Diet guidance system and method for medical interaction of hyperlipemia patient

Abstract

The invention provides a diet guidance system and method for medical interaction of a hyperlipemia patient, and is applied to the technical field of medical data processing. The method comprises the following steps: firstly, collecting demographic, clinical index, medical history and risk factor data for a hyperlipemia patient, and constructing a training sample set; dividing risk levels by using a Python model based on a guide, and screening intervention crowds; according to consensus and guide, energy intake is customized according to BMI and physical labor intensity, a dietary structure is constructed and verified, and a personalized dietary scheme is generated; furthermore, VR education content is developed, and fat visualization is realized through 3D modeling, physical attribute configuration and a rendering scheme; meanwhile, standardizing 9 types of food materials by taking 5g of fat as a unit, converting intake exchange parts and calculating the total fat amount; and finally, the compliance of the patient is improved by means of a VR gamification module.

Description

Technical Field

[0001] This invention relates to the field of medical data processing technology, and in particular to a dietary guidance system and method for medical interaction in patients with hyperlipidemia. Background Technology

[0002] Hyperlipidemia is a significant risk factor for cardiovascular and cerebrovascular diseases such as atherosclerosis, angina pectoris, myocardial infarction, and cerebral thrombosis. Currently, dietary control is one of the effective means of lowering blood lipids, emphasizing a low-salt, low-fat diet and limiting the intake of high-salt, high-fat, and high-cholesterol foods. While some hyperlipidemia patients can benefit from daily dietary management devices or apps, there are still many shortcomings. For example, many devices only provide simple dietary reminders or basic nutritional education, lacking a standardized dietary assessment system that matches clinical guidelines. Patients find it difficult to accurately control their daily nutrient intake, easily leading to fluctuations in blood lipid levels due to improper diet. Existing wearable devices with body indicator detection functions are limited by size, sensor accuracy, and cost, and their results are insufficient for medical evaluation, thus hindering their substantial role in clinical dietary guidance. Furthermore, for most patients, especially the elderly, relying solely on text or digital recording tools is often difficult to maintain long-term, and patients are prone to forgetting or misunderstanding doctors' dietary requirements, resulting in poor adherence to dietary management and affecting the effectiveness of blood lipid control.

[0003] Furthermore, current solutions for dyslipidemia have limitations. Most focus only on building assessment models, or solely on health management, or only emphasize dietary therapy and exercise, lacking a system that organically integrates dyslipidemia assessment models, dietary therapy, and health management. The standardization and quantification system is incomplete: there is a lack of standardized measurement methods for the fat content of food ingredients, making it difficult for patients to accurately control their daily total fat intake, and the effectiveness of dietary plans lacks convenient quantitative assessment methods. Education and adherence improvement methods are simplistic: dietary management education relies heavily on static information delivery, lacking immersive and interactive teaching scenarios, resulting in low patient participation and insufficient long-term adherence to dietary plans, thus affecting management effectiveness.

[0004] It should be noted that the information disclosed in the background section above is only used to enhance the understanding of the background of this disclosure, and therefore may include information that does not constitute prior art known to those skilled in the art. Summary of the Invention

[0005] The purpose of this application is to provide a dietary guidance system and method for medical interaction for patients with hyperlipidemia, which at least to some extent overcomes the problems existing in the prior art. For patients with hyperlipidemia, after collecting data, risk levels are classified according to guidelines using a Python model, intervention populations are screened and personalized dietary plans are generated, and an innovative fat exchange fraction calculation model is developed. Through a virtual interactive medical device, combined with front-end and back-end modules, real-time interaction between patients and the system is realized, thereby improving patients' health management capabilities and compliance.

[0006] Other features and advantages of this application will become apparent from the following detailed description, or may be learned in part by practice of the invention.

[0007] According to one aspect of this application, a method for dietary guidance in medical interaction for patients with hyperlipidemia is provided, comprising: obtaining a training sample set consisting of relevant data of patients with hyperlipidemia, including demographic data, clinical indicator data, medical history and risk factor data, wherein the clinical indicator data includes LDL-C, HDL-C, TG, TC, blood glucose, and renal function, and the risk factor data includes history of ASCVD events, high-risk factors, and history of lipid-lowering drug use; processing the relevant data, constructing a risk stratification database based on the Chinese guidelines for lipid management, dynamically classifying patient risk levels through a Python-driven multidimensional stratification model, and generating intervention population screening results; performing dietary calculations on the intervention population screening results, and customizing daily intake according to BMI and physical activity intensity based on the consensus on medical nutrition for dyslipidemia and patient risk levels. This approach involves inputting energy, constructing a basic dietary structure, and adjusting fat intake. It also incorporates the Chinese Dietary Guidelines to verify nutritional balance from multiple dimensions, generating personalized dietary plans. Based on these personalized plans, VR interactive educational content is developed, constructing a database containing food nutritional information, using 3D modeling to generate food model scenes, configuring physical attributes and dynamic rendering schemes to achieve fat visualization and interactive optimization. Based on a fat exchange unit system, nine types of food are standardized with 5g of fat as one exchange unit. Actual exchange units are calculated by converting the actual food intake weight to the corresponding exchange unit weight, and a correction equation is used to calculate the total daily dietary fat intake. A gamified patient education module based on fat exchange units is implemented in a VR environment, providing virtual character explanations, diverse food pairing tutorials, and interactive gamified experiences to improve patients' understanding and adherence to the dietary plan.

[0008] Another aspect of this application discloses a dietary guidance device for medical interaction in patients with hyperlipidemia, comprising: an acquisition module for acquiring a training sample set consisting of relevant data of hyperlipidemia patients, including demographic data, clinical indicator data, medical history, and risk factor data; a processing module for processing the relevant data, constructing a risk stratification database based on the Chinese Guidelines for Lipid Management, dynamically classifying patient risk levels using a Python-driven multidimensional stratification model, and generating intervention population screening results; and performing dietary calculation processing on the intervention population screening results, customizing daily energy intake according to BMI and physical activity intensity based on the consensus on medical nutrition for dyslipidemia and patient risk levels, constructing a basic dietary structure and adjusting fat intake, and combining it with the Chinese Dietary Guidelines for Residents. The guidelines verify nutritional balance from multiple dimensions and generate personalized dietary plans. Based on these personalized plans, VR interactive educational content is developed, a database containing food nutritional information is constructed, food model scenes are generated using 3D modeling, and physical properties and dynamic rendering schemes are configured to achieve fat visualization and interactive optimization. Based on the fat exchange unit system, nine types of food are standardized with 5g of fat as one exchange unit. The actual intake of exchange units is calculated by converting the actual food intake weight to the corresponding exchange unit weight, and the total daily dietary fat intake is calculated using a correction equation. Based on the fat exchange unit gamified patient education module, virtual character explanations, diversified food pairing teaching, and interactive gamified experiences are conducted in a VR environment to improve patients' understanding and adherence to the dietary plan.

[0009] According to another aspect of this application, an electronic device includes: a first processor; and a memory for storing executable instructions of the first processor; wherein the first processor is configured to execute the dietary guidance method for implementing the above-described medical interaction for patients with hyperlipidemia by executing the executable instructions.

[0010] According to another aspect of this application, a computer-readable storage medium is provided having a computer program stored thereon, which, when executed by a second processor, implements the above-described dietary guidance method for medical interaction in patients with hyperlipidemia.

[0011] This application provides a medical interactive dietary guidance system and method for patients with hyperlipidemia. It collects demographic and clinical data on hyperlipidemia patients, classifies risk levels using a Python model based on Chinese lipid management guidelines, and screens intervention populations. Daily energy intake is customized for the intervention population, constructing a dietary structure primarily based on olive oil and plant-based foods, and generating personalized dietary plans in conjunction with relevant guidelines. VR educational content is developed to visualize fat and standardize the calculation of total fat content from nine food categories. A VR gamification module is used to improve patient compliance, forming a comprehensive management process.

[0012] It should be understood that the above general description and the following detailed description are exemplary and explanatory only, and are not intended to limit this disclosure. Attached Figure Description

[0013] Figure 1 This document illustrates a flowchart of a dietary guidance method for medical interaction in patients with hyperlipidemia, provided in an embodiment of this application. Figure 2 This illustration shows a schematic diagram of a dietary guidance device for patients with hyperlipidemia, provided in one embodiment of this application. Detailed Implementation

[0014] The preferred embodiments of the present invention will be described below with reference to the accompanying drawings. It should be understood that the preferred embodiments described herein are for illustration and explanation only and are not intended to limit the present invention.

[0015] The following is combined Figure 1 This application describes a dietary guidance method for medical interaction in patients with hyperlipidemia according to exemplary embodiments of the present application. It should be noted that the following application scenarios are shown only to facilitate understanding of the spirit and principles of the present application, and the embodiments of the present application are not limited in any way. Rather, the embodiments of the present application are applicable to any suitable scenario.

[0016] In one embodiment, this application also proposes a dietary guidance system and method for medical interaction in patients with hyperlipidemia. Figure 1 A schematic flowchart of a dietary guidance method for medical interaction in patients with hyperlipidemia, according to an embodiment of this application, is shown.

[0017] S101: Obtain a training sample set consisting of relevant data from patients with hyperlipidemia, including demographic data, clinical indicator data, medical history, and risk factor data.

