Traditional Chinese medicine physique diet therapy optimization system based on dynamic feedback and data tracking

Abstract

The invention discloses a traditional Chinese medicine physique diet therapy optimization system based on dynamic feedback and data tracking, which comprises a data layer, an evaluation layer, an analysis layer, a recommendation layer, a feedback optimization layer and a user interaction layer, and is characterized in that the data layer constructs a standardized traditional Chinese and western medicine disease name-syndrome type-medicine and food taboo database, and the evaluation layer verifies and optimizes a traditional Chinese medicine physique evaluation scale; the analysis layer performs physique judgment on the physique evaluation scale, integrates health data input by a user, matches the health data with each database in the data layer, and screens out taboo medicine food and suitable medicine food according to a matching result, the recommendation layer is based on a result of the analysis layer, and the feedback optimization layer is used for constructing a dynamic data tracking and algorithm optimization module. The user interaction layer is a core carrier and is used for integrating all the layers, the data layer is used for constructing a database, the evaluation layer is used for verifying a physique evaluation scale, the analysis layer is used for integrating health data and matching medicine and food screening, the recommendation layer is used for generating a targeted diet scheme, and scientific, personalized and sustainable optimized physique diet therapy guidance is provided for a user.

Description

Technical Field

[0001] This invention relates to the fields of healthcare information and intelligent dietary therapy health management, and in particular to a TCM constitution dietary therapy optimization system based on dynamic feedback and data tracking. Background Technology

[0002] With the accelerating pace of modern life, approximately 35% to 50% of the global population is in a state of "sub-health." The World Health Organization has listed the prevention of sub-health as a global health strategy, and developed countries are investing heavily in related research, posing a challenge to the international development of my country's sub-health industry.

[0003] The TCM concepts of "prevention before disease occurs" (preventing disease before it manifests, preventing its progression once it has occurred, and preventing recurrence after recovery) and "medicine and food sharing the same origin" provide unique advantages for sub-health intervention. In 2007, the State Administration of Traditional Chinese Medicine incorporated TCM constitution identification into the TCM preventive healthcare service system. However, the existing TCM preventive healthcare service system still faces the following key problems: 1. Insufficient information standardization: The correspondence between TCM disease names, Western medicine disease names, and drug and food taboos is confusing (for example, the description of diseases in the "Chinese Materia Medica" contains different types of terms such as "spleen deficiency and seminal emission" and "glaucoma"), lacking a unified standard of comparison, which makes it difficult for users to search and is easy to confuse; 2. Lack of convenience and accuracy in physical fitness assessment: Existing physical fitness assessment tools mostly rely on professional operation, making it difficult for ordinary users to quickly obtain their own physical fitness information. In addition, the assessment scales lack large-scale population validation and iterative optimization, resulting in insufficient accuracy. 3. Static dietary therapy plans: Existing dietary therapy recommendations are mostly "one-size-fits-all" and cannot be dynamically adjusted according to users' real-time health data (such as changes in symptoms, dietary feedback, tongue coating / weight indicators), making it difficult to meet personalized needs; 4. Lack of closed-loop management: It only implements a one-way process of "assessment-recommendation", without data tracking and feedback mechanisms, making it impossible to continuously optimize the dietary therapy plan and making it difficult to guarantee the health management effect; 5. High barriers to implementation: Traditional Chinese medicine dietary therapy knowledge is highly specialized, making it difficult for ordinary users to translate theory into practical dietary plans, and there is a lack of connecting channels from "plan recommendations" to "ingredient acquisition".

[0004] Therefore, a TCM constitution-based dietary therapy optimization system based on dynamic feedback and data tracking is proposed. This system can achieve standardized information, convenient assessment, dynamic solutions, and closed-loop management, thereby promoting the popularization and precise application of TCM dietary therapy. Summary of the Invention

[0005] This invention overcomes the shortcomings of the prior art and provides a TCM constitution-based dietary therapy optimization system based on dynamic feedback and data tracking.

[0006] To achieve the above objectives, the technical solution adopted by the present invention is: a TCM constitution diet therapy optimization system based on dynamic feedback and data tracking, including a data layer, an assessment layer, an analysis layer, a recommendation layer, a feedback optimization layer, and a user interaction layer; The data layer is used to construct a standardized database of TCM and Western medicine disease names, syndrome types, and food and drug contraindications. The assessment layer is used to verify and optimize the TCM constitution assessment scale. The analysis layer is used to determine the constitution of the constitution assessment scale and integrate the health data input by the user, match it with the various databases in the data layer, and filter out contraindicated and suitable food and drug based on the matching results. The recommendation layer generates food and drug recommendations based on the analysis layer results from two aspects: dietary contraindications for common diseases and dietary therapy plans for constitution. The feedback optimization layer is used to construct a dynamic data tracking and algorithm optimization module. The user interaction layer is the core carrier, including the online mini-programs built to integrate the above layers.

[0007] In a preferred embodiment of the present invention, the database of Chinese and Western medicine disease names-syndromes-drug and food contraindications includes the following databases: Database 1: A database comparing Chinese and Western medicine disease names, syndrome types, and symptoms; Database 2: Information and Contraindications on Food and Medicine Homologous; Database 3: Constitution-Dietary Therapy Recipe Database; Database 3 is built upon Database 1, with Database 2 as its hub, to correlate symptoms with corresponding dietary restrictions and to establish data based on the assessment results of the evaluation layer.

[0008] In a preferred embodiment of the present invention, the first database contains the Western medicine names and traditional Chinese medicine names of 114 common diseases, 380 traditional Chinese medicine syndrome types and typical symptoms, the second database contains the synonyms, indications, pharmacology and contraindications of 798 kinds of medicinal foods, and the third database contains medicinal food homologous ingredients, supplementary foods and their effects corresponding to various traditional Chinese medicine constitutions.

[0009] In a preferred embodiment of the present invention, the evaluation layer includes taste evaluation and physical fitness evaluation. The evaluation questionnaires are designed using the Likert 5-point method. The physical fitness evaluation questionnaires are validated in three stages: accuracy verification stage, satisfaction optimization stage, and deployment stage. When the user satisfaction rate is ≥98% in the satisfaction optimization stage, the questionnaires are deployed to the use stage.

