Enteral nutrition multi-source data fusion analysis system

Abstract

The invention relates to the technical field of data processing, in particular to an enteral nutrition multi-source data fusion analysis system which comprises an encryption processing module, a data fusion module, a nutrition demand prediction module and a matching degree evaluation module. According to the method, the multi-source data set is provided by encrypting the enteral nutrition associated data and integrating the intake and physiological parameter results, the security and privacy protection of the data are remarkably improved, cross-mechanism data can be analyzed and compared when encrypted data sets of different mechanisms are fused, nutritional requirements can be judged more accurately, and the safety of the data is improved. The consistency of analysis results is ensured, by analyzing the change trend of nutritional requirements, the evaluation of nutritional conditions is more scientific and dynamic, the medical service quality is effectively improved, accurate matching of individualized nutritional plans is ensured, making of a more appropriate nutritional supply strategy is facilitated, and by means of accurate nutritional requirement evaluation, the accuracy of nutritional requirement evaluation is improved. The recovery effect and the life quality of the patient are improved.

Description

Technical Field

[0001] The present invention relates to the technical field of data processing, and in particular to an enteral nutrition multi-source data fusion analysis system. Background Art

[0002] The field of data processing technology includes information collection, storage, management, analysis and optimization. The core content is to improve the availability, reliability and processing speed of data while ensuring data security and privacy protection. Data processing technology plays a vital role in the modern information age and is widely used in many industries such as finance, medical care, education, and government. It involves the application of basic data entry and verification to complex data mining and machine learning algorithms, aiming to extract valuable information from large and complex data sets to support the decision-making process.

[0003] Among them, the enteral nutrition multi-source data fusion analysis system refers to a system for processing nutrition-related data from different data sources and providing a comprehensive analysis of enteral nutrition status by fusing data. The technical matters targeted by the subject of this patent include data fusion and analysis, which is achieved by integrating multiple data inputs, including patients' health data, dietary intake records, and laboratory test results, and applying a unified analysis framework. This system does not rely on complex model adjustments, but instead uses proven data processing and analysis techniques to ensure the accuracy of the data and the consistency of the analysis results.

[0004] Existing technologies lack adequate data security and privacy protection measures when processing complex and multi-source data, which is particularly important in the medical industry because it is related to the protection of sensitive patient information. Existing technologies cannot effectively integrate data from different institutions, resulting in data inconsistencies in cross-institutional analysis, which directly affects the accuracy of data analysis and the effectiveness of decision-making. In actual operations, there is a lack of dynamic and personalized nutritional needs assessment, and conventional static data processing methods are difficult to adapt to the reality of large individual differences, resulting in inaccurate nutritional plans and difficulty in meeting patients' actual nutritional needs, affecting the recovery effect and patient rehabilitation process. Summary of the invention

[0005] The purpose of the present invention is to solve the shortcomings of the prior art and to propose an enteral nutrition multi-source data fusion analysis system.

[0006] In order to achieve the above object, the present invention adopts the following technical scheme: an enteral nutrition multi-source data fusion analysis system comprises: The encryption processing module performs encryption transformation processing on the values ​​of protein intake, carbohydrate intake, and fat intake based on the input enteral nutrition-related data source, and performs encryption calculation processing on the values ​​of blood glucose concentration, human serum albumin concentration, and urea nitrogen concentration to obtain a multi-source encrypted data set; The data fusion module uses the multi-source encrypted data set to extract the encrypted intake data values ​​from the medical institution data and the individual patient data, calculate the total amount of differentiated source data, calculate the encrypted mean for multiple parameters, merge the encrypted mean and the encrypted total amount, and perform merging and comparison with the parameters to obtain a cross-institutional fused encrypted data set; The nutritional demand prediction module uses the cross-institutional fused encrypted data set to perform normalization calculations on the encrypted total intake value and the encrypted physiological mean value, analyzes the relationship between protein intake and human serum albumin concentration, generates dynamic nutritional demand identification results, calculates dynamic offset values, analyzes offset trend identification ranges, evaluates nutritional demand change trends, and obtains enteral nutritional demand prediction values; The matching degree assessment module compares and calculates the difference between the protein intake and the predicted value item by item based on the predicted value of enteral nutrition demand, obtains the parameter deviation relationship analysis result by combining the blood glucose concentration, calculates the degree of deviation, analyzes the degree of nutrition supply adaptation, and obtains the matching degree of enteral nutrition demand.

[0007] As a further solution of the present invention, the step of acquiring the multi-source encrypted data set is specifically as follows: Based on the input enteral nutrition-related data source, the protein intake, carbohydrate intake and fat intake values ​​are recorded, standardized, the deviation between the intake value and the average value is calculated, and the three types of intake deviation values ​​are summed to obtain the intake weighted result; The weighted results of the intake are used, and the values ​​of blood glucose concentration, human albumin concentration, and urea nitrogen concentration are standardized and matched one by one with the weighted results of the intake, using the formula: ; Calculate the intake and physiological parameters to match the deviation value , generating physiological parameter matching results; in, is the weight parameter, represents the intake-weighted value, Represents the corresponding physiological parameters The numerical value of is the number of matches between intake parameters and physiological parameters; The deviation analysis is performed on the physiological parameter matching results, the matching deviation value is compared with the set threshold, the results with deviation values ​​exceeding the threshold are deleted, the weight parameters of the correlation between the intake and the physiological parameters are adjusted, and a multi-source encrypted data set is generated.

[0008] As a further solution of the present invention, the step of obtaining the encrypted intake data value is specifically as follows: The data fusion module extracts the medical institution data and the individual patient data from the multi-source encrypted data set, and during the extraction process, screens and separates the encrypted data entries of the medical institution and the patient to obtain an encrypted patient data set; Identify encrypted entries associated with enteral nutrition intake from the encrypted patient data set, extract intake data from the encrypted entries through keyword matching, and obtain an encrypted data entry set; Decrypting the encrypted data entry set item by item, extracting the intake data value, and obtaining the weight information associated with the intake from the data entry, and combining it with the patient individual difference data to obtain the decrypted intake information; The intake values ​​in the decrypted intake information are weighted according to the weights, and the patient individual difference parameters and the credibility parameters are combined to adopt the formula: ; Calculate the encrypted intake data value ; in, Represents the intake value of each item, represents the intake item weight, Represents the individual difference parameter of patients, Represents the credibility parameter of medical institutions, represents the encryption strength correction factor, Represents the number of entries, Represents the entry index.