[0018] In one implementation, demographic data includes the age, sex, height, and weight of patients with hyperlipidemia. Age is used to differentiate the risk of early-onset coronary heart disease (e.g., men <55 years and women <65 years are considered early-onset), and sex influences risk stratification details (e.g., sex-specific thresholds for certain risk factors). Height and weight are calculated using the formula "BMI = weight (kg) ÷ height (m)²" for customized energy intake (e.g., obese individuals with a BMI ≥ 24 kg / m² require stricter energy restrictions). For example, a 55-year-old male patient (not meeting the criteria for early-onset coronary heart disease), 175 cm tall, weighing 80 kg, has a BMI ≈ 26.1 kg / m² (overweight). According to energy calculation rules, his daily energy intake should be customized according to the overweight standard (1400-1500 kcal / day for men).

[0019] Clinical indicators include LDL-C, HDL-C, TG, TC, blood glucose, and renal function. LDL-C, HDL-C, TG, and TC are core indicators for risk stratification (e.g., LDL-C ≥ 1.4 mmol / L requires intervention in very high-risk individuals). Blood glucose is used to identify diabetes (one of the high-risk factors). Renal function indicators (such as creatinine, combined with age and sex, and calculated using the CKD-EPI formula to determine glomerular filtration rate (eGFR)) are used to determine the presence of CKD stage 3-4 (one of the high-risk factors). The patient's LDL-C was 3.2 mmol / L (above the high-risk threshold of 2.6 mmol / L), HDL-C was 0.9 mmol / L (< 1.0 mmol / L, indicating a high-risk factor), TG was 2.1 mmol / L (≥ 1.7 mmol / L, requiring intervention), TC was 6.5 mmol / L, fasting blood glucose was 5.6 mmol / L (not meeting the criteria for diabetes), and renal function indicators were normal (eGFR ≥ 60 ml / min / 1.73 ml / min). 2 (excluding CKD stages 3-4), the overall assessment suggests that a low-fat diet intervention should be included.

[0020] A history of ASCVD events, including ACS within the past year and previous myocardial infarction, directly affects the risk level (e.g., one event plus two high-risk factors qualifies as "ultra-high risk"). If a patient had an acute myocardial infarction 5 years ago (a history of myocardial infarction), and also has hypertension and smoking (two high-risk factors), they meet the criteria for "ultra-high risk." All factors are key to risk stratification; for example, diabetic patients aged ≥40 years are directly classified as high-risk; patients with CKD stages 3-4 plus LDL-C ≥2.6 mmol / L are also classified as high-risk. A patient with a history of hypertension (blood pressure 145 / 90 mmHg) and a history of smoking (10 cigarettes per day), a total of two high-risk factors, and also with one ASCVD event, will have their risk level upgraded to ultra-high risk. Regardless of current lipid levels, the use of lipid-lowering medications (current or past) is considered an independent intervention indication, requiring the initiation of a low-fat diet. The patient is currently taking atorvastatin (20 mg / day). Even though their LDL-C is already below the target value, they still need to incorporate a low-fat diet intervention to synergize with the medication and enhance its effects.

[0021] S102. The relevant data is processed, a risk stratification database is constructed based on the Chinese guidelines for lipid management, and a multi-dimensional stratification model driven by Python is used to dynamically classify the risk level of patients and generate the intervention population screening results.

[0022] In one implementation, relevant data is organized, and a risk stratification database is constructed based on the Chinese guidelines for lipid management. This database includes basic patient information, clinical biochemical indicators, detailed ASCVD event history, multidimensional high-risk factors, and history of lipid-lowering drug use. Standardized data storage results are generated. Multidimensional high-risk factors include early-onset coronary heart disease, familial hypercholesterolemia, diabetes, hypertension, CKD stage 3-4, and smoking. Basic patient information (age, gender, height, weight), clinical biochemical indicators (LDL-C, HDL-C, TG, TC, blood glucose, renal function), detailed ASCVD event history (ACS within 1 year, previous myocardial infarction, etc.), multidimensional high-risk factors (eight items including early-onset coronary heart disease), and history of lipid-lowering drug use are integrated and stored in a unified format.

[0023] For example, patient Li, female, 60 years old, height 160cm, weight 70kg (BMI=27.3kg / m²); clinical indicators: LDL-C=3.5mmol / L, HDL-C=0.8mmol / L, TG=2.0mmol / L, TC=6.2mmol / L, blood glucose 6.8mmol / L (diabetes), renal function indicators suggest CKD stage 3 (creatinine 180μmol / L, eGFR 37 ml / min / 1.73m²). 2 ASCVD event history: None; High-risk factors: Diabetes mellitus, CKD stage 3; History of lipid-lowering drug use: Currently taking rosuvastatin (10mg / day). All data is stored in the database after field standardization.

[0024] The model processes data from the risk stratification database, dynamically classifying patient risk levels according to target judgment logic using a Python-driven multi-dimensional stratification model, and generating risk level classification results. The Python-driven multi-dimensional stratification model is a rule-based reasoning model (constructed based on the deterministic logic rules of the "Chinese Guidelines for Lipid Management (2023)"), including a five-level stratification architecture, ordered from highest to lowest risk as "Very High Risk → Extremely High Risk → High Risk → Intermediate Risk → Low Risk," with each level corresponding to specific judgment rules. Specifically, the model includes the following structure: Input layer: Receives structured data from the risk stratification database (ASCVD event history, number of high-risk factors, clinical indicators such as LDL-C, age, and disease history). Inference layer: Implements stratification logic through nested conditional statements, prioritizing the determination of high-risk levels (e.g., first checking if it is extremely high risk, then checking extremely high risk, high risk, etc.). Output layer: Returns the patient's corresponding risk level label (string format, such as "Very High Risk," "High Risk," etc.).

[0025] High-risk factor quantification: Each high-risk factor is counted as 1 unit (e.g., hypertension = 1, smoking = 1, ≥2 triggering ultra-high-risk conditions). ASCVD event count: Severe ASCVD events (ACS within 1 year, previous myocardial infarction, etc.) are counted by the number of occurrences (≥2 events directly determine ultra-high risk). Lipid thresholds: LDL-C (e.g., high risk corresponds to ≥2.6mmol / L), TC (e.g., ≥7.2mmol / L corresponds to high risk), etc., are used as auxiliary judgment parameters.

[0026] The data in the risk stratification database is processed, and a Python-driven multi-dimensional stratification model is used to dynamically classify patients' risk levels according to the target judgment logic of "very high risk → extremely high risk → high risk → intermediate risk → low risk", generating risk level classification results: Example 1: The patient is classified as "extremely high risk." The model input data is a 65-year-old male patient who has experienced one ACS event (ASCVD event) within the past year, and has hypertension (one high-risk factor) + smoking (one high-risk factor), with LDL-C = 2.0 mmol / L. The inference layer first checks if the extremely high risk criteria are met: one ASCVD event + two high-risk factors (cumulative ≥ 2). If the criteria are met, the patient is directly classified as "extremely high risk," and no further checks are performed on lower risk levels. The risk level classification result is "extremely high risk."

[0027] Example 2: The patient is classified as "Very High Risk." The input data for the model is a 60-year-old female patient with a history of ischemic stroke (1 ASCVD event) and only one high-risk factor (diabetes), with LDL-C = 1.9 mmol / L. Verification of "Very High Risk": 1 ASCVD event + 1 high-risk factor (less than 2), which does not meet the criteria. Verification of "Very High Risk": A confirmed ASCVD (ischemic stroke) but not reaching the "Very High Risk" level, which meets the criteria. The risk level classification result is "Very High Risk."

[0028] Example 3: The patient is classified as "high-risk." The model input data is a 45-year-old male patient with no history of ASCVD events, suffering from diabetes (age ≥ 40 years), and an LDL-C of 3.0 mmol / L. Validation of very high and extremely high risk: no ASCVD events were found, neither of which were met. Validation of high risk: a diabetic patient aged ≥ 40 years met the criteria. The risk level classification result is "high-risk."

[0029] Example 4: The patient is classified as "Intermediate Risk." The input data for the model is a 50-year-old female patient with no history of ASCVD events, no hypertension, LDL-C = 2.8 mmol / L, and two high-risk factors (smoking, HDL-C < 1.0 mmol / L). Validation of very high to high risk: None of the criteria are met. Validation of intermediate risk: No hypertension, LDL-C 2.6~3.4 mmol / L, and two risk factors are met. The risk level classification is "Intermediate Risk."

[0030] Example 5: The patient is classified as "low risk." The model input data is a 55-year-old male patient with no history of ASCVD events, hypertension, LDL-C = 2.0 mmol / L, and one high-risk factor (age ≥ 45 years). Validation of very high to intermediate risk: None of the criteria are met. Validation of low risk: Hypertension, LDL-C 1.8~2.6 mmol / L, and one risk factor are present, meeting the criteria. The risk level classification result is "low risk."

