Method for processing home care monitoring data of old patient with lung cancer

Abstract

The invention discloses a method for processing home care monitoring data of an old lung cancer patient, which comprises the following steps of: firstly, monitoring physiological parameter data, mastering the health state of the patient in real time, and timely discovering abnormal conditions, such as body temperature rise, blood oxygen level drop and the like; various physiological parameters are integrated through multi-dimensional data analysis, comprehensive physical state evaluation information evaluation is provided, health problems possibly unable to be revealed by single parameters can be found, for example, the nutritional status of a patient can be better understood through weight data analysis; furthermore, through the data monitoring and analysis mechanism, nursing personnel can make more accurate nursing decisions and adjust nursing measures and treatment schemes according to the specific data of the patient, and the treatment effect of the patient is improved.

Description

Technical Field

[0001] The present invention relates to the medical field, and in particular to a method for processing home care monitoring data of elderly lung cancer patients. Background Art

[0002] At present, the physical condition monitoring and evaluation of elderly lung cancer patients mainly rely on regular hospital examinations and simple home monitoring equipment. Traditional monitoring methods mainly include the following aspects: 1. Regular physical examinations, that is, patients need to go to the hospital regularly for comprehensive examinations, including blood tests, imaging examinations, etc.; 2. Implementation of single parameter monitoring, that is, at home, patients may use equipment such as sphygmomanometers and thermometers to monitor a single physiological parameter; 3. Static data analysis: Medical personnel are usually based on the examination results at a certain point in time and are single parameters, so this single-dimensional physiological data indicator monitoring and evaluation and the monitoring time of each parameter indicator are different, which will lead to mismatched parameter indicator data during static data analysis. For example, at time j, the patient's heart rate is measured to be 75bpm, while the blood pressure measurement result is 120mmHg (that is, data collection and analysis are not timely, so sometimes the blood pressure measurement result does not truly reflect the actual blood pressure value at the current moment);

[0003] Therefore, it is found that this traditional method has the following technical defects: 1. Data discontinuity: Regular physical examinations cannot provide continuous physiological data and may miss important changes in health status. 2. Parameter limitation: Single parameter monitoring cannot fully reflect the overall health status of the patient. 3. Lack of dynamic analysis and inability to detect abnormalities in time: Static data analysis cannot capture the dynamic changes in the patient's health status, that is, due to untimely data collection and analysis, the discovery and treatment of abnormal conditions may be delayed. Summary of the invention

[0004] The purpose of the present invention is to provide a method for processing home care monitoring data of elderly lung cancer patients, thereby solving the above-mentioned technical problems pointed out in the prior art.

[0005] The present invention provides a method for processing home care monitoring data of elderly lung cancer patients, comprising the following steps:

[0006] Real-time monitoring of multiple physiological parameter data of elderly lung cancer patients;

[0007] Acquire all the physiological parameter data monitored above;

[0008] Based on all the physiological parameter data, combined with their multi-dimensional data correlation and time development attenuation, processing and analysis are performed to obtain the current physical condition assessment information of the elderly lung cancer patients.

[0009] Preferably, the physiological parameter data includes body temperature data, respiratory rate data, heart rate data, blood oxygen saturation data SpO2, and weight data.

[0010] Preferably, the processing and analysis based on all the physiological parameter data combined with their multi-dimensional data correlation and time development decay to obtain the current physical condition assessment information of the elderly lung cancer patient includes the following steps:

[0011] Preprocessing the physiological parameter data to obtain target physiological parameter data;

[0012] Based on the target physiological parameter data, combined with its multidimensional data correlation and time development attenuation, processing and analysis are performed, and the current physical condition assessment information of the elderly lung cancer patients is obtained through fuzzy comprehensive evaluation processing.

[0013] Preferably, the target physiological parameter data is processed and analyzed in combination with its multidimensional data correlation and time development decay, and the current physical condition assessment information of the elderly lung cancer patient is obtained through fuzzy comprehensive evaluation, including the following steps:

[0014] Establishing a body state assessment information matrix based on the target physiological parameter data;

[0015] Based on each of the target physiological parameter data, a correlation factor corresponding to each target physiological parameter data is obtained through correlation analysis; the correlation factor represents the correlation relationship value between every two target physiological parameter data;

[0016] Calculate the physiological parameter weight w based on the correlation factor;

[0017] Obtain a fuzzy comprehensive evaluation value Q based on the body state assessment information matrix R and the physiological parameter weights w corresponding to each of the physiological parameter data;

[0018] Based on the fuzzy comprehensive evaluation value, analysis and judgment are performed to obtain the current physical condition assessment information of the elderly lung cancer patient.

[0019] Preferably, obtaining the correlation factor corresponding to each target physiological parameter data through correlation analysis based on each target physiological parameter data specifically includes the following operation steps:

[0020] Establish a correlation matrix F based on the target physiological parameter data corresponding to each moment j ;

[0021] Calculate the mean value of each target physiological parameter data collected at all times And the standard deviation Sr of the target physiological parameter data j ;

[0022] Based on the mean of the target physiological parameter data And the standard deviation Sr of the target physiological parameter data j The first correlation factor k is calculated by calculating each sub-parameter in the correlation matrix F. j ;

[0023] Based on the first correlation factor k j The correlation factor k is calculated.

[0024] Preferably, obtaining the correlation factor corresponding to each target physiological parameter data through correlation analysis based on each target physiological parameter data specifically includes the following operation steps:

[0025] Initialize iteration parameters;

[0026] The iteration parameters include a target physiological parameter data set P, a time series set J, an initialization adaptive time decay matrix D, an iteration counter, a maximum number of iterations Cmax, and a convergence threshold ε;

[0027] Constructing a time correlation matrix R(tj) for each time point j based on the target physiological parameter data set and the time series set;

[0028] Add 1 to the iteration number of the iteration counter to obtain the current iteration number; calculate and obtain the parameter importance vector Ia based on each element in the time correlation matrix R(tj); update the adaptive time decay matrix D based on the parameter importance vector Ia to obtain an updated adaptive time decay matrix D';

[0029] Constructing a multidimensional correlation factor tensor K based on the updated adaptive time decay matrix D'; calculating a comprehensive correlation factor matrix L based on the multidimensional correlation factor tensor K;

[0030] After performing singular value decomposition on the comprehensive correlation factor matrix L, the left singular vectors corresponding to the first three largest singular values ​​are selected to form a characteristic matrix T;

[0031] Clustering the feature matrix T to obtain multiple clusters; screening each of the clusters based on a preset discreteness threshold and an inter-cluster distance threshold to obtain abnormal clusters and normal clusters;

[0032] Calculate and obtain the average correlation factor in each of the normal clusters, and based on the average correlation factor, modify the abnormal clusters through principal component analysis to obtain modified cluster clusters; based on the modified cluster clusters, update the comprehensive correlation factor matrix L to obtain an updated comprehensive correlation factor matrix L';

[0033] Calculate and obtain the norm difference of the comprehensive correlation factor matrix L and the updated comprehensive correlation factor matrix L'; determine whether the norm difference is greater than or equal to the convergence threshold ε; if so, calculate and obtain the eigenvalue Λ and the eigenvector V based on the comprehensive correlation factor matrix L; construct the correlation factor vector U based on the eigenvalue Λ and the eigenvector V; if not, further determine whether the current number of iterations is greater than or equal to the maximum number of iterations Cmax; if so, calculate and obtain the eigenvalue Λ and the eigenvector V based on the comprehensive correlation factor matrix L; construct the correlation factor vector U based on the eigenvalue Λ and the eigenvector V; if not, return to the above operation until the norm difference is greater than or equal to the convergence threshold ε or the current number of iterations is greater than or equal to the maximum number of iterations Cmax, and obtain the correlation factor vector U.