[0009] As a further solution of the present invention, the steps of obtaining the cross-institutional fused encrypted data set are specifically as follows: Extracting encrypted intake values ​​across medical institutions from the encrypted intake data, grouping and classifying the data by identifying data tags of differentiated medical institution sources, and obtaining institution grouping records; According to the institutional grouping records, the encrypted intake mean and the corresponding data coverage of the differentiated institutions are calculated, and the data coverage parameter set is obtained by calculating the proportion of the intake data values ​​in each institution; Using the data coverage parameter set, the mean encrypted intake of each institution is weighted according to the data coverage, using the formula: ; Calculate the normalized value for each institution , perform weighted summation on the normalized values ​​of the institutions to generate a cross-institutional fused encrypted data set; in, Representative The data coverage of each institution, Representative The average intake of each institution, Represents the total number of medical institutions involved in the calculation, The maximum intake mean in the representative institution.

[0010] As a further solution of the present invention, the steps of obtaining the dynamic nutritional requirement identification result are specifically as follows: The cross-institutional fused encrypted data set is used to extract the encrypted total intake parameter and the encrypted physiological mean parameter, and the values ​​between the two are normalized to obtain a normalized parameter value set; By using the normalized parameter value set, the normalized encrypted intake and the encrypted physiological mean parameter values ​​are extracted, and the correlation change trend of the two parameters within the dynamic range is analyzed, using the formula: ; Calculate the dynamic correlation coefficient , analyze the time series deviation amplitude of protein intake and human albumin concentration in enteral nutrition to generate dynamic nutritional demand identification results; in, Represents the encrypted intake data values, Represents the encrypted physiological mean data values, represents the mean of crypto intake, represents the mean of the encrypted physiological means, Represents the total number of data, Represents an accumulation operation.

[0011] As a further solution of the present invention, the step of obtaining the predicted value of enteral nutrition demand is specifically as follows: Extracting the dynamic offset value range parameter of the dynamic nutritional requirement identification result, combining the correlation between protein intake and human albumin concentration, analyzing the mapping relationship between the dynamic offset value and the intake adjustment coefficient, and obtaining an adjusted dynamic offset value set; Based on the adjusted dynamic offset value set, combined with the change parameters of the short-term offset trend and the long-term offset trend in the time series, the formula is adopted: ; Calculate dynamic weighted adjustment values , generate a predicted offset value; in, Representative The adjusted dynamic offset value, Represents the short-term deviation trend parameter values, Represents the long-term deviation trend parameter values, Represents the total number of time series data points; According to the predicted offset value, the encrypted intake parameters and encrypted physiological mean parameters in enteral nutrition are called, and the dynamic change range of intake demand is analyzed in combination with the amplitude range of the offset in the time series, and the impact of short-term and long-term trends on enteral nutrition demand is evaluated to obtain the predicted value of enteral nutrition demand.

[0012] As a further solution of the present invention, the step of obtaining the parameter deviation relationship analysis result is specifically: The matching degree evaluation module stores the difference data between the intake of the data point and the predicted value in a vector form according to the predicted value of the enteral nutrition demand called, combined with the protein intake parameter of the enteral nutrition, and analyzes the difference direction by comparing and analyzing the difference direction item by item, identifies the matching relationship between the direction and the absolute value, and generates a protein intake difference comparison result; The protein intake difference comparison results were used to extract the protein intake difference comparison results and blood glucose concentration parameters, and a multidimensional deviation relationship model between the two was constructed using the formula: ; Calculate the deviation between calculation parameters , generate parameter deviation relationship analysis results; in, Indicates the total number of data. Represents protein intake, represents the predicted value of enteral nutrition requirement, represents the weight coefficient of blood glucose concentration, Represents the blood glucose concentration parameter.

[0013] As a further solution of the present invention, the step of obtaining the enteral nutrition requirement matching degree is specifically as follows: Using the parameter deviation relationship analysis results, dividing the data according to the threshold of the differentiated deviation interval, classifying the divided interval data and recording the frequency, and generating the enteral nutrition deviation distribution results by counting the deviation points in the differentiated interval; According to the enteral nutrition deviation distribution results, combined with the enteral nutrition demand prediction value, the formula is used: ; Calculate the overall matching degree and generate the enteral nutrition requirement matching degree; in, Represents the matching degree of enteral nutrition requirements, Representative The weight coefficient of the deviation interval, Representative The normalized frequency value of the interval, represents the sum of the frequencies of the deviation intervals, The number representing the deviation interval, Indicates the total number of intervals.

[0014] Compared with the prior art, the advantages and positive effects of the present invention are: In the present invention, by encrypting the enteral nutrition-related data, the results of the intake and physiological parameters are integrated to provide a multi-source data set, which significantly improves the security and privacy protection of the data. When integrating the encrypted data sets of different institutions, it is possible to analyze and compare cross-institutional data, which helps to more accurately judge nutritional needs and ensure the consistency of the analysis results. By normalizing the total intake value and the physiological mean, and analyzing the changing trend of nutritional needs, the assessment of nutritional status is made more scientific and dynamic, and the quality of medical services is effectively improved. The use of difference comparison and parameter deviation analysis further ensures the accurate matching of individualized nutrition plans, helps to formulate more appropriate nutritional supplementation strategies, and improves the patient's recovery effect and quality of life through accurate nutritional needs assessment. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] Figure 1 is a system flow chart of the present invention; Figure 2 It is a flowchart of the multi-source encrypted data set in the present invention; Figure 3 A flow chart of encrypting intake data values ​​in the present invention; Figure 4 A flowchart of cross-institutional encrypted data set fusion in the present invention; Figure 5 A flow chart of the dynamic nutritional requirement identification results in the present invention; Figure 6 A flow chart of the enteral nutrition requirement prediction value in the present invention; Figure 7 It is a flow chart of the parameter deviation relationship analysis results in the present invention; Figure 8 The figure is a flow chart of the enteral nutrition requirement matching degree in the present invention. DETAILED DESCRIPTION

[0016] In order to make the purpose, technical solution and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with 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 intended to limit the present invention.