[0031] Based on the risk level classification results and corresponding LDL-C and TG thresholds, combined with the history of lipid-lowering drug use as an independent intervention indicator, a population requiring low-fat dietary intervention was screened, generating the intervention population screening results. The selected intervention population must simultaneously meet two core conditions: the lipid threshold corresponding to the risk level and a history of lipid-lowering drug use. The specific rules are as follows: Lipid threshold standards are: Low-risk group: LDL-C ≥ 3.4 mmol / L or TG ≥ 1.7 mmol / L; Intermediate / High-risk group: LDL-C ≥ 2.6 mmol / L or TG ≥ 1.7 mmol / L; Very High-risk group: LDL-C ≥ 1.8 mmol / L or TG ≥ 1.7 mmol / L; Extremely High-risk group: LDL-C ≥ 1.4 mmol / L or TG ≥ 1.7 mmol / L. (Note: Exceeding either LDL-C or TG qualifies as a lipid threshold). For the independent intervention indicator, regardless of whether lipid indicators meet the standards, anyone currently or previously using lipid-lowering drugs (such as statins or fibrates) is directly included in the intervention population.

[0032] Example 1: Patient Zhang was included in the intervention program due to high blood lipid levels and medication use. Specifically, Zhang's risk level was "very high risk," with LDL-C = 2.0 mmol / L (≥1.8 mmol / L, meeting the very high risk threshold), and he was taking atorvastatin (20 mg / day). The screening logic required meeting both the criteria of "very high risk LDL-C ≥1.8 mmol / L" and "using lipid-lowering medication." Since both conditions were met, this patient was included in the low-fat diet intervention program.

[0033] Example 2: Inclusion in the intervention due to medication alone. Specifically, patient information: Mr. Wang, risk level "low risk," LDL-C = 3.0 mmol / L (<3.4 mmol / L, below the low-risk threshold), but had previously taken rosuvastatin (discontinued 3 months ago). The screening logic was that although blood lipids were not excessive, the "history of lipid-lowering drug use" served as an independent indicator, triggering the intervention criteria. Therefore, this patient was included in the low-fat diet intervention group.

[0034] Example 3: The patient was included in the intervention program due to simple hyperlipidemia. Specifically, the patient's information was Zhao, with a risk level of "intermediate," a TG level of 2.1 mmol / L (≥1.7 mmol / L, meeting the intermediate-risk threshold), and no history of lipid-lowering drug use. The screening logic was that only those meeting the "intermediate-risk population with TG ≥1.7 mmol / L" target were eligible for the intervention. Therefore, this patient was included in the low-fat diet intervention program.

[0035] Example 4: Not included in the intervention. Specifically, the patient information is Chen, with a risk level of "low risk," LDL-C = 3.2 mmol / L (<3.4 mmol / L), TG = 1.5 mmol / L (<1.7 mmol / L), and no history of lipid-lowering drug use. The screening logic was that since blood lipids were not within the target range and there was no history of medication use, no intervention criteria were met. Therefore, this patient was not included in the low-fat diet intervention population.

[0036] S103 involves processing the dietary calculations based on the screening results of the intervention population, customizing daily energy intake according to BMI and physical labor intensity based on the consensus on medical nutrition for dyslipidemia and the patient's risk level, constructing a basic dietary structure and adjusting fat intake, and verifying nutritional balance from multiple dimensions in conjunction with the Chinese Dietary Guidelines to generate personalized dietary plans.

[0037] In one implementation, patient risk level, BMI, and physical labor intensity information from the intervention population screening results are extracted and classified to generate population characteristic classification information. Five levels—"very high risk, extremely high risk, high risk, intermediate risk, and low risk"—are extracted from the intervention population screening results, each corresponding to a clear clinical judgment criterion (e.g., very high risk is ≥2 ASCVD events or 1 event + ≥2 high-risk factors). After calculation using the formula "BMI = weight (kg) ÷ height (m)²", a dichotomy is used to classify individuals into "<24kg / m² (normal / underweight)" and "≥24kg / m² (overweight / obese)" for differentiated energy intake (e.g., stricter energy restrictions for overweight individuals). Physical labor intensity is classified according to occupation type: bedridden: long-term bedridden patients; light physical labor: clerks, teachers, etc. (low daily activity); moderate physical labor: students, drivers, etc. (moderate activity); heavy physical labor: porters, construction workers, etc. (high-intensity activity). Combining Dimensional Classification Logic: Combining risk level, BMI classification, and physical labor intensity to form a unique classification label, ensuring that subsequent dietary plans are accurately matched to individual characteristics.

[0038] Example 1: Extremely High Risk + Overweight + Moderate Physical Activity. Specifically, the patient information is Mr. Wang, male, 60 years old, with two ACS events within one year (extremely high risk), BMI = 28 kg / m² (≥24), and occupation: bus driver (moderate physical labor). The classification result is "extremely high risk + overweight + moderate physical activity". The classification basis is that the risk level meets the criteria of "≥2 ASCVD events", the BMI is in the overweight range, and the occupation corresponds to moderate physical labor. The combination of these three factors clearly indicates that his diet needs to strictly control fat intake (LDL-C ≥1.4 mmol / L) and energy intake needs to be appropriate for moderate physical activity.

[0039] Example 2: Intermediate risk + normal weight + heavy physical labor. Specifically, the patient information is Ms. Liu, female, 50 years old, with no history of ASCVD, LDL-C = 3.0 mmol / L and two risk factors (intermediate risk), BMI = 22 kg / m² (<24), and occupation: manual laborer. The classification result is "intermediate risk + normal weight + heavy physical labor". The classification is based on the risk level meeting the intermediate risk criteria, normal BMI, occupation corresponding to heavy physical labor, and the need to control fat (LDL-C ≥ 2.6 mmol / L) while ensuring a high energy intake (energy coefficient 45~50 kcal / kg / day).

[0040] Example 3: Very High Risk + Overweight + Bedridden. Specifically, the patient information is Mr. Zhang, male, 70 years old, with a history of ischemic stroke (not reaching the very high risk level, but classified as very high risk), BMI = 25 kg / m² (≥24), and long-term bedridden due to hemiplegia. The classification result is "Very High Risk + Overweight + Bedridden". The classification basis is a very high risk level (requiring LDL-C control ≥1.8 mmol / L), overweight BMI and extremely low activity level, requiring strict restriction of fat and reduction of total energy intake (1300-1400 kcal / day for bedridden overweight men).

[0041] The energy parameters and food nutrient data required for dietary calculations are extracted and statistically analyzed to generate basic dietary nutritional characteristics information. The food nutrient data includes fat, cholesterol, and saturated fat. Energy parameters are determined based on a combination of BMI and physical labor intensity. The energy calculation coefficients are as follows: BMI < 24 kg / m² (normal / underweight): Bedridden: 25~30 kcal / kg / day; Light physical labor (e.g., office worker): 35 kcal / kg / day; Moderate physical labor (e.g., driver): 40 kcal / kg / day; Heavy physical labor (e.g., porter): 45~50 kcal / kg / day. BMI ≥ 24 kg / m² (overweight / obese): Bedridden / light physical labor: Adult men 1300~1400 kcal / day, adult women 1100~1200 kcal / day (fixed values, not calculated based on weight coefficients); Moderate / heavy physical labor: Refer to the coefficients for normal weight individuals, but with a slightly lower upper limit (e.g., moderate physical labor at 30~35 kcal / kg / day). The extraction of energy parameters needs to be combined with the individual's weight to calculate the actual needs (e.g., for a person of normal weight with light physical activity, daily energy = weight × 25~30 kcal / kg) to provide a quantitative basis for the energy allocation of the subsequent dietary plan.

[0042] Based on nine food categories (grains and tubers, eggs, meat, poultry, dairy, beans, vegetables and fruits, nuts, and oils), key nutritional components were extracted as follows: Fat content: the ratio of grams of fat to actual weight for each food category (e.g., 5g fat / 150g whole milk, 5g fat / 50g lean pork, beef, or lamb); Cholesterol content: cholesterol levels in animal-based foods (e.g., approximately 200mg cholesterol per egg, approximately 200-300mg cholesterol per 100g of animal organs); Saturated fat content: the percentage of saturated fat from different fat sources (e.g., 5g saturated fat in 5g butter, 1g saturated fat in 5g olive oil). During extraction, the data must be labeled according to the specific type of food (e.g., whole / low-fat dairy, lean / fatty meat) to ensure that the nutritional data matches the actual food.

[0043] Example 1: Extraction of nutritional components from dairy and eggs. Specifically, the extracted nutrients are: "Whole milk: 5g / 150g fat, 15mg / 150g cholesterol, 3g / 150g saturated fat"; "Low-fat milk: 5g / 500g fat, 10mg / 500g cholesterol, 2g / 500g saturated fat"; and "Egg: 5g / 50g fat (1 egg), 200mg / 50g cholesterol, 1.5g / 50g saturated fat". The generated feature information is: "Whole milk: 5g / 150g fat, 15mg / 150g cholesterol, 3g / 150g saturated fat; Egg: 5g / 50g fat, 200mg / 50g cholesterol, 1.5g / 50g saturated fat".

[0044] Example 2: Extraction of nutrients from oils and nuts. Specifically, the extracted nutrients are: "Olive oil: 5g / 5g fat (100% fat), 1g / 5g saturated fat, 4g / 5g unsaturated fat"; "Walnuts: 5g / 10g fat, 0mg / 10g cholesterol, 0.8g / 10g saturated fat". The generated feature information is: "Olive oil: 5g / 5g fat, 1g / 5g saturated fat; Walnuts: 5g / 10g fat, 0.8g / 10g saturated fat".

