A method and related device for intelligent recommendation of diet for patients with chronic kidney disease

Abstract

The application discloses a kind of chronic kidney disease patient diet intelligent recommendation method and related equipment, it is related to health management field, the method includes: the multi-source health data of target patient is collected;Based on the multi-source health data described above, medical rule matching is executed, and the fusion chronic kidney disease stage of patient is mapped to nutrient permissible amount calculation model with each test value, to generate phase nutrition constraint parameter;Based on patient complication information, construct multi-disease constraint model, when detecting diabetes, hypertension or hyperuricemia complication, corresponding restriction rule is superimposed to form multidimensional constraint set;Under the premise that the above nutrition constraint parameter and multidimensional constraint set are satisfied, personalized recommendation algorithm is executed in combination with patient diet preference, generates dynamic diet scheme and outputs recipe result.The present application improves the medical precision and real-time of CKD diet guidance, and also improves the clinical safety of patient diet.

Description

Technical Field

[0001] This specification relates to the field of health management, and more specifically, this application relates to a method and related equipment for intelligent dietary recommendation for patients with chronic kidney disease. Background Technology

[0002] Chronic kidney disease (CKD) is a group of chronic, progressive diseases caused by continuous damage to the structure and function of nephrons. It is characterized by high morbidity, high relapse rate, and high disability rate. As the disease progresses, patients often develop azotemia, electrolyte disturbances, acid-base imbalances, and nutritional metabolic disorders. Among these, controlling serum potassium, serum phosphorus, and protein intake is particularly crucial for slowing the deterioration of kidney function.

[0003] In clinical practice, dietary management is a crucial component of CKD patient treatment plans. However, current dietary guidance models primarily rely on manual calculations and experience-based judgment by healthcare professionals, often presented in the form of paper manuals or generic recipes. This makes it difficult to dynamically adjust nutritional intake based on patients' real-time laboratory data. Traditional methods fail to comprehensively consider eGFR stage, fluctuations in laboratory indicators, complication status, and individual dietary preferences, resulting in guidance plans that are generally generalized, outdated, and poorly adhered to.

[0004] In recent years, some medical institutions and third-party health management platforms have attempted to introduce digital diet management tools, but these systems often have significant shortcomings: First, most applications lack a medically compliant nutrition rule engine and lack protein and electrolyte control logic that matches CKD stages; second, while electronic medical record systems can record patient test data, they lack proactive recommendation functions; third, existing diet recommendation apps are generally based on general nutritional models and fail to link test data with medical contraindications (such as fruit selection restrictions in cases of hyperkalemia). Furthermore, CKD patients often have complications such as diabetes, hypertension, or gout, and the dietary restrictions for these diseases are often contradictory, making it difficult for traditional solutions to balance safety and feasibility.

[0005] Therefore, it is necessary to provide a method and related equipment for intelligent dietary recommendations for patients with chronic kidney disease, in order to at least solve some of the above-mentioned problems. Summary of the Invention

[0006] The summary section introduces a series of simplified concepts, which will be further explained in detail in the detailed description section. This summary section is not intended to limit the key and essential technical features of the claimed technical solution, nor is it intended to determine the scope of protection of the claimed technical solution.

[0007] Firstly, this application proposes a method for intelligent dietary recommendations for patients with chronic kidney disease, including: Collect multi-source health data from the target patients. The multi-source health data includes laboratory test data, electronic medical record data, anthropometric data and dietary preference data. Among them, the laboratory test data includes key indicators such as serum potassium, serum phosphorus, eGFR and blood urea nitrogen, and the electronic medical record data includes chronic kidney disease staging, complication diagnosis and current medication information. Based on the above multi-source health data, medical rule matching is performed to map the patient's fused chronic kidney disease stage and various test values ​​to the nutrient allowance calculation model to generate phased nutrient constraint parameters. A multi-disease constraint model is constructed based on patient complication information. When complications such as diabetes, hypertension, or hyperuricemia are detected, corresponding constraint rules are superimposed to form a multi-dimensional constraint set. Under the premise of satisfying the above nutritional constraints and multidimensional constraint set, a personalized recommendation algorithm is executed in combination with the patient's dietary preferences to generate dynamic dietary plans and output the recipe results.

[0008] In one feasible implementation, the above-mentioned personalized recommendation algorithm is executed in conjunction with the patient's dietary preferences to generate a dynamic dietary plan and output the recipe results: Using collaborative filtering and preference learning algorithms, we search for food combinations similar to the patient's historical scores within the range of the patient's set taste preferences, list of foods to avoid, and cooking conditions. The prohibited ingredients are replaced with safe alternative foods that are nutritionally equivalent by using a nutritional equivalence mapping function. When a key indicator is detected to be out of limit, a protective recommendation mechanism is automatically triggered to remove risky ingredients and recalculate the nutrient allocation in order to output the above-mentioned recipe results.

[0009] In one feasible implementation, the personalized recommendation algorithm that incorporates patient dietary preferences includes a two-layer optimization framework consisting of a medical safety constraint layer and a personalized preference layer. The above-mentioned medical safety constraint layer implements constraint screening based on safety thresholds to eliminate food ingredients with high potassium, high phosphorus or high sodium risks; The preference maximization function is executed at the personalized preference layer described above: U(x) = w1 × T(x) + w2 × F(x) w3×R(x), where T(x) is the taste matching degree, F(x) is the familiarity with the ingredients, R(x) is the health risk penalty item, and w1, w2, and w3 are dynamic weights; By using gradient iterative search, a set of candidate dishes that maximizes U(x) is obtained, achieving a balance between individualization and safety.

