A method, system and medical device for monitoring vital signs and physiological data of patients in nephrology department

Abstract

The present invention relates to the field of medical data mining technology, and in particular to a method, system and medical equipment for monitoring vital signs and physiological data of patients in the Department of Nephrology. The present invention first obtains the standard vital signs vectors of patients in the Department of Nephrology every day; further clusters the standard vital signs vectors to obtain time series sub-segments, and screens out abnormal sub-segments; further obtains the abnormal factors of each abnormal sub-segment according to the number and abnormal characteristics of the standard vital signs vectors in each abnormal sub-segment; further merges adjacent similar time series sub-segments to obtain normal sequence segments and abnormal sequence segments; further obtains the corrected deterioration factor of each abnormal sequence segment according to the number of abnormal sub-segments in each abnormal sequence segment, the fluctuation characteristics and statistical characteristics of the abnormal factors, and the difference characteristics between each abnormal sequence segment and the adjacent normal sequence segment; finally obtains the weight of daily vital signs and physiological data, and uses the ARIMA model to obtain predicted physiological data.

Description

Technical Field

[0001] The present invention relates to the technical field of medical data mining, and in particular to a method, system and medical equipment for monitoring vital signs and physiological data of patients in a nephrology department. Background Art

[0002] Currently, changes in patients' physiological data are often predicted to provide early warning of possible risks, such as acute events or worsening of the disease. For example, the ARIMA model (Autoregressive Integrated Moving Average Model) can be combined with the patient's personal historical data for prediction, providing patients with more personalized medical services and treatment recommendations.

[0003] However, in the collected physiological data, abnormal data of certain problems or complications of patients are often less than normal data. If predictions are made directly based on historical data, the model may be biased towards more common normal data, thereby ignoring abnormal situations, resulting in unreliable predicted data and affecting the effect of monitoring the vital signs and physiological data of nephrology patients. Therefore, improving the sensitivity and prediction accuracy of the ARIMA model to abnormal data of nephrology patients has become a technical problem that needs to be solved urgently. Summary of the invention

[0004] In order to solve the technical problem that the ARIMA model has low sensitivity to abnormal data of patients in the Department of Nephrology and the prediction results are not ideal, the purpose of the present invention is to provide a method, system and medical equipment for monitoring the vital signs and physiological data of patients in the Department of Nephrology. The technical solutions adopted are as follows:

[0005] A method for monitoring vital signs and physiological data of a patient in a nephrology department, the method comprising:

[0006] Obtaining the physical sign and physiological data of all preset physiological index dimensions of patients in the Department of Nephrology every day; obtaining the standard physical sign vector every day according to the distribution characteristics of the physical sign and physiological data compared with the normal range of the preset physiological index;

[0007] Clustering the standard vital sign vectors, segmenting the time domain according to the clustering results to obtain time series subsegments; obtaining abnormal subsegments according to the overall characteristics of the standard vital sign vectors in each of the time series subsegments; obtaining the abnormal factor of each of the abnormal subsegments according to the number and abnormal characteristics of the standard vital sign vectors in each of the abnormal subsegments;

[0008] Merging adjacent time sequence sub-segments of the same type to obtain normal sequence segments and abnormal sequence segments; obtaining a corrected abnormal factor for each abnormal sequence segment according to the number of abnormal sub-segments in each abnormal sequence segment, the fluctuation characteristics and statistical characteristics of the abnormal factor; obtaining a deterioration factor for each abnormal sequence segment according to the difference characteristics between the standard eigenvector in each abnormal sequence segment and the standard eigenvector in the adjacent normal sequence segment, combined with the corrected abnormal factor; obtaining a corrected deterioration factor for each abnormal sequence segment according to the difference characteristics between each abnormal sequence segment and the adjacent normal sequence segment, combined with the deterioration factor;

[0009] According to the type of sequence segment at the end of the time domain, combined with the modified deterioration factor and the similarity and time difference between different abnormal sequence segments, the weight of the vital sign physiological data of each day is obtained; based on the weight of the vital sign physiological data of each day, the ARIMA model is used to obtain the predicted physiological data.

[0010] Furthermore, the method for obtaining the standard vital signs vector includes:

[0011] Select any vital sign physiological data of any physiological indicator dimension as the data to be processed; based on the preset normal range of the physiological indicator, obtain the maximum value and the median value of the range; use the absolute value of the difference between the data to be processed and the corresponding median value of the range as the numerator, use the difference between the maximum value and the median value of the range corresponding to the data to be processed as the denominator, and use the fraction as the processed vital sign data of the data to be processed;

[0012] The processed vital sign data of all dimensions of each day are used to form a standard vital sign vector for each day; the order of the physiological indicator dimensions in all the standard vital sign vectors is the same.

[0013] Furthermore, the method for obtaining the abnormal sub-segment includes:

[0014] The average value vector of each of the time series sub-segments is obtained as a comprehensive sub-vector of each of the time series sub-segments; when data greater than a first preset threshold value exists in the comprehensive sub-vector, the corresponding time series sub-segment is determined as an abnormal sub-segment.

[0015] Furthermore, the method for obtaining the abnormal factor includes:

[0016] According to the number of data greater than a first preset threshold in the comprehensive sub-vector of the abnormal sub-segment, combined with the maximum data value in the comprehensive sub-vector, an abnormal characteristic coefficient corresponding to the abnormal sub-segment is obtained; the number of data greater than the first preset threshold and the maximum data value are both positively correlated with the abnormal characteristic coefficient;

[0017] The product of the number of the standard eigenvectors in the abnormal sub-segment and the abnormal characteristic coefficient is used as the abnormal factor corresponding to the abnormal sub-segment.