[0045] Example 3: Energy parameter extraction (for specific populations). Specifically, patient Li, BMI = 22 kg / m² (normal), light physical labor (teacher): extracted "energy coefficient 35 kcal / kg / d, daily energy = 60 kg × 35 kcal / kg = 2100 kcal / d"; patient Zhang, BMI = 26 kg / m² (overweight), light physical labor (clerk): extracted "fixed energy 1400 kcal / d (male)". The generated feature information is: "Li: energy coefficient 35 kcal / kg / d, daily energy 2100 kcal; Zhang: daily energy 1400 kcal".

[0046] The system processes population characteristic classification information and dietary nutrition baseline information, customizes daily energy intake according to risk level, BMI, and physical labor intensity, constructs a basic dietary structure based on olive oil and plant-based foods, adjusts fat intake, and generates preliminary dietary plan information. Detailed rules for customized daily energy intake are established: For BMI ≥ 24 kg / m² (overweight / obese): Light physical labor (e.g., office workers, teachers): Men 1300-1400 kcal / day, Women 1200 kcal / day (fixed values, not fluctuating based on weight); Moderate physical labor (e.g., drivers): Men 1600 kcal / day, Women 1400 kcal / day; Bedridden: Uniformly 1200-1400 kcal / day for Men, Women 1100-1200 kcal / day. BMI < 24 kg / m² (normal / underweight): Calculated as "weight × energy coefficient" (e.g., 35 kcal / kg / day for light physical activity, 40 kcal / kg / day for moderate physical activity). For example, a 60 kg underweight individual engaged in light physical activity would have a daily energy intake of 60 × 35 = 2100 kcal. Energy customization needs to be combined with risk level; for very high / extremely high-risk patients, the energy intake should be reduced by 5%-10% from the corresponding standard (e.g., 25-30 kcal / kg / day for normal weight, light physical activity, and extremely high-risk individuals).

[0047] The basic dietary structure is built by matching food intake to energy levels. Specifically: core food proportions: plant-based foods (grains, vegetables, fruits, and legumes) should account for more than 70% of the total daily food intake, of which vegetables should be ≥500g / day (dark-colored vegetables should account for 1 / 2), and whole grains should account for more than 1 / 4 of the grains; animal-based foods (fish, poultry, eggs, and dairy) should be consumed in moderation, with fish at 50g / day (deep-sea fish preferred), skinless poultry at 50-75g / day, and one egg / day (those with high cholesterol can reduce this by half); the main oil should be olive oil, with the amount adjusted according to energy levels (1200kcal corresponds to 10g / day, 1400kcal corresponds to 15g / day).

[0048] The specific standards for adjusting fat intake are as follows: Total fat percentage: Strictly controlled at 20%-25% of total energy (e.g., 30-39g / day of fat for 1400kcal / day); Saturated fat restriction: Low / intermediate risk patients: <10% of total energy (<15.6g / day for 1400kcal / day); High / very high / extremely high risk patients: <7% (<10.9g / day for 1400kcal / day); Cholesterol restriction: Basal <300mg / day, Intermediate and above <200mg / day (e.g., avoiding animal organs, limiting the number of egg yolks).

[0049] Example 1: High-risk + Overweight + Low physical activity (Mr. Zhao). Specifically, the customized daily energy intake is 1400 kcal / day (meeting the standard for overweight men with low physical activity). The basic dietary structure is: 175g of grains (45g whole grains), 500g of vegetables (250g dark green vegetables), 200g of fruit; 75g of skinless poultry, 50g of fish, 25g of soybeans, 250ml of skim milk; and 15g of olive oil (as the main source of fat). The fat intake is adjusted to 30-35g of total fat (accounting for 21%-25% of total energy); saturated fat <10.9g (<7% standard for high-risk patients), and cholesterol <200mg / day (avoiding animal organs, 1 egg / day).

[0050] Example 2: Extremely high risk + normal weight + moderate physical strength (Sun). Specifically, the customized daily energy intake is: 65kg × 40kcal / kg / d × 90% (reduced by 10% for extremely high risk) = 2340kcal / d. The basic dietary structure is: 250g grains (65g whole grains), 750g vegetables, 200g fruit; 50g lean meat, 50g fish, 1 egg, 500ml skim milk; 20g olive oil. Fat intake is adjusted to total fat 52-64g (20%-25%); saturated fat <14.0g (<7%, 2340kcal × 7% ≈ 14.0g), cholesterol <200mg / d.

[0051] Example 3: Intermediate risk + Obesity + Bedridden (Ms. Zhou, female). Specifically, the customized daily energy intake is 1200 kcal / day (standard for obese bedridden women). The basic dietary structure is: 150g of grains (38g of whole grains), 500g of vegetables, 200g of fruit; 50g of skinless poultry, 250ml of skim milk; and 10g of olive oil. Fat intake is adjusted to 25-30g of total fat (20%-25%); saturated fat <13.3g (intermediate risk <10%, 1200kcal × 10% ≈ 13.3g), and cholesterol <300mg / day.

[0052] Based on the initial dietary plan information and in conjunction with the Chinese Dietary Guidelines, nutritional balance was verified across five dimensions: food diversity, vegetable and fruit intake, high-quality protein intake, energy distribution, and fatty acid balance. This resulted in personalized dietary plans tailored to the needs of patients at different risk levels. These personalized plans clearly defined daily intake limits for various food groups, as well as restrictions on fat and cholesterol. The food diversity verification required a daily intake of at least 12 different food types and at least 25 different types per week, covering nine food groups (grains and tubers, eggs, meat, poultry, dairy, beans, vegetables and fruits, nuts, and oils). Verification included checking for any missing food categories (e.g., lack of whole grains or legumes) and supplementing them by substitution or addition (e.g., replacing a single grain with a combination of whole grains and refined grains). For vegetable and fruit intake verification, the recommended daily intake was 300-500g of vegetables, with dark green vegetables (such as spinach and broccoli) comprising at least half to ensure adequate intake of dietary fiber and antioxidants. The recommended daily intake of fruits was 200-350g, prioritizing low-GI fruits (such as apples and pears) and avoiding excessive consumption of high-sugar fruits (such as lychees and mangoes). If the test fails to meet the standard, the types of vegetables need to be adjusted or the amount of fruit needs to be increased (e.g., from 200g to 300g).

[0053] Verify high-quality protein intake: Daily total animal-based food intake should be 120-200g, with fish (containing unsaturated fatty acids) comprising at least 1 / 3, poultry (skinless) preferred over red meat, and eggs limited to 1 (this number can be halved for those with high cholesterol). If protein sources are limited (e.g., only pork), increase fish or soy products (e.g., replace 25g pork with 50g fish). Verify energy distribution: Carbohydrates should account for 50-65% of total energy, fat 20-25%, and protein 10-15%. If carbohydrates are too high (e.g., >65%), reduce refined grains and increase high-quality protein (e.g., replace 25g rice with 25g soybeans); if fat is excessive, reduce oil intake (e.g., reduce from 20g to 15g). Fatty acid balance check: Saturated fat: <7% of total energy for patients with intermediate or higher risk, <10% for low / intermediate risk; Unsaturated fat: ≥70% of total fat, preferably obtained from olive oil, deep-sea fish, and nuts (e.g., replacing 5g of butter with 10g of walnuts); Avoid trans fats (fried foods, non-dairy creamer).

[0054] Example 1: Optimization of a personalized dietary plan for a high-risk, overweight, and low-activity individual (Mr. Zhao). The initial plan had the following issues: vegetables were mainly light-colored (cabbage, cucumber) (dark-colored vegetables accounted for less than half), grains were refined rice (no whole grains), and protein consisted only of skinless poultry (no fish). Optimization and validation: Specifically, food diversity was improved by increasing whole grains (45g oats), deep-sea fish (50g salmon), and soybeans (25g), bringing the daily intake to 13 different foods; vegetables and fruits were adjusted to 250g dark-colored (spinach, purple cabbage) + 250g light-colored (cucumber) vegetables, and 200g apples were retained as fruit; high-quality protein was increased to 125g of total animal-based foods (75g skinless chicken breast + 50g salmon), meeting the 120-200g standard; energy distribution was 175g of carbohydrates (grains). The diet provides 50% of energy (175 x 4 ÷ 1400 ≈ 50%), 19% of fat (30 x 9 ÷ 1400 ≈ 19%), and 21% of protein (75 x 4 ÷ 1400 ≈ 21%). After adjustment, carbohydrates are 55%, fat is 22%, and protein is 23% (within the acceptable range). Fatty acid balance: 15g olive oil (12g unsaturated fat) + 50g salmon (3g unsaturated fat), total saturated fat < ​​10.9g (high-risk standard). The final regimen is 1400kcal / day, including 175g grains (45g whole grains), 500g vegetables (250g dark green vegetables), 200g fruit, 75g skinless chicken breast, 50g salmon, 25g soybeans, 15g olive oil, total fat 32g (9.5g saturated fat), and cholesterol 180mg (< 200mg).

