process, different health risk dimensions have corresponding physical risk benchmark sets for different disease risks. Each physical risk benchmark set has a risk benchmark value corresponding to various candidate physical constitutions. User personal data is acquired, including one or more of gender, age, and body mass index. Different health risk dimensions have corresponding physical risk correction sets for different disease risks. Physical risk correction sets for different disease risks are matched based on user personal data. Each physical risk correction set has a risk correction value corresponding to various candidate physical constitutions. Under different health risk dimensions, the corrected risk level data corresponding to the corresponding health risk dimension is calculated and synthesized based on the fusion discrimination degree data in the physical constitution discrimination degree set and the risk correction values ​​in the corresponding candidate physical constitutions. The disease risk level data and the corrected risk level data are then combined to obtain physical health assessment data.

[0007] The tongue image-based physical health assessment method according to embodiments of the present invention has at least the following beneficial effects: This invention relates to a tongue image-based physical health assessment method. It normalizes and fuses first matching degree data derived from tongue image recognition results and second matching degree data derived from questionnaire screening data to obtain fused discrimination degree data corresponding to at least one candidate physical constitution. Furthermore, it constructs a physical constitution discrimination degree set. Since the fused discrimination degree data indicates the degree of correlation with the candidate physical constitution, under different health risk dimensions, disease risk degree data can be calculated using each fused discrimination degree data and the corresponding risk benchmark value in the candidate physical constitution. This data is then combined with corrected risk degree data matched from the user's personal data for comprehensive calculation, resulting in physical health assessment data. This application combines tongue image data, questionnaire screening data, and user personal data, allowing for flexible correction of multi-dimensional data and improving the accuracy of the assessment.

[0008] According to some embodiments of the present invention, the process of analyzing the tongue image to obtain a tongue image recognition result, mapping at least one candidate constitution based on the tongue image recognition result, and a first matching degree data corresponding to the candidate constitution includes: Tongue image analysis yields a set of tongue image features as the tongue image recognition result. This set of tongue image features includes various tongue image features and corresponding feature confidence levels. Based on the tongue image feature set, the rule matching score corresponding to different candidate constitutions is calculated according to different constitution rules; the multiple candidate constitutions are sorted from high to low according to the rule matching score, and the rule matching score of the candidate constitution is multiplied by its corresponding position correction coefficient to obtain the corrected matching score. The corrected matching scores of each corrected candidate constitution constitute at least part of the first matching degree data. The higher the rule matching score of the candidate constitution, the larger the corresponding position correction coefficient.

[0009] According to some embodiments of the present invention, the step of calculating the rule matching score corresponding to different candidate constitutions based on different constitution rules according to the tongue image feature set includes: the rule matching score of the candidate constitution is... ,in, For the first in the corresponding constitution rules Feature weights of tongue image features Does the tongue image feature set contain the element corresponding to the constitution rule? Tongue features consistent with those of the previous tongue image, possessing the following characteristics. If it is 1, then it does not have 1. =0, For the first in the corresponding constitution rules Confidence of features of tongue image The process of multiplying the rule-matching score of the candidate physical constitution by its corresponding position correction coefficient to obtain the corrected matching score includes: the corrected matching score of the candidate physical constitution. ,in, This is the positional correction coefficient corresponding to the candidate physique.

[0010] According to some embodiments of the present invention, the second matching degree data includes at least one candidate physique and a questionnaire screening score corresponding to the candidate physique. The process of fusing the first matching degree data and the second matching degree data to obtain the fusion discrimination degree data corresponding to at least one candidate physique includes: calculating the corrected matching score within the same candidate physique. and questionnaire screening The difference is used to determine the degree of consistency. ; will correct matching score Questionnaire screening and consistency Weighted calculations are used to derive the fusion discriminant score corresponding to the candidate constitution. This is used as data to determine the degree of fusion.