[0017] In the description of the present invention, it should be understood that the terms "length", "width", "up", "down", "front", "back", "left", "right", "vertical", "horizontal", "top", "bottom", "inside", "outside" and the like indicate positions or positional relationships based on the positions or positional relationships shown in the drawings, and are 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 cannot be understood as limiting the present invention. In addition, in the description of the present invention, "multiple" means two or more, unless otherwise clearly and specifically defined.

[0018] See also Figure 1 , an enteral nutrition multi-source data fusion analysis system comprises: The encryption processing module performs encryption transformation processing on the values ​​of protein intake, carbohydrate intake, and fat intake based on the input enteral nutrition-related data source, performs encryption calculation processing on the values ​​of blood glucose concentration, human albumin concentration, and urea nitrogen concentration, integrates the encryption results of intake and physiological parameters, and obtains a multi-source encrypted data set; The data fusion module uses multi-source encrypted data sets to extract encrypted intake data values ​​from medical institution data and individual patient data, calculate the total amount of differentiated source data, calculate the encrypted mean for multiple parameters, merge the encrypted mean and the encrypted total, and perform merging and parameter comparison to obtain a cross-institutional fused encrypted data set; The nutritional demand prediction module uses a cross-institutional fused encrypted data set to perform normalization calculations on the encrypted total intake value and the encrypted physiological mean value, analyze the relationship between protein intake and human albumin concentration, generate dynamic nutritional demand identification results, calculate dynamic offset values, analyze offset trends to identify the range of changes, evaluate the changing trend of nutritional demand, and obtain the enteral nutritional demand prediction value; The matching degree assessment module compares and calculates the difference between protein intake and the predicted value item by item based on the predicted value of enteral nutrition demand, and obtains the parameter deviation relationship analysis results in combination with the blood glucose concentration, calculates the degree of deviation, analyzes the degree of nutrition supply adaptation, and obtains the matching degree of enteral nutrition demand; The multi-source encrypted data sets include encrypted data of protein intake, encrypted data of carbohydrate intake, encrypted data of fat intake, encrypted data of blood glucose concentration, encrypted data of human albumin concentration, and encrypted data of urea nitrogen concentration. The results of the encrypted intake data values ​​include encrypted values ​​of total intake and encrypted means of multiple nutrients. The cross-institutional fusion encrypted data sets include encrypted intake data, encrypted physiological mean data, encrypted mean of parameters, and encrypted total of parameters. The results of dynamic nutritional demand identification include analysis of the relationship between protein and human albumin concentration, calculation of dynamic offset values, analysis of offset trend, and trend of nutritional demand changes. The predicted values ​​of enteral nutritional demand include protein demand prediction, carbohydrate demand prediction, and fat demand prediction. The results of parameter deviation relationship analysis include deviation values ​​of blood glucose concentration, deviation values ​​of human albumin concentration, and deviation values ​​of urea nitrogen concentration. The matching degree of enteral nutritional demand includes analysis of the difference between protein intake and predicted values, deviation of blood glucose concentration parameters, and analysis results of nutritional supply adaptation degree.

[0019] See also Figure 2 ,The specific steps for obtaining multi-source encrypted data sets are: Based on the input enteral nutrition-related data source, the protein intake, carbohydrate intake and fat intake values ​​are recorded, standardized, the deviation between the intake value and the average value is calculated, and the three types of intake deviation values ​​are summed to obtain the intake weighted result; Based on the input protein intake, carbohydrate intake, and fat intake values, the original data are decomposed into individual items, and each type of intake value is normalized by the maximum and minimum values ​​to obtain a set of standardized intake values. The average of each intake value is calculated, and the deviation of each intake value from the average is calculated. The deviation is taken as an absolute value and then added to the weight factor. The weight factor is set according to the importance of the impact on physiological metabolism. Each weighted deviation value is added and summed, and the accumulated result is normalized again to unify the numerical range. All normalized weighted deviation values ​​are integrated to form a comprehensive characteristic index of intake, and a weighted result of intake is generated.

[0020] The intake weighted results were used, and the values ​​of blood glucose concentration, human albumin concentration, and urea nitrogen concentration were standardized and matched one by one with the intake weighted results, using the formula: ; Calculate the intake and physiological parameters to match the deviation value , generating physiological parameter matching results; in, is the weight parameter, represents the intake-weighted value, Represents the corresponding physiological parameters The numerical value of is the number of matches between intake parameters and physiological parameters; The formula introduces the weight parameter Assigning different weights to the deviations between each intake and physiological parameters can more accurately reflect the contribution of each parameter to the overall result. At the same time, the calculation method of matching the square of the deviation can strengthen the influence of the larger deviation value and improve the sensitivity and accuracy of the calculation. Weight Parameters The values ​​are 0.4, 0.3, and 0.3, corresponding to protein intake, carbohydrate intake, and fat intake, respectively; Set the intake weighted value , , ; Corresponding physiological parameter values , , ; Substituting into the formula: ; The results showed that the matching deviation value was 5. The lower the deviation value, the higher the matching degree between the intake and the physiological parameters. This value was used as the physiological parameter matching result for subsequent analysis.

