Online support system for cancer patients

Abstract

The invention relates to the technical field of health management, in particular to an online support system for cancer patients. In the invention, through disease course similarity identification and health index cross analysis, the law of individual health change can be extracted by comparing disease course tracks and physiological changes of similar patients, so that customized treatment suggestions and diet guidance can be more effectively provided for the patients, doctors and the patients can be helped to master disease progression more accurately, and the patient experience is improved. And timely intervention is carried out according to different health events. The nutritional requirements of each patient are adjusted in a layered mode, unbalanced nutritional conditions can be recognized by analyzing the nutrient intake proportion, adjustment can be made in time, and health risks caused by improper diet are avoided. By means of the personalized adjustment, daily nursing of the patient is optimized, the treatment effect can be improved, and deterioration of the illness state is delayed. In addition, based on the prediction result of the health trend analysis, the physiological fluctuation of the cancer patient is effectively monitored.

Description

Technical Field

[0001] The present invention relates to the technical field of health management, and particularly to an online support system for cancer patients. Background Art

[0002] The technical field of health management includes aspects such as health status monitoring, disease prevention, rehabilitation support, and medical information interaction for individuals or groups. The core content of this technical field includes collecting, analyzing, and managing patients' health data through information technology means to provide personalized health guidance and medical assistance services.

[0003] Among them, an online support system for cancer patients refers to a technical solution that provides information sharing, disease course management, remote communication, and psychological support for cancer patients based on an Internet platform. It includes content such as online medical record recording, symptom monitoring, remote consultation, and patient community interaction. Specifically, through data collection and analysis, the health indicators uploaded by patients are sorted out, and personalized health suggestions are provided.

[0004] The existing technology mainly relies on static health data records and conventional treatment suggestions for the online support of cancer patients, and fails to make dynamic adjustments according to the disease course trajectories and changes in health indicators of individual patients. Due to the lack of targeted health prediction and analysis in the existing technology, the potential health fluctuations of patients cannot be identified in advance, resulting in the inability to provide appropriate treatment or dietary suggestions in a timely manner when the condition changes rapidly. In addition, most existing systems only monitor based on a single health indicator, and fail to cross-analyze the mutual influence of multiple health data, ignoring the complex correlations between various physiological parameters in the patient's body, resulting in the failure of treatment or dietary suggestions to achieve the best match. Dietary recommendations are often general, without fully considering the individualized nutritional needs of patients, and without being optimized according to the patient's metabolic status, which may lead to nutritional imbalance and further affect the treatment effect or quality of life. Therefore, the existing technology fails to make full use of big data analysis and personalized models, lacks forward-looking judgments on the changing trends of cancer patients, and fails to comprehensively improve the accuracy and efficiency of patient health management. Summary of the Invention

[0005] The purpose of the present invention is to solve the deficiencies existing in the prior art, and to propose an online support system for cancer patients.

[0006] To achieve the above purpose, the present invention adopts the following technical solutions: An online support system for cancer patients includes:

[0007] A disease course similarity recognition module obtains the examination result data of each cancer patient, screens out groups of cancer patients of the same type and matches the disease course trajectories of the corresponding cancer patients, and establishes a set of disease course similarity features;

[0008] The health indicator cross - analysis module calls the health indicators of patients with the same type of cancer in the set of similar disease course characteristics, screens the index combinations with cross - influence, and obtains the health indicator interaction information;

[0009] The nutritional requirement hierarchical adjustment module obtains the diet records of cancer patients based on the health indicator interaction information, calculates the intake proportion of each type of nutrient, compares the intake proportion with the dietary recommendation standard, adjusts the dietary data level, and generates dietary data recommendation information;

[0010] The health trend analysis module calculates the change rate of the specified physiological parameters of cancer patients within a specified time, predicts the possibility of future physiological parameter fluctuations of cancer patients, and generates a health trend prediction result;

[0011] The dietary recommendation module matches the corresponding dietary data recommendation information according to the changes in the physiological parameters of cancer patients in the health trend prediction result, recommends the dietary data matching the current physiological parameters during the remote support process, and generates an online dietary support plan for cancer patients.

[0012] As a further solution of the present invention, the set of similar disease course characteristics is specifically a health event category, an intervention time point, and a disease course trajectory matching result. The health indicator interaction information includes a cross - influence weight, an interaction - influencing health indicator combination, and an adjusted health indicator hierarchical structure. The dietary data recommendation information includes a nutrient intake proportion, a category of abnormally - ingested nutrients, and a dietary adjustment level. The health trend prediction result specifically refers to a physiological parameter change rate, a health fluctuation over - limit parameter, and a prediction of future physiological parameter fluctuations. The online dietary support plan for cancer patients includes personalized dietary data, a nutrition plan matching the current physiological parameters, and remote dietary support suggestions.

[0013] As a further solution of the present invention, the disease course similarity recognition module includes:

[0014] The similarity parameter calculation sub - module obtains the examination result data of cancer patients and uses the formula:

[0015]

[0016] Calculate the similarity S of patient i at time point t i,t , and obtain the disease course similarity parameter;

[0017] Among them, X i,t is a certain examination parameter value of the current patient i at time point t, and H i,t is the corresponding examination parameter value of the historical patient at the same time point t;

[0018] Based on the disease course similarity parameter, the disease course trajectory matching sub-module compares the similarity parameters of multiple cancer patients, screens out the group of cancer patients whose similarity exceeds the set similarity threshold, extracts the disease course event sequences of the group of cancer patients, obtains the key disease course nodes and health change patterns of the group of cancer patients, and matches them with the disease course trajectory of the current cancer patient to obtain the disease course matching result;

[0019] The disease course similar feature extraction sub-module extracts the health events that occur multiple times in the disease course matching result, obtains the occurrence time, duration of the health events and the corresponding intervention time points, and establishes a disease course similar feature set.