[0010] In a preferred embodiment of the present invention, the accuracy verification stage of the physical fitness assessment questionnaire is as follows: a large amount of clinical data of people with different physical fitness is collected and integrated to obtain a simple and easy-to-use physical fitness assessment scale. 100-200 test subjects are randomly selected from people with certain knowledge of traditional Chinese medicine theory, and feedback is obtained through survey. After sorting out the data, the accuracy of the assessment questionnaire is judged. The satisfaction optimization phase: Based on the survey results and feedback from the accuracy verification phase, the content of the evaluation questionnaire is further improved in a reasonable way. The scope of the test population is no longer limited, and the number of test subjects is expanded. In the same way as before, after collecting feedback information, the data is sorted out to obtain the user's satisfaction with the evaluation questionnaire. Based on the satisfaction results, the questionnaire is randomly distributed to the population again so that the satisfaction reaches a stable expected value of ≥98%. Use phase: After the satisfaction rate reaches a stable expected value of ≥98%, a Traditional Chinese Medicine (TCM) constitution assessment scale is generated and put into use.

[0011] In a preferred embodiment of the present invention, the taste evaluation questionnaire is operated as follows: the various medicinal and food ingredients are archived and classified according to their four natures and five flavors, efficacy, taste, and modern pharmacology; the public selects relevant ingredients according to their own tastes; the questionnaire is then filtered based on the public's acceptance data of the selected ingredients; and finally, a taste evaluation form is generated for use.

[0012] In a preferred embodiment of the present invention, the analysis layer includes health data analysis and physical condition assessment; The health data analysis uses NLP parsing to combine the user-inputted free text symptoms with Chinese and Western medicine disease names into standardized symptoms; The constitution determination is based on the combination of health data analysis results and the TCM constitution assessment scale. The combination logic is to combine the symptoms-syndrome types in the TCM constitution assessment scale with the standardized symptoms obtained. Based on the combination results, the user's constitution is output and an analysis table is generated.

[0013] In a preferred embodiment of the present invention, the recommendation layer supports dual-path retrieval of symptom keywords or Chinese and Western medicine disease names, and generates drug and food recommendations using a two-level constraint retrieval mechanism; The first level is based on a standardized knowledge base of disease / syndrome type, medicine / food, and contraindications, and implements hard constraint filtering to eliminate conditions that cannot be violated, such as allergies, drug-food incompatibilities, contraindications for chronic diseases, and drug combination conflicts. The second level, based on soft constraints such as user taste preferences, processing time, cost availability, and cultural taboos, outputs a set of safe and highly feasible candidate food and medicine formulas through multi-indicator scoring and ranking with configurable weights, and provides links to purchase ingredients.

[0014] In a preferred embodiment of the present invention, the feedback optimization layer dynamically adjusts the dietary therapy plan by tracking, triggering, updating, and freezing the closed-loop control law; Within the follow-up window set in the user interaction layer, the system continuously tracks multiple indicators of the user, including diet, symptoms, weight, and sleep duration, and generates status data. Set trigger conditions and priority judgments, and determine whether to enter the optimization process based on preset data; when the trigger conditions are met, update the dietary therapy plan through AI algorithm; when the trigger conditions are not met or the plan reaches the preset stable effect, freeze the plan.

[0015] In a preferred embodiment of the present invention, the triggering condition determination is based on three types of quantitative triggering conditions: security triggering conditions, effect triggering conditions, and experience triggering conditions; The priority is set according to P1 level security category > P2 level effect category > P3 level experience category. When multiple conditions conflict, the higher priority and the item with the earlier number within the same priority are processed first.

[0016] In a preferred embodiment of the present invention, the online mini-program is built using Qiaotuo Cloud technology and includes a medicine and food query, physical fitness assessment, dietary calendar, health record, feedback module and food ingredient mall, and supports PDF export of assessment reports and data visualization.

[0017] This invention addresses the shortcomings of the prior art and has the following beneficial effects: (1) The system of the present invention constructs a standardized database of Chinese and Western medicine disease names, syndrome types and drug and food taboos through the data layer. It unifies and normalizes the syndrome terms such as spleen deficiency and seminal emission in Chinese medicine, disease names such as glaucoma in Western medicine and drug and food taboo information, and establishes a cross-system comparison standard. This solves the problem of the fragmentation of Chinese and Western medicine terms and the disordered correspondence of taboos in the traditional dietary therapy field. Users can query the matching relationship between symptoms and drugs and foods through the unified standard, avoiding query difficulties or misunderstandings caused by terminology confusion, and providing a precise and unified benchmark for subsequent data interaction at each layer.

[0018] Meanwhile, the assessment layer ensures that the scale is both scientific and easy to use through systematic optimization. At the same time, the user interaction layer uses an online mini-program as a carrier to present the assessment function in a lightweight way, which solves the problem that existing technologies rely on professional operation and are difficult for ordinary users to access. Ordinary users can complete the assessment independently without the guidance of professionals. Moreover, the optimized scale is more accurate than unverified tools, achieving the dual goals of independent assessment and reliable results, laying a precise foundation for personalized dietary therapy.

[0019] Furthermore, the dynamic data tracking and algorithm optimization module built within the feedback optimization layer can integrate user-input health data (such as symptom changes and weight fluctuations) in real time and dynamically match and adjust it in conjunction with the analysis and recommendation layers. Compared to the traditional static recommendation of "one prescription for everyone," this architecture can update the food and medicine recommendation logic based on the user's real-time health status, allowing the dietary therapy plan to continuously adapt to individual health changes, accurately meeting personalized needs, and aligning with the traditional Chinese medicine principle of "adapting to the three factors" in dietary therapy.

[0020] (2) The present invention has a hierarchical linkage of assessment layer, analysis layer, recommendation layer and feedback optimization layer, forming a complete closed loop of physical fitness assessment → scheme recommendation → data tracking → optimization. Compared with the traditional scheme that only has a one-way process of assessment-recommendation, the dynamic tracking and optimization function of the feedback optimization layer can continuously collect user data and feed back to the scheme iteration, which solves the problems of no feedback and difficulty in optimization, ensuring that the dietary therapy scheme can adapt to the user's health needs in the long term, and significantly improving the stability and sustainability of health management effect.