[0018] Furthermore, the method for obtaining the corrected abnormal factor includes:

[0019] Select any of the abnormal sequence segments as a target sequence segment; obtain a fluctuation intensity parameter according to the fluctuation intensity characteristics of the abnormal factors of the abnormal sub-segment in the target sequence segment; obtain the sum of all the abnormal factors of the target sequence segment as a statistical sum value;

[0020] According to the number of the abnormal sub-segments in the target sequence segment, the statistical sum value and the violent fluctuation parameter, a corrected abnormal factor of the target sequence segment is obtained; the number of the abnormal sub-segments, the statistical sum value and the violent fluctuation parameter are all positively correlated with the corrected abnormal factor.

[0021] Furthermore, the method for obtaining the deterioration factor includes:

[0022] Obtaining the average value vector of each of the normal sequence segments as the comprehensive vector of each of the normal sequence segments; obtaining the average value vector of each of the abnormal sequence segments as the comprehensive vector of each of the abnormal sequence segments;

[0023] According to the difference characteristics of the comprehensive vector of each abnormal sequence segment and the comprehensive vector of the adjacent normal sequence segment, combined with the corresponding corrected abnormal factor, the deterioration factor of each abnormal sequence segment is obtained; the difference characteristics of the comprehensive vector and the corrected abnormal factor are both positively correlated with the deterioration factor.

[0024] Furthermore, the method for obtaining the modified deterioration factor includes:

[0025] Marking the normal sequence segments adjacent to the target sequence segment as a first sequence segment and a second sequence segment respectively;

[0026] In the comprehensive vector of the target sequence segment and the comprehensive vector of the first sequence segment, the absolute value of the difference of each preset physiological indicator dimension is calculated to obtain a first difference vector; in the first difference vector, the physiological indicators with the largest values ​​of the second preset number are selected as the first target physiological indicators, and the others are selected as the first reference physiological indicators; according to the difference between the numerical mean of the first target physiological indicator in the first difference vector and the numerical mean of the first reference physiological indicator, the numerical mean of the first target physiological indicator in the first difference vector and the data mean of the first target physiological indicator in the comprehensive vector of the target sequence segment, the first deterioration correction coefficient of the target sequence segment is obtained; the difference between the numerical mean of the first target physiological indicator in the first difference vector and the numerical mean of the first reference physiological indicator, the numerical mean of the first target physiological indicator in the first difference vector and the data mean of the first target physiological indicator in the comprehensive vector of the target sequence segment are all positively correlated with the first deterioration correction coefficient;

[0027] acquiring a second degradation correction coefficient of the target sequence segment according to the comprehensive vector of the second sequence segment and the target sequence segment;

[0028] acquiring a modified degradation factor of the target sequence segment according to the first degradation correction coefficient, the second degradation correction coefficient and the degradation factor of the target sequence segment; the first degradation correction coefficient, the second degradation correction coefficient and the degradation factor are all positively correlated with the modified degradation factor;

[0029] The second preset number is smaller than the total number of preset physiological indicator dimensions.

[0030] Furthermore, the method for obtaining the weight of the daily vital sign physiological data includes:

[0031] Setting the weight of the physical sign physiological data of each day in the normal sequence segment to a third preset positive value;

[0032] When the end of the time domain is a normal sequence segment, the sum of the normalized value of the modified deterioration factor of the abnormal sequence segment and the third preset positive value is used as the weight of the vital sign physiological data of each day in the abnormal sequence segment;

[0033] When the end of the time domain is an abnormal sequence segment, the abnormal sequence segment at the end of the time domain is used as the current state sequence segment; according to the similar characteristics of the vital sign physiological data of each abnormal sequence segment and the current state sequence segment in all the same physiological indicator dimensions, combined with the time difference between the abnormal sequence segment and the current state sequence segment, and the corrected deterioration factor, a weight factor of each abnormal sequence segment is obtained; the normalized value of the weight factor and the sum of the third preset positive value are used as the weight of the vital sign physiological data of each day in the abnormal sequence segment;

[0034] The similar features of the vital sign physiological data in all the same physiological indicator dimensions, the time difference and the corrected deterioration factor are all positively correlated with the weight factor.

[0035] The present invention also proposes a system for monitoring vital signs and physiological data of patients in nephrology. The system includes a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it implements any step of the method for monitoring vital signs and physiological data of patients in nephrology.

[0036] The present invention also proposes a medical device for monitoring vital signs and physiological data of patients in the Department of Nephrology, the medical device comprising a data acquisition module, a preliminary analysis module, a deep analysis module and a prediction module;

[0037] The data acquisition module is used to obtain the physiological data of all preset physiological index dimensions of the patients in the Department of Nephrology every day; according to the distribution characteristics of the physiological data of the daily signs compared with the normal range of the preset physiological indexes, the standard physiological vector of each day is obtained;

[0038] A preliminary analysis module is used to cluster the standard vital sign vectors, segment the time domain according to the clustering results, and obtain time series subsegments; obtain abnormal subsegments according to the overall characteristics of the standard vital sign vectors in each of the time series subsegments; and obtain the abnormal factor of each abnormal subsegment according to the number and abnormal characteristics of the standard vital sign vectors in each of the abnormal subsegments;

[0039] A depth analysis module is used to merge adjacent time sequence sub-segments of the same type to obtain normal sequence segments and abnormal sequence segments; obtain a corrected abnormal factor for each abnormal sequence segment according to the number of abnormal sub-segments in each abnormal sequence segment, the fluctuation characteristics and statistical characteristics of the abnormal factor; obtain a deterioration factor for each abnormal sequence segment according to the difference characteristics between the standard eigenvector in each abnormal sequence segment and the standard eigenvector in the adjacent normal sequence segment, combined with the corrected abnormal factor; obtain a corrected deterioration factor for each abnormal sequence segment according to the difference characteristics between each abnormal sequence segment and the adjacent normal sequence segment, combined with the deterioration factor;

[0040] The prediction module is used to obtain the weight of the daily vital sign physiological data according to the type of sequence segment at the end of the time domain, combined with the modified deterioration factor and the similarity and time difference between different abnormal sequence segments; based on the weight of the daily vital sign physiological data, the ARIMA model is used to obtain predicted physiological data.