[0020] As a further solution of the present invention, the health index cross-analysis module includes:

[0021] The health monitoring data extraction sub-module calls the health indexes of similar cancer patients in the disease course similar feature set, including the monitoring data of heart rate, creatinine, and albumin, and sorts them according to the time series to obtain the health monitoring data set;

[0022] The health index cross-influence calculation sub-module is based on the health monitoring data set and uses the formula:

[0023]

[0024] Calculate the health index And the health index The cross-influence weight between To obtain the health index cross-influence weight information;

[0025] Among them, Represents the health index of similar cancer patients at the t1-th time point Monitoring value, Represents the health index of similar cancer patients at the t1-th time point Monitoring value, Represents the health index The mean value within the time window, Represents the health index The mean value within the time window, T represents the total number of data points within the time window;

[0026] The cross-influence index screening and classification sub-module is based on the health index cross-influence weight information, screens out the index combinations with cross-influence weights higher than the preset influence threshold, adjusts the hierarchical structure of the health indexes, constructs a health index association network according to the influence weight level, and classifies the health indexes with strong interaction to obtain the health index interaction information.

[0027] As a further solution of the present invention, the nutritional requirement hierarchical adjustment module includes:

[0028] Based on the health index interaction information, the health data collection sub-module obtains the diet records of cancer patients. Among them, the diet records include the intake of protein, fat, carbohydrates, and vitamins, and combine with the patient's metabolic status, including BMI and blood glucose fluctuations; obtain the health data of cancer patients;

[0029] Based on the health data of the cancer patient, the nutrient intake calculation sub-module uses the formula:

[0030]

[0031] Calculate the intake proportion of the i2th type of nutrient Obtain the intake proportion of each type of nutrient;

[0032] Among them, represents the daily intake of the i2th type of nutrient, represents the energy coefficient of the i2th type of nutrient, E total represents the daily total energy intake value;

[0033] The nutritional balance adjustment sub-module compares the intake proportion of each type of nutrient with the dietary recommendation standard, screens out the nutrient categories with abnormal intake, determines the nutrient categories with excessive or insufficient intake, adjusts the dietary data level, and generates dietary data recommendation information.

[0034] As a further solution of the present invention, the health trend analysis module includes:

[0035] The change rate calculation sub-module obtains the blood pressure, blood glucose, and heart rate data of cancer patients within a specified time, and uses the formula:

[0036]

[0037] Calculate the change rate of the i3th type of physiological parameter

[0038] Among them, represents the value of the i3th type of physiological parameter at the end moment, represents the value of the i3th type of physiological parameter at the start moment, T end represents the time end moment, T start represents the time start moment;

[0039] The health trend prediction sub-module compares the change rate with the health fluctuation threshold, screens out the physiological parameters that exceed the health fluctuation threshold, and uses the formula:

[0040]

[0041] Calculate the trend change degree T of the screened physiological parameters trend, predicting the possibility of future physiological parameter fluctuations of cancer patients according to the trend change degree, and generating a health trend prediction result;

[0042] Among them, represents the change rate of the i3th type of physiological parameter, is the historical average change rate of the i3th type of physiological parameter, is the standard deviation of the change rate of the i3th type of physiological parameter, and m is the number of types of physiological parameters.

[0043] As a further solution of the present invention, the dietary recommendation module includes:

[0044] The metabolic status analysis sub-module analyzes the metabolic status and nutritional needs of the current patient according to the changes in the physiological parameters of the cancer patient in the health trend prediction result, matches the corresponding dietary data recommendation information, and obtains a patient status evaluation result;

[0045] The dietary support recommendation sub-module combines the patient status evaluation result and the health management needs of cancer patients, and recommends dietary data matching the current physiological parameters during the remote support process to generate an online dietary support plan for cancer patients.

[0046] Compared with the prior art, the advantages and positive effects of the present invention are as follows:

[0047] In the present invention, through the recognition of disease course similarity and the cross-analysis of health indicators, the law of individual health changes can be extracted by comparing the disease course trajectories and physiological changes of similar patients, so as to more effectively provide customized treatment suggestions and dietary guidance for patients, help doctors and patients more accurately grasp the disease progress, and perform timely interventions according to different health events. The nutritional needs of each patient are adjusted in layers. By analyzing the nutrient intake ratio, an unbalanced nutritional status can be identified and adjusted in time to avoid health risks caused by improper diet. These personalized adjustments not only optimize the daily care of patients, but also improve the treatment effect and delay the deterioration of the disease. In addition, based on the prediction results of health trend analysis, the physiological fluctuations of cancer patients are effectively monitored, and preparations are made in advance and corresponding adjustments and interventions are carried out when health abnormalities may occur in the future. The dietary recommendation for patients can be matched in real time with the change trend of their physiological parameters to form a personalized health support plan, minimizing the risks that patients may face during the treatment process and improving the quality of life of patients. Description of the Drawings

[0048] Figure 1 is the system flow chart of the present invention;

[0049] Figure 2 is the flow chart of the disease course similarity recognition module of the present invention;

[0050] Figure 3 This is the flowchart of the health index cross-analysis module of the present invention;

[0051] Figure 4 This is the flowchart of the nutritional requirement hierarchical adjustment module of the present invention;

[0052] Figure 5 This is the flowchart of the health trend analysis module of the present invention;

[0053] Figure 6 This is the flowchart of the dietary recommendation module of the present invention. Detailed implementation manners

[0054] In order to make the objectives, technical solutions and advantages of the present invention more clear and understandable, the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention.