[0010] In one feasible implementation, the above-mentioned medical rule matching based on the multi-source health data maps the patient's fused chronic kidney disease stage and various test values ​​to a nutrient allowance calculation model to generate phased nutritional constraint parameters, including: The corresponding protein restriction factor k1 and basal phosphorus tolerance Q0 are determined based on the patient's chronic kidney disease stage. Collect the target patient's weight W, recent serum phosphorus level Ps, and target threshold Pt; Calculate the daily allowable protein intake using the formula P=W×k1; According to formula Q p =Q0 (Ps The daily phosphorus budget is calculated by Pt)×λ, where λ is a dynamic correction factor. The above-mentioned daily allowable protein intake P and the above-mentioned ingestible phosphorus budget Q are used to calculate the daily allowable protein intake P and the ingestible phosphorus budget Q. p The nutritional constraints are allocated according to the ratio factor β1:β2:β3 for breakfast, lunch, and dinner, and are automatically generated based on the patient's eating habits.

[0011] In one feasible implementation, the specific steps for determining the proportion factors of breakfast, lunch, and dinner include: The activity monitoring data and continuous blood glucose monitoring results of the target patients over the past seven days were obtained, and the energy consumption intensity E(t) and blood glucose fluctuation amplitude G(t) at different time periods were calculated. Based on the above patient medication schedule and dialysis period, an intraday metabolic weight curve M(t) was constructed, where M(t) = α×E(t) + β×G(t) + γ×D(t), α, β, and γ are the regulatory weights of energy, blood glucose, and drug efficacy, respectively, and D(t) is the drug effect intensity function. Integrating the above intraday metabolic weight curve M(t) over the breakfast, lunch and dinner time intervals, we obtain the energy demand integrals A1, A2 and A3, respectively. The portioning ratio factor is determined by normalizing the above energy demand integral.

[0012] In one feasible implementation, the above method further includes: When the target patients experience nocturnal blood glucose fluctuations or increased metabolic load after dialysis, the weight of the corresponding time interval is automatically increased based on the abnormal period, and the above-mentioned meal proportion factor is adjusted in real time.

[0013] In one feasible implementation, the above method further includes: Collect patients' subjective ratings of the recommended dishes and laboratory test results at follow-up visits; When a patient's subjective score falls below a set threshold, the corresponding taste weight is automatically reduced and the recommendation strategy is adjusted. When an abnormally high level of blood potassium or phosphorus is detected, the nutritional rule matching operation is re-triggered based on the magnitude of the change.

[0014] Secondly, this invention also proposes a smart dietary recommendation system for patients with chronic kidney disease, comprising: The data acquisition unit is used to collect multi-source health data of the target patient. The multi-source health data includes laboratory test data, electronic medical record data, anthropometric data and dietary preference data. Among them, the laboratory test data includes key indicators such as serum potassium, serum phosphorus, eGFR and blood urea nitrogen, and the electronic medical record data includes chronic kidney disease staging, complication diagnosis and current medication information. The generation unit is used to perform medical rule matching based on the above multi-source health data, mapping the patient's fused chronic kidney disease stage and various test values ​​to the nutrient allowance calculation model to generate phased nutrient constraint parameters. Constraint units are used to construct multi-disease constraint models based on patient complication information. When complications such as diabetes, hypertension, or hyperuricemia are detected, corresponding constraint rules are superimposed to form a multi-dimensional constraint set. The recommendation unit is used to execute a personalized recommendation algorithm based on the patient's dietary preferences, under the premise of satisfying the above nutritional constraints and multidimensional constraint set, to generate dynamic dietary plans and output recipe results.

[0015] Thirdly, the present invention also proposes an electronic device comprising: a memory and a processor, characterized in that the processor is used to execute a computer program stored in the memory to implement the steps of the intelligent dietary recommendation method for patients with chronic kidney disease as described in any of the first aspects.

[0016] Fourthly, the present invention also proposes a computer-readable storage medium having a computer program stored thereon, characterized in that, when the computer program is executed by a processor, it implements the steps of the intelligent dietary recommendation method for patients with chronic kidney disease as described in any one of the first aspects.

[0017] In summary, this invention proposes a personalized intelligent dietary recommendation method for chronic kidney disease (CKD) patients based on multi-dimensional health data. Through the fusion of multi-source health data, it achieves dynamic medical management of CKD patients' diets. This invention can access laboratory test data and electronic medical records in real time, dynamically adjusting permissible nutritional intake based on key indicators such as eGFR stage, serum potassium, serum phosphorus, and blood urea nitrogen. It also automatically updates the dietary plan when the patient's condition fluctuates, thus achieving synchronous linkage between dietary prescriptions and the patient's disease progression. This invention introduces a multi-disease constraint mechanism based on rule-based and model-based collaboration. When a patient has diabetes, hypertension, or hyperuricemia, the system automatically overlays corresponding dietary constraint rules and solves for the safe intake range through a multi-objective optimization algorithm, avoiding food conflicts or nutritional imbalances caused by multiple diseases in traditional solutions. This invention achieves personalized taste matching and enhanced compliance through collaborative filtering and preference learning algorithms. It can automatically optimize recommendation results based on the patient's taste rating, list of foods to avoid, and historical dietary records, using nutritional equivalence mapping to replace high-risk foods, enabling patients to achieve higher dietary satisfaction within a safe range. The intelligent closed-loop feedback mechanism of this invention can automatically optimize the recommendation model based on changes in patient subjective scores and follow-up indicators. The system uses a long short-term memory network to predict blood potassium and phosphorus trends and generates early warning instructions before indicators become abnormal, achieving proactive management. This invention not only improves the medical accuracy and real-time performance of CKD dietary guidance but also enhances patient dietary adherence and clinical safety. It can effectively reduce the risk of acute hyperkalemia, hyperphosphatemia, and nutritional imbalance caused by improper diet, providing kidney disease patients with a scientific, personalized, and sustainably optimized dietary management solution.