[0041] The present invention has the following beneficial effects:

[0042] The present invention first obtains the standard vital sign vectors of patients in the Department of Nephrology every day to provide a data basis for subsequent analysis and remove the interference of differences in data ranges and units of different physiological indicators; further clusters the standard vital sign vectors, segments the time domain according to the clustering results, and obtains time series sub-segments, which is convenient for identifying different vital sign physiological states of patients; further obtains abnormal sub-segments, identifies abnormal sub-segments that are significantly different from normal patterns, reduces the interference of single data abnormalities, and obtains abnormal factors for each abnormal sub-segment, preliminarily quantifies the health status of patients in the time period corresponding to the abnormal sub-segment, and provides a basis for the subsequent integration of multiple continuous abnormal sub-segments to obtain the corrected abnormal factors; further considering that merging adjacent similar time series sub-segments is beneficial to analyzing the changes in the patient's vital sign physiological data from a longer time perspective, adjacent similar time series sub-segments are merged And, obtain normal sequence segments and abnormal sequence segments; further obtain the modified abnormal factor of each abnormal sequence segment, characterize the condition of the patient in the abnormal sequence segment, and provide a basis for the subsequent analysis of the patient's condition deterioration characteristics; further by comparing the difference between adjacent abnormal sequence segments and normal sequence segments, obtain the deterioration factor of each abnormal sequence segment, identify the change of the patient's physiological state, so as to effectively monitor the health risk, and prepare for the subsequent assignment of appropriate weights to daily physical sign physiological data; further to avoid the difference between the abnormal sequence segment and the adjacent normal sequence segment is a normal small fluctuation of multiple parameters of the human body, and according to the difference characteristics between each abnormal sequence segment and the adjacent normal sequence segment, combined with the deterioration factor, obtain a more accurate modified deterioration factor for each abnormal sequence segment; finally obtain the weight of daily physical sign physiological data, and use the ARIMA model to obtain predicted physiological data. The present invention identifies the abnormal physical sign physiological data of the patient's history and analyzes the abnormal degree, analyzes the deterioration of the patient at different historical stages, and finally assigns a greater weight to the abnormal data, improves the sensitivity of the ARIMA model to the abnormal data, and provides more accurate predicted physiological data for relevant personnel. BRIEF DESCRIPTION OF THE DRAWINGS

[0043] In order to more clearly illustrate the technical solutions and advantages in the embodiments of the present invention or the prior art, the drawings required for use in the embodiments or the prior art descriptions are briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0044] Figure 1 A flow chart of a method for monitoring vital signs and physiological data of a nephrology patient provided by one embodiment of the present invention;

[0045] Figure 2 The present invention is a flowchart of a method for obtaining a modified degradation factor provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0046] In order to further explain the technical means and effects adopted by the present invention to achieve the predetermined invention purpose, the following is a detailed description of a method, system and medical device for monitoring the physiological data of a patient in nephrology according to the present invention, its specific implementation, structure, features and effects, in conjunction with the accompanying drawings and preferred embodiments. In the following description, different "one embodiment" or "another embodiment" does not necessarily refer to the same embodiment. In addition, specific features, structures or characteristics in one or more embodiments may be combined in any suitable form.

[0047] Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs.

[0048] The following is a detailed description of a method, system and medical device for monitoring vital signs and physiological data of patients in nephrology provided by the present invention in conjunction with the accompanying drawings.

[0049] See also Figure 1 , which shows a flow chart of a method for monitoring vital signs and physiological data of a patient in a nephrology department provided by an embodiment of the present invention, specifically comprising:

[0050] Step S1: Obtain the daily physiological sign data of all preset physiological index dimensions of the nephrology patients; obtain the daily standard physiological sign vector according to the distribution characteristics of the daily physiological sign data compared with the normal range of the preset physiological indexes.

[0051] The embodiment of the present invention improves the sensitivity of the ARIMA model to abnormal data by identifying abnormal data and weighting abnormal vital sign and physiological data, thereby ultimately improving the accuracy of the prediction results and providing more accurate prediction data for relevant personnel.

[0052] First, the daily vital sign physiological data of all preset physiological indicator dimensions of nephrology patients are obtained to provide a data basis for subsequent analysis; considering that the data ranges and units of different physiological indicators are different, the daily standard vital sign vector is obtained based on the distribution characteristics of the daily vital sign physiological data compared with the normal range of preset physiological indicators.

[0053] In one embodiment of the present invention, urine routine data monitoring, hemodialysis monitoring, and renal function monitoring are performed on nephrology patients every day through a urine analyzer, a hemodialysis machine, and a biochemical analyzer to obtain physiological data of urine protein, urine sugar, urine occult blood, blood potassium, blood sodium, blood creatinine, and urea nitrogen, wherein urine protein, urine sugar, urine occult blood, blood potassium, blood sodium, blood creatinine, and urea nitrogen are each used as a preset physiological indicator dimension.

[0054] It should be noted that the processing process and prediction method of the vital signs and physiological data of each nephrology patient are the same, and only any patient is taken as an example for analysis here; in other embodiments of the present invention, the implementer can set other preset physiological indicator dimensions.

[0055] Preferably, in one embodiment of the present invention, considering that the median value usually represents the center position of the normal range, the data is compared with the median value of the normal range, so that the deviation distribution characteristics of the data can be easily obtained. At the same time, in order to ensure that the changes of different physiological indicators are compared on a unified scale, it is also necessary to use the maximum deviation range of the data, that is, the difference between the maximum value of the range corresponding to the data and the median value of the range;

[0056] Based on this, any physiological sign data of any physiological indicator dimension is selected as the data to be processed; based on the preset normal range of the physiological indicator, the maximum value and the median value of the range are obtained; the absolute value of the difference between the data to be processed and the corresponding median value of the range is used as the numerator, the difference between the maximum value and the median value of the range corresponding to the data to be processed is used as the denominator, and the fraction is used as the processed physiological sign data of the data to be processed;

[0057] The processed vital sign data of all dimensions of each day are used to form a standard vital sign vector for each day; the order of the physiological indicator dimensions in all standard vital sign vectors is the same.