[0055] In the description of the present invention, it should be understood that the orientation or positional relationship indicated by the terms "length", "width", "upper", "lower", "front", "rear", "left", "right", "vertical", "horizontal", "top", "bottom", "inner", "outer", etc. is based on the orientation or positional relationship shown in the accompanying drawings, and is only for the convenience of describing the present invention and simplifying the description, rather than indicating or implying that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore should not be construed as a limitation of the present invention. In addition, in the description of the present invention, the meaning of "a plurality of" is two or more unless otherwise specifically defined.

[0056] Please refer to Figure 1 , an online support system for cancer patients includes:

[0057] The disease course similarity recognition module obtains the examination result data of each cancer patient, screens similar cancer patient groups and matches the disease course trajectories of the corresponding cancer patients, and establishes a disease course similarity feature set;

[0058] The health index cross-analysis module calls the health indexes of similar cancer patients in the disease course similarity feature set, screens the index combinations with cross-influence, and obtains the health index interaction information;

[0059] The nutritional requirement hierarchical adjustment module, based on the health index interaction information, obtains the diet records of cancer patients, calculates the intake proportion of each type of nutrient, compares the intake proportion with the dietary recommendation standard, adjusts the dietary data level, and generates dietary data recommendation information;

[0060] The health trend analysis module calculates the change rate of specified physiological parameters of cancer patients within a specified time period, predicts the possibility of future physiological parameter fluctuations of cancer patients, and generates a health trend prediction result;

[0061] The dietary recommendation module matches the corresponding dietary data recommendation information according to the changes in the physiological parameters of cancer patients in the health trend prediction result, recommends dietary data suitable for the current physiological parameters during the remote support process, and generates an online dietary support plan for cancer patients;

[0062] The set of similar disease course characteristics specifically includes health event categories, intervention time points, and disease course trajectory matching results. The health indicator interaction information includes cross - influence weights, interactive influence health indicator combinations, and adjusted health indicator hierarchical structures. The dietary data recommendation information includes nutrient intake ratios, abnormal nutrient intake categories, and dietary adjustment levels. The health trend prediction result specifically refers to the physiological parameter change rate, health fluctuation over - limit parameters, and future physiological parameter fluctuation predictions. The online dietary support plan for cancer patients includes personalized dietary data, a nutrition plan suitable for the current physiological parameters, and remote dietary support suggestions.

[0063] Please refer to Figure 2 , the disease course similarity recognition module includes:

[0064] The similarity parameter calculation sub - module obtains the examination result data of cancer patients. The examination result data includes hemoglobin, white blood cell, and platelet information. The data source can be the patient's medical record, hospital database, or laboratory report. For example, a certain patient's hemoglobin is 130 g / L, white blood cell count is 6.5×10 9 / L, and platelet count is 250×10 9 / L. At the same time, it obtains the patient's chemotherapy cycle information, such as chemotherapy time points and intervals. For example, the chemotherapy interval of this patient is 21 days, and each cycle is detected at a specific time point to form time - series data. Then, it extracts the blood examination data of historical patients during the same chemotherapy cycle. For example, the average hemoglobin values of a certain historical patient in the same 1st, 2nd, and 3rd cycles are 135 g / L, 132 g / L, and 128 g / L, and the white blood cell counts are 5.8×10 9 / L, 5.5×10 9 / L, 5.2×10 9 / L, and the platelet counts are 240×10 9 / L, 238×10 9 / L, 235×10 9 / L. On this basis, the formula:

[0065]

[0066] is used to calculate the similarity S of patient i at time point ti,t , the disease course similarity parameter is obtained;

[0067] Among them, X i,t is a certain examination parameter value of the current patient i at time point t, such as hemoglobin, white blood cell count or platelet count, and this data is sourced from hospital laboratory test results or electronic health records (EHR). H i,t is the corresponding examination parameter value of the historical patient at the same time point t, and the historical mean value can be calculated by statistically analyzing the historical medical record data or databases of a large number of patients. max(X i,t , H i,t ) is used for normalization so that the calculated similarity is maintained between 0 and 1, and this value is directly obtained from the maximum value of the current patient value and the historical patient value without additional calculation.

[0068] For example, if the hemoglobin of the current patient in the first chemotherapy cycle is 130 g / L while the average value of the historical patients is 135 g / L, the similarity calculation is as follows:

[0069]

[0070] Similarly, calculate the white blood cell count:

[0071]

[0072] Calculate the platelet count:

[0073]

[0074] Calculate the similarity at all time points in the same way, and finally obtain the set of disease course similarity parameters at all time points, which is used to screen similar patient groups subsequently.