[0011] According to some embodiments of the present invention, the step of constructing a body constitution discrimination set from the fusion discrimination score data corresponding to various candidate body constitutions includes a screening step for each candidate body constitution, the screening step including: when the fusion discrimination score of the candidate body constitution... The score is less than the first discrimination threshold, and the questionnaire screening score of the candidate's physical condition is... If the score is below the questionnaire threshold, the candidate constitution is removed from the constitution discrimination set; the remaining candidate constitutions in the constitution discrimination set are then categorized according to the fusion discrimination score. Sort from largest to smallest; calculate the difference between the fusion discrimination scores of the candidate constitution with the largest fusion discrimination score and the candidate constitution with the second largest fusion discrimination score; when the difference is greater than the difference threshold, the candidate constitution with the largest fusion discrimination score is defined as the dominant constitution in the constitution discrimination degree set; when the difference is less than the difference threshold, and the fusion discrimination scores of the largest and second largest candidate constitutions are both greater than the second discrimination threshold, the two candidate constitutions with the largest and second largest fusion discrimination scores are defined as the mixed constitutions in the constitution discrimination degree set.

[0012] According to some embodiments of the present invention, the step of calculating and synthesizing the disease risk level data corresponding to the corresponding health risk dimension based on the fusion discrimination level data of each fusion discrimination level data in the body constitution discrimination level set and the risk benchmark value in the corresponding candidate body constitution includes: for a body constitution discrimination level set with a dominant body constitution, matching the body constitution risk correction set of the dominant body constitution under different health risk dimensions to obtain the corresponding risk benchmark value, using the risk benchmark value as the disease risk score of the corresponding disease risk, and using the disease risk score as the disease risk level data of the health risk dimension; for a body constitution discrimination level set with mixed body constitutions, calculating the first fusion discrimination level data of each fusion discrimination level data in the body constitution discrimination level set... Body weight of candidate body types ,in, For the first Fusion discriminant score of candidate constitutions; disease risk classification of constitution discriminant set with mixed constitutions under different health risk dimensions. Furthermore, disease risk is categorized as the degree of disease risk within this health risk dimension. For the health risk dimension The risk benchmark value corresponding to the candidate physical condition of the item.

[0013] According to some embodiments of the present invention, some of the candidate constitutions are defined as high-risk constitutions. For a constitution discrimination set containing both constitutions, the set further includes: when the combined constitution is a high-risk constitution, the disease risk under different health risk dimensions is divided into... ,in, This represents the maximum risk benchmark value corresponding to the candidate's physical condition under the health risk dimension.

[0014] According to some embodiments of the present invention, user personal data includes at least one of gender data, age data, and body mass index (BMI) data. Risk correction values ​​are matched into a body risk correction set based on at least one of the gender data, age data, and BMI data. Corrected risks are categorized under different health risk dimensions. ,in, The risk correction value is the user's personal data; the process of combining disease risk level data and corrected risk level data to obtain physical health assessment data includes: weighting the disease risk score and the corrected risk score of different user personal data packages to obtain physical health assessment data.

[0015] The assessment system according to a second aspect of the present invention is used to perform the tongue image physical health assessment method disclosed in any of the above embodiments.

[0016] According to a third aspect of the present invention, a computer-readable storage medium stores a computer program, characterized in that, when executed by a processor, the computer program implements the tongue image physical health assessment method disclosed in any of the above embodiments.

[0017] Additional aspects and advantages of the invention will be set forth in part in the description which follows, and in part will be obvious from the description, or may be learned by practice of the invention. Attached Figure Description

[0018] The above and / or additional aspects and advantages of the present invention will become apparent and readily understood from the description of the embodiments taken in conjunction with the following drawings, in which: Figure 1 This is a first flowchart of one embodiment of the tongue image physical health assessment method of the present invention; Figure 2 This is a second flowchart of one embodiment of the tongue image physical health assessment method of the present invention. Detailed Implementation

[0019] To make the objectives, technical solutions, and advantages of this application clearer, the following detailed description is provided in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application.

[0020] It should be noted that although functional modules are divided in the device schematic diagram and a logical order is shown in the flowchart, in some cases, the steps shown or described may be performed in a different order than the module division in the device or the order in the flowchart. The terms "first," "second," etc., in the specification, claims, and the aforementioned drawings are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence.

[0021] In the description of this invention, "several" means one or more, "multiple" means two or more, "greater than", "less than", "exceeding" are understood to exclude the number itself, and "above", "below", "within" are understood to include the number itself.

[0022] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application belongs. The terminology used herein is for the purpose of describing embodiments of this application only and is not intended to limit this application.