[0058] Among them, the absolute value of the difference between the data to be processed and the corresponding median of the range represents the absolute deviation degree of the data to be processed from the corresponding median of the range. The larger the absolute value of the difference, the more the data to be processed deviates from the corresponding median of the range. By taking the difference between the maximum value of the range corresponding to the data to be processed and the median of the range as the denominator, the maximum deviation range of the data to be processed is expressed. When the fraction ratio is less than or equal to 1, the data to be processed are all in the normal range. When the fraction ratio is greater than 1, the data to be processed exceeds the normal range and is abnormal data, thereby expressing the distribution characteristics of the vital sign physiological data compared with the normal range of the preset physiological indicators. The processed vital sign data represents the deviation degree of the data to be processed compared with the median of the range. The larger the processed vital sign data, the greater the deviation degree. When the processed vital sign data exceeds 1 and is larger, the greater the abnormality.

[0059] It should be noted that using the difference between the median value and the minimum value of the range corresponding to the data to be processed as the denominator, or using half of the difference between the maximum value and the minimum value of the range as the denominator, has the same effect as using the difference between the maximum value and the median value of the range corresponding to the data to be processed as the denominator, and the implementer can replace it at will; the sorting order of the physiological indicator dimensions in the vector does not affect the final processing result, and the implementer can set a fixed sorting order at will; the preset normal range of physiological indicators is the normal range of the monitored physiological indicator dimensions, such as the normal range of urea nitrogen is 2.9-8.2mmol / L, the normal range of blood potassium is 3.5-5.1mmol / L, and the normal range of blood sodium is 135-145mmol / L. The normal ranges of other physiological indicators can be obtained by consulting the hospital database or existing information, which will not be repeated here.

[0060] Step S2: Cluster the standard vital sign vectors, segment the time domain according to the clustering results, and obtain time series subsegments; obtain abnormal subsegments according to the overall characteristics of the standard vital sign vectors in each time series subsegment; obtain the abnormal factor of each abnormal subsegment according to the number and abnormal characteristics of the standard vital sign vectors in each abnormal subsegment.

[0061] Under normal circumstances, the physiological data of the vital signs of patients in the Department of Nephrology should show a steady change over time. If these data suddenly change, that is, suddenly increase or decrease, it usually indicates that the patient may have some health problems or complications, so it is necessary to first distinguish the abnormal time period when the complications occur. Considering that clustering can classify similar data into the same category and distinguish different data patterns, the standard vital sign vectors are clustered, and the time domain is segmented according to the clustering results to obtain time series sub-segments, which is convenient for identifying different vital sign physiological states of patients; considering the overall characteristics of the standard vital sign vectors in the time series sub-segments, abnormal sub-segments that are significantly different from the normal mode can be identified to reduce the interference of single data abnormalities, so according to the overall characteristics of the standard vital sign vectors in each time series sub-segment, abnormal sub-segments are obtained.

[0062] Preferably, in one embodiment of the present invention, the Euclidean distance between standard vital sign vectors is used as the clustering distance, and the DBSCAN clustering algorithm is used to cluster the time-series continuous and similar standard vital sign vectors into one category, and finally a cluster is regarded as a time-series sub-segment.

[0063] It should be noted that the Euclidean distance between vectors and the DBSCAN clustering algorithm are both existing technologies. In other embodiments of the present invention, the implementer may also use other distance methods such as Manhattan distance to obtain clustering distance for clustering, or may use other clustering algorithms such as K-means clustering algorithm to segment the time domain and obtain time series sub-segments.

[0064] Preferably, in one embodiment of the present invention, considering that the average value represents the overall concentrated characteristics of the data, the average value vector of each time series sub-segment is obtained as the comprehensive sub-vector of each time series sub-segment to represent the overall characteristics of the standard vital sign vector in the time series sub-segment; considering that the larger the data in the comprehensive sub-vector, the greater the degree of deviation between the vital sign physiological data and the median of the corresponding range, so when there is data greater than the first preset threshold in the comprehensive sub-vector, the corresponding time series sub-segment is judged as an abnormal sub-segment.

[0065] As an example, the first preset threshold is a constant of 1; when the data in the comprehensive sub-vector is less than or equal to 1, it indicates that the overall vital sign and physiological data in the time series sub-segment are within the normal range, and it is a normal sub-segment; when the data in the comprehensive sub-vector is greater than 1, the overall vital sign and physiological data in the time series sub-segment deviate from the normal range, and it is determined to be an abnormal sub-segment.

[0066] It should be noted that the average value vector of multiple vectors is a vector composed of the average values ​​of each dimension of the vectors; in other embodiments of the present invention, the implementer may also obtain the weighted result of the average value, median and mode of each dimensional data as a comprehensive sub-vector, such as weighted summing the average value, median and mode with weights of 0.5, 0.25, and 0.25. Similarly, when obtaining the comprehensive vector subsequently, the weighted result of the average value, median and mode of each dimensional data may also be used as the comprehensive vector.

[0067] Considering that the number of standard sign vectors represents the duration of the patient's physical abnormality, and the abnormal characteristics can analyze the severity of the patient's physical abnormality, the abnormal factor of each abnormal sub-segment is obtained according to the number and abnormal characteristics of the standard sign vectors in each abnormal sub-segment, and the patient's health status in the time period corresponding to the abnormal sub-segment is preliminarily quantified, providing a basis for the subsequent integration of multiple consecutive abnormal sub-segments to obtain the corrected abnormal factor.