[0034] Preferably, the adaptive time decay matrix D is updated based on the parameter importance vector Ia, and the updated adaptive time decay matrix D' is obtained by updating the decay coefficient dab of the adaptive time decay matrix D to obtain a new decay coefficient dab new Then, the parameter importance vector of the target physiological parameter data a and the parameter importance vector of the target physiological parameter data b are obtained to update the new attenuation coefficient dab new , and finally according to the new attenuation coefficient dab new Get the updated adaptive time decay matrix D';

[0035] The updated adaptive time decay matrix D'=[dab new ]n×n;

[0036] Among them, dab new represents the new attenuation coefficient of the target physiological parameter data pa to the target physiological parameter data pb; n represents the number of target physiological parameter data.

[0037] The new attenuation coefficient dab new The update calculation method is:

[0038] dab new =dab*(1-α*Ia*Ib);

[0039] Wherein, α is the learning rate (learning rate controls the update amplitude); Ia is the parameter importance vector of the target physiological parameter data a; lb is the parameter importance vector of the target physiological parameter data b;

[0040] Compared with the prior art, the embodiments of the present invention have at least the following technical advantages:

[0041] By analyzing the processing method of the home care monitoring data of the elderly lung cancer patients provided by the present invention, it can be known that in specific application, the patient's health status is first grasped in real time through the monitoring of physiological parameter data, and abnormal conditions, such as increased body temperature, decreased blood oxygen level, etc., are discovered in time; then, various physiological parameters are integrated through multidimensional data analysis to provide a comprehensive physical condition assessment information assessment, which helps to discover health problems that may not be revealed by a single parameter. For example, through weight data analysis, the patient's nutritional status can be better understood; further, through the above-mentioned data monitoring and analysis mechanism, nursing staff can make more accurate nursing decisions based on the patient's specific data, adjust nursing measures and treatment plans, and improve the patient's treatment effect. BRIEF DESCRIPTION OF THE DRAWINGS

[0042] Figure 1 This is a schematic diagram of the overall operation steps of a method for processing home care monitoring data of elderly lung cancer patients in Example 1;

[0043] Figure 2 This is an example data diagram of correlation factors of target physiological parameter data of a method for processing home care monitoring data of elderly lung cancer patients in embodiment 1;

[0044] Figure 3 This is an example data diagram of collecting target physiological parameter data of elderly lung cancer patients in accordance with a method for processing home care monitoring data of elderly lung cancer patients in Example 2. DETAILED DESCRIPTION

[0045] The technical solution of the present invention will be described clearly and completely below in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0046] The present invention is further described in detail below through specific embodiments in conjunction with the accompanying drawings.

[0047] Embodiment 1

[0048] like Figure 1 As shown, the first embodiment of the present invention provides a method for processing home care monitoring data of elderly lung cancer patients, including the following steps:

[0049] Step S10: real-time monitoring of multiple physiological parameter data of the current elderly lung cancer patient;

[0050] The physiological parameter data include body temperature data, respiratory rate data, heart rate data, blood oxygen saturation data SpO2, and weight data;

[0051] It should be noted that the above-mentioned embodiment of the present application monitors the physiological parameter data of elderly lung cancer patients through monitoring equipment. These physiological parameter data comprehensively reflect the current physical condition assessment information of elderly lung cancer patients. Through these physiological parameter data obtained by monitoring, the physical condition assessment information of elderly lung cancer patients can be intuitively viewed; the presence of fever or abnormal body temperature is detected through body temperature data, which is common in infection or inflammation; the patient's breathing condition is viewed through respiratory rate data, and abnormal respiratory rate may indicate respiratory system problems; the heart health status is displayed through heart rate data, and abnormal heart rate may indicate heart problems or overall health problems; the oxygen level in the blood is measured through blood oxygen saturation data SpO2, and low oxygen saturation may indicate an inadequate respiratory system or other health problems; monitoring weight changes through weight data helps to assess the patient's nutritional status and disease progression; real-time monitoring of these physiological parameters can fully understand the patient's physical condition assessment information and provide intuitive health data; these data provide caregivers or medical teams with basic information for assessing the patient's health and help to identify potential health problems in a timely manner.

[0052] Step S20: Acquire all the physiological parameter data monitored above;

[0053] Step S30: Based on all the physiological parameter data, combined with their multi-dimensional data correlation and time development decay, processing and analysis are performed to obtain the physical condition assessment information of the current elderly lung cancer patient.

[0054] It should be noted that the above-mentioned embodiment of the present application comprehensively considers all monitored physiological parameters through multi-dimensional data analysis, and evaluates the patient's overall physical condition assessment information by analyzing the relationship between these parameters; integrates data such as body temperature, respiratory rate, heart rate, blood oxygen saturation, weight change, etc. for comprehensive assessment. By comprehensively analyzing physiological parameters, it is possible to fully understand the patient's health status, identify potential problems, and provide a well-founded health assessment to help formulate or adjust personalized care and treatment plans;

[0055] The above-mentioned embodiments of the present application first monitor the physiological parameter data to grasp the patient's health status in real time and promptly detect abnormal conditions, such as increased body temperature, decreased blood oxygen level, etc.; then, various physiological parameters are integrated through multi-dimensional data analysis to provide a comprehensive physical condition assessment information assessment, which helps to discover health problems that may not be revealed by individual parameters. For example, through weight data analysis, the patient's nutritional status can be better understood; further, through the above-mentioned data monitoring and analysis mechanism, nursing staff can make more accurate nursing decisions based on the patient's specific data, adjust nursing measures and treatment plans, and improve the patient's treatment effect.

[0056] Specifically, in step S30, based on all the physiological parameter data, combined with their multi-dimensional data correlation and time development attenuation, processing and analysis are performed to obtain the current physical state assessment information of the elderly lung cancer patient, including the following operation steps:

[0057] Step S31: preprocessing the physiological parameter data to obtain target physiological parameter data;

[0058] The above preprocessing includes outlier detection, data interpolation, and data scoring in sequence;

[0059] It should be noted that the above-mentioned embodiment of the present application first detects the outliers in each physiological parameter data through outlier detection (standard deviation method, isolation forest, support vector machine, time series analysis, etc. can be used. The above-mentioned outlier detection methods are common knowledge of technicians in this field and will not be repeated in this application). Then, the outliers are interpolated by data interpolation, and then the physiological parameter data after interpolation is scored (data statistics, weighted scoring, decision tree, random forest or support vector machine can be used. The above-mentioned data scoring methods are common knowledge of technicians in this field and will not be repeated in this application), so as to obtain unified scoring data that is convenient for subsequent multi-dimensional data processing and improve the efficiency of subsequent data analysis and processing.