[0075] Based on the disease course similarity parameter, the disease course trajectory matching sub-module compares the similarity parameters of multiple cancer patients, screens the cancer patient group with a similarity exceeding the set similarity threshold, extracts the disease course event sequence of the cancer patient group, obtains the key disease course nodes and health change patterns of the cancer patient group, and matches them with the disease course trajectory of the current cancer patient to obtain the disease course matching result;

[0076] First, set the similarity threshold, which can be determined according to statistical analysis or clinical experience. For example, set the similarity threshold to 0.95. When screening for similar patients, it is necessary to calculate the average similarity of each patient and compare the set of disease course similarity parameters of the current patient with those of the historical patients. For example, the average similarity of the hemoglobin, white blood cells, and platelets of the current patient is calculated as follows: If the disease course similarities of a certain historical patient are 0.96, 0.92, and 0.95 respectively, then its average similarity is: Since 0.943 is greater than the similarity threshold of 0.95, this historical patient is selected into the group of similar cancer patients. Subsequently, the sequence of disease course events of these similar patients is extracted, and these sequences may contain information such as blood test data during chemotherapy and the occurrence of complications. For example, after the third chemotherapy cycle, multiple similar patients had their white blood cell count drop below 3.5×10 9 / L, and received granulocyte colony-stimulating factor (G-CSF) intervention on the 10th day. Then, these disease course events are matched with the disease course trajectory of the current patient to finally obtain the matching disease course event sequence. Among them, S avg : The average similarity of all examination parameters of the current patient, representing the similarity level of the overall disease course of the patient. This value is obtained by calculating the mean of the similarities of all examination parameters. S avg,hist : The average similarity of historical patients, obtained by calculating the mean of the similarities of multiple examination parameters. When the value is higher than the set threshold (0.95), the patients are regarded as similar patients.

[0077] The disease course similarity feature extraction sub-module extracts the health events that occur multiple times in the disease course matching results, obtains the occurrence time, duration of the health events and the corresponding intervention time points, and establishes a disease course similarity feature set;

[0078] First, count the health events that frequently occur in the group of similar patients. For example, if more than 80% of the similar patients have leukopenia (WBC < 3.5×10 9 / L) within 7 to 10 days after the third chemotherapy, then this event is considered a disease course feature with statistical significance. Then, determine the intervention time point of this health event. For example, on the 7th day after the third chemotherapy, more than 75% of the patients received granulocyte colony-stimulating factor (G-CSF) treatment and returned to normal (WBC > 4.0×10 9 / L) on the 14th day. Therefore, this intervention event can be recorded and classified into the disease course similarity feature set. Finally, this set will provide individualized reference for the current patient, such as intervening G-CSF in advance on the 7th day after his third chemotherapy to prevent the occurrence of leukopenia. Among them, WBC: white blood cell count, unit is ×10 9 / L, obtained through blood tests. Intervention time point: The determined time window, such as 7 days after chemotherapy, obtained by statistically analyzing the patient's disease course data and extracting health events with high frequency. Disease course similarity feature set: Generated by matching the disease course event sequence and extracting frequently occurring health events and their intervention time points, determined based on statistical laws.

[0079] Please refer to Figure 3 , the health index cross-analysis module includes:

[0080] The health monitoring data extraction submodule calls the health indicators of similar cancer patients in the disease course similar feature set, including heart rate, creatinine, and albumin monitoring data, and sorts them according to time series to obtain the health monitoring data set;

[0081] Medical data collection requires access to health monitoring data from similar cancer patients with similar disease course characteristics. This data is sourced from multiple independent healthcare systems, such as hospital electronic health record (EHR) systems, follow-up databases, and patient self-monitoring devices. First, core health indicators such as heart rate, creatinine, and albumin are extracted from these sources and organized chronologically to ensure time series integrity. For example, in a study where health data collection is weekly, within a one-month window, the patient's heart rate data might be 72, 75, 78, and 74 beats per minute, creatinine levels might be 0.9, 1.1, 1.0, and 1.2 mg / dL, and albumin levels might be 4.0, 3.8, 3.9, and 4.1 g / dL. Because the collection process may contain missing data or outliers, such as a measurement failure at a certain time point due to instrument malfunction, a data cleaning step is required. First, missing data is detected and the missing pattern is determined. If the missing data percentage exceeds 30%, the data entry is removed. Otherwise, linear interpolation is used to fill the missing values. For example, if the creatinine value at the third measurement is missing, the mean of the previous and subsequent data can be used to fill in the missing value, that is, (1.1 + 1.2) / 2 = 1.15 mg / dL. For outlier detection, the 3σ principle is usually used to determine abnormal data points. For example, if the standard deviation of heart rate data is σ = 2.5 beats / minute, then values outside the mean ± 3σ range will be marked as abnormal. For example, if the measured heart rate value is 85 beats / minute, it will be marked as an abnormal point because it exceeds the range of [67.25, 82.25] and replaced with the mean of the previous and subsequent values. Finally, after data cleaning, all patient data are organized into a time series data structure and stored in the database to ensure data integrity, thereby obtaining a health monitoring data set.