[0021] Perform deviation analysis on the physiological parameter matching results, compare the matching deviation value with the set threshold, delete the results with deviation values ​​exceeding the threshold, adjust the weight parameters of the correlation between intake and physiological parameters, and generate a multi-source encrypted data set; Based on the set matching deviation threshold, the sizes of the deviation values ​​are compared one by one, and the results higher than the set threshold are filtered out according to the numerical range of the deviation value, and only the set with matching deviation values ​​lower than or equal to the threshold is retained. The weight ratio of each deviation value in the retained set is calculated according to the original input data source, and the corresponding weight coefficient is generated. The weighted average value is calculated by multiplying the weight coefficient with the matching deviation value, and then the sum is calculated. The weighted matching deviation result of the original intake and physiological parameters corresponding to each weight is further called, and the parameter set with smaller deviation values ​​is integrated. The integrated result is compared with the weighted average value, and multiple matching conditions are optimized. The rationality of the result is ensured by adjusting the matching logic layer by layer to generate a multi-source encrypted data set.

[0022] See also Figure 3 , the specific steps for obtaining the encrypted intake data value are: The data fusion module extracts medical institution data and individual patient data from multi-source encrypted data sets, and during the extraction process, screens and separates the encrypted data entries of medical institutions and patients to obtain an encrypted patient data set; Medical institution data and individual patient data are extracted from multi-source encrypted data sets. Data entries are classified by calling medical data files. Medical institution data entries and patient data entries are split according to source labels, and field identifiers of medical institution data sources are extracted. The institutional attributes and data integrity of the data sources are analyzed in turn. For each data record, the preset rules are called to verify the integrity of the entry fields. If a field is found to be missing, the field is completed according to the rules of historical data. After completion, the entry is marked as completed data for subsequent tracing and tracing. For entries from patient data sources, they are grouped according to patient identity identifiers, and entries with duplicate identity identifiers are merged into a comprehensive record. At the same time, the data source priority is marked and the merging rule is recorded. After completing the classification processing of medical institution data entries and patient entries, the classification algorithm is called to store the data set in layers, and it is stored as an independent data set according to the data source. In each independent data set, a subdirectory is generated according to the data field category for fast retrieval to obtain an encrypted patient data set.

[0023] identifying encrypted entries associated with enteral nutrition intake from the encrypted patient data set, extracting intake data from the encrypted entries through keyword matching, and obtaining an encrypted data entry set; Retrieve the intake-related encrypted entries from the encrypted medical institution data set and the encrypted patient data set, extract the field names in the data set entries according to the keyword rules, check one by one whether the "intake" field exists in the entries, filter out the entries containing the relevant fields, call the rule matching model to perform data integrity check on the filtered entries, if the entry contains empty value fields or abnormal characters, the entry will be eliminated and the reason for elimination will be recorded, and the field format of the complete entry will be checked to ensure that the field content conforms to the expected format rules (such as: the uniformity of the intake unit, the legality of the field length, etc.), and the entries that have completed the check will be stored as an independent data set, and then the intake field value carried by each entry in the independent data set will be decrypted and marked, and the encrypted value will be converted to a state to be parsed, providing identifiable intake data for subsequent operations, and obtaining an encrypted data entry set.

[0024] Decrypt the encrypted data entry set item by item, extract the intake data value, obtain the weight information associated with the intake from the data entry, and combine it with the patient's individual difference data to obtain the decrypted intake information; The encrypted data entry set containing the intake information is decrypted item by item, the key field carried by the entry is called to parse the data encryption status, the data field decoding operation is completed by calling the encryption and decryption rules, the intake field value is converted from the encrypted format to a recognizable numerical form, and the feature extraction algorithm is called for each entry to extract the data information related to the weight in the entry, and the weight parameter is generated according to the weight rule in combination with the institution identification of the source of the entry. Next, the historical data related to the patient's identity in the entry is analyzed, and the patient's individual difference parameters related to the intake are generated by extracting features such as the fluctuation range of the patient's intake in the previous data and the intake time, combined with the normalization method. On this basis, according to the historical scores of the source institution of the data entry, the quality parameters in the scoring record are extracted and the credibility parameters of the current entry are calculated. According to the encryption strength level recorded in the data entry, the encryption strength correction coefficient of the corresponding entry is generated by parsing the strength parameter value to obtain the decrypted intake information.

[0025] The intake values ​​in the decrypted intake information are weighted according to the weights, combined with the patient individual difference parameters and the credibility parameters, using the formula: ; Calculate the encrypted intake data value ; in, Represents the intake value of each item, represents the intake item weight, Represents the individual difference parameter of patients, Represents the credibility parameter of medical institutions, represents the encryption strength correction factor, Represents the number of entries, Represents the entry index; The formula weights the intake data by adding individual patient difference parameters and medical institution credibility parameters, and adjusts the weight calculation logic in combination with the encryption strength correction coefficient, which can significantly improve the accuracy of the results and the adaptability of data fusion. Detailed explanation of the formula and the process of formula calculation and derivation: Set the number of data entries , intake data value , , , , ; Weight Parameters , , , , ; Patient individual difference parameters , , , , ; Institutional Credibility Parameters , , , , ; Encryption Strength Correction Factor , , , , ; Weighted intake calculation:

[0026] ; Modified weight sum calculation: ; Substitute the data for calculation: ; The results show that the encrypted intake data value is 176.37, which represents the balanced weighted average of intake after multi-source data fusion. This value is the core basis for subsequent processing and represents the comprehensive result of data fusion.

[0027] See also Figure 4 ,The specific steps for obtaining the cross-institutional fused encrypted data set are: Extract encrypted intake values ​​across medical institutions from the encrypted intake data, group and classify the data by identifying data tags of differentiated medical institution sources, and obtain institution group records; Extracting encrypted intake values ​​across medical institutions involves detailed identification and labeling of data sets from different medical institutions, and accurate classification of intake data using institution-specific data labels. By analyzing the data labels in each institution's data set, the data records of each institution can be identified and separated. This stage includes key steps such as confirming data labels, checking data integrity, and deleting duplicate records. It is necessary to verify that each data record is complete and ensure that all necessary fields are filled. By comparing the data labels with the known institution identification list, confirm that the attribution of each data record is correct. Ensuring the accuracy and reliability of the data set lays a solid foundation for subsequent data processing and analysis. The correct classification of each data set is a prerequisite for effective data analysis, providing important original materials for further data analysis and application, and obtaining institutional grouping records.