[0055] Example 2: Optimization of a personalized dietary plan for a person with extremely high risk, normal weight, and moderate physical strength (Mr. Chen). The initial plan had the following issues: carbohydrates accounted for 70% (excessive), no nuts (insufficient unsaturated fat), and vegetables were 300g (less than 500g). The optimization included: food diversity: adding 10g almonds, 50g buckwheat, and 50g tofu, for a total of 14 different ingredients daily; vegetables and fruits: increasing vegetables to 500g (300g dark-colored + 200g light-colored), and fruits to 300g (blueberries + pears); high-quality protein: 200g of animal-based foods (100g sea bass + 75g lean beef + 25g eggs); energy distribution: reducing refined rice by 50g, increasing beef by 25g, reducing carbohydrates to 60%, fat to 22%, and protein to 18%; fatty acid balance: 20g olive oil + 10g almonds (total unsaturated fat 22g), saturated fat <12g (extremely high risk standard: 1800kcal × 7% ≈ 14g). The final plan is 1800kcal / day, 250g of grains (60g of whole grains), 500g of vegetables, 300g of fruit, 100g of sea bass, 75g of lean beef, 1 egg, 50g of tofu, 20g of olive oil, 10g of almonds, 45g of total fat (11g of saturated fat), and 190mg of cholesterol.

[0056] S104 develops VR interactive educational content based on personalized dietary plans, constructs a database containing food nutrition information, uses 3D modeling to generate food model scenes, configures physical properties and dynamic rendering schemes, and realizes fat visualization and interactive optimization.

[0057] In one implementation, information on food types and nutritional components in a personalized dietary plan is extracted and statistically analyzed to generate dietary nutritional characteristic information. Food types (such as grains, vegetables, and fish) and corresponding nutritional components (fat, cholesterol, and saturated fat content) are extracted from the personalized dietary plan, and the average daily intake and nutritional percentage of each food are calculated. For example, Zhao's personalized plan includes "15g olive oil, 50g salmon, and 45g whole grains." Extraction and statistical analysis revealed: "Olive oil: 15g fat (3g saturated fat); Salmon: 7.5g fat (6g unsaturated fat); Whole grains: 180kcal carbohydrates," generating dietary nutritional characteristic information.

[0058] The data on food appearance, scene layout, and interaction logic required for the development of VR interactive educational content are extracted and statistically analyzed to generate virtual scene construction feature information. Specifically, this includes the color of ingredients (e.g., orange-red for salmon, dark green for spinach) and shape (e.g., liquid for olive oil, block for walnuts); scene layout including the spatial structure of a virtual kitchen (including stove and ingredient shelves) and a VR supermarket (divided into fruit and vegetable sections and meat sections); and trigger actions when users select food (e.g., gesture dragging) and nutritional information pop-up logic (e.g., clicking on food to display fat content).

[0059] For example, extracting "salmon has an orange-red, long strip shape, the virtual kitchen food rack has 3 layers (grains on the top layer, fruits and vegetables on the middle layer, and meat on the bottom layer), and a nutrition label pops up 3 seconds after clicking on the food" to generate virtual scene construction feature information.

[0060] By combining dietary nutritional characteristics and virtual scene construction features with 3D modeling technology, three-dimensional feature information of food models and dietary pairing scenarios is generated. The types and amounts of food in the dietary nutritional characteristics (e.g., 15g olive oil, 50g salmon) are matched one-to-one with the appearance (color, shape) of the food and the scene layout (food shelf partitions, space dimensions) in the virtual scene features, ensuring that the 3D model conforms to both actual nutritional parameters and the interactive requirements of the virtual scene.

[0061] Blender script-based modeling uses the Python API to batch generate basic food models (e.g., according to the rule "15g olive oil = 5cm³ liquid model" and "50g salmon = 15cm×5cm×2cm solid model"), and associates them with nutritional labels (e.g., the model has a built-in "15g fat" attribute). Maya handles the details, specifically adding subdivided surfaces (e.g., salmon scale texture, olive oil liquid sheen) and UV mapping (e.g., spinach dark green vein texture) to the models to enhance visual realism.

[0062] Size parameters are calculated based on actual food intake (e.g., 15g olive oil ≈ 15cm³, corresponding to a virtual model diameter of 3cm and a height of 2cm; 50g lean pork ≈ 50cm³, with a model size of 5cm × 5cm × 2cm). Texture parameters match the realistic texture of the food (e.g., the rough surface of whole grains, the smooth eggshell, and the transparent liquid texture of olive oil). Spatial coordinates are positioned according to scene layout rules (e.g., a virtual kitchen food rack has 3 layers: the top layer [z=120cm] for grains, the middle layer [z=90cm] for fruits and vegetables, and the bottom layer [z=60cm] for meat; each food model corresponds to a unique coordinate value).

[0063] Example 1: Personalized treatment plan model for Zhao (high-risk + overweight + low physical fitness). Specifically, the dietary nutritional characteristics are: 15g olive oil (15g fat, 3g saturated fat), 75g skinless chicken breast (5g fat, 1g saturated fat), and 250g spinach (0g fat). The virtual scene features a virtual kitchen food rack (30cm high, lower layer [z=60cm] for meat, middle layer [z=90cm] for fruits and vegetables, and upper layer [z=120cm] for oils); olive oil is a pale yellow liquid, chicken breast is a pink solid, and spinach is a dark green leaf.

[0064] Blender generates the base models: Olive oil: a cylinder with a diameter of 3cm and a height of 2cm (volume 15cm³), with a built-in "fat 15g" attribute; Chicken breast: a cuboid of 7cm×3cm×3cm (volume 63cm³, matching 75g weight), with a built-in "fat 5g" attribute; Spinach: a cluster of leaves of 10cm×8cm×2cm, with a built-in "fat 0g" attribute. Maya is used for detailing: a transparent liquid texture is added to the olive oil, muscle fiber texture to the chicken breast, and leaf vein texture to the spinach.

[0065] Spatial coordinate positioning: Specifically, olive oil is placed in the top layer, cell 1 (coordinates x=20cm, y=30cm, z=120cm); chicken breast is placed in the bottom layer, cell 3 (coordinates x=60cm, y=30cm, z=60cm); spinach is placed in the middle layer, cell 2 (coordinates x=40cm, y=30cm, z=90cm). 3D feature information is generated: "Olive oil model: dimensions 3cm×3cm×2cm, texture transparent liquid, coordinates (20, 30, 120); Chicken breast model: dimensions 7cm×3cm×3cm, texture muscle fibers, coordinates (60, 30, 60); Spinach model: dimensions 10cm×8cm×2cm, texture leaf veins, coordinates (40, 30, 90)."

[0066] Example 2: Personalized treatment plan modeling for Mr. Chen (extremely high risk + normal weight + moderate physical strength). Specifically, his dietary nutritional characteristics are: 100g of sea bass (3g fat, 2g unsaturated fat), 10g of almonds (5g fat, 4g unsaturated fat), and 50g of buckwheat (1g fat). The virtual scene features are a VR supermarket seafood section (simulated water temperature), a nut section (grid shelves), and a grain section (sealed containers); the sea bass is silver-gray with scales, the almonds are brownish-yellow ovals, and the buckwheat is light brown granules. Blender generates the basic model: sea bass: 20cm×8cm×5cm spindle shape (volume 800cm³, matching 100g weight), with a built-in "3g fat" attribute; almonds: 10 oval granules (each 1cm×0.8cm×0.5cm), with a built-in "5g fat" attribute; buckwheat: 50 granule clusters (each 0.3cm in diameter), with a built-in "1g fat" attribute. Maya's details include adding silver scales for reflective texture to the sea bass, adding shell texture to the almonds, and adding a rough grain texture to the buckwheat.

[0067] The spatial coordinates are as follows: sea bass is placed on the ice platform in the feeding area (coordinates x=100cm, y=50cm, z=80cm); almonds are placed on the second layer of the nut area (coordinates x=200cm, y=50cm, z=100cm); buckwheat is placed in the first row of the grain area (coordinates x=300cm, y=50cm, z=90cm). Three-dimensional feature information is generated, specifically: "The sea bass model has dimensions of 20cm×8cm×5cm, with reflective scales and texture, coordinates (100, 50, 80); almond model: 10 almonds, 1cm×0.8cm×0.5cm, with textured shell pattern, coordinates (200, 50, 100); buckwheat model: 50 grains, 0.3cm in diameter, with a coarse texture, coordinates (300, 50, 90)."

[0068] Based on 3D feature information, physical properties and dynamic rendering schemes are configured, and CNN gesture recognition and Azure voice interaction functions are integrated to generate VR interactive educational content. This VR interactive educational content is used to intuitively display dietary plans and achieve real-time interaction. The physical properties are a density of 0.3-0.9 g / cm³ and a coefficient of friction of 0.2-0.8. The dynamic rendering scheme includes red particle streams for saturated fats and a blue semi-transparent coating layer for unsaturated fats. The physical properties of the food model are defined by setting parameters according to food category. Specifically, the density is: 0.8-0.9 g / cm³ for liquid foods (such as olive oil and salad dressing); and 0.6-0.7 g / cm³ for meats (salmon and chicken breast), 0.5-0.6 g / cm³ for grains (whole grains and buckwheat), and 0.4-0.5 g / cm³ for nuts (almonds and walnuts). Coefficient of friction: 0.2-0.3 for smooth-surfaced foods (such as fish and oil); 0.4-0.5 for rough-surfaced foods (such as grains and vegetables); 0.6-0.8 for sticky foods (such as peanut butter). Physical properties need to match the actual texture of the food (e.g., olive oil has a higher density than water to simulate the feeling of liquid flow; nuts have a lower density to simulate the feeling of light particles).