[0021] (3) The online mini-program of the user interaction layer of the present invention serves as the core carrier, integrating functions such as physical fitness assessment, food and medicine inquiry, and plan viewing. It transforms professional TCM dietary therapy knowledge into an easy-to-understand operation process, solving the difficulty of transforming professional knowledge. At the same time, the modular integration of functions at each layer provides a carrier support for the connection between plan recommendation and ingredient acquisition (such as the subsequent expansion of ingredient purchase entry), breaking the barrier between traditional dietary therapy recommendation and implementation, and allowing ordinary users to easily realize the transformation from knowing the plan to implementing it. Attached Figure Description

[0022] The present invention will be further described below with reference to the accompanying drawings and embodiments; Figure 1 This is a schematic diagram of a TCM constitution dietary therapy optimization system module based on dynamic feedback and data tracking, according to a preferred embodiment of the present invention. Figure 2 This is a schematic diagram of the workflow of a TCM constitution dietary therapy optimization system based on dynamic feedback and data tracking, according to a preferred embodiment of the present invention. Figure 3 This is a schematic diagram of the data layer logic of a preferred embodiment of the present invention; Figure 4 This is a schematic diagram of the evaluation layer process of a preferred embodiment of the present invention; Figure 5 This is a schematic diagram of the analysis layer flow of a preferred embodiment of the present invention; Figure 6 This is a schematic diagram of the recommendation layer process according to a preferred embodiment of the present invention; Figure 7 This is a schematic diagram of the feedback optimization layer process according to a preferred embodiment of the present invention; Figure 8 This is a schematic diagram of the user interaction layer logic according to a preferred embodiment of the present invention. Detailed Implementation

[0023] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0024] 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 invention pertains. The terminology used herein in the description of the invention is for the purpose of describing particular embodiments only and is not intended to be limiting of the invention. The term "and / or" as used herein includes any and all combinations of one or more of the associated listed items.

[0025] The TCM constitution-based dietary therapy optimization system based on dynamic feedback and data tracking includes a data layer, an assessment layer, an analysis layer, a recommendation layer, a feedback optimization layer, and a user interaction layer. Specifically, the data layer is used to construct a standardized database of TCM and Western medicine disease names, syndrome types, and food and drug contraindications. This database includes the following databases: Database 1: TCM and Western medicine disease names, syndrome types, and symptom comparison database; Database 2: Food and drug homology information and contraindications database; Database 3: Constitution-dietary therapy formula database. Database 3 is built on the data from the assessment results of the assessment layer, with Database 1 using Database 2 as the hub, matching symptoms with corresponding dietary contraindications. Database 1 includes the Western medicine disease names and TCM disease names of 114 common diseases, 380 TCM syndrome types, and typical symptoms. Database 2 includes the synonyms, indications, pharmacology, and contraindications of 798 kinds of food and drug. Database 3 includes food and drug homology ingredients, supplementary foods, and their effects corresponding to various TCM constitutions.

[0026] Example 1: Preliminary Generation of Database 3 Background: A 35-year-old user sought medical attention for daily dizziness and flushed face and eyes. The Western medicine diagnosis was hypertension (grade 1). The user hoped to use traditional Chinese medicine dietary therapy to assist in conditioning, but was unclear about the corresponding TCM syndrome, suitable medicines and foods, and contraindications for his symptoms. He needed to quickly obtain accurate information through the standardized association of the system's data layer. At this time, the three databases of the data layer worked together.

[0027] Database 1: Construction of a Comparison Database of Disease Names, Syndromes, and Symptoms in Traditional Chinese and Western Medicine Data source: Based on authoritative sources such as the "Standards for Diagnosis and Efficacy of Diseases and Syndromes in Traditional Chinese Medicine", "International Classification of Diseases (ICD-11)", and "Dictionary of Traditional Chinese Medicine", hypertension was selected from 114 common diseases as the target Western medicine disease name, and the corresponding TCM disease names dizziness and headache were simultaneously associated.

[0028] Syndrome and Symptom Matching: For hypertension (Western medicine) - dizziness / headache (Traditional Chinese medicine), the three most relevant syndrome types out of 380 TCM syndrome types are included, with typical symptoms marked for each: Liver Yang Rising Syndrome: Dizziness and headache, flushed face and red eyes, irritability, insomnia and dreaminess, red tongue with yellow coating (corresponding to the user's symptoms of dizziness, flushed face and red eyes); Phlegm-dampness obstruction syndrome: dizziness, heaviness in the head, chest tightness, nausea, poor appetite, excessive sleepiness, and a white, greasy tongue coating; Kidney essence deficiency syndrome: long-term dizziness, soreness and weakness of the lower back and knees, tinnitus and forgetfulness, pale tongue with thin coating.

[0029] Data format: The data is stored in a structured format, consisting of Western medicine disease name, traditional Chinese medicine disease name, syndrome ID, and symptom list, such as "hypertension-dizziness-001-headache, flushed face and red eyes, irritability". This ensures that the terms are correctly matched.

[0030] Database 2: Construction of a Database of Information and Contraindications on Food and Medicinal Herbs Data Collection: 798 kinds of medicinal and edible substances were selected from the "List of Medicinal and Edible Substances" and the "National Food Safety Standards". 15 frequently used medicinal and edible substances were specifically included to address hypertension-related management needs, as shown in Table 1 below: Table 1; Contraindication association logic: Clearly label the contraindications of medicine and food based on syndrome type and underlying disease, such as chrysanthemum - contraindicated for those with spleen and stomach deficiency and cold, and hawthorn - contraindicated for those with gastric ulcers, to avoid users misusing due to confusion of contraindications.

[0031] Database 3: Construction of a Constitution-Dietary Therapy Formula Database (Preliminary generation based on Databases 1 and 2) Construction Process: Step 1: Using Database 1 as the source of symptoms, extract the core correlation between hypertension, dizziness (Traditional Chinese Medicine), and liver yang hyperactivity syndrome; Step 2: Using Database 2 as the hub, screen for medicinal foods (chrysanthemum, celery) in Database 2 that treat liver yang hyperactivity syndrome and have no user contraindications (users do not have spleen and stomach deficiency, low blood pressure, etc.), and exclude contraindicated medicinal foods (e.g., if the user has excessive stomach acid, exclude hawthorn); Step 3: Design multiple formulas by combining the physical constitution data and taste preferences of the assessment layer.