[0028] According to the records of institutional grouping, the encrypted intake mean and corresponding data coverage of differentiated institutions are calculated, and the data coverage parameter set is obtained by calculating the proportion of intake data values ​​in each institution; Calculate the mean of each institution's encrypted intake and the corresponding data coverage. The calculation process involves a detailed statistical analysis of each institution's data set. Statistics will be performed on each institution's data set, including the calculation of the average intake, standard deviation and related statistical indicators of the data set. The statistical results are obtained by checking and calculating the data set record by record. Evaluate the coverage of the data set, that is, analyze the time range and data integrity covered in the data set. This process is not just a mathematical calculation, but also requires an assessment of data quality, including checking the consistency and accuracy of the data. By comparing the consistency of data in different time periods, the stability and reliability of the data set can be evaluated. All statistical analyses are completed with the support of strict data processing protocols and algorithms to ensure that the data of each institution can accurately reflect the characteristics, provide a scientific basis for subsequent data fusion and application, and obtain a set of data coverage parameters.

[0029] Using the data coverage parameter set, the mean encrypted intake of each institution is weighted according to the data coverage, using the formula: ; Calculate the normalized value for each institution , perform weighted summation on the normalized values ​​of the institutions to generate a cross-institutional fused encrypted data set; in, Representative The data coverage of each institution, Representative The average intake of each institution, Represents the total number of medical institutions involved in the calculation, The largest mean intake in the representative institution; The formula adjusts the ratio of the mean encrypted intake of each institution to the data coverage and standardizes it by the maximum intake mean. The formula can more accurately calculate the relative importance of each institution's data in the overall data set and more reasonably allocate the influence weight of each institution when merging data. Detailed explanation of the formula and the process of formula calculation and derivation: The data coverage of three institutions is set as follows: , , ; The total coverage is , the mean intake of each institution is assumed to be , , ; The maximum intake is , then the normalized data for each institution Calculated as: ; ; ; The results show that by normalizing the data of each institution, it is possible to ensure that the data quality and coverage of each institution are properly considered during the data fusion process, thereby obtaining a balanced and representative fusion data set, providing accurate basic data for subsequent analysis.

[0030] See also Figure 5 ,The specific steps for obtaining the dynamic nutritional demand identification results are: By using a cross-institutional fusion encrypted data set, the encrypted total intake parameters and encrypted physiological mean parameters were extracted, and the values ​​between the two were normalized to obtain a normalized parameter value set; By calling the encrypted total intake parameters and the encrypted physiological mean parameters, the normalization operation is completed based on the standardized formula. The mean and standard deviation of the encrypted intake data and the encrypted physiological mean data are calculated respectively. The mean is obtained by summing all data items and dividing by the total number of data. The standard deviation is obtained by calculating the sum of the squares of the deviations of each data item from the mean and dividing it by the total number of data and then taking the square root. The normalization operation is completed by subtracting the mean from each data item and dividing it by the standard deviation. At the same time, in order to avoid interference from extreme values, the data set is cleaned up. Specifically, the abnormal data that exceeds the upper and lower limit range is eliminated by setting upper and lower limit values. The limit value can be set according to the distribution range of the normalized data by a statistical method. The values ​​corresponding to the 1st percentile and the 99th percentile are taken as the limit. The normalization operation is performed again on the data set after eliminating the extreme values ​​to ensure the standardization of the normalized data distribution, provide basic data support for subsequent dynamic correlation calculations and time series analysis, and obtain a normalized parameter value set.

[0031] Using the normalized parameter value set, the normalized encrypted intake and encrypted physiological mean parameter values ​​are extracted, and the correlation change trend of the two parameters within the dynamic range is analyzed using the formula: ; Calculate the dynamic correlation coefficient , analyze the time series deviation amplitude of protein intake and human albumin concentration in enteral nutrition to generate dynamic nutritional demand identification results; in, Represents the encrypted intake data values, Represents the encrypted physiological mean data values, represents the mean of crypto intake, represents the mean of the encrypted physiological means, Represents the total number of data, Represents an accumulation operation; The formula quantifies the correlation between the encrypted intake data and the encrypted physiological mean data into a value that can be used for trend judgment and offset analysis by calculating the dynamic correlation coefficient. At the same time, combined with the offset amplitude extraction of time series data, it can accurately analyze the dynamic change trend and provide quantitative results, providing a strong basis for the prediction of nutritional needs. Detailed explanation of the formula and the process of formula calculation and derivation: Suppose the encrypted intake data is , the encrypted physiological mean data is , calculate the mean of the two groups of data respectively, and , calculate the mean deviation for each group of data, that is and , the covariance term is calculated by summing the mean deviation product of all data points , calculate the sum of squares of mean deviations for the two sets of data respectively, that is and , the ratio of the covariance term to the square root of the product of the sum of squares is the dynamic correlation coefficient; Substitute into the example: set up , ,but , ; Calculating covariance ; The squared deviation is ; ; Substitute into the formula to calculate: ; The results show that the dynamic correlation between the two sets of data is completely consistent. Based on this correlation coefficient, the time series offset amplitude of protein intake and human serum albumin concentration can be extracted, and dynamic nutritional demand identification results can be generated by dynamically analyzing the offset amplitude, providing quantitative support for the prediction of the next nutritional demand.