[0018] Other advantages, objectives and features of this application will be apparent in part from the description which follows, and in part from what those skilled in the art will understand through study and practice of this application. Attached Figure Description

[0019] Various other advantages and benefits will become apparent to those skilled in the art upon reading the following detailed description of preferred embodiments. The accompanying drawings are for illustrative purposes only and are not intended to limit this specification. Furthermore, the same reference numerals denote the same parts throughout the drawings. In the drawings: Figure 1 This is a schematic flowchart of a method for intelligent dietary recommendation for patients with chronic kidney disease, provided in an embodiment of this application. Figure 2 A structural schematic diagram of a smart diet recommendation system for patients with chronic kidney disease provided in this application embodiment; Figure 3 This is a schematic diagram of an electronic device structure provided in an embodiment of this application. Detailed Implementation

[0020] The terms "first," "second," "third," "fourth," etc. (if present) in the specification, claims, and accompanying drawings of this application are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments described herein can be implemented in a sequence other than that illustrated or described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus. The technical solutions of the embodiments of this application will now be clearly and completely described in conjunction with the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of this application, and not all of them.

[0021] Please see Figure 1 This is a flowchart illustrating a method for intelligent dietary recommendations for patients with chronic kidney disease, provided in an embodiment of this application. Specifically, it may include: Firstly, this application proposes a method for intelligent dietary recommendations for patients with chronic kidney disease, including: S110. Collect multi-source health data of the target patient. The multi-source health data includes laboratory test data, electronic medical record data, anthropometric data and dietary preference data. Among them, the laboratory test data includes key indicators such as serum potassium, serum phosphorus, eGFR and blood urea nitrogen. The electronic medical record data includes chronic kidney disease staging, complication diagnosis and current medication information. S120. Based on the above multi-source health data, perform medical rule matching to map the patient's fused chronic kidney disease stage and various test values ​​to the nutrient allowance calculation model to generate phased nutrient constraint parameters. S130. Construct a multi-disease constraint model based on patient complication information. When complications such as diabetes, hypertension, or hyperuricemia are detected, superimpose corresponding constraint rules to form a multi-dimensional constraint set. S140. Under the premise of satisfying the above nutritional constraint parameters and multidimensional constraint set, a personalized recommendation algorithm is executed in combination with the patient's dietary preferences to generate a dynamic diet plan and output the recipe results.

[0022] For example, the system first automatically collects multi-source health data from patients through the Hospital Information System (HIS) interface, including laboratory test data, electronic medical record information, anthropometric parameters, and dietary preference data filled in by patients on mobile devices. Laboratory test data, from the Laboratory Information System (LIS), covers key biochemical indicators such as serum potassium, serum phosphorus, eGFR, and blood urea nitrogen; electronic medical record data includes chronic kidney disease (CKD) staging information, diagnoses of accompanying complications (such as diabetes, hypertension, gout, etc.), and current medication use; anthropometric data is used to calculate weight, BMI, and dry weight; patient preference data includes taste preferences (salty, sweet, and spicy ratings), a list of foods to avoid (such as seafood and dairy products), and cooking conditions (whether steaming, baking, and stir-frying equipment are available). Through the fusion and collection of the above multi-source data, the system establishes a multi-dimensional profile of the patient's current health status and behavioral habits.

[0023] The system executes medical rule matching based on the collected data. First, it determines the corresponding protein restriction standard based on the patient's eGFR stage (e.g., when eGFR < 30 ml / min, protein intake is limited to 0.6 g / kg / day), and then combines this with indicators such as serum potassium and phosphorus to determine the permissible range of corresponding nutrients. When serum phosphorus levels are detected to be higher than the target value, the system triggers a low-phosphorus diet rule, automatically adjusting the daily phosphorus intake limit. By mapping chronic kidney disease staging and real-time laboratory test data to a nutrient permissible intake calculation model, the system calculates the core nutritional parameters such as protein, sodium, potassium, and phosphorus that the patient can safely ingest at the current stage, forming stage-specific nutritional constraint parameters and providing a medical basis for subsequent recommendations.

[0024] A multi-disease constraint model is established based on complication information in the patient's electronic medical record. When a patient is detected to have diabetes, hypertension, or hyperuricemia, the system will apply corresponding nutritional restriction rules. For example, a low GI (glycemic index) constraint is automatically applied to diabetic patients, a sodium restriction constraint is applied to hypertensive patients, and a low-purine diet constraint is applied to hyperuricemic patients. If a patient has multiple complications, the system uses a multi-objective constraint solving algorithm to integrate various constraints and dynamically form a multi-dimensional constraint set to ensure that the recommended plan simultaneously meets the medical safety requirements for managing multiple diseases.

[0025] Under the premise of meeting the aforementioned medical constraints, a personalized recommendation algorithm is invoked to generate dietary plans. This algorithm combines collaborative filtering with a preference learning model, taking the patient's dietary taste characteristics, historical scores, and contraindications as input. By calculating the nutritional equivalence and risk weights among different ingredients, a set of candidate ingredients that meet safety thresholds is selected. Subsequently, the system performs multi-objective optimization based on the patient's taste preferences to generate personalized meal plans. For example, for a patient who prefers Sichuan cuisine and has hyperphosphatemia, the system will automatically replace traditional fermented bean paste with low-potassium fermented bean paste and recommend low-phosphorus, high-protein combinations such as chicken breast and tofu, while proactively blocking high-risk dishes containing phosphate additives (such as pickled fish and ham sausage). Finally, the system outputs illustrated recipes with nutritional annotations and risk warnings, where each dish is labeled with its protein, sodium, potassium, and phosphorus content for reference by patients and medical staff.