[0082] The health indicator cross-impact calculation submodule is based on the health monitoring data set and uses the formula:

[0083]

[0084] Calculating health indicators and health indicators Cross-influence weight Obtain cross-impact weight information of health indicators;

[0085] in, Represents the health indicators of similar cancer patients at time point t1 Monitoring values,are obtained from the health monitoring dataset; Represent the health indicators of cancer patients similar to the t1-th time point The monitoring value, obtained from the health monitoring dataset; Represent the health indicator The mean value within the time window; Represent the health indicator The mean value within the time window; T represents the total number of data points within the time window, which is determined by the data acquisition time span and the sampling frequency. For example, if the monitoring data is sampled once per hour and the time window is 7 days, then T = 24×7 = 168; Represent the sum of all data points within the time window to ensure the cumulative effect of all time points is calculated.

[0086] If in the database of a certain medical institution, the health data of a certain patient within a continuous week is as follows:

[0087] Table 1 Patient health monitoring data;

[0088]

[0089]

[0090] As shown in Table 1, the data is collected at 6-hour intervals. If the data at a certain time point is missing, for example, the creatinine value at 00:00 is not collected, then the mean value of the previous and subsequent measurement values is used for interpolation, that is, (1.1 + 1.0) / 2 = 1.05 mg / dL. If the proportion of missing data exceeds 30%, then the data point of this patient is directly excluded. To ensure the comparability of numerical calculations, standardization is required to map health indicators with different units to the range of [0, 1].

[0091] Taking heart rate and creatinine data as an example, substitute into the formula to calculate the cross-influence weight:

[0092]

[0093] The calculated cross-influence weight between heart rate and creatinine is 0.723.

[0094] The cross-influence indicator screening and classification sub-module, based on the cross-influence weight information of health indicators, screens out the indicator combinations with cross-influence weights higher than the preset influence threshold, adjusts the hierarchical structure of health indicators, constructs a health indicator association network according to the high and low influence weights, classifies the health indicators with strong interaction, and obtains the health indicator interaction information;

[0095] Based on the cross - influence weights of health indicators, select the indicator combinations whose cross - influence weights are higher than the preset influence threshold. Suppose the preset threshold is set to 0.5, then select the indicator pairs with a correlation coefficient greater than 0.5. For example, the calculated cross - influence weight of heart rate and creatinine is 0.723, which exceeds the set threshold, so it is regarded as an indicator combination with strong cross - influence. Next, adjust the hierarchical structure of health indicators. First, group the health indicators to determine the relevance between different health indicators. For example, if the cross - influence weight of creatinine - albumin is only 0.32, its relevance is weak and can be classified into different categories, while the cross - influence weight of heart rate - creatinine is 0.723, indicating that there may be a strong physiological correlation between them, so they should be grouped into the same category. Finally, by classifying all health indicators, establish the hierarchical structure of health indicators and obtain the interaction information of health indicators. Among them, the preset influence threshold: used to screen indicators with strong cross - influence, usually set according to historical data, such as 0.5; classification of health indicators: classify indicators according to the cross - influence weight matrix, and those with higher values are grouped into one category. For example, heart rate and creatinine are grouped into the same category, and albumin is grouped into another category.

[0096] Please refer to Figure 4 , the nutritional requirement hierarchical adjustment module includes:

[0097] The health data collection sub - module obtains the diet records of cancer patients based on the interaction information of health indicators. Among them, the diet records include the intake of protein, fat, carbohydrates, and vitamins, combined with the patient's metabolic status, including BMI and blood glucose fluctuations; obtain the health data of cancer patients;

[0098] First, it is necessary to collect the daily diet records of cancer patients, including the intake of various nutrients such as protein, fat, carbohydrates, and vitamins consumed by the patient every day. At the same time, the patient's weight changes, exercise conditions (such as the number of steps per day and exercise time), and metabolic status (including BMI and blood glucose fluctuations, etc.) should also be recorded. The collection of these data provides the basic data for subsequent analysis. In actual implementation, the patient's exercise volume can be automatically recorded by intelligent devices (such as smart bracelets or APPs), and their diet, blood glucose fluctuations, and weight changes can be collected through questionnaires or electronic recording systems. Specifically, assume that a patient intakes 60 grams of protein, 50 grams of fat, 200 grams of carbohydrates, and 800 μg of vitamin A in a day, the patient's weight change is - 0.5 kg, the exercise volume is 8000 steps, the BMI is 24, and the blood glucose fluctuation is within the normal range. Based on these data, it can provide the necessary basic data for subsequent nutrient intake calculation and nutritional balance adjustment.

[0099] The nutrient intake calculation sub - module is based on the health data of cancer patients and uses the formula:

[0100]

[0101] Calculate the intake proportion of the i2-th type of nutrient Obtain the intake proportion of each type of nutrient;

[0102] Among them, represents the daily intake of this type of nutrient, represents the energy coefficient of this nutrient (i.e., the energy provided by this nutrient per unit mass), E total represents the daily total energy intake value.

[0103] First, it is necessary to quantify the daily intakes of various nutrients. In this step, the specific parameters involved include the daily intakes of each type of nutrient (such as 60 grams of protein, 50 grams of fat, 200 grams of carbohydrates, etc.), and the energy coefficients of each type of nutrient (for example, protein provides 4 kcal per gram, fat provides 9 kcal per gram, carbohydrates provide 4 kcal per gram, etc.). Taking this calculation method as an example, first, according to the provided intakes and energy coefficients, the total energy provided by each nutrient can be calculated. For example: 60 grams of protein × 4 kcal / gram = 240 kcal, 50 grams of fat × 9 kcal / gram = 450 kcal, 200 grams of carbohydrates × 4 kcal / gram = 800 kcal. Then, through these calculation results, the patient's daily total energy intake can be obtained. Assuming the total energy intake of the patient on that day is 240 + 450 + 800 = 1490 kcal. Then, according to the energy proportion formula of each type of nutrient, calculate the energy proportion of each type of nutrient as follows:

[0104] The energy proportion of protein is

[0105] The energy proportion of fat is

[0106] The energy proportion of carbohydrates is

[0107] The final result of this step is to obtain the intake proportion of each type of nutrient, which provides a necessary reference basis for subsequent nutritional balance adjustment.