[0023] like Figure 1 As shown, the tongue image physical health assessment method according to a first aspect embodiment of the present invention includes: S110. Acquire a tongue image, analyze the tongue image to obtain a tongue image recognition result, and map at least one candidate constitution and a first matching degree data corresponding to the candidate constitution based on the tongue image recognition result. S120. Obtain questionnaire screening data and normalize the questionnaire screening data to obtain second matching degree data corresponding to at least one candidate physique. S130. The first matching degree data and the second matching degree data are fused and calculated to obtain the fusion discrimination degree data corresponding to at least one candidate constitution, and the fusion discrimination degree data corresponding to various candidate constitutions are used to construct a constitution discrimination degree set. S140. Under different health risk dimensions, calculate and synthesize the disease risk level data corresponding to the corresponding health risk dimension based on the fusion discrimination degree data of each physique discrimination degree set and the risk benchmark value in the corresponding candidate physique. Among them, different health risk dimensions have physique risk benchmark sets corresponding to different disease risks, and each physique risk benchmark set has risk benchmark values ​​corresponding to various candidate physiques. S150. Obtain user personal data, which includes one or more of gender, age, and body mass index. Different health risk dimensions have corresponding body risk correction sets for different disease risks. Match body risk correction sets for different disease risks based on user personal data. Each body risk correction set has risk correction values ​​corresponding to various candidate body types. Under different health risk dimensions, calculate and synthesize the corrected risk level data corresponding to the corresponding health risk dimension based on the fusion discrimination degree data in the body type discrimination degree set and the risk correction values ​​in the corresponding candidate body types. S160. Combine disease risk level data and adjusted risk level data to obtain physical health assessment data.

[0024] In step S110, the tongue image can be input into an artificial intelligence model to obtain the tongue image recognition result. The tongue image usually includes various tongue feature data such as tongue color and coating color. Conventional artificial intelligence modules include image recognition models, such as real-time target detection models, classification models based on convolutional neural networks, and tongue region segmentation and feature recognition models based on segmentation networks, so as to analyze the feature data such as tongue color and coating color for subsequent candidate constitution judgment.

[0025] Questionnaire screening data is used to provide users with options to select corresponding answers for different questions based on their own situation. Questionnaire screening scores The score range can be 0-100. Both the first and second match scores are normalized to a score of 0-100. The questionnaire screening score... ,in, Let J be the score given by the user for answering the j-th question. As the weight of the problem, This is the highest score for a single question.

[0026] It is understandable that different health risk dimensions refer to risk dimensions associated with different diseases. For example, health risk dimensions may include risks related to the neck, shoulders, and lower back, thoracic spine, cervical spine, lumbar spine, endocrine system, gastrointestinal system, cardiopulmonary system, immune system, cardiovascular system, musculoskeletal system, reproductive and urinary system, neurological system, mental stress, and sleep. Different candidate constitutions have different risks of developing these diseases. Therefore, in the constitution risk benchmark set corresponding to different health risk dimensions, the risk benchmark values ​​corresponding to different candidate constitutions, and the fused discriminant data in the constitution discrimination set can characterize the user's bias towards which candidate constitutions they belong to. By calculating the disease risk level data corresponding to the health risk dimension using the fused discriminant data and the risk benchmark values ​​in the corresponding candidate constitution, the user's risk level score for that disease can be obtained.

[0027] This application not only assesses the risk of the disease based on tongue images and questionnaire screening data, but also takes into account the user's personal data, including gender, age, and body mass index, among other factors.

[0028] This invention relates to a tongue image-based physical health assessment method. It normalizes and fuses first matching degree data derived from tongue image recognition results and second matching degree data derived from questionnaire screening data to obtain fused discrimination degree data corresponding to at least one candidate physical constitution. Furthermore, it constructs a physical constitution discrimination degree set. Since the fused discrimination degree data indicates the degree of correlation with the candidate physical constitution, under different health risk dimensions, disease risk degree data can be calculated using each fused discrimination degree data and the corresponding risk benchmark value in the candidate physical constitution. This data is then combined with corrected risk degree data matched from the user's personal data for comprehensive calculation, resulting in physical health assessment data. This application combines tongue image data, questionnaire screening data, and user personal data, allowing for flexible correction of multi-dimensional data and improving the accuracy of the assessment.