[0032] See also Figure 6 , the specific steps for obtaining the predicted value of enteral nutrition demand are: Extract the dynamic offset value range parameters of the dynamic nutritional demand identification results, combine the correlation between protein intake and human albumin concentration, analyze the mapping relationship between the dynamic offset value and the intake adjustment coefficient, and obtain the adjusted dynamic offset value set; The dynamic offset value range parameters of the dynamic nutritional demand identification results are called to analyze the relationship between protein intake and human albumin concentration, and the intake adjustment factor of each mapping point is calculated. The intake adjustment factor is calculated by operating with the components of the dynamic offset value to obtain the specific adjustment parameter value of each data point in different ranges. In the mapping process, the time series data of protein intake is called, and a linear weighted analysis is performed on each point in the dynamic offset value range in combination with the relational model. A corresponding weight parameter matrix is ​​established according to the dynamic offset adjustment factor mapping value of each point. The dynamic offset value is gradually adjusted to the ideal range that conforms to the relational model through the weight matrix, and the abnormal fluctuation data of the intake is further identified and eliminated. The weighted average of the dynamic offset value is calculated after summing all the intake adjustment factors. The average value is used to adjust the points of abnormal deviation in the offset value set. Through further recursive calls, the stable state of the adjusted set is confirmed, and the adjusted dynamic offset value set is generated.

[0033] Based on the adjusted dynamic offset value set, combined with the changing parameters of the short-term offset trend and the long-term offset trend in the time series, the formula is adopted: ; Calculate dynamic weighted adjustment values , generate a predicted offset value; in, Representative The adjusted dynamic offset value, Represents the short-term deviation trend parameter values, Represents the long-term deviation trend parameter values, Represents the total number of time series data points; The formula performs absolute value weighted calculation on the superposition relationship between short-term offset trend and long-term offset trend, so that the offsets of different time periods contribute to the dynamic adjustment value in a balanced manner, avoiding the excessive influence of a single trend parameter on the overall adjustment value, and improving the reliability and representativeness of the adjustment value. Detailed explanation of the formula and the process of formula calculation and derivation: Get the adjusted dynamic offset value set , set the value of the first point in the collection to , the second point is , the third point is , a total of three points; Short-term deviation trend collection middle , , ; Long-term deviation trend collection middle , , ; Total points ; Substitute the above parameters into the formula for calculation: Calculate the superposition of short-term and long-term trends: ; ; ; Calculate the absolute value of the product of the offset value and the superimposed trend for each point: ; ; ; Sum and calculate the average: ; ; Take the square root of the result: ; The results show that the dynamic weighted adjustment value This value, as an important parameter for predicting the offset value, can directly affect the next step of dynamic change range calculation and the prediction results of enteral nutrition needs.

[0034] According to the predicted offset value, the encrypted intake parameters and encrypted physiological mean parameters in enteral nutrition are called, and the dynamic range of intake demand is analyzed in combination with the amplitude range of the offset in the time series, and the impact of short-term and long-term trends on enteral nutrition demand is evaluated to obtain the predicted value of enteral nutrition demand; According to the predicted offset value, the encrypted intake parameters and the encrypted physiological mean parameters are further called to calculate the intake demand value of each time series point. The time series data is split into short-term data and long-term data. The offset amplitude of each point in the short-term data is calculated, and the data with an offset amplitude higher than the upper limit or lower than the lower limit is marked as an abnormal point, and similar data in the long-term data is called as a substitute. Each demand value is filtered by calling the parameters of the upper and lower limits of the offset amplitude in the time series to eliminate abnormal high or low values. At the same time, linear interpolation is performed on the alternative data points to ensure data continuity, and the offset mean of all remaining data points is further calculated. The overall intake demand fluctuation range in the time series is calculated by the mean, and the linear weighted sum of the short-term trend and long-term trend parameters is solved by calling the filtered demand value set. Combined with the overall change range in the time series, the dynamic change parameters of the intake demand are recorded, and the weighted calculation results of the dynamic change parameters are used to evaluate the comprehensive impact of short-term and long-term trends on enteral nutrition demand to obtain the predicted value of enteral nutrition demand.

[0035] See also Figure 7 , the specific steps for obtaining the parameter deviation relationship analysis results are: The matching degree evaluation module calls the enteral nutrition demand prediction value and combines the enteral nutrition protein intake parameter to organize the difference data between the data point intake and the prediction value into vector form for storage, analyzes the difference direction by item-by-item comparison, identifies the matching relationship between the direction and the absolute value, and generates the protein intake difference comparison result; Call the predicted value of enteral nutrition demand and protein intake parameters, calculate the difference between the intake value and the demand value at each time point one by one based on the actual value of daily protein intake and the predicted demand value, arrange the calculated differences in time point order to generate difference sequence data, obtain the degree of deviation by calculating the absolute value of the difference, judge whether the deviation of each difference exceeds the statistical fluctuation range, set the standard deviation interval to quantitatively determine the normal fluctuation range, mark the difference within the deviation range as "normal deviation", and mark the difference outside the range as "abnormal deviation", record the time of the point outside the range, and mark the direction as positive deviation or negative deviation, store the deviation sequence data and add a time mark table of the deviation direction, extract the data with larger absolute deviation in the mark table again, record the deviation data exceeding a certain threshold as extreme points, integrate the difference sequence, deviation sequence data, time mark table and extreme deviation points to generate protein intake difference comparison results.

[0036] Using the protein intake difference comparison results, the protein intake difference comparison results and blood glucose concentration parameters were extracted to construct a multidimensional deviation relationship model between the two, using the formula: ; Calculate the deviation between calculation parameters , generate parameter deviation relationship analysis results; in, Indicates the total number of data. Represents protein intake, represents the predicted value of enteral nutrition requirement, represents the weight coefficient of blood glucose concentration, represents the blood glucose concentration parameter; The formula can quantify the deviation between protein intake and nutritional requirements and the stability of blood glucose concentration by comprehensively calculating the square of the intake deviation ratio and the weighted square of the blood glucose concentration fluctuation. It can more accurately reflect the comprehensive impact of supply and demand imbalance on patient health when analyzing the adaptability of nutritional supply. Detailed explanation of the formula and the process of formula calculation and derivation: (Protein intake: the average protein intake at each time point is obtained by sampling and counting the actual protein intake data of the patient every day); (Enteral nutrition demand prediction value is obtained by combining the patient's long-term historical data with the dynamic prediction of the nutrition demand model); (Blood sugar concentration weight parameter, referring to the relative importance of blood sugar concentration fluctuations, is set by observing the actual impact of blood sugar fluctuations on protein metabolism and can be adjusted according to the range of blood sugar fluctuations); (Blood glucose concentration parameter, collected by the patient's continuous blood glucose monitoring device at fixed time intervals and calibrated to obtain the average value); (Total number of data, indicating the number of data points collected during the statistical analysis); Substituting into the formula: ; Calculation steps: ; ; ; The results showed that the degree of deviation was 6.04. The high value indicated that there was a significant deviation in the matching of protein intake and demand. It also reflected that the impact of blood glucose concentration fluctuations on the adaptability analysis of nutritional supply was more obvious. This value will be further used as an input parameter in subsequent adaptability analysis.