[0060] Step S32: Based on the target physiological parameter data, combined with its multi-dimensional data correlation and time development attenuation, processing and analysis are performed to obtain the current physical condition assessment information of the elderly lung cancer patient through fuzzy comprehensive evaluation processing.

[0061] It should be noted that the above-mentioned embodiment of the present application performs a comprehensive evaluation based on the processed target physiological parameter data to obtain the physical status evaluation information of the elderly lung cancer patients, which can process complex and ambiguous data and provide a more comprehensive health status evaluation;

[0062] The above-mentioned embodiment of the present application first identifies and removes abnormal data through outlier detection to ensure the accuracy and reliability of the data; then fills the outliers through data interpolation to restore the continuity and integrity of the data, making the data more suitable for analysis. Further, the processed data is scored through data scoring and converted into a unified standardized data form for further processing and analysis, effectively improving the quality and consistency of the data, and ensuring the accuracy and efficiency of subsequent analysis; then, based on the processed target physiological parameter data, a comprehensive evaluation is performed to obtain the physical status assessment information of elderly lung cancer patients, which can process complex and fuzzy data and provide a more comprehensive health status evaluation; through the system preprocessing steps (outlier detection, data interpolation, data scoring), the accuracy and consistency of the data can be ensured, providing a reliable basis for multi-dimensional analysis; the fuzzy comprehensive evaluation method is used to convert the processed data into a comprehensive evaluation of the patient's physical status assessment information, which can effectively improve the accuracy and practicality of the health status analysis.

[0063] Specifically, in step S32, based on the target physiological parameter data, combined with its multidimensional data correlation and time development attenuation, processing and analysis are performed, and the physical state assessment information of the current elderly lung cancer patient is obtained through fuzzy comprehensive evaluation processing, including the following operation steps:

[0064] Step S321: establishing a body state assessment information matrix based on the target physiological parameter data;

[0065] The physical state assessment information matrix is:

[0066]

[0067] In the formula, r i,j is the i-th target physiological parameter data at time j;

[0068] It should be noted that the above-mentioned embodiment of the present application integrates the target physiological parameter data into an evaluation matrix as basic data for further analysis.

[0069] Step S322: obtaining a correlation factor corresponding to each target physiological parameter data through correlation analysis based on each target physiological parameter data; the correlation factor represents the correlation relationship value between every two target physiological parameter data;

[0070] It should be noted that in the process of nursing monitoring of lung cancer patients, because various physiological parameter data may affect each other, for example, difficulty breathing will lead to high respiratory rate, and low blood oxygen saturation will lead to high frequency, etc. Therefore, in the process of analyzing the physical state assessment information, the correlation between various physiological parameter data (that is, the above-mentioned correlation factors) must also be comprehensively considered to obtain physical state assessment information that is more in line with the patient's physical condition.

[0071] Step S323: Calculating based on the correlation factor to obtain the physiological parameter weight w;

[0072] It should be noted that, for example, through the operation of step S222, the correlation factor corresponding to each target physiological parameter data is obtained, such as Figure 2 As shown;

[0073] exist Figure 2 In , each value represents the correlation between indicators. For example, if the correlation between body temperature and respiratory rate is 0.6, it means that the influence of respiratory rate on body temperature (i.e., correlation factor) is 0.6. Further, the weight of each physiological parameter is obtained by calculation, which is calculated by using the average value of the sum of the correlation factors of each indicator, for example:

[0074]

[0075] Step S324: Calculate and obtain a fuzzy comprehensive evaluation value Q based on the body state assessment information matrix R and the physiological parameter weights w corresponding to each of the physiological parameter data;

[0076] The calculation method of the fuzzy comprehensive evaluation value Q is:

[0077] Q = R × w;

[0078] The above physiological parameter weights are obtained through principal component analysis, which extracts the main components of each physiological parameter data through dimensionality reduction technology, and then determines the contribution of the main components; the above principal component analysis is a prior art and will not be described in detail in this application;

[0079] The fuzzy comprehensive evaluation value in the above-mentioned embodiment of the present application is obtained by multiplying the elements in the physical state assessment information matrix with their corresponding physiological parameter weights and then taking the average value. The fuzzy comprehensive evaluation value reflects the evaluation score of the current physical state assessment information of the elderly lung cancer patient. The physical state assessment information of the current elderly lung cancer patient can be determined through the physical state assessment information evaluation score.

[0080] Step S325: Analyze and judge based on the fuzzy comprehensive evaluation value to obtain the physical condition assessment information of the current elderly lung cancer patient.

[0081] It should be noted that the above-mentioned embodiment of the present application determines the physical condition assessment information of the current elderly lung cancer patient by judging the size of the fuzzy comprehensive evaluation value; specifically: by presetting the physical condition assessment threshold, when the fuzzy comprehensive evaluation value is within the physical condition assessment threshold range, the corresponding physical condition assessment information of the elderly lung cancer patient is obtained.

[0082] The above-mentioned embodiment of the present application first integrates data by establishing an evaluation matrix, then simplifies the data and determines the weights through principal component analysis, and then combines historical data to determine the benchmark of physical condition evaluation information, and finally makes a comprehensive judgment; by combining fuzzy comprehensive evaluation values ​​and principal component analysis, it can more accurately handle complex data and uncertainties, making the evaluation of patient physical condition evaluation information more comprehensive and reliable; by using data analysis methods (such as principal component analysis) and historical data for evaluation, the evaluation process is supported by data, reducing the influence of subjective judgment, improving the scientificity and objectivity of the evaluation, avoiding the subjective influence that needs to be determined manually in the process of determining fuzzy comprehensive evaluation indicators in traditional fuzzy comprehensive evaluation methods, and enhancing the accuracy and objectivity of the physical condition evaluation information of elderly lung cancer patients finally determined.

[0083] During the specific implementation of the above-mentioned embodiment of the present application, the technical personnel discovered that because various physiological parameter data may influence each other, for example, difficulty in breathing will lead to high respiratory rate, and low blood oxygen saturation will lead to high mutual frequency, etc. Therefore, in the process of analyzing the physical state assessment information, it is also necessary to comprehensively consider the correlation between various physiological parameter data (that is, the above-mentioned correlation factors), so as to obtain physical state assessment information that is more in line with the patient's physical condition; in the specific operation process, the technical personnel assume that there is a mutual influence relationship between every two target physiological parameter data, and then use the mutual influence relationship to analyze and obtain the correlation factor corresponding to each target physiological parameter data. The specific operation is shown in the following steps.