[0029] In some embodiments of the present invention, such as Figure 2 As shown, the process of analyzing tongue images to obtain tongue image recognition results, mapping at least one candidate constitution based on the tongue image recognition results, and the first matching degree data corresponding to the candidate constitution includes: S210. Based on tongue image analysis, a tongue image feature set is obtained as the tongue image recognition result, wherein the tongue image feature set includes multiple tongue image features and feature confidence levels corresponding to the tongue image features. ; S220. Calculate the rule matching score corresponding to different candidate constitutions based on the tongue image feature set and different constitution rules; S230. Sort multiple candidate physical types from high to low according to the rule matching score. Multiply the rule matching score of each candidate physical type by its corresponding position correction coefficient to obtain the corrected matching score. The corrected matching scores of each corrected candidate physical type constitute at least part of the first matching degree data. The higher the rule matching score of the candidate physical type, the larger the corresponding position correction coefficient.

[0030] Understandably, in step S210, the tongue image is analyzed by an artificial intelligence model to extract tongue image features and the corresponding feature confidence levels. Among them, the tongue image feature of "pale red tongue" has a confidence level of 0.86; the tongue image feature of "yellow coating" has a confidence level of 0.91; and the tongue image feature of "teeth marks" has a confidence level of 0.78.

[0031] Different candidate constitutions have different tongue features, such as damp-heat constitution and yin deficiency constitution. Different candidate constitutions have different constitution rules, and correspondingly, different constitution rules have different tongue features. In step S220, the rule matching score is obtained by judging the degree of conformity between the tongue features in the tongue feature set and the constitution rules. The rule matching score also reflects which types of constitution the user prefers.

[0032] To avoid simply listing multiple candidate constitutions side by side without reflecting their priority relationship, in step S230, the rule matching score of each candidate constitution is multiplied by its corresponding position correction coefficient to obtain the corrected matching score. The higher the rule matching score of a candidate constitution, the larger the corresponding position correction coefficient, thereby widening the score difference between candidate constitutions with similar rule matching scores and making it easier to distinguish them.

[0033] Specifically, the step of calculating the rule matching score corresponding to different candidate constitutions based on the tongue image feature set and different constitution rules includes: Rule matching for candidate physical characteristics is divided into ,in, For the first in the corresponding constitution rules Feature weights of tongue image features Does the tongue image feature set contain the element corresponding to the constitution rule? Tongue features consistent with those of the previous tongue image, possessing the following characteristics. If it is 1, then it does not have 1. =0, For the first in the corresponding constitution rules Confidence of features of tongue image ; The process of multiplying the candidate physical characteristics' rule-matching score by their respective corresponding position correction coefficients to obtain the corrected matching score includes: Corrected matching score of candidate physique ,in, This is the positional correction coefficient corresponding to the candidate physique.

[0034] For example, among multiple candidate physical traits, the position correction coefficient corresponding to the highest rule matching score. It can be 1, which is the position correction coefficient corresponding to the second highest rule matching score. It can be 0.8, the position correction coefficient corresponding to the third highest rule matching score. It can be 0.7, a specific location correction factor. The value can be set according to the actual situation.

[0035] To improve the accuracy of the discriminant data obtained by fusing the first and second matching degree data in assessing disease risk and reduce the impact of analytical errors, in some embodiments of the present invention, the second matching degree data includes at least one candidate constitution and a questionnaire screening score corresponding to that candidate constitution. ; The process of fusing the first matching degree data and the second matching degree data to obtain the fusion discrimination degree data corresponding to at least one candidate physique includes: Calculate the corrected matching score in the same candidate physique and questionnaire screening The difference is used to determine the degree of consistency. ; Correct the matching score Questionnaire screening and consistency Weighted calculations are used to derive the fusion discriminant score corresponding to the candidate constitution. This is used as data to determine the degree of fusion.

[0036] This section introduces a degree of consistency. Used to characterize the corrected matching score and questionnaire screening The degree of proximity, correcting the matching score and questionnaire screening The closer the two samples are, the better the fusion discrimination score can be in the weighted calculation. Conversely, it reduces the fusion discrimination score. For closer corrected matching scores and questionnaire screening It can more accurately characterize which candidate constitution a user prefers.