[0026] This invention proposes a personalized intelligent dietary recommendation method for chronic kidney disease (CKD) patients based on multi-dimensional health data. Through the fusion of multi-source health data, it achieves dynamic medical management of CKD patients' diets. This invention can access laboratory test data and electronic medical records in real time, dynamically adjusting permissible nutritional intake based on key indicators such as eGFR stage, serum potassium, serum phosphorus, and blood urea nitrogen. It also automatically updates the dietary plan when the patient's condition fluctuates, thus achieving synchronous linkage between dietary prescriptions and the patient's disease progression. This invention introduces a multi-disease constraint mechanism based on rule-based and model-based collaboration. When a patient has diabetes, hypertension, or hyperuricemia, the system automatically overlays corresponding dietary constraint rules and solves for a safe intake range through a multi-objective optimization algorithm, avoiding food conflicts or nutritional imbalances caused by multiple diseases in traditional solutions. This invention achieves personalized taste matching and enhanced compliance through collaborative filtering and preference learning algorithms. It can automatically optimize recommendation results based on the patient's taste rating, list of foods to avoid, and historical dietary records, using nutritional equivalence mapping to replace high-risk foods, enabling patients to achieve higher dietary satisfaction within a safe range. The intelligent closed-loop feedback mechanism of this invention can automatically optimize the recommendation model based on changes in patient subjective scores and follow-up indicators. The system uses a long short-term memory network to predict blood potassium and phosphorus trends and generates early warning instructions before indicators become abnormal, achieving proactive management. This invention not only improves the medical accuracy and real-time performance of CKD dietary guidance but also enhances patient dietary adherence and clinical safety. It can effectively reduce the risk of acute hyperkalemia, hyperphosphatemia, and nutritional imbalance caused by improper diet, providing kidney disease patients with a scientific, personalized, and sustainably optimized dietary management solution.

[0027] In one feasible implementation, the above-mentioned personalized recommendation algorithm is executed in conjunction with the patient's dietary preferences to generate a dynamic dietary plan and output the recipe results: Using collaborative filtering and preference learning algorithms, we search for food combinations similar to the patient's historical scores within the range of the patient's set taste preferences, list of foods to avoid, and cooking conditions. The prohibited ingredients are replaced with safe alternative foods that are nutritionally equivalent by using a nutritional equivalence mapping function. When a key indicator is detected to be out of limit, a protective recommendation mechanism is automatically triggered to remove risky ingredients and recalculate the nutrient allocation in order to output the above-mentioned recipe results.

[0028] For example, after receiving multi-source health data from a target patient, the system first determines the safe range of edible ingredients based on medical rules and inputs these results into the personalized recommendation algorithm module. This module uses the patient's taste preferences, dietary restrictions, and home cooking conditions filled out on the mobile device as input constraints. The algorithm employs a combination of collaborative filtering and preference learning to construct a dietary profile for each patient. Specifically, by analyzing the patient's historical ratings of dishes, the system calculates the similarity of the patient's food preferences with other registered patients, thus obtaining a reference group with similar preferences. Subsequently, the algorithm searches within the patient's allowed food set for combinations of ingredients that overlap with commonly used recipes by the highly similar group, forming a preliminary personalized recommendation list.

[0029] After generating preliminary results, the system further invokes a nutritional equivalence mapping function to replace ingredients with medical contraindications. This mapping function automatically selects alternative foods that are similar to the original ingredients in protein, energy, potassium, and phosphorus content but pose lower risks, based on macronutrient and micronutrient content information from a nutritional database. For example, if a patient has hyperphosphatemia and the original diet includes pork liver, the system will automatically replace it with low-phosphorus, high-protein foods such as chicken breast or egg whites; if a patient is allergic to seafood, the system will automatically recommend combinations of poultry or soy products with equivalent protein value, thus ensuring both nutritional balance and safety.

[0030] Furthermore, during dynamic execution, the system continuously monitors the patient's latest laboratory test results. When key indicators (such as serum potassium and serum phosphorus) exceed safe thresholds, the algorithm automatically triggers a protective recommendation mechanism. This mechanism first identifies high-risk food categories that increase risk (such as high-potassium fruits and processed meat products containing phosphorus additives) and recalculates the upper limit of nutrient allocation based on the latest test values. After correcting the nutritional parameters, the system re-executes collaborative filtering calculations to generate a new dietary recommendation list, ensuring that the recommendations are matched in real time with the patient's current physiological state.

[0031] Through the above implementation methods, the system achieves dynamic personalization of patient dietary recommendations. It can fully consider patients' dietary preferences within safe thresholds, improving the acceptability and adherence of recipes, while also adjusting recommendation strategies promptly based on changes in medical indicators, avoiding metabolic risks caused by delayed adjustments, thus achieving a balance between nutritional precision and medical safety.

[0032] In one feasible implementation, the personalized recommendation algorithm that incorporates patient dietary preferences includes a two-layer optimization framework consisting of a medical safety constraint layer and a personalized preference layer. The above-mentioned medical safety constraint layer implements constraint screening based on safety thresholds to eliminate food ingredients with high potassium, high phosphorus or high sodium risks; The preference maximization function is executed at the personalized preference layer described above: U(x) = w1 × T(x) + w2 × F(x) w3×R(x), where T(x) is the taste matching degree, F(x) is the familiarity with the ingredients, R(x) is the health risk penalty item, and w1, w2, and w3 are dynamic weights; By using gradient iterative search, a set of candidate dishes that maximizes U(x) is obtained, achieving a balance between individualization and safety.

[0033] For example, the personalized recommendation algorithm first enters the medical safety constraint layer, using the patient's current laboratory test results, CKD stage, and complication status as input to perform the first round of screening of the candidate food set. The system eliminates potentially high-risk foods based on the safety threshold ranges of key indicators such as blood potassium, blood phosphorus, and blood sodium. For instance, when a patient has hyperkalemia, the algorithm automatically excludes high-potassium foods such as bananas, orange juice, and spinach; when a patient has high blood phosphorus, it removes animal organs, beans, and processed foods containing phosphate additives. The role of the medical safety constraint layer is to ensure that the output results meet the medical safety standards for CKD patients under all circumstances, eliminating metabolic risks caused by inappropriate food choices at the source.