[0084] Specifically, in step S323, obtaining the correlation factor corresponding to each target physiological parameter data through correlation analysis based on each target physiological parameter data includes the following operation steps:

[0085] Step S3231: Establish a correlation matrix F based on the target physiological parameter data corresponding to each moment j ;

[0086] The incidence matrix F j It is expressed as:

[0087]

[0088] In the formula, (rm j ,rm j) represents the association relationship between the mth target physiological parameter data and the mth physiological parameter data at time j;

[0089] For example: See Figure 2 In the figure, the data corresponding to the lower right corner, i.e., the value "1" corresponding to the 6th row of the 6th column, i.e., (dyspnea, dyspnea) = 1, indicates the correlation between the dyspnea data and the physiological parameter data of dyspnea at a certain moment;

[0090] See also Figure 2 For example, a certain value, that is, the value "0.6" corresponding to the 6th column and the 5th row, can be understood as (pain level, dyspnea) = 0.6, which represents the correlation between the pain level data and the physiological parameter data of dyspnea at a certain moment, and no further details are given.

[0091] The above embodiment of the present application uses the basic setting that there is a mutual influence relationship between every two target physiological parameter data to establish the correlation matrix between every two physiological parameter data at each moment, such as (r1 j ,r2 j ) is the influence relationship (or association relationship) of the first target physiological parameter data on the second target physiological parameter data at time j;

[0092] For example: A technician monitors the physiological parameters of a group of patients according to the technical solution adopted in the embodiment of the present application. Assume that the data collected at the first moment (j=1) are as follows:

[0093] HR = 75 bpm;

[0094] BP = 120 / 80 mmHg (we used systolic 120 as a representative value);

[0095] T = 37.0 °C;

[0096] We can create the following 3x3 incidence matrix

[0097]

[0098] Among them, (HR1, BP1) represents the influence of heart rate on blood pressure at the first moment.

[0099] Step S3232: Calculate the mean of each target physiological parameter data collected at all times And the standard deviation Sr of the target physiological parameter data j ;

[0100] Continuing with the above example, suppose we collect the following data at 4 consecutive moments:

[0101] Time 1: HR=75, BP=120, T=37.0;

[0102] Time 2: HR=78, BP=122, T=37.1;

[0103] Time 3: HR=73, BP=118, T=36.9;

[0104] Time 4: HR=76, BP=121, T=37.0;

[0105] Calculate the mean:

[0106] Average HR = (75 + 78 + 73 + 76) / 4 = 75.5 bpm;

[0107] Average BP value = (120 + 122 + 118 + 121) / 4 = 120.25 mmHg;

[0108] T average value = (37.0 + 37.1 + 36.9 + 37.0) / 4 = 37.0 ° C;

[0109] Calculate the standard deviation:

[0110] HR standard deviation ≈2.08 bpm;

[0111] BP standard deviation ≈1.71 mmHg;

[0112] T standard deviation ≈ 0.08°C;

[0113] Step S3233: Based on the mean of the target physiological parameter data And the standard deviation Sr of the target physiological parameter data j The first correlation factor k is calculated by calculating each sub-parameter in the correlation matrix F. j ;

[0114] The first correlation factor k j The calculation method is:

[0115]

[0116] Continuing with the above example: Taking heart rate (HR) as an example, calculate the first correlation factor k1 at the first moment

[0117] K1=|HR1-HR mean|×HR standard deviation

[0118] =|75-75.5|×2.08

[0119] ≈1.04;

[0120] The first correlation factor between blood pressure and body temperature was calculated in the same way;

[0121] Step S3234: Based on the first correlation factor k j The correlation factor k is calculated;

[0122] The correlation factor k is calculated as follows:

[0123] In the formula, k j represents the first correlation factor, j represents the number of moments or j moments;

[0124] It should be noted that, continuing with the above example, assuming that we have calculated all the first correlation factors at four moments, we now calculate the final correlation factor k:

[0125] k=(√k1+√k2+√k3+√k4) / 4;

[0126] Take heart rate as an example:

[0127] Assumptions

[0128] Then the final correlation factor of heart rate k_HR=(1.02+1.14+0.91+1.06) / 4≈1.03;

[0129] The final correlation factor between blood pressure and body temperature was calculated in the same way;

[0130] The above-mentioned embodiment of the present application comprehensively considers the influence relationship between every two target physiological parameter data, establishes an association matrix for every two target physiological parameter data at each moment, and each element in the matrix represents the degree of mutual influence between the two parameters at a specific moment, providing a basic data structure for subsequent analysis; further calculates the mean and standard deviation of each target physiological parameter data at the current moment, understands the basic statistical characteristics of the parameter data, and helps evaluate the central trend and fluctuation range of the data; further combines the mean, standard deviation and association matrix to calculate the first association factor, which can reflect the correlation strength between the target physiological parameter data at the j moment; finally, the final association factor is obtained by taking the mean of the sum of the square roots of the first association factors at the j moments; the above-mentioned embodiment of the present application considers the mutual influence between the target physiological parameter data through a systematic method, so that the physical state assessment is not only based on the absolute value of a single parameter, but comprehensively considers the correlation between the parameters; thereby more accurately reflecting the actual physical condition of the patient, especially when processing complex physiological data, can improve the accuracy and reliability of the assessment results.

[0131] Embodiment 2

[0132] Embodiment 2 of the present invention provides a method for processing home care monitoring data of elderly lung cancer patients, including the following steps:

[0133] Step S100: real-time monitoring of multiple physiological parameter data of the current elderly lung cancer patient;

[0134] The physiological parameter data include body temperature data, respiratory rate data, heart rate data, blood oxygen saturation data SpO2, and weight data;

[0135] Step S200: Acquire all the physiological parameter data monitored above;

[0136] Step S300: Based on all the physiological parameter data, combined with their multi-dimensional data correlation and time development decay, processing and analysis are performed to obtain the physical condition assessment information of the current elderly lung cancer patient.

[0137] Specifically, in step S300, based on all the physiological parameter data, combined with their multi-dimensional data correlation and time development attenuation, processing and analysis are performed to obtain the current physical state assessment information of the elderly lung cancer patient, including the following operation steps:

[0138] Step S301: preprocessing the physiological parameter data to obtain target physiological parameter data;

[0139] The above preprocessing includes outlier detection, data interpolation, and data scoring in sequence;

[0140] Step S302: Based on the target physiological parameter data, combined with its multi-dimensional data correlation and time development attenuation, processing and analysis are performed to obtain the current physical condition assessment information of the elderly lung cancer patient through fuzzy comprehensive evaluation processing.

[0141] Specifically, in step S302, based on the target physiological parameter data, combined with its multidimensional data correlation and time development attenuation, processing and analysis are performed, and the physical state assessment information of the current elderly lung cancer patient is obtained through fuzzy comprehensive evaluation processing, including the following operation steps:

[0142] Step S3021: establishing a body state assessment information matrix based on the target physiological parameter data;

[0143] The physical state assessment information matrix is:

[0144]

[0145] In the formula, r i,j is the i-th target physiological parameter data at time j;

[0146] Step S3022: obtaining a correlation factor corresponding to each target physiological parameter data through correlation analysis based on each target physiological parameter data; the correlation factor represents the correlation relationship value between every two target physiological parameter data;

[0147] Step S3023: Calculating based on the correlation factor to obtain a physiological parameter weight w;

[0148] Step S3024: Calculate and obtain a fuzzy comprehensive evaluation value Q based on the body state assessment information matrix R and the physiological parameter weights w corresponding to each of the physiological parameter data;

[0149] The calculation method of the fuzzy comprehensive evaluation value Q is:

[0150] Q = R × w;

[0151] Step S3025: Analyze and judge based on the fuzzy comprehensive evaluation value to obtain the physical condition assessment information of the current elderly lung cancer patient.