[0037] Specifically, ; ； in, To correct the matching score The weighting coefficients, Questionnaire screening score The weighting coefficients, To be divided into consistency The weighting coefficients, + + =1. Specifically, 0.4 is acceptable. 0.5 is acceptable. A value of 0.1 can be used; it can also be dynamically adjusted based on the average confidence level of the tongue image model and the completeness of the questionnaire.

[0038] In some embodiments of the present invention, the step of constructing a body constitution discrimination set by fusing discrimination degree data corresponding to various candidate body constitutions includes a screening step for each candidate body constitution, the screening step including; When the candidate constitution is fused, the discrimination score The score is less than the first discrimination threshold, and the questionnaire screening score of the candidate's physical condition is... If the score is below the questionnaire threshold, the candidate constitution will be removed from the constitution discrimination set. The remaining candidate constitutions are then grouped according to the fusion discrimination criteria. Sort from largest to smallest; Calculate the difference in the fusion discriminant scores between the candidate constitution with the largest fusion discriminant score and the candidate constitution with the second largest fusion discriminant score; When the discrimination difference value is greater than the discrimination threshold, the candidate constitution with the largest fusion discrimination score is defined as the dominant conductor constitution in the constitution discrimination degree set; When the discrimination difference is less than the discrimination threshold, and the fusion discrimination score of the largest candidate constitution and the fusion discrimination score of the second largest candidate constitution are both greater than the second discrimination threshold, then the two candidate constitutions with the largest and second largest fusion discrimination scores are defined as the mixed constitutions in the constitution discrimination degree set.

[0039] Specifically, the difference threshold, the first discrimination threshold, and the second discrimination threshold can be adjusted according to the actual situation. It is understandable that for fusion discrimination... and questionnaire screening Candidate constitutions with relatively small values ​​have little correlation with the user's actual constitution. To prevent affecting the accuracy of subsequent judgments, these candidate constitutions are removed from the constitution discrimination set.

[0040] Then the fusion discrimination will be performed. Sort from largest to smallest. If the difference between the largest candidate constitution and the second largest candidate constitution's fusion discrimination score is large, it proves that the largest candidate constitution is more consistent with the user's actual constitution. The largest candidate constitution can be directly used to judge the disease risk in the future.

[0041] Specifically, the calculation and synthesis of disease risk level data corresponding to the health risk dimension based on the fusion discrimination level data of each fusion discrimination level data in the body constitution discrimination level set and the risk benchmark value in the corresponding candidate body constitution includes: For a set of constitution discrimination degree with dominant constitution, the constitution risk correction set under different health risk dimensions is matched with the corresponding risk benchmark value. The risk benchmark value is used as the disease risk score of the corresponding disease risk, and the disease risk score is used as the disease risk degree data of the health risk dimension. That is, the dominant constitution is directly used to match the risk benchmark value of the corresponding disease risk, and then used as the disease risk score.

[0042] If the difference in the discrimination score is less than the difference threshold, and the fusion discrimination score of the largest candidate constitution and the fusion discrimination score of the second largest candidate constitution are both greater than the second discrimination threshold, it proves that the user's actual constitution may be a mixture of multiple constitutions. Therefore, the two candidate constitutions are defined as a mixed constitution with a concentration of constitution discrimination, and the mixed constitution is subsequently used to judge the disease risk.

[0043] Specifically, the calculation and synthesis of disease risk level data corresponding to the health risk dimension based on the fusion discrimination level data of each fusion discrimination level data in the body constitution discrimination level set and the risk benchmark value in the corresponding candidate body constitution includes: For a set of constitution discrimination levels that includes both mixed constitutions, calculate the first-order discrimination level in the set of constitution discrimination levels respectively. Body weight of candidate body types ,in, For the first Fusion discrimination score of candidate physical characteristics; The disease risk of individuals with mixed constitutions is categorized under different health risk dimensions. Furthermore, disease risk is categorized as the degree of disease risk within this health risk dimension. For the health risk dimension The risk benchmark value corresponding to the candidate physical condition of the item.