[0034] After the medical constraint screening is completed, the system enters the personalized preference layer. This layer uses the patient's taste preferences, eating habits, and historical rating data recorded in the mobile questionnaire as the optimization objective to construct a preference maximization function: U(x) = w1×T(x) + w2×F(x) w3×R(x), Here, T(x) represents the taste matching degree of the candidate dishes, used to measure the similarity between the dish's flavor profile (such as salty, spicy, sour, etc.) and the patient's taste preferences; F(x) represents the familiarity with the ingredients, reflecting the patient's acceptance of this type of dish or cooking method; R(x) represents the health risk penalty, measuring the potential risk of the dish to the patient's kidneys or electrolyte balance. w1, w2, and w3 are dynamically adjustable weights, adjusted in real time by the system based on the patient's compliance score and medical safety level. For example, when the system detects low recent patient compliance, it will appropriately increase the weight of w1 to prioritize satisfying taste preferences; while when elevated blood phosphorus or potassium is detected, the weight of w3 will be increased to strengthen medical constraints.

[0035] Based on this, the system uses a gradient iterative search method to continuously optimize the value of U(x) in the candidate ingredient set until the global maximum solution is obtained, thereby selecting the set of dishes that best suits individual preferences and meets safety constraints. This optimization process not only considers the attributes of individual dishes, but also coordinates the nutritional balance between meals globally. For example, if protein intake is increased at lunch, the corresponding protein content is automatically reduced at dinner to maintain the medical rationality of the overall daily intake.

[0036] Through this two-layer optimization architecture, the system achieves a multi-objective balance between prioritizing medical safety and adapting to taste preferences at the algorithmic level. Compared with traditional single recommendation mechanisms, this method can provide higher personalization and patient satisfaction while ensuring dietary safety for CKD patients, thereby significantly improving long-term dietary adherence and clinical management outcomes.

[0037] In one feasible implementation, the above-mentioned medical rule matching based on the multi-source health data maps the patient's fused chronic kidney disease stage and various test values ​​to a nutrient allowance calculation model to generate phased nutritional constraint parameters, including: The corresponding protein restriction factor k1 and basal phosphorus tolerance Q0 are determined based on the patient's chronic kidney disease stage. Collect the target patient's weight W, recent serum phosphorus level Ps, and target threshold Pt; Calculate the daily allowable protein intake using the formula P=W×k1; According to formula Q p =Q0 (Ps The daily phosphorus budget is calculated by Pt)×λ, where λ is a dynamic correction factor. The above-mentioned daily allowable protein intake P and the above-mentioned ingestible phosphorus budget Q are used to calculate the daily allowable protein intake P and the ingestible phosphorus budget Q. p The nutritional constraints are allocated according to the ratio factor β1:β2:β3 for breakfast, lunch, and dinner, and are automatically generated based on the patient's eating habits.

[0038] For example, the process of performing medical rule matching based on multi-source health data involves merging the patient's chronic kidney disease stage, laboratory test indicators, and vital signs data into a nutrient allowance calculation model to dynamically generate stage-specific nutritional constraint parameters that are consistent with the current condition.

[0039] First, the corresponding protein restriction standards and basal phosphorus allowances are determined based on the patient's CKD stage information. Patients at different stages have different glomerular filtration rates (eGFR), and their tolerance to protein metabolites and phosphorus excretion varies considerably. For example, when a patient is in CKD stage 3 (eGFR 30–59 ml / min), the system automatically sets the protein restriction coefficient k1 to 0.8 g / kg / day; while when entering CKD stage 4 (eGFR 15–29 ml / min), it is automatically lowered to 0.6 g / kg / day to reduce the burden on the kidneys. Simultaneously, the basal phosphorus allowance Q0 is set according to medical guidelines as 70%–80% of the intake of healthy individuals, serving as a baseline value for subsequent adjustments.

[0040] The system retrieves the patient's latest test results in real time from the hospital's Laboratory Information System (LIS), including serum phosphorus levels (Ps) and the corresponding target threshold (Pt) (e.g., a target threshold of 1.45 mmol / L). Combined with the patient's weight (W), the system calculates the patient's daily allowable protein intake using the formula P = W × k1. The system also calculates the daily phosphorus budget (Q). p =Q0 (Ps Pt)×λ, where λ is a dynamic correction coefficient used to reflect the impact of deviations in test data on phosphorus intake. When a patient's serum phosphorus level is higher than the target threshold, λ is positive and the deduction is increased to strengthen phosphorus restriction; when the serum phosphorus level is lower than the target threshold, λ is negative, allowing for a moderate relaxation of intake limits, thus achieving bidirectional regulation of nutritional balance.

[0041] After obtaining the daily allowable protein intake (P) and daily phosphorus budget (Q) p Then, based on the patient's daily routine and eating habits, the system will assign P and Q. p The diet is allocated based on a ratio of β1:β2:β3 for breakfast, lunch, and dinner. This ratio is determined not only based on traditional dietary patterns but also on the patient's daily activity level and medication administration time. For example, for patients with high daytime work intensity and who take phosphate binders in the evening, the system might set the ratios to β1=0.25, β2=0.45, and β3=0.30 to ensure sufficient energy intake at lunchtime and to help the medication bind excess phosphorus at dinner. When the system detects nighttime blood glucose fluctuations or increased metabolic load after dialysis, it automatically increases the weighting of the corresponding time periods, achieving adaptive optimization of nutrient allocation.

[0042] The allocated nutritional structure of the three meals is combined with medical constraint parameters to generate a phased nutritional constraint table, which serves as the input for subsequent diet recommendation algorithms. Through this dynamic modeling approach based on clinical indicators, staging standards, and individual behavioral characteristics, the system can achieve precise control over the nutritional intake of CKD patients, ensuring that the diet plan meets medical safety requirements while also possessing real-time and individualized characteristics.