[0108] The nutritional balance adjustment sub-module compares the intake proportion of each type of nutrient with the dietary recommendation standards, screens out the nutrient categories with abnormal intakes, determines the nutrient categories with excessive or insufficient intakes, adjusts the dietary data level, and generates dietary data recommendation information;

[0109] First, it is necessary to set a reasonable range for the recommended intake ratios of various nutrients. For example, according to relevant dietary guidelines, the recommended energy proportion of protein is 10%-35%, that of fat is 20%-35%, and that of carbohydrates is 45%-65%. Then, judge the actual energy intake ratio of the patient. If the intake ratio of a certain nutrient exceeds the recommended range, adjustment is required. For example, according to the aforementioned calculation results, the protein intake ratio of the patient is 16.1%, the fat ratio is 30.2%, and the carbohydrate ratio is 53.7%. Since these values are all within the dietary recommendation standards, no adjustment is required for the time being. If the intake ratio of a certain nutrient is not within the recommended range, nutritional balance adjustment needs to be made according to the patient's health condition. Suppose the fat intake ratio of a certain patient is 40%, exceeding the recommended range, then the nutritional balance can be optimized by adjusting the diet or advising the patient to reduce fat intake. Finally, based on the comparison between the nutrient intake ratio and the recommended standard, adjust the dietary plan to optimize the dietary recommendation information.

[0110] Please refer to Figure 5 , the health trend analysis module includes:

[0111] The change rate calculation sub-module obtains the blood pressure, blood sugar, and heart rate data of cancer patients within a specified time, and uses the formula:

[0112]

[0113] Calculate the change rate of the i3th type of physiological parameter

[0114] Among them, represents the value of this type of physiological parameter at the end time, represents the value of this type of physiological parameter at the start time, T end represents the end time of the time, T start represents the start time of the time.

[0115] First, obtain the blood pressure, blood sugar, heart rate and other data of cancer patients within a specified time. These data are usually collected in real time through physiological monitoring devices such as blood glucose meters, sphygmomanometers, and heart rate monitors, and record the start time T start and the end time T end . Suppose a patient's blood sugar data starts from the time point T start =0, and the initial blood sugar value is continuously monitored until T end =10 hours, and the blood sugar value is At this time, the blood sugar change rate can be calculated Substitute the data into the formula to get the blood sugar change rate as: Here, the rate of change of blood glucose is 0.22 mmol / L / h, indicating that the blood glucose increases by 0.22 mmol / L per hour within 10 hours. Similarly, other physiological parameters such as blood pressure and heart rate can also be calculated in the same way. The initial value of blood pressure and the end value are 120 mmHg and 130 mmHg respectively, and the continuous monitoring time is 10 hours. Substitute these values into the formula to calculate the rate of change The calculation method for heart rate is similar. Determine the rate of change based on the change in the patient's heart rate, so as to obtain the rate of change of each physiological parameter, providing data support for subsequent analysis of health fluctuations. Through this process, we obtain the rate of change of each physiological parameter (blood glucose, blood pressure, heart rate, etc.), providing basic data for the health fluctuation trend prediction module.

[0116] The health trend prediction sub-module compares the rate of change with the health fluctuation threshold, screens out the physiological parameters that exceed the health fluctuation threshold, and uses the formula:

[0117]

[0118] Calculate the trend change degree T of the screened physiological parameters trend , and predict the possibility of future physiological parameter fluctuations in cancer patients based on the trend change degree, generating the health trend prediction result;

[0119] Among them, represents the rate of change of the i3th type of physiological parameter, is the historical average rate of change of the i3th type of physiological parameter, is the standard deviation of the rate of change of the i3th type of physiological parameter, and m is the number of types of physiological parameters.

[0120] First, compare the calculated rate of change of each physiological parameter with its historical fluctuation range. The historical fluctuation range is obtained by analyzing the patient's historical data. For example, the average change and the standard deviation of physiological parameters such as blood glucose, blood pressure, and heart rate over a certain period of time in the past can be analyzed. Suppose the historical average rate of change of blood glucose The standard deviation According to the currently calculated rate of change Calculate its abnormality relative to historical fluctuations. Substitute the known rate of change of blood glucose and historical data into the formula to obtain: This result indicates that the deviation degree of the current blood glucose change rate relative to the historical fluctuations is 1.4 standard deviations, indicating that there is a certain degree of abnormality in the blood glucose change compared to the historical fluctuations. If this deviation degree value exceeds a certain threshold (such as 2 standard deviations), it can be regarded as an abnormal health fluctuation and further monitoring is required. Similarly, for physiological parameters such as blood pressure and heart rate, the same calculation can be performed. Assuming that the average change rate of historical blood pressure is standard deviation Current blood pressure change rate Substitute into the formula: This result shows that the deviation of the blood pressure change rate compared to the historical fluctuations is small and the fluctuations are relatively stable. Based on the calculated values of these fluctuation trends, and further through comprehensive analysis of the change trends of various physiological parameters, an overall prediction of the patient's health status can be obtained. By continuously monitoring and comparing the change trends of current and historical data, it can provide a real-time reference basis for medical decision-making.