[0152] During the specific implementation of the above-mentioned embodiment of the present application, the technicians also found that in the process of obtaining each target physiological parameter data, due to the influence of time factors, the data of each target physiological parameter is not accurate enough. For example, the heart rate measurement is directly measured by the sensor, and the patient's blood pressure is affected by the long measurement time period, resulting in the final data not being at the same time point. For example, the patient's heart rate is 75bpm measured at time j, and the blood pressure measurement result is 120mmHg; however, due to the long blood pressure measurement time period, in fact, the patient's heart rate is 75bpm in the data obtained at time j, and the blood pressure measurement result of 120mmHg in the data obtained at time j is measured at time j-1, and the actual blood pressure measurement result obtained by monitoring at time j will be displayed at time j+1 due to the blood pressure measurement time period, which will cause the result to have a time deviation, thereby resulting in inaccurate analysis of the patient's physical state assessment information; therefore, in the subsequent analysis of the correlation factors corresponding to each target physiological parameter data, it is also necessary to add the time attenuation factor of each target physiological parameter data, so that the patient's physical state assessment information obtained by the final analysis is more accurate.

[0153] Specifically, in step S3023, obtaining the correlation factor corresponding to each target physiological parameter data through correlation analysis based on each target physiological parameter data includes the following operation steps:

[0154] Step S30231: Initialize iteration parameters;

[0155] The iteration parameters include a target physiological parameter data set P, a time series set J, an initialization adaptive time decay matrix D, an iteration counter, a maximum number of iterations Cmax, and a convergence threshold ε;

[0156] Target physiological parameter data set P = {p1, p2, p3, ..., pn}, where pn is the nth target physiological parameter data; time series set J = {j1, j2, j3, ..., jm}; where jm is the mth time point (i.e., the mth time point j at which the target physiological parameter data is collected); initialize the adaptive time decay matrix D = [dab]n×n; where dab represents the decay coefficient of the target physiological parameter data pa to the target physiological parameter data pb; n represents the number of target physiological parameter data, especially the first n in the initialized adaptive time decay matrix D represents the number of rows in the matrix D and the second n represents the number of columns in the matrix D; iteration counter, the number of iterations of the iteration counter is initially 0; Cmax is recorded as the maximum number of iterations and ε represents the convergence threshold;

[0157] For example: According to the technical solution adopted in the above-mentioned embodiment of the present application, the technicians monitor and obtain the target physiological parameter data of elderly lung cancer patients: heart rate (HR), blood pressure (BP), respiratory rate (RR), blood oxygen saturation (SpO2), and body temperature (T); initialize the iteration parameters according to each target physiological parameter data and the monitoring time point, including the target physiological parameter data set P = {HR, BP, RR, SpO2, T}, the time series set J = {j1, j2, j3, j4, j5, j6}; initialize the adaptive time decay matrix D as a 5×5 unit matrix; set the maximum number of iterations Cmax to 10 and the convergence threshold ε to 0.001;

[0158] Step S30232: constructing a time correlation matrix R(tj) for each time point j based on the target physiological parameter data set and the time series set;

[0159] The time correlation matrix R(tj) is expressed as:

[0160] R(j) = [rab(j)]n×n;

[0161] Wherein, rab(j) is a single element in the time correlation matrix, which indicates the correlation strength between the target physiological parameter data a and the target physiological parameter data b at time point j; n is the number of target physiological parameter data, and n×n indicates that the time correlation matrix R(tj) is an n×n matrix;

[0162] Among them, the calculation method of a single element rab(j) of the time correlation matrix is:

[0163] rab(j)=w(j)*corr(pa(j),pb(j))*(1+|Δpa(j) / pa(j)|)*(1+|Δpb(j) / pb(j)|);

[0164] Wherein, w(j) is the time weight; corr is the time correlation coefficient, corr(pa(j),pb(j) represents the time correlation coefficient between the target physiological parameter data a and the target physiological parameter data b at time point j; Δpa represents the parameter change rate of the target physiological parameter data a; Δpb represents the parameter change rate of the target physiological parameter data b; Δpa(j) represents the parameter change rate of the target physiological parameter data a at time point j; Δpb(j) represents the parameter change rate of the target physiological parameter data b at time point j; pa(j) represents the value of the target physiological parameter data a at time point j (i.e., real-time parameter); pb(j) represents the value of the target physiological parameter data b at time point j (i.e., real-time parameter);

[0165] It should be noted that in the above-mentioned embodiment of the present application, the time weight w(j) is used to give a higher weight to the most recent time point; the temporal correlation coefficient corr(pa(j), pb(j) of the target physiological parameter data a and the target physiological parameter data b at time point j is used to measure the linear relationship strength between the two parameters; further, by considering the parameter change rate of the target physiological parameter data a and the target physiological parameter data b, the influence of the parameter with a larger change is amplified; the above-mentioned embodiment of the present application introduces the time weight through w(j), so that the model can pay more attention to the most recent data, and uses the temporal correlation coefficient corr(pa(j), pb(j) to capture the basic correlation between the parameters; by introducing the parameter change rate term, the model can be more sensitive to the dynamic changes of the parameters, and when the parameters change significantly, their influence will be amplified. At the same time, the formula takes into account the changes of both the target physiological parameter data a and the target physiological parameter data b to ensure that the two-way influence is captured; finally, the relative change rate is used to make parameters of different magnitudes comparable;

[0166] Continuing with the above example, in the above embodiment of the present application, target physiological parameter data of elderly lung cancer patients is collected as follows: Figure 3 As shown;

[0167] Furthermore, the time correlation matrix R(tj) of each time point j is calculated as follows:

[0168] R(t6)=

[0169] [1.000 0.985 0.975 -0.950 0.990]

[0170] [0.985 1.000 0.980 -0.960 0.995]

[0171] [0.975 0.980 1.000 -0.970 0.985]

[0172] [-0.950-0.960-0.970 1.000-0.955]

[0173] [0.990 0.995 0.985 -0.955 1.000].

