Remote home dialysis data management cloud platform for nephropathy and data analysis method

Abstract

The invention discloses a remote home dialysis data management cloud platform for nephropathy and a data analysis method, and relates to the technical field of home medical treatment, the remote home dialysis data management cloud platform comprises a data acquisition module, a data cloud end and an interaction terminal; the data acquisition module comprises a dialysis process data acquisition unit used for acquiring real-time dialysis data in the household dialysis process of a user; the home behavior data acquisition unit is used for acquiring user home behavior data; the data cloud comprises a cloud data storage module used for storing real-time dialysis data and user home behavior data in the user home dialysis process collected by a data collection module; the cloud data analysis module is used for analyzing whether the home dialysis treatment of the user is abnormal or not; and the virtual diagnosis and treatment module is used for constructing a virtual diagnosis and treatment room. The system has the advantages that an intelligent management mode is adopted, so that the treatment effect and the management efficiency of home dialysis are remarkably improved, and a safer and more convenient medical service solution is provided for patients with chronic kidney diseases.

Description

Technical Field

[0001] The present invention relates to the field of home medical technology, and in particular to a remote home dialysis data management cloud platform and a data analysis method for kidney disease. Background Art

[0002] As the number of patients with chronic kidney disease (CKD) continues to rise year by year, the traditional hospital-based centralized dialysis treatment model faces severe challenges. Patients need to frequently travel to and from the hospital, which not only increases the economic burden, but may also lead to a decrease in treatment compliance due to problems such as inconvenient transportation and uneven distribution of medical resources. In addition, during home dialysis, patients' self-management capabilities for dialysis parameters (such as ultrafiltration volume, electrolyte balance) and daily diet (such as protein, potassium and phosphorus intake) vary, which can easily lead to abnormal treatment and even life-threatening complications. Existing remote monitoring systems mostly focus on a single data dimension and lack the ability to integrate and analyze multi-source heterogeneous data (such as dialysis process data, dietary behavior, and physical signs and symptoms), making it difficult to achieve early warning and precise intervention.

[0003] Existing data analysis models typically use fixed weights or simple statistical methods, failing to dynamically adapt to individual differences and changes in condition, resulting in insufficient sensitivity and specificity in abnormality identification. For example, some systems focus solely on dialysate flow or electrolyte concentration threshold alarms, while ignoring the cumulative impact of behavioral factors such as the patient's diet and water intake habits on treatment effectiveness. Furthermore, issues such as inefficient doctor-patient communication and the difficulty in distributing diagnostic and treatment resources to the grassroots level further exacerbate the management difficulties of home dialysis. Summary of the Invention

[0004] In order to solve the above technical problems, a remote home dialysis data management cloud platform and data analysis method for kidney disease are provided. This technical solution solves the problems raised in the above background technology.

[0005] In order to achieve the above purpose, the technical solution adopted by the present invention is:

[0006] A remote home dialysis data management cloud platform for kidney disease, including: data collection module, data cloud and interactive terminal;

[0007] The data acquisition module includes:

[0008] A dialysis process data acquisition unit, which is used to collect real-time dialysis data during the user's home dialysis process;

[0009] A home behavior data collection unit, which is used to collect user home behavior data, including daily diet data and user self-reported physical sign data;

[0010] The data cloud includes:

[0011] A cloud data storage module, which is used to store the real-time dialysis data of the user during the home dialysis process and the user's home behavior data collected by the data collection module;

[0012] A cloud data analysis module, which is used to analyze whether there are any abnormalities in the user's home dialysis treatment by combining the user's real-time dialysis data during the home dialysis process and the user's home behavior data;

[0013] A virtual diagnosis and treatment module, which is used to build a virtual diagnosis and treatment room and send an access request to the interactive terminal when there is an abnormality in the user's home dialysis treatment;

[0014] The interactive terminal includes a user terminal and a doctor terminal. The interactive terminal is used to respond to the access request sent by the virtual diagnosis and treatment module and access the user terminal and the doctor terminal to enter the virtual diagnosis and treatment room.