[0037] See also Figure 8 , the specific steps for obtaining the matching degree of enteral nutrition requirements are as follows: Using the parameter deviation relationship analysis results, the data is divided according to the threshold of the differentiated deviation interval, the divided interval data is classified and the frequency is recorded, and the enteral nutrition deviation distribution results are generated by counting the deviation points in the differentiated interval; By calculating the absolute value of the deviation of each data point and analyzing its correspondence with the distribution interval, the data points are preliminarily divided according to the range of the deviation value and classified into high deviation, medium deviation or low deviation intervals. For the data points in each interval after the division, the relative frequency ratio of each interval is calculated through frequency statistics. At the same time, according to the numerical value of each frequency ratio, the data points in the interval are further adjusted to ensure the consistency of the frequency distribution and the characteristics of the data points. In the normalization process, for the overlapping parts between intervals, the calculation accuracy of the overlapping point frequency is corrected through multiple iterations to eliminate the confusion effect between different deviation intervals and generate the enteral nutrition deviation distribution results.

[0038] According to the enteral nutrition deviation distribution results and the predicted value of enteral nutrition demand, the formula is used: ; Calculate the overall matching degree and generate the enteral nutrition requirement matching degree; in, Represents the matching degree of enteral nutrition requirements, Representative The weight coefficient of the deviation interval, Representative The normalized frequency value of the interval, represents the sum of the frequencies of the deviation intervals, The number representing the deviation interval, Indicates the total number of intervals; The formula is introduced by introducing the deviation interval frequency ratio With interval weight The weighted calculation of can quantify the contribution of each deviation interval to the overall nutritional matching, and at the same time combine the total frequency value Normalization can accurately reflect the matching status of overall nutrient supply and demand, which helps to optimize supply adjustment strategies and reduce deviations in subsequent processes; How to obtain each parameter and its value description: Total number of intervals , determined by statistical analysis of the enteral nutrition deviation distribution results; Deviation interval frequency ratio : The frequency statistics of data points from each interval are obtained by normalization. Set , , , , ; Weight Parameters :Based on the degree of deviation of each interval and the degree of contribution to the overall nutritional needs matching, set , , , , ; Total frequency value : Calculated ; Calculation process: Calculate the numerator part: ; Calculate the denominator: ; Substituting into the formula: ; The results show that the overall matching degree of current nutrition supply to demand is 0.825, which is close to 1, indicating that the overall nutrition supply situation is relatively good. The specific numerical results show that the overall nutrition supply deviation is within a reasonable range, but there is still room for adjustment in some intervals. The calculated results can be used as an important basis for further optimizing the nutrition supply strategy and improving key issues in the deviation interval.

[0039] The above are only preferred embodiments of the present invention and are not intended to limit the present invention in other forms. Any technician familiar with the profession may use the technical contents disclosed above to change or modify them into equivalent embodiments with equivalent changes and apply them to other fields. However, any simple modification, equivalent change and modification made to the above embodiments based on the technical essence of the present invention without departing from the technical solution of the present invention still falls within the protection scope of the technical solution of the present invention.

Claims

1. An enteral nutrition multi-source data fusion analysis system, characterized in that: The system comprises: The encryption processing module performs encryption transformation processing on the values ​​of protein intake, carbohydrate intake, and fat intake based on the input enteral nutrition-related data source, and performs encryption calculation processing on the values ​​of blood glucose concentration, human serum albumin concentration, and urea nitrogen concentration to obtain a multi-source encrypted data set; The data fusion module uses the multi-source encrypted data set to extract the encrypted intake data values ​​from the medical institution data and the individual patient data, calculate the total amount of differentiated source data, calculate the encrypted mean for multiple parameters, merge the encrypted mean and the encrypted total amount, and perform merging and comparison with the parameters to obtain a cross-institutional fused encrypted data set; The nutritional demand prediction module uses the cross-institutional fused encrypted data set to perform normalization calculations on the encrypted total intake value and the encrypted physiological mean value, analyzes the relationship between protein intake and human serum albumin concentration, generates dynamic nutritional demand identification results, calculates dynamic offset values, analyzes offset trend identification ranges, evaluates nutritional demand change trends, and obtains enteral nutritional demand prediction values; The matching degree assessment module compares and calculates the difference between the protein intake and the predicted value item by item based on the predicted value of enteral nutrition demand, obtains the parameter deviation relationship analysis result in combination with the blood glucose concentration, calculates the degree of deviation, analyzes the degree of nutrition supply adaptation, and obtains the matching degree of enteral nutrition demand.

2. The enteral nutrition multi-source data fusion analysis system according to claim 1, characterized in that: The steps of obtaining the multi-source encrypted data set are specifically as follows: Based on the input enteral nutrition-related data source, the protein intake, carbohydrate intake and fat intake values ​​are recorded, standardized, the deviation between the intake value and the average value is calculated, and the three types of intake deviation values ​​are summed to obtain the intake weighted result; The weighted results of the intake are used, and the values ​​of blood glucose concentration, human albumin concentration, and urea nitrogen concentration are standardized and matched one by one with the weighted results of the intake, using the formula: ； Calculate the intake and physiological parameters to match the deviation value , generating physiological parameter matching results; in, is the weight parameter, represents the intake-weighted value, Represents the corresponding physiological parameters The numerical value of is the number of matches between intake parameters and physiological parameters; The physiological parameter matching results are subjected to deviation analysis, the matching deviation value is compared with the set threshold, the results whose deviation value exceeds the threshold are deleted, the weight parameters of the correlation between the intake and the physiological parameters are adjusted, and a multi-source encrypted data set is generated.