[0032] The assessment layer is used to validate and optimize the TCM constitution assessment scale. Specifically, the assessment layer includes taste assessment and constitution assessment. The assessment questionnaires are designed using the Likert 5-point method. The constitution assessment questionnaire is validated in three stages: accuracy verification stage, satisfaction optimization stage, and implementation stage. When the user satisfaction rate is ≥98% in the satisfaction optimization stage, it enters the implementation stage.

[0033] Specifically, the accuracy verification stage of the physical fitness assessment questionnaire involves: collecting a large amount of clinical data from people with different physical constitutions, integrating it into a simple and easy-to-use physical fitness assessment scale, randomly selecting 100-200 test subjects from people with certain knowledge of traditional Chinese medicine theory, obtaining feedback through surveys, and judging the accuracy of the assessment questionnaire after organizing the data. Satisfaction optimization phase: Based on the survey results and feedback from the accuracy verification phase, further improve the content of the evaluation questionnaire, no longer limit the scope of the test population, expand the number of test subjects, and use the same method as before. After collecting feedback information, organize the data to obtain the user's satisfaction with the evaluation questionnaire, and then randomly distribute the satisfaction results to the population again to make the satisfaction reach a stable expected value of ≥98%. Use phase: After the satisfaction rate reaches a stable expected value of ≥98%, a Traditional Chinese Medicine (TCM) constitution assessment scale is generated and put into use.

[0034] The specific operation of the taste evaluation questionnaire is as follows: the various medicinal and food ingredients are archived and classified according to the four natures and five flavors, efficacy, taste and modern pharmacology. The public selects relevant ingredients according to their own taste. The questionnaire is then filtered based on the public's acceptance of the selected ingredients, and finally a taste evaluation form is generated for use.

[0035] The assessment layer ensures that the scale is both scientific and easy to use through systematic optimization. Meanwhile, the user interaction layer uses an online mini-program to present the assessment function in a lightweight way, which solves the problem that existing technologies rely on professional operation and are difficult for ordinary users to access. Ordinary users can complete the assessment independently without the guidance of professionals. Moreover, the optimized scale is more accurate than unverified tools, achieving the dual goals of independent assessment and reliable results, laying a precise foundation for personalized dietary therapy.

[0036] The analysis layer is used to determine the constitution of the constitution assessment scale and integrate the health data input by the user. It matches the data with various databases in the data layer and filters out the contraindicated and suitable medicinal foods based on the matching results. The analysis layer includes health data analysis and constitution determination. Health data analysis uses NLP parsing to combine the free text symptoms input by the user with the names of diseases in Chinese and Western medicine into standardized symptoms. Constitution determination is based on the combination of health data analysis results and the Chinese medicine constitution assessment scale. The combination logic is to combine the symptoms-syndrome types in the Chinese medicine constitution assessment scale with the standardized symptoms obtained. Based on the combination results, the user's constitution is output and an analysis table is generated.

[0037] Based on the results of the analysis layer, the recommendation layer generates food and medicine recommendations from two aspects: dietary taboos for common diseases and dietary therapy plans for body constitution. The recommendation layer supports dual-path retrieval by symptom keywords or Chinese and Western medicine disease names, and uses a two-level constraint retrieval mechanism to generate food and medicine recommendations. The first level is based on a standardized knowledge base of disease / syndrome type, medicine / food, and contraindications, and implements hard constraint filtering to eliminate conditions that cannot be violated, such as allergies, drug-food incompatibilities, contraindications for chronic diseases, and drug combination conflicts. The second level, based on soft constraints such as user taste preferences, processing time, cost availability, and cultural taboos, outputs a set of safe and highly feasible candidate food and medicine formulas through multi-indicator scoring and ranking with configurable weights, and provides links to purchase ingredients.

[0038] Example 2: A preliminary dietary therapy formula was obtained based on the improved database content from Example 1. Based on the analysis layer's comparison and analysis of the user's constitution data with the Traditional Chinese Medicine constitution assessment scale, it was determined that the user's constitution is balanced with a tendency towards liver yang hyperactivity. According to the taste assessment scale, the customer is sensitive to unusual flavors and dislikes celery; therefore, celery should be avoided. The following formula was designed: Recipe Name: Chrysanthemum and Cassia Seed Porridge Suitable for: Hypertension (Western medicine) → Dizziness and Liver Yang Rising Syndrome (Traditional Chinese Medicine) → Balanced Constitution with Liver Yang Rising (Constitution Assessment) → No Celery (User Preference) Ingredients: 10g chrysanthemum, 5g cassia seed, 50g rice.

[0039] Efficacy: Soothes the liver and lowers blood pressure, clears heat and improves eyesight, and also moistens the intestines and promotes bowel movement. Contraindications: Those with spleen and stomach deficiency should reduce the amount of chrysanthemum to 5g, and those with diarrhea should discontinue the use of cassia seeds.

[0040] Flavor compatibility notes: It has no peculiar odor of celery and has a subtle fragrance of cassia seeds, which is suitable for user needs.

[0041] The feedback optimization layer is used to build a dynamic data tracking and algorithm optimization module. The feedback optimization layer dynamically adjusts the diet therapy plan by tracking, triggering, updating, and freezing the closed-loop control law. Within the follow-up window set in the user interaction layer, the system continuously tracks multiple indicators of the user, including diet, symptoms, weight, and sleep duration, and generates status data. Set trigger conditions and priority judgments, and determine whether to enter the optimization process based on preset data; when the trigger conditions are met, update the dietary therapy plan through AI algorithm; when the trigger conditions are not met or the plan reaches the preset stable effect, freeze the plan.

[0042] Trigger condition judgment is based on three types of quantitative trigger conditions: safety trigger conditions, effect trigger conditions, and experience trigger conditions. The priority is set as P1 level safety > P2 level effect > P3 level experience. When multiple conditions conflict, the higher priority and the item with the earlier number within the same priority are processed first.

[0043] Example 3: Feedback and optimization of dietary therapy plans for patients with hypertension and liver yang hyperactivity syndrome Background: Based on Examples 1 and 2, after a 35-year-old hypertensive user (with a balanced constitution but a tendency towards liver yang hyperactivity and who should avoid celery) enabled the dietary therapy plan in the user interaction layer applet, the system feedback optimization layer initiated a tracking-triggering-updating-freezing closed-loop control to verify the effectiveness of the plan and dynamically adapt to changes in the user's health.