[0015] Furthermore, a remote home dialysis data analysis method for kidney disease is proposed, including:

[0016] A personalized dataset is constructed based on the user's real-time dialysis data, daily diet data, and self-reported vital signs data during home dialysis. The real-time dialysis data includes the adequacy index of a single dialysis session, ultrafiltration volume deviation, and changes in blood potassium and blood phosphorus concentrations before and after dialysis. The daily diet data includes the daily intake imbalance index of protein, phosphorus, potassium, and water. The self-reported vital signs data includes urine volume classification, edema degree score, and text descriptions of the patient's self-reported symptoms.

[0017] Determine the focus weights of different indicator data and construct the focus weight matrix of indicator data;

[0018] Analyze the indicator data types in the personalized data set that deviate from the standard indicators and record them as abnormal indicator data;

[0019] Determine the abnormal attention index by combining the deviation of abnormal indicator data and the attention focus weight matrix of indicator data;

[0020] Determine whether the abnormal attention index exceeds the attention threshold. If so, trigger the abnormal identification mechanism; if not, do not respond;

[0021] The abnormality identification mechanism is specifically as follows:

[0022] Based on the deviation of abnormal indicator data, the abnormal indicators of home dialysis are analyzed to determine whether the abnormal indicators of home dialysis exceed the abnormal threshold. If so, a virtual diagnosis and treatment room is constructed through the virtual diagnosis and treatment module. If not, the abnormal indicator data is output to the user end to remind the user to pay attention to the abnormal indicator index.

[0023] Preferably, the construction of a personalized data set based on the user's real-time dialysis data, daily diet data, and self-reported vital sign data during home dialysis specifically includes:

[0024] Gaussian process regression interpolation was used for real-time dialysis data;

[0025] Use the BERT model to extract nutritional entities from daily diet data, map them to standard food composition tables, and calculate the daily cumulative nutritional values of protein, phosphorus, potassium, and water based on the standard food composition tables;

[0026] Daily nutrient accumulation values based on protein, phosphorus, potassium and water, as well as daily imbalance index of protein, phosphorus, potassium and water;

[0027] The user's self-reported physical sign data is graded and quantified, and symptom entity recognition is performed on the text description of the patient's self-reported symptoms. The standardized physical sign code is output, where the physical sign code 1 represents the presence of the symptom entity, and the physical sign code 0 represents the absence of the symptom entity.

[0028] Aggregate all feature data to obtain the user's personalized data set.

[0029] Preferably, determining the focus weights of different indicator data and constructing the focus weight matrix of the indicator data specifically includes:

[0030] Principal component analysis and / or multiple linear regression analysis were used to analyze the weight of each indicator data in relation to the user's focus on kidney disease care;

[0031] Summarize the weights of all indicator data related to the user's kidney disease maintenance to obtain the indicator data's focus weight matrix.

[0032] Preferably, the determining of the abnormal attention index by combining the deviation of the abnormal indicator data and the attention focus weight matrix of the indicator data specifically includes:

[0033] Extract the focus weight of each abnormal indicator from the focus weight matrix of indicator data;

[0034] The abnormal attention index is obtained by weighted summing the deviation value of each abnormal indicator data and the attention focus weight of each abnormal indicator.

[0035] Preferably, the abnormal indicators of home dialysis based on the deviation analysis of abnormal indicator data specifically include:

[0036] Constructing an abnormal risk model for home dialysis based on logistics regression or neural network, wherein the abnormal risk model takes the deviation value of all indicator data as input and uses the abnormal risk probability of home dialysis as the abnormal indicator of home dialysis output;

[0037] Substitute the deviation values of all the user's indicator data into the abnormal risk model to obtain the abnormal indicators of the user's home dialysis. The deviation value of the abnormal indicator data is recorded as the absolute difference between the abnormal indicator data value and the corresponding standard indicator data, and the deviation value of the normal indicator data is recorded as 0.