[0043] In one feasible implementation, the specific steps for determining the proportion factors of breakfast, lunch, and dinner include: The activity monitoring data and continuous blood glucose monitoring results of the target patients over the past seven days were obtained, and the energy consumption intensity E(t) and blood glucose fluctuation amplitude G(t) at different time periods were calculated. Based on the above patient medication schedule and dialysis period, an intraday metabolic weight curve M(t) was constructed, where M(t) = α×E(t) + β×G(t) + γ×D(t), α, β, and γ are the regulatory weights of energy, blood glucose, and drug efficacy, respectively, and D(t) is the drug effect intensity function. Integrating the above intraday metabolic weight curve M(t) over the breakfast, lunch and dinner time intervals, we obtain the energy demand integrals A1, A2 and A3, respectively. The portioning ratio factor is determined by normalizing the above energy demand integral.

[0044] For example, the activity monitoring data and continuous glucose monitoring results of the target patient over the past seven days are first obtained. The energy expenditure intensity E(t) at different time periods is calculated using parameters such as exercise volume, step frequency, and metabolic equivalent recorded by wearable devices or smart terminals. Simultaneously, the intraday glucose fluctuation curve is extracted using a continuous glucose monitoring device to obtain the glucose fluctuation amplitude G(t), which reflects the differences in insulin sensitivity and energy requirements of the patient at different time periods.

[0045] Based on this, the system combines the patient's medication schedule and dialysis time information to construct an intraday metabolic weight curve M(t). M(t) is determined by the formula M(t) = α × E(t) + β × G(t) + γ × D(t), where α, β, and γ are the regulatory weights for energy consumption, blood glucose fluctuations, and drug efficacy, respectively; D(t) represents the drug effect intensity function, used to reflect the impact of hypoglycemic drugs, phosphate binders, or diuretics on metabolism at specific times. For example, when a patient takes hypoglycemic drugs before breakfast, D(t) takes a higher value in the morning to reflect the drug's inhibitory effect on energy metabolism.

[0046] The system then integrates M(t) over the three time intervals of breakfast, lunch, and dinner to obtain energy requirement integrals A1, A2, and A3. A larger integral value indicates a higher metabolic load and energy demand during that time period. The system normalizes A1, A2, and A3 to calculate the meal allocation ratio factors β1:A1 / (A1+A2+A3), β2:A2 / (A1+A2+A3), and β3:A3 / (A1+A2+A3). This allows for the acquisition of a dynamic dietary allocation ratio that reflects the individual metabolic characteristics of the patient.

[0047] In further adaptive optimization, the system continuously monitors changes in the patient's blood glucose and activity data. When significant nighttime blood glucose fluctuations or abnormally high energy consumption after dialysis are detected, the system automatically increases the weight γ for the corresponding time period or adjusts the β coefficient, thereby correcting the meal distribution ratio in real time. This method allows the energy distribution ratio of each patient's three meals to change dynamically according to their actual physiological state, avoiding nutritional imbalances caused by fixed ratios and achieving personalized and precise control of dietary structure.

[0048] Through this intelligent meal proportion determination mechanism, the system can not only improve the scientific nature and suitability of nutrient allocation, but also achieve a balance between drug metabolism, energy consumption and blood sugar control in CKD patients, effectively improving the clinical safety and long-term adherence of the diet plan.

[0049] In one feasible implementation, the above method further includes: When the target patients experience nocturnal blood glucose fluctuations or increased metabolic load after dialysis, the weight of the corresponding time interval is automatically increased based on the abnormal period, and the above-mentioned meal proportion factor is adjusted in real time.

[0050] For example, when a target patient is detected to have significant blood glucose fluctuations at night or increased metabolic load after dialysis, a dynamic correction mechanism is automatically triggered to adaptively adjust the metabolic weights for the corresponding time interval, thereby achieving real-time optimization of the meal proportion factor.

[0051] The system acquires the patient's nighttime blood glucose curve using a continuous glucose monitoring device. When the monitoring results show that the blood glucose fluctuation exceeds a preset threshold (e.g., >2.5 mmol / L), it indicates that the patient is at risk of nighttime metabolic instability. In this case, the algorithm automatically increases the weighting coefficient β of the nighttime period in the metabolic weighting curve M(t) to enhance the response to energy demand during that period. By increasing the integral value A3 of the corresponding nighttime interval, the system can appropriately increase the proportion of energy and carbohydrate intake at dinner in the next round of nutrient allocation, preventing patients from experiencing hypoglycemia or metabolic disorders due to insufficient nighttime energy.

[0052] For dialysis patients, the system also automatically analyzes the metabolic recovery curve after dialysis. When a significant increase in metabolic load is detected after dialysis (manifested as a sudden increase in energy consumption E(t) or short-term fluctuations in blood phosphorus and urea nitrogen levels), the algorithm increases the drug efficacy weight γ or energy weight α and recalculates the metabolic weight curve M(t). This adjustment increases the energy demand integral in the post-dialysis period, and the system automatically increases the nutritional allocation ratio of the post-dialysis meal in the next round of meal plan generation, such as increasing the proportion of high-quality protein and low-phosphorus carbohydrates, to accelerate metabolic recovery and maintain stable blood glucose levels.

[0053] Through the aforementioned real-time correction mechanism, the system can dynamically adjust the meal proportion factor according to the individual metabolic changes of patients, keeping the dietary structure synchronized with their physiological state. Unlike traditional fixed-proportion diet plans, the adaptive adjustment method of this invention can proactively intervene when there are nocturnal metabolic fluctuations or changes in dialysis load, significantly improving the accuracy and safety of dietary management for CKD patients, reducing the risks of nutritional imbalance and hypoglycemia, thereby achieving truly individualized dynamic nutritional regulation.

[0054] In one feasible implementation, the above method further includes: Collect patients' subjective ratings of the recommended dishes and laboratory test results at follow-up visits; When a patient's subjective score falls below a set threshold, the corresponding taste weight is automatically reduced and the recommendation strategy is adjusted. When an abnormally high level of blood potassium or phosphorus is detected, the nutritional rule matching operation is re-triggered based on the magnitude of the change.