[0174] Step S30233: adding 1 to the iteration number of the iteration counter to obtain the current iteration number; obtaining the parameter importance vector Ia based on each element in the time correlation matrix R(tj); updating the adaptive time decay matrix D based on the parameter importance vector Ia to obtain the updated adaptive time decay matrix D';

[0175] The parameter importance vector Ia is calculated as follows:

[0176] Ia=ΣfΣb|rab(j)| / (f*n);

[0177] In the formula, f represents the number of time points; rab(j) represents the correlation strength between the target physiological parameter data a and the target physiological parameter data b at time point j; n is the number of target physiological parameter data;

[0178] The adaptive time decay matrix D is updated based on the parameter importance vector Ia to obtain the updated adaptive time decay matrix D' by updating the decay coefficient dab of the adaptive time decay matrix D to obtain a new decay coefficient dab new Then, the parameter importance vector of the target physiological parameter data a and the parameter importance vector of the target physiological parameter data b are obtained to update the new attenuation coefficient dab new , and finally according to the new attenuation coefficient dab new Get the updated adaptive time decay matrix D';

[0179] The updated adaptive time decay matrix D'=[dab new ]n×n;

[0180] Among them, dab new represents the new attenuation coefficient of the target physiological parameter data pa to the target physiological parameter data pb; n represents the number of target physiological parameter data. The new attenuation coefficient dab new The update calculation method is:

[0181] dab new =dab*(1-α*Ia*Ib);

[0182] Wherein, α is the learning rate (learning rate controls the update amplitude); Ia is the parameter importance vector of the target physiological parameter data a; lb is the parameter importance vector of the target physiological parameter data b;

[0183] It should be noted that the above embodiment of the present application firstly reflects the average correlation strength of a parameter with all other parameters at all time points by calculating the parameter importance vector Ia, thereby measuring its overall importance. Further, adaptability is introduced by (1-α*Ia*Ib). If parameters i and j are both important (Ia and Ib are larger), then the attenuation coefficient between them will decrease faster; conversely, if one or two parameters are not very important, the attenuation coefficient will decrease more slowly.

[0184] The above-mentioned embodiment of the present application can adapt to different decay rates between different parameter pairs by dynamically adjusting the decay matrix D; more important parameter pairs (i.e., those parameters that have a stronger correlation with other parameters) will receive more attention, and the relationship between them will be considered more persistent (decay slower); by introducing the learning rate α, the sensitivity of the model to new information can be controlled. A larger α will lead to faster adaptation, while a smaller α will make the model more stable; the product Ia*Ib introduces a nonlinear relationship, allowing the model to capture more complex interactions between parameters; since the (1-α*Ii*Ij) term is always less than 1, this ensures that the decay coefficient will gradually decrease over time, reflecting the natural tendency of information to decay over time;

[0185] After analysis, it can be seen that the adaptive time decay matrix D = [dab]n × n; so the updated adaptive time decay matrix D' = [dab new ]n×n;

[0186] Continuing with the above example, the technician obtains the parameter importance vector according to the technical solution adopted in the above embodiment of the present application as: I≈[0.22, 0.23, 0.21, 0.20, 0.22]; further, the adaptive time attenuation matrix D' (assuming α=0.1) is updated as:

[0187] D≈ [1.000 0.995 0.995 0.996 0.995] [0.995 1.000 0.995 0.995 0.995] [0.995 0.995 1.000 0.996 0.995] [0.996 0.995 0.996 1.000 0.996]

[0192] [0.995 0.995 0.995 0.996 1.000].

[0193] Step S30234: constructing a multidimensional correlation factor tensor K based on the updated adaptive time decay matrix D'; and calculating a comprehensive correlation factor matrix L based on the multidimensional correlation factor tensor K;

[0194] The multidimensional correlation factor tensor K is expressed as:

[0195] K = [kabj]n × n × f;

[0196] Wherein, kabj represents the correlation factor between the target physiological parameter data a and the target physiological parameter data b at time point j; n is the number of target physiological parameter data; f is the number of time points; n×n×f represents that the multidimensional correlation factor tensor K is a three-dimensional tensor with a dimension of n×n×f;

[0197] Where kabj = rab(j)*exp(-dab new *(jz-js));

[0198] Where rab(j) is the correlation strength between target physiological parameter data a and target physiological parameter data b at time point j; dab new is the new attenuation coefficient; jz is the new time point; js is the previous time point; exp(-dab new *(jz-js)) is the time decay term (indicating that it decreases with the increase of the time interval (tm-tk));

[0199] The above-mentioned embodiments of the present application not only consider the correlation between parameters, but also consider the decay of this correlation over time;

[0200] The comprehensive correlation factor matrix L is expressed as:

[0201] L = [lab] n × n;

[0202] Wherein, lab represents the comprehensive correlation factor between target physiological parameter data a and target physiological parameter data b; n×n represents that the comprehensive correlation factor matrix L is a two-dimensional matrix of n×n; n is the number of target physiological parameter data;

[0203] Among them, the calculation method of the comprehensive correlation factor lab is:

[0204] lab=(Σfkabj / Σfexp(-2dab new *(jz-js))) 0.5 ;

[0205] In the formula, the numerator Σfkabj represents the square of the correlation factor accumulated at all time points; the denominator Σfexp(-2dab new *(jz-js)) is a normalization term to ensure that the decays at different time scales can be compared;

[0206] The above embodiment of the present application obtains a comprehensive correlation measure by weighted summing of the correlation factors of all time points. Nonlinearity is introduced by square and square root operations, which can better capture complex correlation patterns. The time decay exp(-dij*(tm-tk)) term is further used to ensure that the contribution of earlier time points decreases over time, but the rate of decrease is controlled by dij, which is adaptive. The denominator ensures that the correlation factors between different parameter pairs can be fairly compared, even if they have different decay rates. By considering all time points, the embodiment of the present application can capture correlation patterns on different time scales.

[0207] The above-mentioned embodiments of the present application integrate static association and dynamic time decay, and can capture complex and possibly nonlinear association patterns, while taking into account the temporal changes in parameter importance, and provide an interpretable and comprehensive association metric.

[0208] Step S30235: After performing singular value decomposition on the comprehensive correlation factor matrix L, select the left singular vectors corresponding to the first three largest singular values ​​to form a characteristic matrix T;

[0209] It should be noted that the above embodiment of the present application is a dimensionality reduction operation, which makes the subsequent analysis and calculation easier; it is to decompose the comprehensive correlation factor matrix L by using singular value decomposition (SVD): C = UΣV T ; Select the left singular vectors corresponding to the first r largest singular values ​​to form a characteristic matrix T; wherein the singular value decomposition operation is a prior art and will not be described in detail in this application;

[0210] The above feature matrix T is a single, reduced-dimensional matrix that contains the main features of the original data.

[0211] Step S30236: clustering the feature matrix T to obtain a plurality of clusters; screening each of the clusters based on a preset discreteness threshold and an inter-cluster distance threshold to obtain abnormal clusters and normal clusters;

[0212] It should be noted that the above embodiment of the present application clusters the main features of the original data in the feature matrix T by clustering to obtain multiple cluster clusters, and then screens each cluster cluster by a preset discreteness threshold and an inter-cluster distance threshold. The cluster clusters whose discreteness is greater than the discreteness threshold or whose inter-cluster distance is greater than the inter-cluster distance threshold are identified as abnormal clusters, and the other cluster clusters are identified as normal clusters.