[0038] Preferably, the construction of a virtual diagnosis and treatment room through the virtual diagnosis and treatment module specifically includes:

[0039] Combined with the Pearson correlation coefficient and expert experience analysis, the contribution value of different abnormal indicator data to this abnormality identification is calculated;

[0040] Sort indicators based on contribution values and generate auxiliary diagnosis sheets for users;

[0041] Building a virtual clinic session based on the virtual diagnosis and treatment module, storing the user's auxiliary diagnosis form in the cache space of the virtual clinic session, and locking it;

[0042] Ensure that both the user and doctor are connected to the virtual clinic session, unlock the user's auxiliary diagnosis form, and after the virtual clinic session is completed, delete the user's auxiliary diagnosis form in the cache space and disband the virtual clinic session.

[0043] Compared with the prior art, the present invention has the following beneficial effects:

[0044] The present invention constructs a full-process management system for home dialysis through multi-source data fusion, dynamic weight allocation and intelligent diagnosis and treatment technology, effectively solving the problems of data fragmentation, early warning lag and insufficient doctor-patient collaboration in the traditional model. The platform integrates real-time dialysis parameters, dietary behavior and physical signs and symptoms data, uses Gaussian interpolation, BERT semantic analysis and other technologies to form personalized dynamic portraits, and optimizes the indicator weight matrix through principal component analysis to achieve improved accuracy and sensitivity in abnormal identification. When an abnormality is detected, the system automatically triggers the virtual clinic, generates a structured auxiliary diagnosis form, supports remote collaborative diagnosis and treatment, and reduces emergency hospitalization rates and medical expenses. In addition, visualization tools and standardized physical sign coding help patients manage themselves, improve treatment compliance, and achieve a closed loop from early warning, precise intervention to long-term health management, significantly improving patients' quality of life and treatment compliance. BRIEF DESCRIPTION OF THE DRAWINGS

[0045] Figure 1 This is a flow chart of the remote home dialysis data analysis method for kidney disease proposed in Example 2 of this solution;

[0046] Figure 2 This is a flow chart of the method for constructing a focus weight matrix for indicator data proposed in Example 2 of this solution;

[0047] Figure 3This is a flow chart of the method for determining the abnormal attention index proposed in Example 2 of this solution;

[0048] Figure 4 This is a flow chart of the method for analyzing abnormal indicators of home dialysis proposed in Example 2 of this solution;

[0049] Figure 5 This is a flow chart of the method for constructing a personalized data set proposed in Example 3 of this solution;

[0050] Figure 6 This is a flow chart of the method for constructing a virtual treatment room through a virtual treatment module proposed in the fourth embodiment of this solution. DETAILED DESCRIPTION

[0051] The following description is intended to disclose the present invention so that those skilled in the art can implement the present invention. The preferred embodiments described below are merely examples, and those skilled in the art may conceive of other obvious variations.

[0052] Example 1:

[0053] This embodiment proposes a remote home dialysis data management cloud platform for kidney disease, including: a data acquisition module, a data cloud and an interactive terminal;

[0054] The data acquisition module includes:

[0055] The dialysis process data acquisition unit is used to collect real-time dialysis data during the user's home dialysis process;

[0056] Home behavior data collection unit, which is used to collect user home behavior data, including daily diet data and user self-reported physical sign data;

[0057] The data cloud includes:

[0058] A cloud data storage module, which is used to store the real-time dialysis data of the user during home dialysis and the user's home behavior data collected by the data collection module;

[0059] The cloud data analysis module is used to analyze whether there are any abnormalities in the user's home dialysis treatment by combining the user's real-time dialysis data during the home dialysis process and the user's home behavior data;

[0060] The virtual diagnosis and treatment module is used to build a virtual diagnosis and treatment room and send an access request to the interactive terminal when there is an abnormality in the user's home dialysis treatment;

[0061] The interactive terminal includes a user terminal and a doctor terminal. The interactive terminal is used to respond to the access request sent by the virtual diagnosis and treatment module and access the user terminal and the doctor terminal to enter the virtual diagnosis and treatment room.