[0055] For example, the system first records patients' subjective ratings of the daily recommended dishes in the mobile application. When a patient's rating for a certain type of dish is below a set threshold (e.g., below 3 points), the algorithm determines that the patient is dissatisfied with the taste or has low acceptance. The system then constructs a taste deviation vector based on the rating results, calculates the direction and magnitude of the deviation for each taste dimension (such as saltiness, spiciness, and sweetness), and accordingly reduces the weight of that dimension in the recommendation model. For instance, if a patient repeatedly gives low ratings to "spicy" dishes, the system will reduce the weight of the spiciness factor in the taste matching function T(x), thereby automatically increasing the proportion of "mildly spicy" or "mild" dishes in the next round of recommendations, achieving self-learning and dynamic correction of taste features.

[0056] Simultaneously, the system automatically synchronizes laboratory test results after the patient's follow-up visit, including key indicators such as serum potassium, serum phosphorus, and blood urea nitrogen. When an abnormally high level of serum potassium or phosphorus is detected, the system triggers a medical safety priority mechanism, re-invokes the nutrition rule matching module, and adjusts the corresponding nutritional thresholds according to the magnitude of the increase. For example, when the serum phosphorus level rises more than 0.3 mmol / L above the target threshold, the system strengthens the phosphorus restriction strategy, eliminating foods containing phosphate additives and reducing the daily phosphorus budget Q. p When blood potassium levels exceed the upper limit, high-potassium fruits and vegetables will be automatically blocked.

[0057] Through the aforementioned feedback and correction rules, on the one hand, the recommended results can be continuously optimized to improve patient dietary adherence. On the other hand, it ensures that the dietary plan is matched with changes in the patient's condition in real time, thereby balancing nutritional safety and long-term health benefits.

[0058] Secondly, this invention also proposes an intelligent dietary recommendation system for patients with chronic kidney disease, such as... Figure 2 As shown, it includes: The data acquisition unit 21 is used to collect multi-source health data of the target patient. The multi-source health data includes laboratory test data, electronic medical record data, anthropometric data and dietary preference data. Among them, the laboratory test data includes key indicators such as blood potassium, blood phosphorus, eGFR and blood urea nitrogen, and the electronic medical record data includes chronic kidney disease staging, complication diagnosis and current medication information. The generation unit 22 is used to perform medical rule matching based on the above multi-source health data, and map the patient's fused chronic kidney disease stage and various test values ​​to the nutrient allowance calculation model to generate phased nutrient constraint parameters. Constraint unit 23 is used to construct a multi-disease constraint model based on patient complication information. When complications such as diabetes, hypertension or hyperuricemia are detected, corresponding constraint rules are superimposed to form a multi-dimensional constraint set. Recommendation unit 24 is used to execute a personalized recommendation algorithm based on the patient's dietary preferences, under the premise of satisfying the above nutritional constraints and multidimensional constraint set, to generate dynamic dietary plans and output recipe results.

[0059] In one feasible implementation, a smart dietary recommendation system for patients with chronic kidney disease can also perform any step of the method proposed in the first aspect.

[0060] Thirdly, the present invention also proposes an electronic device 300, such as... Figure 3 As shown, it includes a memory 310, a processor 320, and a computer program 311 stored on the memory 310 and executable on the processor. When the processor 320 executes the computer program 311, it implements the steps of the intelligent dietary recommendation method for patients with chronic kidney disease as described in any of the first aspects.

[0061] Fourthly, the present invention also proposes a computer-readable storage medium having a computer program stored thereon, characterized in that, when the computer program is executed by a processor, it implements the steps of the intelligent dietary recommendation method for patients with chronic kidney disease as described in any one of the first aspects.

[0062] It should be noted that the descriptions of each embodiment in the above embodiments have different focuses. For parts that are not described in detail in a certain embodiment, please refer to the relevant descriptions in other embodiments.

[0063] Those skilled in the art will understand that embodiments of this application can be provided as methods, systems, or computer program products. Therefore, this application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this application can take the form of a computer program product embodied on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0064] This application is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of this application. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded computer, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create a machine for implementing the flowchart illustrations. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0065] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0066] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1The steps of the function specified in one or more boxes.

[0067] This application also provides a computer program product, which includes computer software instructions. When the computer software instructions are executed on a processing device, the processing device performs the voice-based identity recognition process in the corresponding embodiment. A computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on a computer, all or part of the flow or function according to the embodiments of this application is generated. The computer may be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions may be stored in a computer-readable storage medium or transmitted from one computer-readable storage medium to another. For example, computer instructions may be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via wired (e.g., coaxial cable, fiber optic, digital subscriber line (DSL)) or wireless (e.g., infrared, wireless, microwave, etc.) means. The computer-readable storage medium may be any available medium that a computer can store or a data storage device such as a server or data center that integrates one or more available media. The available medium may be a magnetic medium (e.g., floppy disk, hard disk, magnetic tape), an optical medium (e.g., DVD), or a semiconductor medium (e.g., solid-state disk (SSD)).

[0068] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working processes of the systems, devices, and units described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0069] In the several embodiments provided in this application, it should be understood that the disclosed systems, apparatuses, and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces, or indirect coupling or communication connection between apparatuses or units, and may be electrical, mechanical, or other forms.

[0070] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0071] Furthermore, the functional units in the various embodiments of this application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit.

[0072] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods of the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0073] The above embodiments are only used to illustrate the technical solutions of this application, and are not intended to limit them. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of this application.

Claims

1. A method for intelligent dietary recommendation for patients with chronic kidney disease, characterized in that, include: Collect multi-source health data from the target patient, including laboratory test data, electronic medical record data, anthropometric data, and dietary preference data. The laboratory test data includes key indicators such as serum potassium, serum phosphorus, eGFR, and blood urea nitrogen. The electronic medical record data includes chronic kidney disease staging, complication diagnosis, and current medication information. Based on the multi-source health data, medical rule matching is performed to map the patient's fused chronic kidney disease stage and various test values ​​to the nutrient allowance calculation model to generate phased nutrient constraint parameters. A multi-disease constraint model is constructed based on patient complication information. When complications such as diabetes, hypertension, or hyperuricemia are detected, corresponding constraint rules are superimposed to form a multi-dimensional constraint set. Under the premise of satisfying the nutritional constraints and multidimensional constraint set, a personalized recommendation algorithm is executed in combination with the patient's dietary preferences to generate a dynamic diet plan and output the recipe results.