[0044] The follow-up window is set at 2 weeks (14 days), and the core monitoring targets are "blood pressure reaches the target (≤135 / 85mmHg), dizziness symptoms are relieved, and there are no adverse reactions".

[0045] Phase 1: Multidimensional Indicator Tracking (Days 1-7) The feedback optimization layer, through the user interaction layer's "Health Record" and "Dietary Calendar" modules, combines automatic data collection with manual entry to continuously track the following four types of status data: indicator type, specific indicator, data collection method, and monitoring results for days 1-7. Phase Two: Trigger Condition Judgment and Priority Ranking (Day 7) The feedback optimization layer automatically matches and judges the status data from days 1-7 based on three preset quantitative trigger conditions. The results are as follows: Safety Trigger Conditions (Level P1) Judgment Criteria: Occurrence of adverse drug or food reactions (such as diarrhea, rash), or new contraindications. Actual Result: Only one instance of mild abdominal bloating (non-persistent, without other accompanying symptoms), which did not reach the adverse reaction quantitative threshold (requires two consecutive occurrences or a single occurrence lasting ≥30 minutes), and therefore did not trigger the Level P1 condition.

[0046] Effect-based trigger conditions (Level P2) judgment criteria: Core indicators (blood pressure, symptoms) not meeting the target for 3 consecutive days: blood pressure > 135 / 85 mmHg, or dizziness frequency > 1 time / day. Actual results: Average blood pressure on days 5-7 was 140 / 88 mmHg (not meeting the target for 3 consecutive days), and dizziness frequency was 1.5 times / day (not relieved for 3 consecutive days), triggering the Level P2 condition (priority higher than experience-based conditions).

[0047] The criteria for triggering experience-related issues (Level P3) are: a dietary execution rate of <80% (i.e., more than 2.8 days of non-execution within 14 days), or user feedback regarding high processing difficulty or unpleasant taste. Actual results: The execution rate for days 1-7 was 71.4% (days 5 / 7, <80%). A user commented in the feedback module that the porridge took 30 minutes to cook and requested greater convenience, triggering the Level P3 condition. Based on the priority rule of "P1 > P2 > P3," the system prioritizes addressing Level P2 issues before simultaneously optimizing Level P3 experience problems.

[0048] Phase 3: AI Algorithm-Driven Solution Update (Day 8) The feedback optimization layer calls the AI ​​optimization model, combining it with Data Layer Database 1 (Syndrome Type-Symptom Association) and Database 2 (Pharmacology and Food). An update solution is generated for the triggered issues: For P2 level effects not meeting standards, Database 2 is searched: Among the foods and medicines treating liver yang hyperactivity syndrome, ingredients with stronger auxiliary antihypertensive effects and no compatibility conflicts with chrysanthemum / cassia seed are selected: lotus leaf (containing lotus leaf alkaloids, enhancing vasodilatory effects, no contraindications); AI adjustment logic: 5g of dried lotus leaf is added to the original formula (synergistic effect with chrysanthemum and cassia seed), and the dosage ratio is adjusted: chrysanthemum is increased to 12g (strengthening heat-clearing and vision-improving effects), while cassia seed remains at 5g (avoiding excessive diarrhea); Contraindications supplement: Because lotus leaf is cooling in nature, a new suggestion is added that those with spleen and stomach deficiency can add 2g of ginger to cook together (the user does not have spleen and stomach deficiency, so this is not added for now).

[0049] Optimizations for the P3-level experience (low execution rate, time-consuming processing): A simplified preparation method is provided: Boil chrysanthemum, cassia seeds, and lotus leaves in a health pot for 10 minutes to extract the juice, then mix it with pre-cooked rice porridge (total time reduced to 15 minutes). Interaction layer adaptation: A convenient video tutorial has been added to the mini-program's dietary calendar, along with a link to purchase pre-packaged porridge base. The updated recipe is named "Chrysanthemum, Cassia Seed, and Lotus Leaf Porridge," and is simultaneously pushed to the user's mini-program homepage with update instructions. Phase Four: Secondary Tracking and Plan Freeze (Days 8-14). The feedback optimization layer continuously monitors and updates the implementation data of the plan. The results are as follows: Dietary compliance rate: 100% (7 / 7 days, convenient methods improve willingness to comply), average blood pressure: 132 / 83 mmHg (reached the target of ≤135 / 85 mmHg for 5 consecutive days), dizziness symptoms: only occurred once on day 8, and there were no attacks on days 9-14, safety indicators: no adverse reactions such as abdominal distension and diarrhea, user feedback: no strange taste, and easy to prepare.

[0050] Solution freeze determination (day 14): Based on the feedback optimization layer's freeze criteria (no safety / effect / experience-related triggering conditions, core indicators meeting standards for 4 consecutive days, and user satisfaction ≥90%), the system automatically freezes the solution, marking it as stable in the mini-program. The freeze period is 4 weeks. During the freeze period, only basic indicators (such as blood pressure) are continuously tracked, and the formula is not actively adjusted. If P1 / P2 level conditions are triggered during the freeze period, the freeze will be automatically lifted and a new round of optimization will be initiated.

[0051] The user interaction layer serves as the core carrier, including an online mini-program that integrates the aforementioned layers. Specifically, the online mini-program is built using Qiaotuo Cloud technology and includes functions such as medicine and food inquiry, physical fitness assessment, dietary calendar, health record, feedback module, and food ingredient store. It also supports exporting assessment reports as PDFs and data visualization.

[0052] Example 4: Full-process dietary therapy service for patients with hypertension (liver yang hyperactivity syndrome) (including taste rejection adaptation) Background: Mr. Zhang, a 45-year-old male user, sought medical attention for "recurrent dizziness for one month, with flushed face and red eyes." He was diagnosed with "hypertension (grade 1, blood pressure 145 / 90 mmHg)" by Western medicine and "vertigo" by Traditional Chinese Medicine. The user stated that he "did not like the hassle and disliked the smell of celery" and hoped to obtain an "easy-to-use, celery-free" dietary therapy plan through the mini-program and track his blood pressure changes in real time.