[0044] It is understandable that the body constitution weight is calculated by first calculating the proportion of the fusion discrimination score of different candidate body constitutions to the total fusion discrimination score, and then adjusting the calculation of disease risk scores under different health risk dimensions based on the body constitution weight.

[0045] In some embodiments of the present invention, some of the candidate constitutions are defined as high-risk constitutions. The specific constitutions defined as high-risk can be adjusted by staff based on actual circumstances. The set of constitution discrimination levels for individuals with mixed constitutions also includes: When a combination of both health and disease conditions is considered a high-risk condition, the disease risk under different health risk dimensions is categorized as follows: ,in, This serves as the maximum risk benchmark value corresponding to candidate body types under the health risk dimension, thereby avoiding excessive dilution of high-risk body types during the weighted averaging of mixed body types, which would affect the accuracy of the judgment.

[0046] In some embodiments of the present invention, user personal data includes at least one of gender data, age data, and body mass index (BMI) data. Risk correction values ​​are matched to a body risk correction set based on at least one of the gender data, age data, and BMI data. Corrected risks are categorized under different health risk dimensions. ,in, Risk correction value for user's personal data; The process of combining disease risk level data and modified risk level data to derive physical health assessment data includes: The physical health assessment data is obtained by weighting the disease risk score and the modified risk score of different user personal data packages.

[0047] User personal data can include gender and age data. The system can categorize users into preset age ranges based on age estimates, such as teenagers, young adults, middle-aged, and elderly. Furthermore, the system pre-creates a gender-age-physical health risk table, which records the risk correction values ​​for diseases affecting different physical conditions in different genders and age ranges. Specifically, when user personal data consists of gender and age data... , For the first Risk correction values ​​for different sexes and age ranges for different physical constitutions.

[0048] Specifically, the modified risk based on gender and age data under different health risk dimensions is divided into... .

[0049] User personal data may include Body Mass Index (BMI) data, calculated as weight / height². The system can determine a BMI risk benchmark based on the BMI range. For example, BMI can be divided into low, normal, overweight, and obese ranges. Different BMI ranges have varying degrees of influence on different body constitutions. For instance, the risk association weight between a high BMI and a phlegm-dampness constitution or damp-heat constitution may be higher than that between a balanced constitution. When user personal data is BMI data... , For the first Risk adjustment values ​​for different body types in different BMI ranges.

[0050] Specifically, the adjusted risk based on body mass index data under different health risk dimensions is divided into... .

[0051] Therefore, in step S160, the disease risk level data and the modified risk level data are combined to obtain the physical health assessment data, which can be reflected by calculating the total risk score.

[0052] Total Risk Score .

[0053] in, , and These are the weighting coefficients for disease risk scores and adjusted risk scores based on different users' personal data.

[0054] The total risk score can be generated using a weighted fusion method, or it can be obtained by fusing methods such as rule threshold fusion, decision tree model, logistic regression model, gradient boosting tree model, neural network model, hierarchical scoring and normalization, etc. There are no specific limitations here.

[0055] After obtaining the total risk score, the risk of disease can be judged based on the specific risk range in which the total risk score falls, providing guidance to users.

[0056] The assessment system according to a second aspect of the present invention is used to perform the tongue image physical health assessment method disclosed in any of the above embodiments.

[0057] According to a third aspect of the present invention, a computer-readable storage medium stores a computer program, characterized in that, when executed by a processor, the computer program implements the tongue image physical health assessment method disclosed in any of the above embodiments.

[0058] Memory, as a non-transitory computer-readable storage medium, can be used to store non-transitory software programs and non-transitory computer-executable programs. Furthermore, memory may include high-speed random access memory, and may also include non-transitory memory, such as at least one disk storage device, flash memory device, or other non-transitory solid-state storage device. In some embodiments, memory may optionally include memory remotely located relative to the processor, and these remote memories can be connected to the processor via a network. Examples of such networks include, but are not limited to, the Internet, intranets, local area networks, mobile communication networks, and combinations thereof.

[0059] The embodiments described in this application are for the purpose of more clearly illustrating the technical solutions of the embodiments of this application, and do not constitute a limitation on the technical solutions provided by the embodiments of this application. As those skilled in the art will know, with the evolution of technology and the emergence of new application scenarios, the technical solutions provided by the embodiments of this application are also applicable to similar technical problems.