[0062] The cloud data analysis module specifically includes:

[0063] The data collection and integration unit is used to acquire and integrate real-time dialysis data during the user's home dialysis process and the user's home behavior data to analyze the user's home dialysis treatment to form a personalized data set:

[0064] The weight allocation and matrix construction unit is used to determine the focus weight of each indicator data, construct the focus weight matrix of the indicator data, and represent the relative importance of different indicators in anomaly judgment;

[0065] The exception analysis unit is used to execute the exception handling process, which includes: when an exception is identified, triggering the virtual diagnosis and treatment module to build a virtual diagnosis and treatment room based on the abnormal indicator data and provide online diagnosis and treatment services; when no exception is identified, no response is made.

[0066] The virtual diagnosis and treatment module includes:

[0067] Anomaly indicator visualization unit, which displays key anomaly indicators in a graphical manner;

[0068] Intelligent diagnostic unit, which generates personalized intervention recommendations based on pre-set clinical guidelines and machine learning models;

[0069] Remote communication interface, supporting users and doctors to enter the virtual treatment room.

[0070] Example 2:

[0071] Reference Figure 1 As shown, this embodiment combines the remote home dialysis data management cloud platform for kidney disease in Example 1 to propose a remote home dialysis data analysis method for kidney disease, including:

[0072] A personalized data set is constructed based on the user's real-time dialysis data, daily diet data, and self-reported vital signs during home dialysis. Real-time dialysis data includes single dialysis adequacy indicators, ultrafiltration volume deviation, and changes in blood potassium and blood phosphorus concentrations before and after dialysis. Daily diet data includes daily intake imbalance indexes of protein, phosphorus, potassium, and water. Self-reported vital signs data includes urine volume classification, edema severity score, and text descriptions of self-reported symptoms.

[0073] Determine the focus weights of different indicator data and construct the focus weight matrix of indicator data;

[0074] Analyze the indicator data types in the personalized data set that deviate from the standard indicators and record them as abnormal indicator data;

[0075] Determine the abnormal attention index by combining the deviation of abnormal indicator data and the attention focus weight matrix of indicator data;

[0076] Determine whether the abnormal attention index exceeds the attention threshold. If so, trigger the abnormal identification mechanism; if not, do not respond;

[0077] The specific abnormality identification mechanism is as follows:

[0078] Based on the deviation of abnormal indicator data, the abnormal indicators of home dialysis are analyzed to determine whether the abnormal indicators of home dialysis exceed the abnormal threshold. If so, a virtual diagnosis and treatment room is constructed through the virtual diagnosis and treatment module. If not, the abnormal indicator data is output to the user end to remind the user to pay attention to the abnormal indicator index;

[0079] in,

[0080] Reference Figure 2 As shown in the figure, determining the focus weights of different indicator data and constructing the focus weight matrix of indicator data specifically includes:

[0081] Principal component analysis and / or multiple linear regression analysis were used to analyze the weight of each indicator data in relation to the user's focus on kidney disease care;

[0082] Summarize the weights of all indicator data related to the user's kidney disease maintenance to obtain the indicator data's focus weight matrix.

[0083] The specific steps of principal component analysis are:

[0084] Construct a standardized data matrix Z, Z = [y ij ] n×m , where y ij is the jth normalized indicator in the i-th sample, n is the number of samples, and m is the number of indicator types;

[0085] Based on the standardized data matrix Z, construct the covariance matrix R;

[0086]

[0087] Perform eigendecomposition on the covariance matrix to obtain the eigenvalue λ j and the corresponding eigenvector v j ;

[0088] Combined with eigenvalue λ b and the corresponding eigenvector v b , calculate the contribution rate U of each indicator j ;

[0089]

[0090] Take Uj As an indicator, the focus on kidney care is weighted.