[0069] Fat visualization is achieved using UnityShaderGraph. Specifically, saturated fats are dynamically displayed as a red particle stream, with particle density increasing with content (e.g., the particle stream intensity increases by 20% for every 1g of saturated fat in butter), and a simulated blood vessel deposition animation is added (particles attach to virtual blood vessel models). Unsaturated fats are presented as a blue semi-transparent wrapping layer, with transparency decreasing with content (e.g., the wrapping layer transparency is 30% when olive oil contains 80% unsaturated fat), and flowing light effects are added (simulating the protective effect on blood vessels). Rendering performance must meet a frame rate of ≥90fps and support the simultaneous display of 200+ food models.

[0070] CNN gesture recognition, with 5 preset core gestures (clenched fist to select, open fist to cancel, swipe to move, zoom to adjust components, double-tap to view details), has a recognition response time of ≤0.5 seconds and is compatible with VR controller movements (such as Oculus Touch controllers). Azure voice interaction supports 10 types of voice commands (querying nutritional information, replacing ingredients, calculating exchange portions, etc.), with a semantic recognition accuracy of ≥95%. The feedback method is virtual character voice + text pop-up (e.g., when querying "egg cholesterol", the virtual nutritionist simultaneously announces "200mg / egg" and displays the value).

[0071] Example 1: VR interactive content configuration for Zhao (high-risk + overweight + low physical strength). Specifically, the food models and physical properties are as follows: 15g olive oil: density 0.9g / cm³, coefficient of friction 0.2 (smooth liquid texture); 75g skinless chicken breast: density 0.6g / cm³, coefficient of friction 0.3 (slightly rough meat surface); 45g whole grains: density 0.5g / cm³, coefficient of friction 0.5 (rough granular texture).

[0072] The olive oil contains 3g of saturated fat distributed as a red particle stream (low intensity), and 12g of unsaturated fat is wrapped in a blue translucent layer (40% transparency). Chicken breast contains 5g of fat, with 1g of saturated fat (red particle stream) and 4g of unsaturated fat (blue layer), dynamically displayed as the model rotates. When a user selects olive oil by clenching their fist and slides it to a virtual plate, the portion automatically matches to 15g (the model size scales synchronously). By asking "How much saturated fat?" via voice, the system responds, "Olive oil contains 3g of saturated fat, accounting for 27.5% of the daily limit (10.9g)," with the virtual character nodding in acknowledgment.

[0073] Example 2: VR interactive content configuration for Mr. Chen (extremely high risk + normal weight + moderate physical strength). Specifically, the food models and physical properties are as follows: 100g sea bass: density 0.7g / cm³, coefficient of friction 0.3 (fish surface is smooth); 10g almonds: density 0.4g / cm³, coefficient of friction 0.4 (nut shell is slightly rough); 20g olive oil: density 0.9g / cm³, coefficient of friction 0.2.

[0074] The 3g fat from the sea bass is entirely unsaturated fat, encapsulated in a blue translucent layer (30% transparency), with no red particles. The 5g fat from the almond contains 0.8g saturated fat (red particle flow) and 4.2g unsaturated fat (blue layer), with particles splashing during chewing (simulating digestion). Double-clicking the almond model prompts "1 exchange serving (5g fat)". Zooming the model adjusts the intake (fat values ​​are updated synchronously when going from 10g to 15g). A voice query to "calculate total fat" returns within 3 seconds: "3g sea bass + 5g almond + 20g olive oil = 28g, meeting the daily 45g limit," and displays a progress bar (62%).

[0075] S105, based on the fat exchange unit system, standardizes 9 types of food with 5g of fat as 1 exchange unit. It calculates the actual intake of exchange units by converting the actual food intake mass into the corresponding exchange unit food mass, and calculates the total daily dietary fat intake by combining the correction equation.

[0076] In one implementation, fat content data for nine categories of food are extracted and statistically analyzed to generate characteristic information on the fat content of the food. The fat content of each of the nine categories (grains and tubers, eggs, meat, poultry, dairy, beans, fruits and vegetables, nuts, and oils) is extracted, and the grams of fat per 100g of food are calculated (e.g., lean pork 10g / 100g, whole milk 3.3g / 100g). Specifically, the fat content of whole milk (3.3g / 100g) and low-fat milk (1.0g / 100g) are extracted from dairy products, and the fat content of lean pork, beef, and lamb (10g / 100g), and skinless chicken breast (5g / 100g) are extracted from meat products, generating characteristic information as follows: "Whole milk: 3.3g / 100g fat; Lean pork: 10g / 100g fat; Skinless chicken breast: 5g / 100g fat."

[0077] The food mass data corresponding to 1 exchange unit required by the fat exchange unit system were extracted and statistically analyzed to generate standard characteristic information for exchange units. Based on the rule of "5g fat = 1 exchange unit", the mass required for each type of food to provide 5g of fat was determined (e.g., 150g of whole milk provides 5g of fat, corresponding to 1 exchange unit), and the exchange unit mass of 9 types of food was statistically analyzed and standardized. In dairy products, "150g of whole milk = 1 exchange unit, 500g of low-fat milk = 1 exchange unit"; in meat products, "50g of lean pork, beef, and lamb = 1 exchange unit, 250g of skinless chicken breast = 1 exchange unit", and the characteristic information generated was "whole milk: 150g / 1 exchange unit; lean pork: 50g / 1 exchange unit; skinless chicken breast: 250g / 1 exchange unit".

[0078] The fat content characteristics of food ingredients and the standard exchange portion characteristics are combined with the actual food intake mass. Actual exchange portion characteristic information is generated by converting the actual intake mass to the corresponding exchange portion mass. The actual food intake mass is divided by the mass corresponding to 1 exchange portion of that food ingredient to obtain the actual number of exchange portions (e.g., if 300g of whole milk was consumed, 300 ÷ 150 = 2 exchange portions). For example, if a patient consumes "300g of whole milk, 100g of lean pork, and 250g of skinless chicken breast" on a certain day, the conversion is: Whole milk: 300g ÷ 150g / exchange portion = 2 exchange portions; Lean pork: 100g ÷ 50g / exchange portion = 2 exchange portions; Skinless chicken breast: 250g ÷ 250g / exchange portion = 1 exchange portion; The generated characteristic information is "Whole milk: 2 exchange portions; Lean pork: 2 exchange portions; Skinless chicken breast: 1 exchange portion".

[0079] Based on the actual exchange portion intake characteristics, the total fat exchange portions are accumulated. Combined with a correction equation, the total daily dietary fat intake is calculated. This total daily dietary fat intake is used to accurately quantify the patient's daily fat intake level. The correction equation is: Total daily dietary fat intake ≈ 1.972 + 1.020 × Total fat exchange portions × 5g. The actual exchange portions are accumulated and substituted into the correction equation "Total daily dietary fat intake ≈ 1.972 + 1.020 × Total fat exchange portions × 5g" to calculate the total daily fat intake. The patient's total exchange portions = 2 + 2 + 1 = 5. Substituting this into the equation, we get: Total daily dietary fat intake ≈ 1.972 + 1.020 × 5 × 5 ≈ 1.972 + 25.5 = 27.472g, generating the result "27.5g / d".

[0080] S106, a gamified patient education module based on fat exchange portions, conducts virtual character explanations, teaches diverse food pairings, and provides interactive gamified experiences in a VR environment, improving patients' understanding and adherence to dietary plans.

[0081] In one implementation, data such as the actual exchange portions ingested and the total daily dietary fat intake in the fat exchange portion system are quantitatively analyzed to generate a fat intake quantification factor. The core of this quantification factor is to quantify the deviation between the patient's actual fat intake and the target value of the personalized dietary plan through multi-dimensional indicators. Specifically, these include: Exchange portion achievement rate: Actual total exchange portions ÷ Target total exchange portions × 100% (reflecting the effectiveness of exchange portion control); Exchange portion deviation: Actual total exchange portions - Target total exchange portions (positive indicates excess, negative indicates deficiency). Fat intake percentage: Actual fat intake ÷ Target fat intake × 100% (reflecting the effectiveness of total fat control); Fat deviation: Actual fat intake - Target fat intake (positive indicates excess, negative indicates deficiency). Saturated fat achievement rate: Actual saturated fat intake ÷ Target saturated fat intake × 100% (a specific indicator for patients at intermediate or higher risk). These indicators are generated by combining the "total daily dietary fat intake" calculated using a correction equation with the target values ​​in the personalized plan (such as target exchange portions and target fat amount), directly reflecting the patient's degree of control over fat intake.

[0082] Example 1: High-risk patient Zhao (with a slight deviation between target and actual intake). Specifically, the personalized plan target is 32g of fat per day (target exchange portion 6, since 1 exchange portion = 5g fat, 6 × 5 = 30g, finely adjusted to 32g using the correction equation), and saturated fat <10.9g. Actual exchange portion 5. The actual fat intake calculated using the correction equation is 1.972 + 1.020 × 5 × 5 ≈ 27.5g, and the actual saturated fat is 9.5g. Quantitative analysis results in the following factors: Exchange portion achievement rate = 5 ÷ 6 × 100% ≈ 83%; Exchange portion deviation = 5 - 6 = -1 (less than 1 exchange portion); Fat intake percentage = 27.5 ÷ 32 × 100% ≈ 86%; Fat deviation = 27.5 - 32 = -4.5g (less than 4.5g); Saturated fat achievement rate = 9.5 ÷ 10.9 × 100% ≈ 87%.