[0060] Those skilled in the art will understand that the technical solutions shown in the figures do not constitute a limitation on the embodiments of this application, and may include more or fewer steps than shown, or combine certain steps, or different steps.

[0061] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs.

[0062] Those skilled in the art will understand that all or some of the steps in the methods disclosed above, as well as the functional modules / units in the systems and devices, can be implemented as software, firmware, hardware, or suitable combinations thereof.

[0063] The terms “first,” “second,” “third,” “fourth,” etc. (if present) in the specification and accompanying drawings of this application are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of this application described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms “comprising” and “having,” and any variations thereof, are intended to cover non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0064] The preferred embodiments of the present application have been described above with reference to the accompanying drawings, but this does not limit the scope of the claims of the present application. Any modifications, equivalent substitutions, and improvements made by those skilled in the art without departing from the scope and substance of the embodiments of the present application shall be within the scope of the claims of the present application.

[0065] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0066] Although embodiments of the invention have been shown and described, those skilled in the art will understand that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the claims and their equivalents.

Claims

1. A method for assessing physical health through tongue imaging, characterized in that, include: Acquire a tongue image, analyze the tongue image to obtain a tongue image recognition result, and map at least one candidate constitution and a first matching degree data corresponding to the candidate constitution based on the tongue image recognition result; Obtain questionnaire screening data and normalize the questionnaire screening data to obtain second matching degree data corresponding to at least one candidate physique; The first matching degree data and the second matching degree data are fused and calculated to obtain the fusion discrimination degree data corresponding to at least one candidate constitution, and the fusion discrimination degree data corresponding to various candidate constitutions are used to construct a constitution discrimination degree set; Under different health risk dimensions, the disease risk level data corresponding to the corresponding health risk dimension is calculated and synthesized based on the fusion discrimination degree data of each physique discrimination degree set and the risk benchmark value in the corresponding candidate physique. Among them, different health risk dimensions have physique risk benchmark sets corresponding to different disease risks, and each physique risk benchmark set has risk benchmark values ​​corresponding to various candidate physiques. Acquire user personal data, which includes one or more of gender, age, and body mass index. Different health risk dimensions have corresponding body risk correction sets for different disease risks. Match body risk correction sets for different disease risks based on user personal data. Each body risk correction set has risk correction values ​​corresponding to various candidate body types. Under different health risk dimensions, calculate and synthesize the corrected risk level data corresponding to the corresponding health risk dimension based on the fusion discrimination degree data of each body type discrimination degree set and the risk correction values ​​in the corresponding candidate body types. The data on disease risk level and the adjusted risk level are combined to obtain physical health assessment data.

2. The tongue image physical health assessment method according to claim 1, characterized in that, The process of analyzing tongue images to obtain tongue image recognition results, mapping at least one candidate constitution based on the tongue image recognition results, and obtaining first matching degree data corresponding to the candidate constitutions includes: Tongue image analysis yields a set of tongue image features as the tongue image recognition result. This set of tongue image features includes various tongue image features and corresponding feature confidence levels. ; Based on the tongue image feature set, rule matching scores corresponding to different candidate constitutions are calculated according to different constitution rules; Multiple candidate physical traits are sorted from high to low according to their rule matching scores. The rule matching scores of the candidate physical traits are multiplied by their respective position correction coefficients to obtain the corrected matching scores. The corrected matching scores of each corrected candidate physical trait constitute at least part of the first matching degree data. The higher the rule matching score of the candidate physical trait, the larger the corresponding position correction coefficient.

3. The tongue image physical health assessment method according to claim 2, characterized in that, The step of calculating rule matching scores for different candidate constitutions based on tongue image feature sets and different constitution rules includes: Rule matching for candidate physical characteristics is divided into ,in, For the first in the corresponding constitution rules Feature weights of tongue image features Does the tongue image feature set contain the element corresponding to the constitution rule? Tongue features consistent with those of the previous tongue image, possessing the following characteristics. If it is 1, then it does not have 1. =0, For the first in the corresponding constitution rules Confidence of features of tongue image ; The process of multiplying the candidate physical characteristics' rule-matching score by their respective corresponding position correction coefficients to obtain the corrected matching score includes: Corrected matching score of candidate physique ,in, This is the positional correction coefficient corresponding to the candidate physique.