2. The intelligent dietary recommendation method for patients with chronic kidney disease according to claim 1, characterized in that, The process involves combining the patient's dietary preferences with a personalized recommendation algorithm to generate dynamic dietary plans and output the resulting recipes. Using collaborative filtering and preference learning algorithms, we search for food combinations similar to the patient's historical scores within the range of the patient's set taste preferences, list of foods to avoid, and cooking conditions. The prohibited ingredients are replaced with safe alternative foods that are nutritionally equivalent by using a nutritional equivalence mapping function. When a key indicator is detected to be out of limit, a protective recommendation mechanism is automatically triggered to remove risky ingredients and recalculate the nutritional allocation in order to output the resulting recipe.

3. The intelligent dietary recommendation method for patients with chronic kidney disease according to claim 2, characterized in that, The personalized recommendation algorithm, which incorporates patient dietary preferences, includes a two-layer optimization framework consisting of a medical safety constraint layer and a personalized preference layer. The medical safety constraint layer performs constraint screening based on safety thresholds to eliminate food ingredients that pose a risk of high potassium, high phosphorus, or high sodium content. The preference maximization function is executed at the personalized preference layer: U(x) = w1 × T(x) + w2 × F(x) w3×R(x), where T(x) is the taste matching degree, F(x) is the familiarity with ingredients, R(x) is the health risk penalty item, and w1, w2, and w3 are dynamic weights; By using gradient iterative search, a set of candidate dishes that maximizes U(x) is obtained, achieving a balance between individualization and safety.

4. The intelligent dietary recommendation method for patients with chronic kidney disease according to claim 1, characterized in that, The process of performing medical rule matching based on the multi-source health data maps the patient's chronic kidney disease fusion stage and various test values ​​to a nutrient allowance calculation model to generate phased nutritional constraint parameters, including: The corresponding protein restriction factor k1 and basal phosphorus tolerance Q0 are determined based on the patient's chronic kidney disease stage. Collect the target patient's weight W, recent serum phosphorus level Ps, and target threshold Pt; Calculate the daily allowable protein intake using the formula P=W×k1; According to formula Q p =Q0 (Ps The daily phosphorus budget is calculated by Pt)×λ, where λ is a dynamic correction factor. The daily allowable protein intake P and the ingestible phosphorus budget Q are used to determine the daily allowable protein intake P and the ingestible phosphorus budget Q. p The nutritional constraints are allocated according to the ratio factor β1:β2:β3 for breakfast, lunch, and dinner, and are automatically generated based on the patient's eating habits.

5. The intelligent dietary recommendation method for patients with chronic kidney disease according to claim 4, characterized in that, The specific steps for determining the proportion factors for breakfast, lunch, and dinner include: The activity monitoring data and continuous blood glucose monitoring results of the target patient over the past seven days were obtained, and the energy consumption intensity E(t) and blood glucose fluctuation amplitude G(t) at different time periods were calculated. Based on the patient's medication schedule and dialysis period, an intraday metabolic weight curve M(t) was constructed, where M(t) = α × E(t) + β × G(t) + γ × D(t), α, β, and γ are the regulatory weights of energy, blood glucose, and drug efficacy, respectively, and D(t) is the drug effect intensity function. The daily metabolic weight curve M(t) is integrated over the time intervals of breakfast, lunch, and dinner to obtain the energy requirement integrals A1, A2, and A3, respectively. The portioning ratio factor is determined by normalizing the integral of the energy demand.

6. The intelligent dietary recommendation method for patients with chronic kidney disease according to claim 5, characterized in that, The method further includes: When the target patient experiences nocturnal blood glucose fluctuations or increased metabolic load after dialysis, the weight of the corresponding time interval is automatically increased according to the abnormal period, and the meal proportion factor is adjusted in real time.

7. The intelligent dietary recommendation method for patients with chronic kidney disease according to claim 1, characterized in that, Also includes: Collect patients' subjective ratings of the recommended dishes and laboratory test results at follow-up visits; When a patient's subjective rating is lower than a set threshold, the corresponding taste weight is automatically reduced and the recommendation strategy is adjusted. When an abnormally high level of blood potassium or phosphorus is detected, the nutritional rule matching operation is re-triggered based on the magnitude of the change.

8. A smart dietary recommendation system for patients with chronic kidney disease, characterized in that, include: The data acquisition unit is used to collect multi-source health data of the target patient. The multi-source health data includes laboratory test data, electronic medical record data, anthropometric data and dietary preference data. The laboratory test data includes key indicators such as serum potassium, serum phosphorus, eGFR and blood urea nitrogen. The electronic medical record data includes chronic kidney disease staging, complication diagnosis and current medication information. The generation unit is used to perform medical rule matching based on the multi-source health data, mapping the patient's fused chronic kidney disease stage and various test values ​​to the nutrient allowance calculation model to generate phased nutrient constraint parameters. Constraint units are used to construct multi-disease constraint models based on patient complication information. When complications such as diabetes, hypertension, or hyperuricemia are detected, corresponding constraint rules are superimposed to form a multi-dimensional constraint set. The recommendation unit is used to execute a personalized recommendation algorithm based on the patient's dietary preferences, under the premise of satisfying the nutritional constraint parameters and multidimensional constraint set, to generate a dynamic diet plan and output the recipe results.

9. An electronic device, comprising: The memory and processor are characterized in that the processor, when executing a computer program stored in the memory, implements the steps of the intelligent dietary recommendation method for patients with chronic kidney disease as described in any one of claims 1-7.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the steps of the intelligent dietary recommendation method for patients with chronic kidney disease as described in any one of claims 1-7.

Images