[0083] Example 2: Patient Chen, a very high-risk individual (target and actual intake exceeded the target). Specifically, the personalized plan target was 45g of fat per day (target exchange portion 9, 9 × 5 = 45g), and saturated fat <14.0g. Actual intake data was 10 exchange portions. Using the correction equation, the actual fat intake was calculated as 1.972 + 1.020 × 10 × 5 ≈ 52.0g, and the actual saturated fat was 15.2g. Quantitative analysis factors were: Exchange portion achievement rate = 10 ÷ 9 × 100% ≈ 111%; Exchange portion deviation = 10 - 9 = 1 (1 unit exceeding the target); Fat intake percentage = 52.0 ÷ 45 × 100% ≈ 116%; Fat deviation = 52.0 - 45 = 7.0g (7.0g exceeding the target); Saturated fat achievement rate = 15.2 ÷ 14.0 × 100% ≈ 109% (exceeding the target).

[0084] Example 3: Patient Liu, a patient at intermediate risk (target and actual intake fully met). Specifically, the personalized plan target was 30g of fat per day (target exchange portion 6, 6×5=30g), and saturated fat <13.3g. Actual intake data was 6 exchange portions. Calculated using the correction equation, actual fat intake = 1.972 + 1.020 × 6 × 5 ≈ 32.6g (due to a slight systematic bias in the correction equation, it is still considered met), and actual saturated fat was 12.0g. Quantitative analysis factors were: exchange portion compliance rate = 6 ÷ 6 × 100% = 100%; exchange portion deviation = 6 - 6 = 0; fat intake percentage = 32.6 ÷ 30 × 100% ≈ 109% (due to the bias in the correction equation, ±10% fluctuation is allowed, and it is considered met); saturated fat compliance rate = 12.0 ÷ 13.3 × 100% ≈ 90%.

[0085] The food pairing rules and fat restriction standards in personalized dietary plans are quantitatively analyzed to generate dietary plan comprehension factors. These factors assess patients' understanding and consistency of the core rules of the personalized dietary plan through quantifiable indicators. Specifically, these include: rule compliance rate (the degree to which the actual food pairings (e.g., proportion of whole grains, proportion of dark green vegetables) match the plan requirements (number of compliant items ÷ total number of rules × 100%)); diversity achievement rate (actual number of food types consumed ÷ number of types required by the plan (≥12 per day) × 100%); restriction awareness rate (number of fat restriction standards that patients can accurately recite ÷ total number of restriction standards × 100%); and restriction execution rate (number of items whose actual intake meets the restriction standards ÷ total number of restriction items × 100%) (e.g., actual saturated fat < ​​target value is considered compliant). These indicators, based on the clearly defined pairing rules (e.g., "whole grains account for 1 / 4 of cereals") and restriction standards (e.g., "cholesterol <200mg / d for high-risk patients") in the personalized dietary plan, directly reflect the patient's understanding and execution of the plan.

[0086] Example 1: High-risk patient Zhao (high understanding and adherence to the plan). Specifically, the core rules of the personalized plan are as follows: the combination rules are: whole grains account for 1 / 4 of the cereals (175g of cereals contains 45g of whole grains), dark-colored vegetables account for 1 / 2 (500g of vegetables contains 250g of dark-colored vegetables), and the daily food types are ≥12 kinds; the restriction standards are: saturated fat <10.9g, cholesterol <200mg, and total fat 30-35g.

[0087] The actual situation is as follows: Whole grains 45g (25.7% ≈ 1 / 4), dark green vegetables 250g (50%), 13 kinds of food consumed daily; Regarding restriction awareness, the student can accurately state "saturated fat < ​​10.9g, cholesterol < 200mg"; Regarding restriction compliance, saturated fat 9.5g, cholesterol 180mg, total fat 32g (all meet the standards). The dietary plan comprehension factors are calculated as follows: Rule compliance rate = 3 items fully compliant ÷ 3 total rules × 100% = 100%; Diversity achievement rate = 13 kinds ÷ 12 kinds × 100% ≈ 108%; Restriction awareness rate = 2 items reciteable ÷ 2 total standards × 100% = 100%; Restriction compliance rate = 3 items fully compliant ÷ 3 total standards × 100% = 100%.

[0088] Based on quantified fat intake factors and dietary plan comprehension factors, and combined with the immersive interactive architecture of a VR environment, the content explained by virtual characters and teaching information on diverse food pairings are integrated to generate teaching features that link fat exchange portions to dietary plans. Combining quantified fat intake factors (e.g., 83% compliance rate) and dietary plan comprehension factors (e.g., 100% compliance rate), the content explained by virtual characters (e.g., "how to adjust if exchange portions are insufficient") and teaching information on food pairings (e.g., "add 1 serving of nuts to supplement fat") are integrated within the VR environment to form targeted teaching features. For example, regarding Mr. Zhao's "83% compliance rate in exchange portions," the virtual nutritionist explains that "adding 10g of walnuts (1 exchange portion) will meet the target" and demonstrates the pairing of walnuts with salad, generating the feature: "Teaching focus: completing exchange portions; Pairing example: walnut + spinach salad."

[0089] This study analyzes the teaching characteristics related to fat exchange portions and dietary plans using a gamified education module. Through interactive experiences including a daily fat challenge and a healthy restaurant simulation, it generates results demonstrating improved patient adherence to dietary management. The VR game includes a "Daily Fat Challenge" (where food purchases must match the target exchange portion) and a "Healthy Restaurant Simulation" (where a meal meets the required standards). The effectiveness of adherence improvement is assessed based on patient performance (e.g., challenge score, meal conformity). In the "Daily Fat Challenge," Mr. Zhao purchased a total of 6 exchange portions (fully meeting the target), and in the "Healthy Restaurant Simulation," the meal contained 9.5g of saturated fat (meeting the standard). The system determined a "20% improvement in adherence," unlocking the "Advanced Knowledge of Fat Exchange Portions" reward.

[0090] This invention proposes a medical interactive dietary guidance system and method for patients with hyperlipidemia, achieving personalized dietary guidance through multiple stages. First, demographic and clinical data are collected. Based on the Chinese Dietary Lipid Management Guidelines, a Python model is used to classify risk levels and screen intervention populations. For the intervention population, daily energy intake is customized according to risk level, BMI, and physical activity level, constructing a dietary structure mainly based on olive oil and plant-based foods. This is verified against the Chinese Dietary Guidelines to generate personalized dietary plans. VR educational content is developed, 3D modeling is used to generate food models, and physical properties and rendering schemes are configured to visualize fat. Nine types of food are standardized using 5g of fat as a unit, converting exchange portions and calculating total fat. Based on a VR gamification module, virtual explanations, food pairing tutorials, and interactive experiences are conducted to improve patient compliance and form a full-process management system from risk identification to behavioral intervention.

[0091] In one implementation, such as Figure 2 As shown, this application also provides a medical interactive dietary guidance device for patients with hyperlipidemia, comprising: The acquisition module 201 is used to acquire a training sample set consisting of relevant data of patients with hyperlipidemia, including demographic data, clinical indicator data, medical history and risk factor data; Processing module 202 is used to process relevant data. Based on the Chinese Guidelines for Lipid Management, a risk stratification database is constructed. A Python-driven multi-dimensional stratification model dynamically classifies patient risk levels and generates intervention population screening results. Dietary calculations are performed on the intervention population screening results. Based on the consensus on medical nutrition for dyslipidemia and patient risk levels, daily energy intake is customized according to BMI and physical activity intensity. A basic dietary structure is constructed and fat intake is adjusted. Nutritional balance is verified from multiple dimensions using the Chinese Dietary Guidelines, generating personalized dietary plans. VR interactive educational content is developed based on these personalized dietary plans. A database containing food nutritional information is constructed. 3D modeling is used to generate food model scenes, configuring physical attributes and dynamic rendering schemes to achieve fat visualization and interactive optimization. Based on a fat exchange unit system, nine types of food are standardized with 5g of fat as one exchange unit. Actual exchange units are calculated by converting the actual food intake mass to the corresponding exchange unit mass, and the total daily dietary fat intake is calculated using a correction equation. A gamified patient education module based on fat exchange units conducts virtual character explanations, diverse food pairing tutorials, and interactive gamified experiences in a VR environment, improving patients' understanding and adherence to the dietary plan.

[0092] The computer-readable storage medium provided in the above embodiments of this application and the dietary guidance method for medical interaction of hyperlipidemia patients provided in the embodiments of this application are based on the same inventive concept and have the same beneficial effects as the methods adopted, run or implemented by the application stored therein.

[0093] The various embodiments in this application are described in a related manner. Similar or identical parts between embodiments can be referred to mutually. Each embodiment focuses on describing the differences from other embodiments. In particular, the embodiments of the dietary guidance method, electronic device, electronic device, and readable storage medium for evaluating medical interactions in patients with hyperlipidemia are basically similar to the embodiments of the dietary guidance method for medical interactions in patients with hyperlipidemia described above, and therefore are described relatively simply. Relevant parts can be referred to in the descriptions of the embodiments of the dietary guidance method for medical interactions in patients with hyperlipidemia described above.