4. The tongue image physical health assessment method according to claim 3, characterized in that, The second matching degree data includes at least one candidate physique and the corresponding questionnaire screening score. ; The process of fusing the first matching degree data and the second matching degree data to obtain the fusion discrimination degree data corresponding to at least one candidate physique includes: Calculate the corrected matching score in the same candidate physique and questionnaire screening The difference is used to determine the degree of consistency. ; Correct the matching score Questionnaire screening and consistency Weighted calculations are used to derive the fusion discriminant score corresponding to the candidate constitution. This is used as data to determine the degree of fusion.

5. The tongue image physical health assessment method according to claim 4, characterized in that, The process of constructing a body constitution discrimination set by fusing discrimination degree data corresponding to various candidate body constitutions includes a screening step for each candidate body constitution, the screening step including: When the candidate constitution is fused, the discrimination score The score is less than the first discrimination threshold, and the questionnaire screening score of the candidate's physical condition is... If the score is below the questionnaire threshold, the candidate constitution will be removed from the constitution discrimination set. The remaining candidate constitutions are then grouped according to the fusion discrimination criteria. Sort from largest to smallest; Calculate the difference in fusion discriminant scores between the candidate constitution with the largest fusion discriminant score and the candidate constitution with the second largest fusion discriminant score; When the discrimination difference value is greater than the discrimination threshold, the candidate constitution with the largest fusion discrimination score is defined as the dominant conductor constitution in the constitution discrimination degree set; When the discrimination difference is less than the discrimination threshold, and the fusion discrimination score of the largest candidate constitution and the fusion discrimination score of the second largest candidate constitution are both greater than the second discrimination threshold, then the two candidate constitutions with the largest and second largest fusion discrimination scores are defined as the mixed constitutions in the constitution discrimination degree set.

6. The tongue image physical health assessment method according to claim 5, characterized in that, The disease risk level data corresponding to this health risk dimension, calculated and synthesized based on the fusion discrimination level data of each fusion discrimination level data in the body constitution discrimination level set and the risk benchmark value in the corresponding candidate body constitution, includes: For the set of constitution discrimination degree with dominant constitution, the constitution risk correction set under different health risk dimensions is matched with the corresponding risk benchmark value. The risk benchmark value is used as the disease risk score of the corresponding disease risk, and the disease risk score is used as the disease risk degree data of that health risk dimension. For a set of constitution discrimination levels that includes both mixed constitutions, calculate the first-order discrimination level in the set of constitution discrimination levels respectively. Body weight of candidate body types ,in, For the first Fusion discrimination score of candidate physical characteristics; The disease risk of individuals with mixed constitutions is categorized under different health risk dimensions. Furthermore, disease risk is categorized as the degree of disease risk within this health risk dimension. For the health risk dimension The risk benchmark value corresponding to the candidate physical condition of the item.

7. The tongue image physical health assessment method according to claim 6, characterized in that, Some of the candidate constitutions are defined as high-risk constitutions, and the constitution discrimination set for those with mixed constitutions also includes: When a combination of both health and disease conditions is considered a high-risk condition, the disease risk under different health risk dimensions is categorized as follows: ,in, This represents the maximum risk benchmark value corresponding to the candidate's physical condition under the health risk dimension.

8. The tongue image physical health assessment method according to claim 6, characterized in that, User personal data includes at least one of gender, age, and body mass index (BMI). Risk correction values ​​are matched against a body risk correction set based on at least one of these three data points. Corrected risks are categorized under different health risk dimensions. ,in, Risk correction value for user's personal data; The process of combining disease risk level data and modified risk level data to derive physical health assessment data includes: The physical health assessment data is obtained by weighting the disease risk score and the modified risk score of different user personal data packages.

9. An evaluation system, characterized in that, Used to perform the tongue image physical health assessment method as described in any one of claims 1 to 8.

10. A computer-readable storage medium storing a computer program, characterized in that, When the computer program is executed by the processor, it implements the tongue image physical health assessment method according to any one of claims 1 to 8.

Images