[0213] Step S30237: Calculate and obtain the average correlation factor in each of the normal clusters, and based on the average correlation factor, correct the abnormal clusters through principal component analysis to obtain a corrected cluster cluster; based on the corrected cluster cluster, update the comprehensive correlation factor matrix L to obtain an updated comprehensive correlation factor matrix L';

[0214] The average correlation factor is calculated as follows:

[0215] Cavg x =Σf∈cluster x Σk lab / |cluster i| 2 ;

[0216] Where Cavg x is the average correlation factor of the xth normal cluster; lab is a single element of the comprehensive correlation factor matrix L, cluster x represents the x-th cluster; |cluster x| represents the number of elements in the x-th cluster;

[0217] The updated comprehensive correlation factor matrix L' is expressed as:

[0218] L'=L*(1-β)+diag(Cavg1,…,Cavg k )*β;

[0219] Where β is the smoothing factor; diag(Cavg1,…,Cavg k ) represents a diagonal matrix with the average correlation factor as the diagonal element;

[0220] Step S30238: Calculate and obtain the norm difference between the comprehensive correlation factor matrix L and the updated comprehensive correlation factor matrix L'; determine whether the norm difference is greater than or equal to the convergence threshold ε; if so, calculate and obtain the eigenvalue Λ and the eigenvector V based on the comprehensive correlation factor matrix L; construct the correlation factor vector U based on the eigenvalue Λ and the eigenvector V; if not, further determine whether the current number of iterations is greater than or equal to the maximum number of iterations Cmax; if so, calculate and obtain the eigenvalue Λ and the eigenvector V based on the comprehensive correlation factor matrix L; construct the correlation factor vector U based on the eigenvalue Λ and the eigenvector V; if not, return to the above operation until the norm difference is greater than or equal to the convergence threshold ε or the current number of iterations is greater than or equal to the maximum number of iterations Cmax, and obtain the correlation factor vector U;

[0221] The eigenvalue Λ and the eigenvector V are calculated as follows:

[0222] Λ,V=eig(L');

[0223] The calculation method of the correlation factor vector U is:

[0224] U=V*diag(exp(Λ))*V j *1;

[0225] The above-mentioned embodiment of the present application realizes personalized time decay between different parameter pairs by dynamically updating the decay matrix D; uses a multidimensional correlation factor tensor K to capture the complex correlation relationship between parameters and considers the time dimension; extracts the main features through SVD to reduce the influence of noise; and continuously optimizes the correlation factor matrix through multiple iterations until convergence, so that the final correlation factor vector U comprehensively considers the mutual influence of all parameters;

[0226] The above-mentioned embodiments of the present application effectively solve the influence of time decay factors through complex matrix operations and iterative processes, and at the same time provide a comprehensive and adaptive correlation factor analysis framework that can process complex physiological parameter data, identify abnormal patterns, and give a comprehensive importance score for each parameter, providing an accurate data basis for subsequent analysis of patient physical status information.

[0227] In summary, the present invention proposes a method for processing home care monitoring data of elderly lung cancer patients. First, by monitoring the physiological parameter data, the health status of the patient can be grasped in real time, and abnormal conditions, such as increased body temperature and decreased blood oxygen level, can be discovered in time. Then, various physiological parameters are integrated through multidimensional data analysis to provide a comprehensive physical status assessment information assessment, which helps to discover health problems that may not be revealed by individual parameters. For example, through weight data analysis, the nutritional status of the patient can be better understood. Furthermore, through the above data monitoring and analysis mechanism, nursing staff can make more accurate nursing decisions based on the patient's specific data, adjust nursing measures and treatment plans, and improve the patient's treatment effect.

[0228] Specifically, in the process of multidimensional data analysis, firstly, outlier detection is used to identify and remove abnormal data to ensure the accuracy and reliability of the data; then, data interpolation is used to fill in outliers to restore the continuity and integrity of the data, making the data more suitable for analysis; further, the processed data is scored through data scoring and converted into a unified standardized data form for further processing and analysis, effectively improving the quality and consistency of the data and ensuring the accuracy and efficiency of subsequent analysis; then, based on the processed target physiological parameter data, a comprehensive evaluation is performed to obtain the physical status evaluation information of elderly lung cancer patients, which can process complex and fuzzy data and provide a more comprehensive health status evaluation; through the preprocessing steps of the system, the accuracy and consistency of the data can be ensured, providing a reliable basis for multidimensional analysis; the fuzzy comprehensive evaluation method is used to convert the processed data into a comprehensive evaluation of the patient's physical status evaluation information, which can effectively improve the accuracy and practicality of health status analysis;

[0229] Furthermore, the data is integrated by establishing an evaluation matrix, and then the data is simplified and the weight is determined by principal component analysis, and then the benchmark of the physical status evaluation information is determined in combination with historical data, and finally a comprehensive judgment is made; by combining the fuzzy comprehensive evaluation value and the principal component analysis, complex data and uncertainty can be processed more accurately, making the evaluation of the patient's physical status evaluation information more comprehensive and reliable; by using data analysis methods and historical data for evaluation, the evaluation process is supported by data, reducing the influence of subjective judgment, improving the scientificity and objectivity of the evaluation, avoiding the subjective influence that needs to be determined manually in the process of determining the fuzzy comprehensive evaluation indicators in the traditional fuzzy comprehensive evaluation method, and enhancing the accuracy and objectivity of the physical status evaluation information of the elderly lung cancer patients finally determined;

[0230] In the specific implementation process, the influence relationship between every two target physiological parameter data is comprehensively considered, and an association matrix is ​​established for every two target physiological parameter data at each moment. Each element in the matrix represents the degree of mutual influence between the two parameters at a specific moment, providing a basic data structure for subsequent analysis; the mean and standard deviation of each target physiological parameter data at the current moment are further calculated to understand the basic statistical characteristics of the parameter data and help evaluate the central trend and fluctuation range of the data; the first association factor calculated by further combining the mean, standard deviation and association matrix can reflect the strength of association between the target physiological parameter data at the j moment; finally, the final association factor is obtained by taking the mean of the sum of the square roots of the first association factors at the j moments; the above-mentioned embodiment of the present application considers the mutual influence between the target physiological parameter data through a systematic method, so that the physical state assessment is not only based on the absolute value of a single parameter, but also comprehensively considers the association between the parameters; thereby more accurately reflecting the actual physical condition of the patient, especially when processing complex physiological data, it can improve the accuracy and reliability of the assessment results;

[0231] On the other hand, the influence of time decay factors on target physiological parameter data is also taken into account. By dynamically updating the decay matrix D, personalized time decay between different parameter pairs is achieved. The multidimensional correlation factor tensor K is used to capture the complex correlation relationship between parameters and consider the time dimension. The main features are extracted through SVD to reduce the influence of noise. Through multiple iterations, the correlation factor matrix is ​​continuously optimized until convergence, so that the final correlation factor vector U comprehensively considers the mutual influence of all parameters.

[0232] Through complex matrix operations and iterative processes, the influence of time decay factors is effectively solved, and a comprehensive and adaptive correlation factor analysis framework is provided, which can process complex physiological parameter data, identify abnormal patterns, and give a comprehensive importance score for each parameter, providing an accurate data basis for subsequent analysis of patient physical status information.