[0053] In the mini-program built by Qiao Tuoyun, Zhang completed the following steps in sequence: Physical constitution assessment: Fill out a questionnaire (such as "Are you often irritable and easily angered?" or "Do you have many dreams during sleep?"), and the system will automatically associate it with the TCM physical constitution assessment scale after verification by the assessment layer. Taste assessment: Select "dislike celery" in the ingredient preference module, and the system will synchronize it to the taste assessment form in the evaluation layer; Symptom input: Enter the free text "dizziness, flushed face and red eyes, blood pressure 145 / 90 mmHg" in the health record module.

[0054] Health data analysis: By using NLP to analyze "dizziness, flushed face and red eyes" and "hypertension", heterogeneous information is normalized into standardized symptoms: dizziness and headache, flushed face and red eyes; Western medicine diagnosis: hypertension; Traditional Chinese medicine diagnosis: vertigo. Constitution determination: Combining the assessment results ("irritability, restless sleep") with the symptom-syndrome association logic, the constitution is output as: balanced constitution with liver yang hyperactivity, syndrome as: liver yang hyperactivity syndrome.

[0055] Database 1: Matching "Hypertension (Western Medicine) → Dizziness (Traditional Chinese Medicine) → Liver Yang Rising Syndrome (Syndrome Type 001) → Typical Symptoms: Dizziness and headache, flushed face and red eyes; Database 2: Screening for medicinal foods that treat liver yang hyperactivity syndrome (chrysanthemum, celery, cassia seed), excluding celery (user objection), and labeling chrysanthemum as: contraindicated for those with spleen and stomach deficiency and cold (Zhang has no such contraindication), and cassia seed as: contraindicated for those with diarrhea (Zhang has no diarrhea). Database 3: Based on the syndrome of liver yang hyperactivity + balanced constitution + no celery, retrieve the pre-stored formula template.

[0056] Recommendation layer: Two-level constraint filtering Hard constraint filtration: Remove high-sodium foods (such as pickled vegetables) and foods that conflict with antihypertensive drugs (such as licorice). Soft constraint scoring: Scoring is based on ease of processing (weight 40%), low cost (30%), and taste acceptance (30%). Chrysanthemum and cassia seed porridge (ingredients: 10g chrysanthemum, 5g cassia seed, 50g rice, processing time 15 minutes) received the highest score, and purchase links for chrysanthemum and cassia seed were pushed out simultaneously.

[0057] Feedback Optimization Layer: Dynamic Tracking and Iteration Tracking: In the dietary calendar of the mini program, a 2-week follow-up window was set up. Zhang recorded his blood pressure, sleep duration and frequency of dizziness every day, and the system automatically generated status data. Trigger: Data on day 10 showed blood pressure of 140 / 88 mmHg (failed to reach the target value of 135 / 85 mmHg for 3 consecutive days), triggering the P2 level effect condition; Update: The AI ​​algorithm was used to adjust the formula based on blood pressure not reaching the target level (cassia seed increased to 8g, lotus leaf added 5g to enhance the blood pressure lowering effect). Freeze: One week after the adjustment, Zhang's blood pressure stabilized at 132 / 82 mmHg, and his dizziness disappeared. The plan was frozen after meeting the criteria of "no safety triggers for four consecutive weeks, core indicators met, and user satisfaction reached 95%".

[0058] Implementation Results: This solution addresses Mr. Zhang's needs regarding his symptom type, his aversion to celery, and the need for blood pressure monitoring. It establishes a closed-loop system encompassing assessment, recommendation, tracking, optimization, and stabilization. The implementation rate of the solution has increased from 60% under traditional static recommendations to 92%, and the time required to achieve target blood pressure has been shortened by 30%.

[0059] Example 5: Contraindication adaptation service (including allergy filtering) for users with type 2 diabetes (diabetes with yin deficiency and dryness-heat syndrome) Background: Ms. Li, a 50-year-old female user, was diagnosed with type 2 diabetes by Western medicine (fasting blood glucose 8.2 mmol / L). According to traditional Chinese medicine, she suffers from thirst and excessive drinking, fatigue, and an allergy to yam. She needs a low-sugar, yam-free, and easily available dietary therapy plan and is worried about the interaction with her hypoglycemic drug (metformin).

[0060] Mr. Li entered his type 2 diabetes, metformin use, and yam allergy into the health record module of the mini-program. He also entered his thirst, excessive drinking, daily water intake of 2000ml, and fatigue into the symptom feedback module to complete the physical constitution assessment (result: Yin deficiency constitution).

[0061] NLP analysis of type 2 diabetes and excessive thirst leads to the following database association: Western medicine disease name: type 2 diabetes → Traditional Chinese medicine disease name: diabetes mellitus → Syndrome type: Yin deficiency and dryness-heat syndrome (syndrome type 028). Constitution determination: Based on the results of the Yin deficiency constitution assessment and the symptoms of Yin deficiency and dryness-heat syndrome, the constitution / symptom type is output as: Yin deficiency constitution + diabetes and dryness-heat syndrome.

[0062] Database 1: Identify typical symptoms of diabetes and yin deficiency with dryness and heat syndrome and appropriate medicinal and dietary directions (nourishing yin and moistening dryness); Database 2: Screening for medicinal foods (Ophiopogon japonicus, Polygonatum odoratum, Pueraria lobata) that are mainly used to treat yin deficiency and dryness, low sugar, and have no conflict with metformin. Yam is labeled as: user allergy (excluded), red dates are labeled as: high sugar (excluded), and licorice is labeled as: no conflict with metformin (retained). Database 3: Retrieve recipes without yam for individuals with Yin deficiency constitution and diabetes / Yin deficiency with dryness and heat syndrome.

[0063] Recommendation layer: Rigid taboos + flexible adaptation Hard constraint filtering: Remove high-sugar medicinal foods (red dates, longan), allergenic foods (yam), and medicinal foods that conflict with metformin (none). Soft constraint ranking: Scoring by low sugar content (50%), availability (30%), and processing time (20%), the Ophiopogon japonicus, Polygonatum odoratum and Pueraria lobata tea (ingredients: 8g Ophiopogon japonicus, 6g Polygonatum odoratum, 10g Pueraria lobata, ready to drink after brewing) was ranked best, and a link to purchase sugar-free packaged Ophiopogon japonicus from online pharmacies was pushed.