[0121] Please refer to Figure 6 , the dietary recommendation module includes:

[0122] The metabolic status analysis sub-module analyzes the current patient's metabolic status and nutritional needs according to the changes in the physiological parameters of cancer patients in the health trend prediction results, matches the corresponding dietary data recommendation information, and obtains the patient status evaluation result;

[0123] First of all, according to the health trend prediction results, the changes in the physiological parameters of cancer patients (such as body temperature, weight, blood glucose, blood lipids, etc.) need to be collected and analyzed through regular monitoring data, which may come from the patient's regular physical examinations, home health monitoring devices or telemedicine systems. For example, if a patient's weight has decreased by 5 kg in the past week, the body temperature has risen by 0.5 °C, and the blood glucose level has fluctuated, the system will first judge whether these changes exceed the normal physiological fluctuation range. Then, calculate the change in the patient's metabolic rate based on these data. For example, if the weight loss is too large, it may indicate malnutrition or disease aggravation, so the system will calculate whether its metabolic requirements have increased, and usually at this time, the patient's energy intake needs to be increased. By judging the change trends of blood glucose and blood lipids, it is inferred whether the patient needs to adjust carbohydrate or fat foods. Combining this information, finally, the patient's nutritional needs will be accurately predicted and a recommended dietary data report will be generated. In practical applications, assuming that a cancer patient A's weight has decreased from 70 kg to 63 kg within 1 month, the body temperature has risen from 36.5 °C to 37 °C, and the blood glucose level has fluctuated from the normal 5.2 mmol / L to 7.5 mmol / L, the system will analyze through these parameters that the patient's metabolic requirements have increased significantly and special attention needs to be paid to their sugar intake. At this time, the system will recommend a dietary plan that increases high-quality protein and low glycemic index foods to supplement nutrition and balance blood glucose.

[0124] The dietary support recommendation sub-module combines the patient status assessment results and the health management needs of cancer patients, and recommends dietary data suitable for the current physiological parameters during the remote support process, generating an online dietary support plan for cancer patients;

[0125] When conducting remote support, considering the comprehensive health management needs of cancer patients, the system will automatically obtain the patient's current physiological parameters and compare them with the information in the dietary recommendation database. By analyzing the patient's physiological data, such as blood sugar, blood lipids, etc., and the patient's physical activity level, an automatic dietary support plan that suits the patient's current condition will be generated. For example, if the patient has a high blood sugar level, the system will reduce the recommendation of high-sugar foods and increase the recommendation of foods rich in dietary fiber and low in sugar, such as green leafy vegetables and whole grains. For patients with significant weight changes, the system will increase the recommendation of high-protein and low-fat foods and ensure that the patient obtains sufficient nutrition while avoiding excessive calorie intake. For instance, assuming that patient B's blood sugar level fluctuates between 7.5 mmol / L and 8.0 mmol / L within a week and the patient loses 3 kg in weight, the system will immediately adjust the patient's dietary support plan, recommend increasing high-quality proteins such as fish and chicken breast, and reducing the intake of high-sugar foods such as rice and desserts. The system will also automatically generate an online dietary plan according to the patient's personal needs, including a detailed daily meal plan, and provide specific food portions and nutritional components to ensure that the patient can obtain timely nutritional support suitable for their physiological state. Through these analyses and recommendations, patients can receive scientific dietary management during the remote support process, effectively helping patients maintain a good nutritional status during cancer treatment and thereby improving the overall treatment effect.

[0126] The above are only the preferred embodiments of the present invention and do not limit the present invention in other forms. Any person skilled in the relevant art may use the disclosed technical content to make changes or modifications into equivalent embodiments with equivalent changes and apply them to other fields. However, as long as it does not depart from the technical solution content of the present invention, any simple modification, equivalent change, and modification made to the above embodiments based on the technical essence of the present invention still fall within the protection scope of the technical solution of the present invention.

Claims

1. An online support system for cancer patients, characterized in that, The system includes: The disease course similarity recognition module obtains the examination result data of each cancer patient, screens out groups of cancer patients of the same type and matches the disease course trajectories of the corresponding cancer patients, and establishes a disease course similarity feature set; The health index cross-analysis module calls the health indexes of cancer patients of the same type in the disease course similarity feature set, screens out combinations of indexes with cross-influence, and obtains health index interaction information; The nutrition requirement hierarchical adjustment module, based on the health index interaction information, obtains the diet records of cancer patients, calculates the intake proportion of each type of nutrient, compares the intake proportion with the dietary recommendation standard, adjusts the dietary data level, and generates dietary data recommendation information; The health trend analysis module calculates the change rate of the specified physiological parameters of cancer patients within a specified time, predicts the possibility of future physiological parameter fluctuations of cancer patients, and generates a health trend prediction result; The dietary recommendation module, according to the changes in the physiological parameters of cancer patients in the health trend prediction result, matches the corresponding dietary data recommendation information, and recommends dietary data matching the current physiological parameters during the remote support process, and generates an online dietary support plan for cancer patients.