3. The enteral nutrition multi-source data fusion analysis system according to claim 2, characterized in that: The steps for obtaining the encrypted intake data value are specifically as follows: The data fusion module extracts the medical institution data and the individual patient data from the multi-source encrypted data set, and during the extraction process, screens and separates the encrypted data entries of the medical institution and the patient to obtain an encrypted patient data set; Identify encrypted entries associated with enteral nutrition intake from the encrypted patient data set, extract intake data from the encrypted entries through keyword matching, and obtain an encrypted data entry set; Decrypting the encrypted data entry set item by item, extracting the intake data value, and obtaining the weight information associated with the intake from the data entry, and combining it with the patient individual difference data to obtain the decrypted intake information; The intake values ​​in the decrypted intake information are weighted according to the weights, and the patient individual difference parameters and the credibility parameters are combined to adopt the formula: ； Calculate the encrypted intake data value ; in, Represents the intake value of each item, represents the intake item weight, Represents the individual difference parameter of patients, Represents the credibility parameter of medical institutions, represents the encryption strength correction factor, Represents the number of entries, Represents the entry index.

4. The enteral nutrition multi-source data fusion analysis system according to claim 3, characterized in that: The steps for obtaining the cross-institutional fused encrypted data set are specifically as follows: Extracting encrypted intake values ​​across medical institutions from the encrypted intake data, grouping and classifying the data by identifying data tags of differentiated medical institution sources, and obtaining institution grouping records; According to the institutional grouping records, the encrypted intake mean and the corresponding data coverage of the differentiated institutions are calculated, and the data coverage parameter set is obtained by calculating the proportion of the intake data values ​​in each institution; Using the data coverage parameter set, the mean encrypted intake of each institution is weighted according to the data coverage, using the formula: ； Calculate the normalized value for each institution , perform weighted summation on the normalized values ​​of the institutions to generate a cross-institutional fused encrypted data set; in, Representative The data coverage of each institution, Representative The average intake of each institution, Represents the total number of medical institutions involved in the calculation, The maximum intake mean in the representative institution.

5. The enteral nutrition multi-source data fusion analysis system according to claim 4, characterized in that: The steps for obtaining the dynamic nutritional requirement identification result are specifically as follows: The cross-institutional fused encrypted data set is used to extract the encrypted total intake parameter and the encrypted physiological mean parameter, and the values ​​between the two are normalized to obtain a normalized parameter value set; By using the normalized parameter value set, the normalized encrypted intake and the encrypted physiological mean parameter values ​​are extracted, and the correlation change trend of the two parameters within the dynamic range is analyzed, using the formula: ； Calculate the dynamic correlation coefficient , analyze the time series deviation amplitude of protein intake and human albumin concentration in enteral nutrition to generate dynamic nutritional demand identification results; in, Represents the encrypted intake data values, Represents the encrypted physiological mean data values, represents the mean of crypto intake, represents the mean of the encrypted physiological means, Represents the total number of data, Represents an accumulation operation.

6. The enteral nutrition multi-source data fusion analysis system according to claim 5, characterized in that: The steps for obtaining the predicted value of enteral nutrition demand are specifically as follows: Extracting the dynamic offset value range parameter of the dynamic nutritional requirement identification result, combining the correlation between protein intake and human albumin concentration, analyzing the mapping relationship between the dynamic offset value and the intake adjustment coefficient, and obtaining an adjusted dynamic offset value set; Based on the adjusted dynamic offset value set, combined with the change parameters of the short-term offset trend and the long-term offset trend in the time series, the formula is adopted: ； Calculate dynamic weighted adjustment values , generate a predicted offset value; in, Representative The adjusted dynamic offset value, Represents the short-term deviation trend parameter values, Represents the long-term deviation trend parameter values, Represents the total number of time series data points; According to the predicted offset value, the encrypted intake parameters and encrypted physiological mean parameters in enteral nutrition are called, and the dynamic change range of intake demand is analyzed in combination with the amplitude range of the offset in the time series, and the impact of short-term and long-term trends on enteral nutrition demand is evaluated to obtain the predicted value of enteral nutrition demand.

7. The enteral nutrition multi-source data fusion analysis system according to claim 6, characterized in that: The steps for obtaining the parameter deviation relationship analysis results are specifically as follows: The matching degree evaluation module stores the difference data between the intake of the data point and the predicted value in a vector form according to the predicted value of the enteral nutrition demand called, combined with the protein intake parameter of the enteral nutrition, and analyzes the difference direction by comparing and analyzing the difference direction item by item, identifies the matching relationship between the direction and the absolute value, and generates a protein intake difference comparison result; The protein intake difference comparison results were used to extract the protein intake difference comparison results and blood glucose concentration parameters, and a multidimensional deviation relationship model between the two was constructed using the formula: ； Calculate the deviation between calculation parameters , generate parameter deviation relationship analysis results; in, Indicates the total number of data. Represents protein intake, represents the predicted value of enteral nutrition requirement, represents the weight coefficient of blood glucose concentration, Represents the blood glucose concentration parameter.

8. The enteral nutrition multi-source data fusion analysis system according to claim 7, characterized in that: The steps for obtaining the enteral nutrition requirement matching degree are specifically as follows: Using the parameter deviation relationship analysis results, dividing the data according to the threshold of the differentiated deviation interval, classifying the divided interval data and recording the frequency, and generating the enteral nutrition deviation distribution results by counting the deviation points in the differentiated interval; According to the enteral nutrition deviation distribution results, combined with the enteral nutrition demand prediction value, the formula is used: ； Calculate the overall matching degree and generate the enteral nutrition requirement matching degree; in, Represents the matching degree of enteral nutrition requirements, Representative The weight coefficient of the deviation interval, Representative The normalized frequency value of the interval, represents the sum of the frequencies of the deviation intervals, The number representing the deviation interval, Indicates the total number of intervals.

Images