[0068] Preferably, in one embodiment of the present invention, considering that in the comprehensive sub-vector, the more the number of data greater than the first preset threshold value is, the more physiological index dimensions of the overall physical sign physiological data exceeding the normal range are in the abnormal sub-segment, and the more obvious the abnormal characteristics of the patient are; considering that the larger the maximum data value in the comprehensive sub-vector is, the greater the degree to which the overall physical sign physiological data exceeds the normal range is, and the more obvious the abnormal characteristics of the patient are; at the same time, the more the number of standard physical sign vectors is, the longer the duration of the abnormal sub-segment is, the more the duration of the patient's physical data abnormality is, and the more obvious the abnormal characteristics of the patient are;

[0069] Based on this, according to the number of data greater than the first preset threshold in the comprehensive sub-vector of the abnormal sub-segment, combined with the maximum data value in the comprehensive sub-vector, the abnormal characteristic coefficient of the corresponding abnormal sub-segment is obtained; the number of data greater than the first preset threshold and the maximum data value are both positively correlated with the abnormal characteristic coefficient;

[0070] The product of the number of standard sign vectors in the abnormal sub-segment and the abnormal characteristic coefficient is taken as the abnormal factor of the corresponding abnormal sub-segment.

[0071] As an example, the calculation formula of the anomaly factor includes:

[0072] ;

[0073] in, Indicates the sequence number of the abnormal sub-segment; Indicates The abnormal factor of the abnormal sub-segment; represents the linear normalization function; Indicates The number of data in the comprehensive sub-vector of the abnormal sub-segments is greater than a first preset threshold; Indicates The maximum data value in the integrated sub-vector of the abnormal sub-segments; Indicates The number of standard sign vectors of abnormal sub-segments; Indicates The abnormal characteristic coefficient of the abnormal sub-segment.

[0074] In the calculation formula of the abnormal factor, the and , and with the help of normalization function Normalization is performed to express the abnormal characteristics of the standard sign vector in the abnormal sub-segment from the perspective of the number of physiological index dimensions that exceed the normal range as a whole and the maximum degree of excess, and the abnormal characteristic coefficient is obtained. Make supplementary adjustments to obtain abnormal factors. The larger the abnormal factor, the more abnormal the patient's vital signs and physiological data.

[0075] It should be noted that in other embodiments of the present invention, the implementer may also use the natural constant Exponential function with base ,Adjustment or The sensitivity, such as .

[0076] Step S3: Merge adjacent time series sub-segments of the same type to obtain normal sequence segments and abnormal sequence segments; obtain the corrected abnormal factor of each abnormal sequence segment according to the number of abnormal sub-segments in each abnormal sequence segment, the fluctuation characteristics and statistical characteristics of the abnormal factor; obtain the deterioration factor of each abnormal sequence segment according to the difference characteristics between the standard eigenvector in each abnormal sequence segment and the standard eigenvector in the adjacent normal sequence segment, combined with the corrected abnormal factor; obtain the corrected deterioration factor of each abnormal sequence segment according to the difference characteristics between each abnormal sequence segment and the adjacent normal sequence segment, combined with the deterioration factor.

[0077] Taking into account the differences between patients' complications at different stages, the same patient's normal vital signs and physiological data may have different health patterns over time. The same type of time domain segments will be divided into different time series sub-segments. Adjacent time series sub-segments of the same type may represent similar physiological states. Merging adjacent time series sub-segments of the same type is conducive to analyzing the changes in patients' vital signs and physiological data from a longer time perspective, which is convenient for ultimately assigning more appropriate weights and obtaining accurate prediction results.

[0078] It should be noted that when the time sequence sub-segments are merged, the single abnormal sub-segment is directly used as the abnormal sequence segment; the single normal sub-segment is directly used as the normal sequence segment.

[0079] Considering that the abnormal factors cannot directly characterize the abnormal degree of the abnormal sequence segment, it is necessary to further process the abnormal factors of the abnormal sub-segments contained in the abnormal sequence segment; considering the number of abnormal sub-segments in the abnormal sequence segment and the fluctuation characteristics of the abnormal factors, the fluctuation of the patient's condition in the abnormal sequence segment is reflected. At the same time, the statistical characteristics of the abnormal factors reflect the patient's condition in the abnormal sequence segment from the overall perspective of all abnormal factors. Therefore, according to the number of abnormal sub-segments in each abnormal sequence segment, the fluctuation characteristics and statistical characteristics of the abnormal factors, the corrected abnormal factors of each abnormal sequence segment are obtained to provide a basis for the subsequent analysis of the patient's condition deterioration characteristics.

[0080] Preferably, in one embodiment of the present invention, considering that the more violent the fluctuation of the corresponding abnormal factor in the abnormal sequence segment is, the more the number of abnormal sub-segments is, indicating that the abnormal sequence segment contains more abnormal patterns of vital sign physiological data, the more factors leading to the change of the disease condition, the more complex the patient's pathology is, and the larger the correction abnormal factor is; at the same time, considering that the larger the sum of all abnormal factors is, the greater the degree to which the vital sign physiological data exceeds the normal range of physiological indicators, and the more serious the patient's disease condition is;

[0081] Based on this, any abnormal sequence segment is selected as the target sequence segment; according to the violent fluctuation characteristics of the abnormal factors of the abnormal sub-segment in the target sequence segment, the violent fluctuation parameter is obtained; and the sum of all abnormal factors of the target sequence segment is obtained as the statistical sum value;

[0082] According to the number, statistical sum and violent fluctuation parameters of abnormal sub-segments in the target sequence segment, the corrected abnormal factor of the target sequence segment is obtained; the number, statistical sum and violent fluctuation parameters of the abnormal sub-segments are positively correlated with the corrected abnormal factor.

[0083] As an example, the calculation formula for the correction anomaly factor includes:

[0084] ;

[0085] in, Indicates the sequence number of the abnormal sequence segment; Indicates Correction anomaly factor for each abnormal sequence segment; represents the linear normalization function; Indicates The parameters of the abnormal sequence segment with sharp fluctuations; Indicates The number of abnormal sub-segments of an abnormal sequence segment; Indicates the sequence number of the abnormal sub-segment; Indicates The first The abnormal factor of the abnormal sub-segment.