2. The online support system for cancer patients according to claim 1, characterized in that, The disease course similarity feature set specifically includes health event categories, intervention time points, and disease course trajectory matching results. The health index interaction information includes cross-influence weights, combinations of interacting health indexes, and adjusted health index hierarchical structures. The dietary data recommendation information includes nutrient intake proportions, categories of abnormally ingested nutrients, and dietary adjustment levels. The health trend prediction result specifically refers to physiological parameter change rates, health fluctuation over-limit parameters, and predictions of future physiological parameter fluctuations. The online dietary support plan for cancer patients includes personalized dietary data, a nutrition plan matching the current physiological parameters, and remote dietary support suggestions.

3. The online support system for cancer patients according to claim 2, wherein The disease course similarity recognition module includes: The similarity parameter calculation sub-module obtains the examination result data of cancer patients and uses the formula: Calculate the similarity S of patient i at time point t i,t , and obtain the disease course similarity parameter; where X i,t is the value of a certain examination parameter of the current patient i at time point t, and H i,t is the corresponding examination parameter value of the historical patient at the same time point t; The disease course trajectory matching sub-module, based on the disease course similarity parameters, compares the similarity parameters of multiple cancer patients, screens out groups of cancer patients with similarity exceeding the set similarity threshold, extracts the disease course event sequences of the cancer patient groups, obtains the key disease course nodes and health change patterns of the cancer patient groups, and matches them with the disease course trajectories of the current cancer patients to obtain a disease course matching result; The disease course similarity feature extraction sub-module extracts the health events that occur multiple times in the disease course matching result, obtains the occurrence time, duration, and corresponding intervention time points of the health events, and establishes a disease course similarity feature set.

4. The online support system for cancer patients according to claim 3, characterized in that, The health index cross-analysis module includes: The health monitoring data extraction sub-module calls the health indexes of similar cancer patients in the disease course similarity feature set, including the monitoring data of heart rate, creatinine, and albumin, and sorts them according to the time series to obtain a health monitoring data set; The health index cross-influence calculation sub-module, based on the health monitoring data set, uses the formula: Calculate health indicators With health indicators The cross - impact weight between Obtain the cross - impact weight information of health indicators; Among them, represents the health indicator of similar cancer patients at the t1-th time point monitoring value, represents the health indicator of similar cancer patients at the t1-th time point monitoring value, represents the health indicator average value within the time window, represents the health indicator Y j1 average value within the time window, T represents the total number of data points within the time window; The cross - impact index screening and classification sub - module, based on the cross - impact weight information of the health indicators, screens out the index combinations with cross - impact weights higher than the preset impact threshold, adjusts the hierarchical structure of the health indicators, constructs a health indicator association network according to the high and low of the impact weights, classifies the health indicators with strong interactive effects, and obtains the health indicator interaction information.

5. The online support system for cancer patients according to claim 4, wherein The nutrition requirement hierarchical adjustment module includes: The health data collection sub - module, based on the health indicator interaction information, obtains the diet records of cancer patients. Among them, the diet records include the intake of protein, fat, carbohydrates, and vitamins, and combines the patient's metabolic status, including BMI and blood glucose fluctuations; obtains the health data of cancer patients; The nutrient intake calculation sub - module, based on the health data of the cancer patients, uses the formula: Calculate the intake proportion of the i2-th type of nutrient Obtain the intake proportion of each type of nutrient; Among them, represents the daily intake of the i2th type of nutrient, represents the energy coefficient of the i2th type of nutrient, E total represents the daily total energy intake value; The nutritional balance adjustment sub - module compares the intake proportion of each type of nutrient with the dietary recommendation standard, screens out the nutrient categories with abnormal intake, determines the nutrient categories with excessive or insufficient intake, adjusts the dietary data hierarchy, and generates dietary data recommendation information.

6. The online support system for cancer patients according to claim 5, characterized in that, The health trend analysis module includes: The change rate calculation sub - module obtains the blood pressure, blood glucose, and heart rate data of cancer patients within a specified time, using the formula: Calculate the change rate of the physiological parameter of the i3-th category Among them, represents the value of the i3-th type of physiological parameter at the end moment, represents the value of the i3-th type of physiological parameter at the start moment, T end represents the end moment of time, T start represents the start moment of time; The health trend prediction sub - module compares the change rate with the health fluctuation threshold, screens out the physiological parameters that exceed the health fluctuation threshold, using the formula: Calculate the trend change degree T of the physiological parameters after screening trend , predict the possibility of future physiological parameter fluctuations of cancer patients based on the trend change degree, and generate a health trend prediction result; Among them, represents the change rate of the i3-th type of physiological parameter, is the historical average change rate of the i3-th type of physiological parameter, is the standard deviation of the change rate of the i3-th type of physiological parameter, and m is the number of types of physiological parameters.

7. The online support system for cancer patients according to claim 6, wherein, The dietary recommendation module includes: The metabolic status analysis sub - module analyzes the current patient's metabolic status and nutritional requirements according to the changes in the physiological parameters of cancer patients in the health trend prediction result, matches the corresponding dietary data recommendation information, and obtains the patient status evaluation result; The dietary support recommendation sub - module combines the patient status evaluation result and the health management requirements of cancer patients, and recommends the dietary data that matches the current physiological parameters during the remote support process to generate an online dietary support plan for cancer patients.

Images