[0064] Follow-up: The follow-up window is set for 3 weeks. Li recorded his fasting blood glucose and the degree of thirst relief every day. The system synchronously linked the database with the pharmacological data on the effect of blood glucose on Ophiopogon japonicus. No trigger: Starting from week 2, Li's fasting blood glucose dropped to 6.8 mmol / L, thirst was relieved, there were no allergies or drug interactions, and no conditions were triggered; Stable output: The system displays the blood sugar decline trend chart through the mini-program data visualization module. Mr. Li confirmed that he was satisfied with the plan and no adjustments were needed.

[0065] Results: The program precisely avoids three major risks: high sugar intake, allergies, and drug interactions. It complies with diabetic dietary guidelines, and Mr. Li's blood sugar control rate has improved by 40%. Due to its convenience of being ready to drink, the daily compliance rate is 100%, solving the pain points of difficulty in controlling professional contraindications and difficulty in obtaining ingredients.

[0066] Example 6: Closed-loop optimization service for users with chronic gastritis (spleen and stomach deficiency syndrome) (unsatisfactory results → solution iteration) Background: Ms. Wang, a 30-year-old female user, was diagnosed with chronic non-atrophic gastritis by Western medicine and with stomach pain and spleen and stomach deficiency syndrome by Traditional Chinese Medicine. She often experienced abdominal distension and dull pain after meals. She initially used the millet and yam porridge program on other platforms, but the abdominal distension worsened because yam is difficult to digest. She hopes that the system can provide a dietary therapy program that can be adjusted and effective.

[0067] Wang submitted feedback in the mini-program's feedback module, stating that he experienced worsening abdominal bloating after meals and difficulty digesting yams. He completed a physical constitution assessment (result: Qi deficiency constitution) and a taste assessment (preferring soft and glutinous foods). He also entered his symptoms of stomach pain after meals and loose stools.

[0068] NLP analysis of chronic non-atrophic gastritis, postprandial bloating, and stomach pain, linking database 1: Western Medicine: Chronic non-atrophic gastritis → Traditional Chinese Medicine: Stomach pain → Syndrome type: Spleen and stomach deficiency syndrome (Syndrome type 015). Constitution determination: Combining Qi deficiency constitution with symptoms of spleen and stomach weakness (loose stools), the constitution / syndrome is output as: Qi deficiency constitution + stomach pain and spleen and stomach weakness syndrome, and it is noted that yam is difficult to digest according to user feedback and needs to be excluded.

[0069] Data Layer and Recommendation Layer: Initial Solution Generation Database 2: Screen for foods and medicines that strengthen the spleen and stomach and are easy to digest (pumpkin, millet, lotus seeds), and exclude yam (difficult to digest) and glutinous rice (greasy); Recommendation layer: Hard constraints filter out irritating ingredients (chili peppers, garlic), soft constraints are scored according to softness (40%), digestibility (30%), and taste (30%). Pumpkin millet lotus seed porridge is recommended (20g pumpkin, 30g millet, 5g lotus seeds, simmered for 30 minutes until soft and easy to chew).

[0070] Follow-up: After one week of follow-up, Wang reported that his postprandial abdominal bloating had subsided, but he still experienced mild, dull pain, and his porridge was slightly bland. Trigger: Abdominal bloating relieved but dull pain persists triggers a P2 level effect condition; bland taste triggers a P3 level experience condition. Prioritize and address effect issues first. Update: Based on the understanding that spleen and stomach deficiency requires strengthening stomach warming, AI has added 2g of ginger (to warm the stomach) to the original formula, and provides an optional option to add a small amount of brown sugar for flavoring.

[0071] Feedback Optimization Layer: Second Iteration and Freezing Follow-up: Two weeks after the adjustment, Wang's records showed that the dull pain disappeared, his taste was acceptable, and his stool was formed; No triggers: No safety / effect / experience-related triggers for 4 consecutive weeks, core indicators (frequency of stomach pain, stool consistency) meet the standards, and user satisfaction is 98%; Freeze: The system automatically freezes the pumpkin, millet, lotus seed, and ginger porridge recipe, marks it as a stable recipe in the mini-program's dietary calendar, can be executed long-term, and exports a 30-day symptom improvement report.

[0072] Results: By implementing an initial plan and one iterative optimization, the problem of Wang's worsening abdominal distension and poor efficacy was resolved. The frequency of stomach pain decreased from 5 times per week to 0 times. The plan's fit increased from the initial 70% to 98%. The closed-loop optimization mechanism demonstrates the advantages of TCM syndrome differentiation and treatment combined with dynamic adjustment. The user's compliance was enhanced due to the visible improvement in symptoms (data visualization).

[0073] This invention features a layered linkage between the assessment layer, analysis layer, recommendation layer, and feedback optimization layer, forming a complete closed loop of physical fitness assessment → solution recommendation → data tracking → optimization. Compared to traditional solutions that only have a one-way process of assessment-recommendation, the dynamic tracking and optimization function of the feedback optimization layer can continuously collect user data and feed it back into the solution iteration, solving the problems of no feedback and difficulty in optimization. This ensures that the dietary therapy plan can adapt to the user's health needs in the long term, significantly improving the stability and sustainability of health management effects.

[0074] Furthermore, the online mini-program in the system's user interaction layer serves as the core carrier, integrating functions such as constitution assessment, food and medicine search, and plan viewing. It transforms professional TCM dietary therapy knowledge into an easy-to-understand operational process, solving the difficulty of translating professional knowledge into practical applications. At the same time, the modular integration of functions at each layer provides a carrier to support the connection between plan recommendations and ingredient acquisition (such as the future expansion of ingredient purchase portals), breaking down the barriers between traditional dietary therapy recommendations and implementation, and allowing ordinary users to easily transform from knowing the plan to implementing it.

[0075] The above embodiments merely illustrate several implementation methods of the present invention, and their descriptions are relatively specific and detailed, but they should not be construed as limiting the scope of the present invention. It should be noted that those skilled in the art can make several modifications and improvements without departing from the concept of the present invention. These are all equivalent modifications and improvements made to the above embodiments based on the essential technology of the present invention, and all of these fall within the protection scope of the present invention.