[0233] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. A person skilled in the art may modify the technical solutions described in the above embodiments, or replace part or all of the technical features therein with equivalents. However, these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for processing home care monitoring data of elderly lung cancer patients, characterized in that: The steps are as follows: Real-time monitoring of multiple physiological parameter data of elderly lung cancer patients; Acquire all the physiological parameter data monitored above; Based on all the physiological parameter data, combined with their multi-dimensional data correlation and time development attenuation, processing and analysis are performed to obtain the current physical condition assessment information of the elderly lung cancer patients.

2. The method for processing home care monitoring data of elderly lung cancer patients according to claim 1 is characterized in that: The physiological parameter data include body temperature data, respiratory rate data, heart rate data, blood oxygen saturation data SpO2, and weight data.

3. The method for processing home care monitoring data of elderly lung cancer patients according to claim 2 is characterized in that: The method of processing and analyzing all the physiological parameter data in combination with the multi-dimensional data correlation and time development decay to obtain the current physical status assessment information of the elderly lung cancer patient includes the following steps: Preprocessing the physiological parameter data to obtain target physiological parameter data; Based on the target physiological parameter data, combined with its multidimensional data correlation and time development attenuation, processing and analysis are performed, and the current physical condition assessment information of the elderly lung cancer patients is obtained through fuzzy comprehensive evaluation processing.

4. The method for processing home care monitoring data of elderly lung cancer patients according to claim 3 is characterized in that: The target physiological parameter data is processed and analyzed in combination with its multi-dimensional data correlation and time development decay, and the current physical condition assessment information of the elderly lung cancer patient is obtained through fuzzy comprehensive evaluation processing, including the following steps: Establishing a body state assessment information matrix based on the target physiological parameter data; Based on each of the target physiological parameter data, a correlation factor corresponding to each target physiological parameter data is obtained through correlation analysis; The correlation factor represents the correlation relationship value between every two target physiological parameter data; Calculate the physiological parameter weight w based on the correlation factor; Obtain a fuzzy comprehensive evaluation value Q based on the body state assessment information matrix R and the physiological parameter weights w corresponding to each of the physiological parameter data; Based on the fuzzy comprehensive evaluation value, analysis and judgment are performed to obtain the current physical condition assessment information of the elderly lung cancer patient.

5. The method for processing home care monitoring data of elderly lung cancer patients according to claim 4 is characterized in that: The step of obtaining the correlation factor corresponding to each target physiological parameter data through correlation analysis based on each target physiological parameter data specifically includes the following steps: Establish a correlation matrix F based on the target physiological parameter data corresponding to each moment j ; Calculate the mean value of each target physiological parameter data collected at all times And the standard deviation Sr of the target physiological parameter data j ; Based on the mean of the target physiological parameter data And the standard deviation Sr of the target physiological parameter data j The first correlation factor k is calculated by calculating each sub-parameter in the correlation matrix F. j ; Based on the first correlation factor k j The correlation factor k is calculated.

6. The method for processing home care monitoring data of elderly lung cancer patients according to claim 4, characterized in that: The step of obtaining the correlation factor corresponding to each target physiological parameter data through correlation analysis based on each target physiological parameter data specifically includes the following steps: Initialize iteration parameters; The iteration parameters include a target physiological parameter data set P, a time series set J, an initialization adaptive time decay matrix D, an iteration counter, a maximum number of iterations Cmax, and a convergence threshold ε; Constructing a time correlation matrix R(tj) for each time point j based on the target physiological parameter data set and the time series set; Add 1 to the iteration number of the iteration counter to obtain the current iteration number; Calculating and obtaining a parameter importance vector Ia based on each element in the time correlation matrix R(tj); updating the adaptive time decay matrix D based on the parameter importance vector Ia to obtain an updated adaptive time decay matrix D'; Constructing a multidimensional correlation factor tensor K based on the updated adaptive time decay matrix D'; Calculate a comprehensive correlation factor matrix L based on the multidimensional correlation factor tensor K; After performing singular value decomposition on the comprehensive correlation factor matrix L, the left singular vectors corresponding to the first three largest singular values ​​are selected to form a characteristic matrix T; Clustering the feature matrix T to obtain multiple clusters; screening each of the clusters based on a preset discreteness threshold and an inter-cluster distance threshold to obtain abnormal clusters and normal clusters; Calculate and obtain the average correlation factor in each of the normal clusters, and correct the abnormal clusters through principal component analysis based on the average correlation factor to obtain a corrected cluster cluster; Based on the modified clustering clusters, the comprehensive correlation factor matrix L is updated to obtain an updated comprehensive correlation factor matrix L'; Calculate and obtain the norm difference of the comprehensive correlation factor matrix L and the updated comprehensive correlation factor matrix L'; determine whether the norm difference is greater than or equal to the convergence threshold ε; if so, calculate and obtain the eigenvalue Λ and the eigenvector V based on the comprehensive correlation factor matrix L; construct the correlation factor vector U based on the eigenvalue Λ and the eigenvector V; if not, further determine whether the current number of iterations is greater than or equal to the maximum number of iterations Cmax; if so, calculate and obtain the eigenvalue Λ and the eigenvector V based on the comprehensive correlation factor matrix L; construct the correlation factor vector U based on the eigenvalue Λ and the eigenvector V; if not, return to the above operation until the norm difference is greater than or equal to the convergence threshold ε or the current number of iterations is greater than or equal to the maximum number of iterations Cmax, and obtain the correlation factor vector U.

7. The method for processing home care monitoring data of elderly lung cancer patients according to claim 6, characterized in that: The initialization adaptive time decay matrix D = [dab] n × n; Wherein, dab represents the attenuation coefficient of the target physiological parameter data pa to the target physiological parameter data pb; n represents the number of target physiological parameter data.

8. The method for processing home care monitoring data of elderly lung cancer patients according to claim 7, characterized in that: The adaptive time decay matrix D is updated based on the parameter importance vector Ia to obtain the updated adaptive time decay matrix D' by updating the decay coefficient dab of the adaptive time decay matrix D to obtain a new decay coefficient dab new Then, the parameter importance vector of the target physiological parameter data a and the parameter importance vector of the target physiological parameter data b are obtained to update the new attenuation coefficient dab new , and finally according to the new attenuation coefficient dab new Get the updated adaptive time decay matrix D'; The updated adaptive time decay matrix D'=[dab new ]n×n; Among them, dab new represents the new attenuation coefficient of the target physiological parameter data pa to the target physiological parameter data pb; n represents the number of target physiological parameter data.

9. The method for processing home care monitoring data of elderly lung cancer patients according to claim 8, characterized in that: The parameter importance vector of the target physiological parameter data a and the parameter importance vector of the target physiological parameter data b are obtained and updated to obtain a new attenuation coefficient dab new ; Wherein, the new attenuation coefficient dab new The update calculation method is: dab new <dab*(1−α*Ib*Ib); Where α is the learning rate; Ia is the parameter importance vector of the target physiological parameter data a; lb is the parameter importance vector of the target physiological parameter data b.

Images