[0091] The specific steps of the multiple linear regression method are as follows:

[0092] Based on the multidimensional indicators related to kidney disease care, a multivariate linear regression equation of kidney disease care-multidimensional indicators was constructed;

[0093] S=β0+β1y1+…+β j y j +…+β m y m

[0094] Among them, S is kidney disease maintenance, β j is the direct impact of the jth indicator on kidney disease care, y j is the jth indicator.

[0095] Based on the multiple linear regression equation of kidney disease care-multidimensional index, the standardized regression coefficient was calculated;

[0096]

[0097] in, is the standardized regression coefficient of the jth indicator, σ S is the standard deviation of the kidney disease care evaluation index, σ yj is the standard deviation of the j-th indicator;

[0098] by As an indicator, the focus on kidney care is weighted;

[0099] Reference Figure 3 As shown in the figure, combining the deviation of abnormal indicator data and the focus weight matrix of indicator data, the abnormal attention index is determined to include:

[0100] Extract the focus weight of each abnormal indicator from the focus weight matrix of indicator data;

[0101] The abnormal attention index is obtained by weighted summing the deviation value of each abnormal indicator data and the attention focus weight of each abnormal indicator.

[0102] It is understandable that different indicators such as urea clearance rate (Kt / V), dialysis ultrafiltration volume, protein intake in dietary structure, potassium and phosphorus intake have different importance for kidney care at different stages of kidney disease, and it is difficult to get professional medical management during home dialysis. Therefore, this program uses a weighted analysis approach to evaluate the importance of different indicators for home kidney care. By weighted summing the deviation values of indicators that deviate from the normal range, a preliminary judgment is made on the importance of abnormal attention during home dialysis. When the importance of abnormal attention is large, it means that the highly important indicators have abnormalities, or some indicators have large deviations. At this time, the risk of worsening of the user's condition is greater, and further analysis and judgment are required.

[0103] Reference Figure 4 As shown, the abnormal indicators of home dialysis based on the deviation analysis of abnormal indicator data specifically include:

[0104] An abnormal risk model for home dialysis is constructed based on logistic regression or neural network. The abnormal risk model takes the deviation value of all indicator data as input and uses the abnormal risk probability of home dialysis as the abnormal indicator of home dialysis output;

[0105] Substitute the deviation values of all the user's indicator data into the abnormal risk model to obtain the abnormal indicators of the user's home dialysis. The deviation value of the abnormal indicator data is recorded as the absolute difference between the abnormal indicator data value and the corresponding standard indicator data, and the deviation value of the normal indicator data is recorded as 0.

[0106] On the basis of judging the deviation of indicators, when making further judgments on risks, logistics regression judgment is used or a neural network is used to construct an abnormal risk model for home dialysis. Through a deeper comprehensive analysis of the deviations of all indicators, the user's current state of kidney disease development is further judged. When the abnormal risk probability of home dialysis is too high, it means that the user's current condition is at high risk of worsening. At this time, the user is connected to the virtual diagnosis room and a doctor performs a professional diagnosis. If the abnormal risk probability of home dialysis is not high, it means that the user's current condition is at low risk of worsening. At this time, the indicator type that deviates from the standard value will be output to the user to remind the user to pay attention.

[0107] Real-time example three:

[0108] Reference Figure 5 As shown, based on the second embodiment, this embodiment further points out that the personalized data set is constructed based on the user's real-time dialysis data, daily diet data and user self-reported vital sign data during home dialysis, specifically including:

[0109] Gaussian process regression interpolation is used for real-time dialysis data. This method effectively handles the time-varying and noise interference of real-time dialysis data, improves the prediction accuracy of hemodynamic parameters, and provides a scientific basis for dynamic treatment plan adjustments.