[0086] In the calculation formula of the modified abnormal factor, the variance is used to express the violent fluctuation characteristics of the abnormal factor, and the variance of the abnormal factor of the abnormal sub-segment in the abnormal sequence segment is used as , the fluctuation characteristics of abnormal factors are expressed by variance. The larger the variance, the more drastic the data fluctuation. At the same time, the more abnormal sub-segments there are, the more abnormal patterns of physiological data contained in the abnormal sequence segment, the more factors leading to changes in the condition, the more complex the patient's pathology, and the larger the correction abnormal factor. Therefore, by right Make supplementary adjustments; by obtaining the sum value, the statistical characteristics of the abnormal factors are expressed. The larger the statistical sum value, the greater the degree to which the physiological data of the physical signs exceed the normal range of the physiological indicators, and the larger the correction abnormal factor.

[0087] It should be noted that when When the term is 0, due to is a positive number, so at this time reaches the minimum value, is 0; in other embodiments of the present invention, the implementer may also obtain fluctuation characteristic parameters such as absolute mean deviation or range as fluctuation severity parameters.

[0088] Taking into account that by comparing the differences between adjacent abnormal sequence segments and normal sequence segments, changes in the patient's physiological state can be identified, thereby effectively monitoring health risks, and the corrected abnormal factor reflects the severity of the patient's condition in the abnormal sequence segment, the deterioration factor of each abnormal sequence segment is obtained based on the difference characteristics between the standard sign vector in each abnormal sequence segment and the standard sign vector in the adjacent normal sequence segment, combined with the corrected abnormal factor, to quantify the actual impact of the abnormality on the patient's health status, and provide data support for subsequently assigning appropriate weights to daily sign physiological data and obtaining accurate predicted physiological data.

[0089] Preferably, in one embodiment of the present invention, firstly, the average value vector of each normal sequence segment is obtained as the comprehensive vector of each normal sequence segment; the average value vector of each abnormal sequence segment is obtained as the comprehensive vector of each abnormal sequence segment; the comprehensive vector is used to represent the overall characteristics of each physiological indicator dimension of the sequence segment;

[0090] Considering that the greater the difference between the comprehensive vector of the abnormal sequence segment and the comprehensive vector of the adjacent normal sequence segment, the greater the abnormal change of the data in the physiological indicator dimension of the abnormal sequence segment, the greater the deviation of the patient's health status, and the greater the deterioration factor; the larger the corrected abnormal factor, the more abnormal patterns the patient has, the more complex the pathology, the more obvious the characteristics of the patient's condition deterioration, and the greater the deterioration factor; therefore, according to the difference characteristics of the comprehensive vector of each abnormal sequence segment and the comprehensive vector of the adjacent normal sequence segment, combined with the corresponding corrected abnormal factor, the deterioration factor of each abnormal sequence segment is obtained; the difference characteristics of the comprehensive vector and the corrected abnormal factor are both positively correlated with the deterioration factor.

[0091] As an example, the calculation formula of the deterioration factor includes:

[0092] ;

[0093] in, Indicates the sequence number of the abnormal sequence segment; Indicates Correction anomaly factor for each abnormal sequence segment; Indicates The deterioration factor of the abnormal sequence segment; Indicates The Euclidean distance between the comprehensive vector of the abnormal sequence segment and the comprehensive vector of the left adjacent normal sequence segment; Indicates The Euclidean distance between the comprehensive vector of the abnormal sequence segment and the comprehensive vector of the right adjacent normal sequence segment.

[0094] In the calculation formula of the deterioration factor, the positive direction of the time domain is on the right and the negative direction of the time domain is on the left; the difference characteristics between the comprehensive vectors are expressed by the Euclidean distance between the vectors. The larger the Euclidean distance, the greater the difference characteristics between the comprehensive vectors, the greater the change in the vital sign and physiological data between the abnormal sequence segment and the adjacent normal sequence segment, and the larger the deterioration factor; the larger the corrected abnormal factor, the more abnormal patterns the patient has, the more complex the pathology, the more obvious the characteristics of the patient's condition deterioration, and the larger the deterioration factor.

[0095] It should be noted that, when an abnormal sequence segment has only one adjacent normal sequence segment, the corresponding Euclidean distance between the abnormal sequence segment and the adjacent normal sequence segment is directly multiplied by the corrected abnormal factor, and the product is used as the deterioration factor; when there is only one abnormal sequence segment, since all the vital sign and physiological data are abnormal data of a certain problem or complication of the patient, the prediction model can be directly used to obtain the prediction data; in other embodiments of the present invention, other distance methods such as Manhattan distance can also be used to represent the difference characteristics between the comprehensive vectors.

[0096] In the embodiment of the present invention, it is normal to have small fluctuations in physiological indicators of multiple dimensions, and the physiological parameters of the human body will have natural fluctuations in a short period of time, and may be affected by factors such as physiological rhythm, diet, drinking water, drug effects and physical activity. Therefore, it is necessary to further analyze the difference characteristics between the abnormal sequence segment and the adjacent normal sequence segment, adjust the deterioration factor, and obtain the final corrected deterioration factor for subsequent weighting of abnormal vital sign physiological data to improve the sensitivity and prediction accuracy of the ARIMA model for abnormal data. Therefore, according to the difference characteristics between each abnormal sequence segment and the adjacent normal sequence segment, combined with the deterioration factor, the corrected deterioration factor of each abnormal sequence segment is obtained to reduce the impact of normal small fluctuations of multiple parameters of the human body.

[0097] Preferably, in one embodiment of the present invention, the method for obtaining the modified deterioration factor includes: Figure 2 , which shows a flow chart of a method for obtaining a modified deterioration factor provided by an embodiment of the present invention, specifically comprising:

[0098] Step S301: Mark normal sequence segments adjacent to the target sequence segment as a first sequence segment and a second sequence segment respectively.