Claims

1. A dietary guidance method for medical interaction in patients with hyperlipidemia, characterized in that, include: A training sample set consisting of relevant data from patients with hyperlipidemia was obtained. The relevant data included demographic data, clinical indicator data, medical history and risk factor data. Among them, the clinical indicator data included LDL-C, HDL-C, TG, TC, blood glucose and renal function, and the risk factor data included history of ASCVD events, high-risk factors and history of lipid-lowering drug use. The relevant data were processed, a risk stratification database was constructed based on the Chinese guidelines for lipid management, and a multi-dimensional stratification model driven by Python was used to dynamically classify the risk level of patients and generate the intervention population screening results. The results of the intervention population screening were processed for dietary calculation. Based on the medical nutrition consensus on dyslipidemia and the patient's risk level, the daily energy intake was customized according to BMI and physical labor intensity. A basic dietary structure was constructed and fat intake was adjusted. Nutritional balance was verified from multiple dimensions in conjunction with the Chinese Dietary Guidelines to generate personalized dietary plans. Based on personalized dietary plans, VR interactive educational content is developed, a database containing food nutrition information is constructed, food model scenes are generated using 3D modeling, physical properties and dynamic rendering schemes are configured, and fat visualization and interactive optimization are achieved. Based on the fat exchange unit system, 9 types of food are standardized with 5g of fat as 1 exchange unit. The actual intake of exchange units is calculated by converting the actual food intake mass into the corresponding exchange unit food mass, and the total daily dietary fat intake is calculated by combining the correction equation. Based on the gamified patient education module of fat exchange units, virtual character explanations, diversified food pairing teaching, and interactive gamified experiences are carried out in a VR environment to improve patients' understanding and adherence to dietary plans.

2. The method as described in claim 1, characterized in that, The relevant data were processed, and a risk stratification database was constructed based on the Chinese guidelines for lipid management. A Python-driven multi-dimensional stratification model was used to dynamically classify patient risk levels and generate intervention population screening results, including: The relevant data were organized and a risk stratification database was constructed based on the Chinese guidelines for lipid management. This database includes basic patient information, clinical biochemical indicators, detailed ASCVD event history, multidimensional high-risk factors, and history of lipid-lowering drug use. Standardized data storage results were generated. Among them, multidimensional high-risk factors include early-onset coronary heart disease, familial hypercholesterolemia, diabetes, hypertension, CKD stage 3-4, and smoking. The system performs calculations on the data in the risk stratification database, uses a Python-driven multi-dimensional stratification model to dynamically classify patient risk levels according to target judgment logic, and generates risk level classification results. Based on the risk level classification results and the corresponding LDL-C and TG thresholds, combined with the history of lipid-lowering drug use as an independent intervention indicator, the population that needs low-fat diet intervention was screened, and the intervention population screening results were generated.

3. The method as described in claim 2, characterized in that, Dietary calculations were performed on the intervention population screening results. Based on the medical nutrition consensus on dyslipidemia and patient risk levels, daily energy intake was customized according to BMI and physical activity intensity. A basic dietary structure was constructed and fat intake was adjusted. Nutritional balance was verified from multiple dimensions in conjunction with the Chinese Dietary Guidelines to generate personalized dietary plans, including: The patient risk level, BMI, and physical labor intensity information in the intervention population screening results are extracted and classified to generate population characteristic classification information; The energy parameters and food nutrient data required for dietary calculation are extracted and statistically analyzed to generate basic dietary nutrient characteristics information, including food nutrient data such as fat, cholesterol, and saturated fat. The system processes population characteristic classification information and dietary nutrition basic characteristic information, customizes daily energy intake according to risk level, BMI and physical labor intensity, constructs a basic dietary structure based on olive oil and plant foods and adjusts fat intake, and generates preliminary dietary plan information. Based on the preliminary dietary plan information, and in conjunction with the Chinese Dietary Guidelines, nutritional balance was verified from five dimensions: food diversity, vegetable and fruit intake, high-quality protein intake, energy distribution, and fatty acid balance. Personalized dietary plans were then generated to meet the needs of patients at different risk levels. The personalized dietary plans were used to specify the daily intake of various foods and the limits for fat and cholesterol.

4. The method as described in claim 1, characterized in that, Developing VR interactive educational content based on personalized dietary plans, constructing a database containing food nutritional information, generating food models and other scenes using 3D modeling, configuring physical properties and dynamic rendering schemes, and achieving fat visualization and interactive optimization, including: Extract and statistically analyze the food types and nutritional components information in personalized dietary plans to generate dietary nutritional characteristic information; Extract and statistically analyze the food appearance, scene layout, and interaction logic data required for VR interactive educational content development to generate virtual scene construction feature information; By combining dietary nutritional characteristics and virtual scene construction features with 3D modeling technology, three-dimensional feature information of food models and dietary combination scenes are generated. Based on 3D feature information, physical properties and dynamic rendering schemes are configured, and CNN gesture recognition and Azure voice interaction functions are integrated to generate VR interactive educational content. The VR interactive educational content is used to intuitively display dietary plans and realize real-time interaction. The physical properties are a density of 0.3-0.9 g / cm³ and a friction coefficient of 0.2-0.

8. The dynamic rendering scheme includes saturated fat as red particle streams and unsaturated fat as a blue semi-transparent coating layer.

5. The method as described in claim 1, characterized in that, Based on the fat exchange unit system, nine types of food were standardized with 5g of fat as one exchange unit. Actual exchange units were calculated by converting the actual food intake mass to the corresponding exchange unit mass. Combined with a correction equation, the total daily dietary fat intake was calculated, including: The fat content data of nine types of food were extracted and statistically analyzed to generate characteristic information on the fat content of the food. Extract and statistically analyze the food quality data corresponding to 1 exchange portion required by the fat exchange portion system to generate standard characteristic information of exchange portions; The fat content characteristics of food ingredients and the standard characteristics of exchange portions are combined with the actual food intake quality. By converting the actual intake quality to the corresponding exchange portion quality, the actual intake exchange portion characteristics are generated. Based on the actual intake of exchange portions, the total fat exchange portions are accumulated. Combined with the correction equation, the total daily dietary fat intake is calculated. The total daily dietary fat intake is used to accurately quantify the patient's daily fat intake level. The correction equation is: Total daily dietary fat intake ≈ 1.972 + 1.020 × Total fat exchange portions × 5g.

6. The method as described in claim 1, characterized in that, Based on a gamified patient education module using fat exchange fractions, this module provides virtual character explanations, diverse food pairing tutorials, and interactive gamified experiences within a VR environment to enhance patients' understanding and adherence to dietary plans. This includes: The data on actual exchange portions and total daily dietary fat intake in the fat exchange system are quantitatively analyzed and processed to generate a fat intake quantification factor. Quantitative analysis and processing of food pairing rules and fat restriction standards in personalized dietary plans are performed to generate dietary plan comprehension factors. Based on the quantification factor of fat intake and the understanding factor of dietary plan, combined with the immersive interactive architecture of VR environment, the content of virtual character explanation and teaching information of diverse food pairings are integrated and processed to generate teaching features that associate fat exchange portions with dietary plans. Based on the gamified education module, the teaching characteristics related to fat exchange portions and dietary plans were analyzed and processed. Through interactive experiences including daily fat challenges and healthy restaurant simulations, results were generated that improved patient dietary management adherence.

7. A dietary guidance device for patients with hyperlipidemia, characterized in that, The device includes: The acquisition module is used to acquire a training sample set consisting of relevant data from patients with hyperlipidemia, including demographic data, clinical indicator data, medical history and risk factor data. The processing module handles relevant data, constructs a risk stratification database based on the Chinese Guidelines for Lipid Management, dynamically classifies patient risk levels using a Python-driven multi-dimensional stratification model, and generates intervention population screening results. It then performs dietary calculations on the intervention population screening results, customizing daily energy intake based on BMI and physical activity intensity according to the medical nutrition consensus on dyslipidemia and patient risk levels. This constructs a basic dietary structure and adjusts fat intake, verifying nutritional balance from multiple dimensions using the Chinese Dietary Guidelines to generate personalized dietary plans. Based on these personalized dietary plans, it develops VR interactive educational content, constructs a database containing food nutritional information, generates food model scenes using 3D modeling, configures physical attributes and dynamic rendering schemes, and achieves fat visualization and interactive optimization. Based on a fat exchange unit system, it standardizes nine types of food with 5g of fat as one exchange unit, calculates actual exchange units by converting actual food intake mass to the corresponding exchange unit mass, and calculates the total daily dietary fat intake using a correction equation. Finally, a gamified patient education module based on fat exchange units provides virtual character explanations, diverse food pairing tutorials, and interactive gamified experiences in a VR environment, improving patients' understanding and adherence to the dietary plan.

8. An electronic device, characterized in that, include: First processor; and memory for storing executable instructions of the first processor; The first processor is configured to execute the dietary guidance method for medical interaction of patients with hyperlipidemia as described in any one of claims 1 to 6 by executing the executable instructions.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the second processor, it implements the dietary guidance method for medical interaction of patients with hyperlipidemia as described in any one of claims 1 to 6.

Images