[0110] The BERT model is used to extract nutritional entities from daily dietary data, mapped to a standard food composition table, and the daily cumulative nutritional values of protein, phosphorus, potassium, and water are calculated based on the standard food composition table. The BERT-based multi-granularity nutritional analysis engine can automatically parse free-text dietary records and accurately measure nutrients in combination with the standard food composition table, significantly improving the efficiency and accuracy of nutritional assessments.

[0111] Daily nutrient accumulation values based on protein, phosphorus, potassium and water, as well as daily imbalance index of protein, phosphorus, potassium and water;

[0112] The calculation formula for the intake imbalance index is:

[0113]

[0114] i∈{P,K,Pr,W} represents phosphorus, potassium, protein, and water, respectively. i (t) is the actual intake, T i (t) is the target intake, σ i is the individualized standard deviation, Z i (t) is the intake imbalance index.

[0115] The system quantifies user-reported vital signs and identifies symptom entities within textual descriptions of self-reported symptoms. Standardized vital sign codes are output, with a sign code of 1 representing the presence of the symptom entity and a sign code of 0 representing its absence. The sign semantic analysis framework utilizes hybrid model technology to achieve standardized recognition and quantification of diverse symptom descriptions, generating structured vital sign assessment reports. The system optimizes individualized patient management through a data-driven approach, reduces the risk of complications, and improves the accuracy of nutritional interventions, providing intelligent solutions for improving patient quality of life and optimizing the allocation of medical resources.

[0116] Real-time example 4:

[0117] Reference Figure 6 As shown, in this embodiment, constructing a virtual treatment room through a virtual treatment module specifically includes:

[0118] Combined with the Pearson correlation coefficient and expert experience analysis, the contribution value of different abnormal indicator data to this abnormality identification is calculated;

[0119] The calculation formula of Pearson correlation coefficient is:

[0120]

[0121] Where r j is the correlation between the jth indicator type and kidney disease care, x ij is the jth index value of the i-th sample value, y i is the kidney disease care result of the i-th sample value (e.g. binary label: 0 = normal, 1 = worsening), and x ij and y i The average value of , n is the number of data point pairs analyzed;

[0122] The specific contribution of abnormal indicator data to this abnormality identification is:

[0123] The contribution value calculation formula of the indicator type is:

[0124]

[0125] Among them, IR j is the contribution value of the j-th indicator type, r j is the correlation between the jth indicator type and kidney disease care, x j is the indicator deviation value of the j-th indicator type, w j is the empirical coefficient of the jth indicator type, b is the bias term, α and β are both weight coefficients, and α+β=1, where, It is the empirical term of logistic regression. The values of α and β are determined by the amount of data. For example, when the data is sufficient, it depends on statistical results, and α increases. When the data is sparse or noisy, it depends on expert experience correction, and β increases.

[0126] Sort indicators based on contribution values and generate auxiliary diagnosis sheets for users;

[0127] Building a virtual clinic session based on the virtual diagnosis and treatment module, storing the user's auxiliary diagnosis form in the cache space of the virtual clinic session, and locking it;

[0128] Ensure that both the user and doctor are connected to the virtual clinic session, unlock the user's auxiliary diagnosis form, and after the virtual clinic session is completed, delete the user's auxiliary diagnosis form in the cache space and disband the virtual clinic session.

[0129] By calculating the contribution value, doctors who access the virtual clinic session can quickly and accurately obtain the user's actual physical condition, thereby improving the accuracy and efficiency of the doctor's diagnosis and treatment decisions.