[0099] It should be noted that, when the target sequence segment has only one adjacent normal sequence segment, only the first sequence segment needs to be marked for acquisition.

[0100] Step S302: In the comprehensive vector of the target sequence segment and the comprehensive vector of the first sequence segment, calculate the absolute value of the difference of each preset physiological indicator dimension to obtain a first difference vector; in the first difference vector, select the physiological indicator with the largest value of the second preset number as the first target physiological indicator, and the others as the first reference physiological indicators; according to the difference between the numerical mean of the first target physiological indicator in the first difference vector and the numerical mean of the first reference physiological indicator, the numerical mean of the first target physiological indicator in the first difference vector and the data mean of the first target physiological indicator in the comprehensive vector of the target sequence segment, obtain the first deterioration correction coefficient of the target sequence segment.

[0101] Among them, the difference between the numerical mean of the first target physiological indicator in the first difference vector and the numerical mean of the first reference physiological indicator, the numerical mean of the first target physiological indicator in the first difference vector, and the data mean of the first target physiological indicator in the comprehensive vector of the target sequence segment are all positively correlated with the first deterioration correction coefficient; the second preset number is less than the total number of preset physiological indicator dimensions.

[0102] In order to avoid the difference between the abnormal sequence segment and the adjacent normal sequence segment being a normal small fluctuation of multiple parameters of the human body, the absolute value of the difference between the comprehensive vector of the target sequence segment and the comprehensive vector of the first sequence segment in each preset physiological indicator dimension is first obtained, and a first difference vector is formed. The data value in the first difference vector is the absolute value of the difference between the comprehensive vector of the target sequence segment and the comprehensive vector of the first sequence segment in each preset physiological indicator dimension. The difference characteristics between the target sequence segment and the first sequence segment are analyzed in more detail through the first difference vector.

[0103] Taking into account the normal small fluctuations of multiple parameters of the human body, the data values ​​in the first difference vector are small and more concentrated. At this time, the data values ​​with the largest values ​​are small, and the data values ​​with the largest values ​​are closer to other data values. Therefore, the first deterioration correction coefficient of the target sequence segment is obtained according to the difference between the numerical mean of the first target physiological indicator in the first difference vector and the numerical mean of the first reference physiological indicator, the numerical mean of the first target physiological indicator in the first difference vector and the data mean of the first target physiological indicator in the comprehensive vector of the target sequence segment.

[0104] As an example, the second preset number is 3, and the calculation formula of the first deterioration correction coefficient includes:

[0105] ;

[0106] in, Indicates the sequence number of the abnormal sequence segment; Indicates The first deterioration correction factor of the abnormal sequence segment; Indicates The numerical mean of the first target physiological index in the first difference vector of the abnormal sequence segments; Indicates The numerical mean of the first reference physiological index in the first difference vector of the abnormal sequence segments; Indicates The data mean of the first target physiological indicator in the comprehensive vector of the abnormal sequence segments.

[0107] In the calculation formula of the first deterioration correction coefficient, due to is the average of the three data with the largest values ​​in the first difference vector, so Not less than , therefore, by taking the difference, the difference between the numerical mean of the first target physiological index in the first difference vector and the numerical mean of the first reference physiological index is expressed. The smaller the difference is, the more concentrated the data values ​​in the first difference vector are; The smaller it is, the smaller the data value in the first difference vector is, which means that the target sequence segment is more likely to be a normal small fluctuation of multiple parameters of the human body, and the smaller the first deterioration correction coefficient is; and The smaller it is, the smaller the deviation of the first target physiological indicator from the normal range in the comprehensive vector of the target sequence segment is, the smaller the abnormality is, and the smaller the first deterioration correction coefficient is.

[0108] Step S303: acquiring a second degradation correction coefficient of the target sequence segment according to the comprehensive vector of the second sequence segment and the target sequence segment.

[0109] Similarly, the second deterioration correction coefficient is obtained: the calculation formula of the second deterioration correction coefficient includes:

[0110] ;

[0111] in, Indicates the sequence number of the abnormal sequence segment; Indicates The second deterioration correction factor of the abnormal sequence segment; Indicates The numerical mean of the second target physiological index in the second difference vector of the abnormal sequence segments; Indicates The numerical mean of the second reference physiological index in the second difference vector of the abnormal sequence segments; Indicates The data mean of the second target physiological index in the comprehensive vector of the abnormal sequence segments.

[0112] Step S304: acquiring a corrected degradation factor of the target sequence segment according to the first degradation correction coefficient, the second degradation correction coefficient and the degradation factor of the target sequence segment.

[0113] Among them, the first deterioration correction coefficient, the second deterioration correction coefficient and the deterioration factor are all positively correlated with the modified deterioration factor;

[0114] As an example, the calculation formula for the modified deterioration factor includes:

[0115] ;

[0116] in, Indicates the sequence number of the abnormal sequence segment; Indicates The deterioration factor of the abnormal sequence segment; No. Corrected deterioration factor for each abnormal sequence segment; Indicates The first deterioration correction factor of the abnormal sequence segment; Indicates The second deterioration correction factor for each abnormal sequence segment.

[0117] In the calculation formula of the corrected deterioration factor, the larger the first deterioration correction coefficient or the second deterioration correction coefficient is, the less likely the target sequence segment is to be a normal small fluctuation of multiple parameters of the human body, the greater the correction weight given to the deterioration factor, and the larger the corrected deterioration factor is.

[0118] It should be noted that, in other embodiments of the present invention, the implementer may also adjust the calculation formula of the first degradation correction coefficient by means of a related mapping function, for example: ; Through the normalization function, the value range of the first deterioration correction coefficient is adjusted, and finally the correction range of the correction deterioration factor is adjusted.