Claims

1. A TCM constitution-based dietary therapy optimization system based on dynamic feedback and data tracking, characterized in that, It includes a data layer, an evaluation layer, an analysis layer, a recommendation layer, a feedback optimization layer, and a user interaction layer; The data layer is used to construct a standardized database of TCM and Western medicine disease names, syndrome types, and food and drug contraindications. The assessment layer is used to verify and optimize the TCM constitution assessment scale. The analysis layer is used to determine the constitution of the constitution assessment scale and integrate the health data input by the user, match it with the various databases in the data layer, and filter out contraindicated and suitable food and drug based on the matching results. The recommendation layer generates food and drug recommendations based on the analysis layer results from two aspects: dietary contraindications for common diseases and dietary therapy plans for constitution. The feedback optimization layer is used to construct a dynamic data tracking and algorithm optimization module. The user interaction layer is the core carrier, including the online mini-programs built to integrate the above layers.

2. The TCM constitution-based dietary therapy optimization system according to claim 1, characterized in that: The database of Chinese and Western medicine disease names, syndrome types, and drug and food contraindications includes the following databases: Database 1: A database comparing Chinese and Western medicine disease names, syndrome types, and symptoms; Database 2: Information and Contraindications on Food and Medicine Homologous; Database 3: Constitution-Dietary Therapy Recipe Database; Database 3, which is composed of Database 1 and uses Database 2 as a hub, links symptoms and corresponding dietary restrictions. It should be established based on the data of the evaluation results at the evaluation level.

3. The TCM constitution-based dietary therapy optimization system according to claim 2, characterized in that: Database 1 contains the Western medicine names and traditional Chinese medicine names of 114 common diseases, 380 traditional Chinese medicine syndrome types and typical symptoms. Database 2 contains the synonyms, indications, pharmacology and contraindications of 798 kinds of medicinal foods. Database 3 contains medicinal food ingredients, supplementary foods and their effects corresponding to various traditional Chinese medicine constitutions.

4. The TCM constitution-based dietary therapy optimization system according to claim 1, characterized in that: The assessment layer includes taste assessment and physical fitness assessment. The assessment questionnaires are designed using the Likert 5-point method. The physical fitness assessment questionnaire is verified in three stages: accuracy verification stage, satisfaction optimization stage, and deployment stage. When the user satisfaction rate is ≥98% in the satisfaction optimization stage, it enters the deployment stage.

5. The TCM constitution-based dietary therapy optimization system according to claim 4, characterized in that: The accuracy verification stage of the physical fitness assessment questionnaire: collect a large amount of clinical data of people with different physical fitness, integrate them to obtain a simple and easy-to-use physical fitness assessment scale, randomly select 100-200 test subjects from people with certain knowledge of traditional Chinese medicine theory, obtain feedback through surveys, and judge the accuracy of the assessment questionnaire after sorting out the data. The satisfaction optimization phase: Based on the survey results and feedback from the accuracy verification phase, the content of the evaluation questionnaire is further improved in a reasonable way. The scope of the test population is no longer limited, and the number of test subjects is expanded. In the same way as before, after collecting feedback information, the data is sorted out to obtain the user's satisfaction with the evaluation questionnaire. Based on the satisfaction results, the questionnaire is randomly distributed to the population again so that the satisfaction reaches a stable expected value of ≥98%. Use phase: After the satisfaction rate reaches a stable expected value of ≥98%, a Traditional Chinese Medicine (TCM) constitution assessment scale is generated and put into use.

6. The TCM constitution-based dietary therapy optimization system according to claim 4, characterized in that: The specific operation of the taste evaluation questionnaire is as follows: the various medicinal and food ingredients included are archived and classified according to the four natures and five flavors, efficacy, taste and modern pharmacology. The public selects relevant ingredients according to their own tastes. The questionnaire is then filtered based on the public's acceptance of the selected ingredients, and finally a taste evaluation form is generated for use.

7. The TCM constitution-based dietary therapy optimization system according to claim 1, characterized in that: The analysis layer includes health data analysis and physical condition assessment; The health data analysis uses NLP parsing to compare the user-inputted free text symptoms with traditional Chinese and Western medicine principles. The disease names are uniformly combined to form standardized symptoms; The constitution determination is based on the combination of health data analysis results and the TCM constitution assessment scale. The combination logic is to combine the symptoms-syndrome types in the TCM constitution assessment scale with the standardized symptoms obtained. Based on the combination results, the user's constitution is output and an analysis table is generated.

8. The TCM constitution-based dietary therapy optimization system according to claim 1, characterized in that: The recommendation layer supports dual-path retrieval using symptom keywords or Chinese and Western medicine disease names, and generates drug and food recommendations using a two-level constraint retrieval mechanism. The first level is based on a standardized knowledge base of disease / syndrome type, medicine / food, and contraindications, and implements hard constraint filtering to eliminate conditions that cannot be violated, such as allergies, drug-food incompatibilities, contraindications for chronic diseases, and drug combination conflicts. The second level, based on user taste preferences, processing time, cost availability, and cultural taboos, uses configurable weighted multi-indicator scoring and ranking to output a set of safe and highly feasible candidate food and medicinal formulas, and provides links to purchase ingredients.

9. The TCM constitution-based dietary therapy optimization system according to claim 1, characterized in that: The feedback optimization layer dynamically adjusts the dietary therapy plan by tracking, triggering, updating, and freezing the closed-loop control law; Within the follow-up window set in the user interaction layer, the system continuously tracks multiple indicators of the user, including diet, symptoms, weight, and sleep duration, and generates status data. Set trigger conditions and priority judgments, and determine whether to enter the optimization process based on preset data; When the triggering conditions are met, the dietary therapy plan is updated through AI algorithms; if the triggering conditions are not met or the plan reaches the preset stable effect, the plan is frozen.

10. The TCM constitution-based dietary therapy optimization system according to claim 9, characterized in that: The triggering condition determination is based on three types of quantitative triggering conditions: security triggering conditions, effect triggering conditions, and experience triggering conditions. The priority is set according to P1 level security category > P2 level effect category > P3 level experience category. When multiple conditions conflict, the higher priority and the item with the earlier number within the same priority are processed first.

Images