[0130] In summary, the advantages of this invention lie in: It proposes a novel home dialysis service that combines multi-dimensional analysis and assessment of individual dietary habits and dialysis parameters, and employs a hierarchical management mechanism that intelligently identifies patient risk levels. For low-risk patients, the system provides real-time health monitoring and early warning alerts; for high-risk patients, a virtual diagnosis and treatment system is established, enabling remote, precise intervention by doctors. This intelligent management model significantly improves the therapeutic effectiveness and management efficiency of home dialysis, providing a safer and more convenient medical service solution for patients with chronic kidney disease.

[0131] The above shows and describes the basic principles, main features, and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The above embodiments and descriptions merely illustrate the principles of the present invention. Various changes and modifications may be made to the present invention without departing from the spirit and scope of the present invention. Such changes and modifications are intended to fall within the scope of the present invention. The scope of protection claimed by the present invention is defined by the appended claims and their equivalents.

Claims

1. A remote home dialysis data management cloud platform for kidney disease, characterized by: include: Data acquisition module, data cloud and interactive terminal; The data acquisition module includes: A dialysis process data acquisition unit, which is used to collect real-time dialysis data during the user's home dialysis process; A home behavior data collection unit, which is used to collect user home behavior data, including daily diet data and user self-reported physical sign data; The data cloud includes: A cloud data storage module, which is used to store the real-time dialysis data of the user during the home dialysis process and the user's home behavior data collected by the data collection module; A cloud data analysis module, which is used to analyze whether there are any abnormalities in the user's home dialysis treatment by combining the user's real-time dialysis data during the home dialysis process and the user's home behavior data; A virtual diagnosis and treatment module, which is used to build a virtual diagnosis and treatment room and send an access request to the interactive terminal when there is an abnormality in the user's home dialysis treatment; The interactive terminal includes a user terminal and a doctor terminal. The interactive terminal is used to respond to the access request sent by the virtual diagnosis and treatment module and access the user terminal and the doctor terminal to enter the virtual diagnosis and treatment room.

2. The remote home dialysis data management cloud platform for kidney disease according to claim 1 is characterized in that: The cloud data analysis module specifically includes: The data collection and integration unit is used to acquire and integrate real-time dialysis data during the user's home dialysis process and the user's home behavior data to analyze the user's home dialysis treatment to form a personalized data set: The weight allocation and matrix construction unit is used to determine the focus weight of each indicator data, construct the focus weight matrix of the indicator data, and represent the relative importance of different indicators in anomaly judgment; The exception analysis unit is used to execute the exception handling process, which includes: when an exception is identified, triggering the virtual diagnosis and treatment module to build a virtual diagnosis and treatment room based on the abnormal indicator data and provide online diagnosis and treatment services; when no exception is identified, no response is made.

3. The remote home dialysis data management cloud platform for kidney disease according to claim 1, characterized in that: The virtual diagnosis and treatment module includes: Anomaly indicator visualization unit, which displays key anomaly indicators in a graphical manner; Intelligent diagnostic unit, which generates personalized intervention recommendations based on pre-set clinical guidelines and machine learning models; Remote communication interface, supporting users and doctors to enter the virtual treatment room.

4. A remote home dialysis data analysis method for kidney disease, characterized in that: The remote home dialysis data management cloud platform for kidney disease according to any one of claims 1 to 3 comprises: A personalized dataset is constructed based on the user's real-time dialysis data, daily diet data, and self-reported vital signs data during home dialysis. The real-time dialysis data includes the adequacy index of a single dialysis session, ultrafiltration volume deviation, and changes in blood potassium and blood phosphorus concentrations before and after dialysis. The daily diet data includes the daily intake imbalance index of protein, phosphorus, potassium, and water. The self-reported vital signs data includes urine volume classification, edema degree score, and text descriptions of the patient's self-reported symptoms. Determine the focus weights of different indicator data and construct the focus weight matrix of indicator data; Analyze the indicator data types in the personalized data set that deviate from the standard indicators and record them as abnormal indicator data; Determine the abnormal attention index by combining the deviation of abnormal indicator data and the attention focus weight matrix of indicator data; Determine whether the abnormal attention index exceeds the attention threshold. If so, trigger the abnormal identification mechanism; if not, do not respond; The abnormality identification mechanism is specifically as follows: Based on the deviation of abnormal indicator data, the abnormal indicators of home dialysis are analyzed to determine whether the abnormal indicators of home dialysis exceed the abnormal threshold. If so, a virtual diagnosis and treatment room is constructed through the virtual diagnosis and treatment module. If not, the abnormal indicator data is output to the user end to remind the user to pay attention to the abnormal indicator index.

5. A remote home dialysis data analysis method for kidney disease according to claim 4, characterized in that: The personalized data set constructed based on the user's real-time dialysis data, daily diet data, and self-reported vital sign data during home dialysis specifically includes: Gaussian process regression interpolation was used for real-time dialysis data; Use the BERT model to extract nutritional entities from daily diet data, map them to standard food composition tables, and calculate the daily cumulative nutritional values of protein, phosphorus, potassium, and water based on the standard food composition tables; Daily nutrient accumulation values based on protein, phosphorus, potassium and water, as well as daily imbalance index of protein, phosphorus, potassium and water; The user's self-reported physical sign data is graded and quantified, and symptom entity recognition is performed on the text description of the patient's self-reported symptoms. The standardized physical sign code is output, where the physical sign code 1 represents the presence of the symptom entity, and the physical sign code 0 represents the absence of the symptom entity. Aggregate all feature data to obtain the user's personalized data set.

6. A remote home dialysis data analysis method for kidney disease according to claim 5, characterized in that: Determining the focus weights of different indicator data and constructing the focus weight matrix of the indicator data specifically includes: Principal component analysis and / or multiple linear regression analysis were used to analyze the weight of each indicator data in relation to the user's focus on kidney disease care; Summarize the weights of all indicator data related to the user's kidney disease maintenance to obtain the indicator data's focus weight matrix.

7. A remote home dialysis data analysis method for kidney disease according to claim 6, characterized in that: Determining the abnormal attention index by combining the deviation of the abnormal indicator data and the attention focus weight matrix of the indicator data specifically includes: Extract the focus weight of each abnormal indicator from the focus weight matrix of indicator data; The abnormal attention index is obtained by weighted summing the deviation value of each abnormal indicator data and the attention focus weight of each abnormal indicator.

8. The remote home dialysis data analysis method for kidney disease according to claim 7, characterized in that: The abnormal indicators of home dialysis based on the deviation analysis of abnormal indicator data specifically include: Constructing an abnormal risk model for home dialysis based on logistics regression or neural network, wherein the abnormal risk model takes the deviation value of all indicator data as input and uses the abnormal risk probability of home dialysis as the abnormal indicator of home dialysis output; Substitute the deviation values of all the user's indicator data into the abnormal risk model to obtain the abnormal indicators of the user's home dialysis. The deviation value of the abnormal indicator data is recorded as the absolute difference between the abnormal indicator data value and the corresponding standard indicator data, and the deviation value of the normal indicator data is recorded as 0.

9. The remote home dialysis data analysis method for kidney disease according to claim 8, characterized in that: The construction of a virtual diagnosis and treatment room through the virtual diagnosis and treatment module specifically includes: Combined with the Pearson correlation coefficient and expert experience analysis, the contribution value of different abnormal indicator data to this abnormality identification is calculated; Sort indicators based on contribution values and generate auxiliary diagnosis sheets for users; Building a virtual clinic session based on the virtual diagnosis and treatment module, storing the user's auxiliary diagnosis form in the cache space of the virtual clinic session, and locking it; Ensure that both the user and doctor are connected to the virtual clinic session, unlock the user's auxiliary diagnosis form, and after the virtual clinic session is completed, delete the user's auxiliary diagnosis form in the cache space and disband the virtual clinic session.

Images