[0119] Step S4: According to the type of sequence segment at the end of the time domain, combined with the modified deterioration factor and the similarity and time difference between different abnormal sequence segments, the weight of daily vital sign physiological data is obtained; based on the weight of daily vital sign physiological data, the ARIMA model is used to obtain predicted physiological data.

[0120] Taking into account the type of sequence segment at the end of the time domain, it reflects the patient's current condition and has a direct impact on the calculation of the weight; and the modified deterioration factor represents the degree of deterioration of the abnormal sequence segment. At the same time, the similarity between different abnormal sequence segments reflects the similar characteristics of the patient's condition deterioration. Taking into account that the data has a certain timeliness, the weight of the daily vital sign physiological data is obtained according to the type of sequence segment at the end of the time domain, combined with the modified deterioration factor and the similarity and time difference between different abnormal sequence segments, to provide a basis for finally obtaining the predicted physiological data.

[0121] Preferably, in one embodiment of the present invention, the weight of each day's physical sign and physiological data in the normal sequence segment is set to a third preset positive value; based on the third preset positive value, a higher weight is given to the data in the abnormal sequence segment;

[0122] Considering that when the end of the time domain is a normal sequence segment, the current patient's body is in a normal state, and the abnormal sequence segments are all historical data of the patient, the weight can be directly obtained by correcting the deterioration factor. Based on this, when the end of the time domain is a normal sequence segment, the normalized value of the corrected deterioration factor of the abnormal sequence segment and the sum of the third preset positive value are used as the weight of the daily physical sign and physiological data in the corresponding abnormal sequence segment;

[0123] Considering that when the end of the time domain is an abnormal sequence segment, the current patient's body is in an abnormal state, and the patient may repeatedly experience worsening of the condition. At this time, the more similar the other abnormal sequence segments are to the abnormal sequence segment at the end of the time series and the shorter the time interval is, the more important they are for predicting the current patient's physical data. Therefore, when the end of the time domain is an abnormal sequence segment, the abnormal sequence segment at the end of the time domain is used as the current state sequence segment; according to the similar characteristics of the vital sign physiological data of each abnormal sequence segment and the current state sequence segment in all the same physiological indicator dimensions, combined with the time difference between the abnormal sequence segment and the current state sequence segment, and the corrected deterioration factor, the weight factor of each abnormal sequence segment is obtained; the normalized value of the weight factor and the sum of the third preset positive value are used as the weight of the vital sign physiological data of each day in the corresponding abnormal sequence segment;

[0124] The similar characteristics, time difference and modified deterioration factor of all physical sign physiological data on the same physiological indicator dimension are positively correlated with the weight factor.

[0125] As an example, the DTW (Dynamic Time Warping) algorithm is used to analyze the similarity of the physiological data of the abnormal sequence segment and the current sequence segment on the same physiological indicator dimension; the calculation formula of the weight factor includes:

[0126] ;

[0127] in, Indicates the sequence number of the abnormal sequence segment; Indicates The weight factor of each abnormal sequence segment; represents the linear normalization function; Expressed as a natural constant The exponential function with base ; represents the number of preset physiological indicator dimensions; n represents the sequence number of preset physiological indicator dimensions; Indicates The first abnormal sequence segment Physiological data of preset physiological indicators; Indicates the current sequence segment Physiological data of preset physiological indicators; represents the DTW distance function; Indicates The time difference between the abnormal sequence segment and the current sequence segment; Indicates Corrected deterioration factor for each abnormal sequence segment.

[0128] In the calculation formula of the weight factor, the average value of the DTW distance between the physiological data of the current sequence segment and the abnormal sequence segment on the same physiological indicator dimension is obtained to indicate the similarity of the physiological data of the abnormal sequence segment and the current sequence segment on all the same physiological indicator dimensions. The smaller the DTW distance, the more similar the two time series data sequences are, the higher the possibility of disease recurrence, the higher the reference value, and the larger the weight factor; the smaller the time difference between the abnormal sequence segment and the current sequence segment, the closer the time of the two disease deteriorations is, the more they are in the same treatment stage, the higher the reference value, and the larger the weight factor; The larger it is, the more serious the deterioration of the abnormal sequence segment is, the more important it is for predicting the patient's condition change, and the larger the weight factor is.

[0129] It should be noted that, in one embodiment of the present invention, the third preset positive value is 1. When the end of the time domain is an abnormal sequence segment, the weight factor of the current sequence segment is 1, and the weight of the daily vital sign physiological data in the current sequence segment is 2. In other embodiments of the present invention, the implementer can also perform negative correlation mapping by taking the inverse, such as ,in Indicates the preset division by zero positive parameter, ; The implementer can also use the Euclidean distance between the comprehensive vectors to replace the DTW distance to obtain the weight factor; the DTW algorithm is already an existing technology and will not be described in detail here.

[0130] Now that we have obtained the weights of the patient's daily vital signs and physiological data, we can use the ARIMA model to obtain predicted physiological data.

[0131] It should be noted that the ARIMA model and weighted processing of the ARIMA model are already existing technologies. In one embodiment of the present invention, by predicting the data of physiological indicators in each dimension, the patient's vital sign physiological data of all preset physiological indicator dimensions in the next day are obtained, and the predicted data is transmitted to relevant personnel to provide more data support for the relevant personnel, thereby improving the effect of monitoring the vital sign physiological data of nephrology patients.

[0132] An embodiment of the present invention also provides a system for monitoring vital signs and physiological data of patients in nephrology, the system comprising a memory, a processor and a computer program, wherein the memory is used to store the corresponding computer program, the processor is used to run the corresponding computer program, and when the computer program runs in the processor, it can implement a method for monitoring vital signs and physiological data of patients in nephrology described in steps S1-S4.

[0133] One embodiment of the present invention further provides a medical device for monitoring vital signs and physiological data of patients in the Department of Nephrology, the medical device comprising a data acquisition module, a preliminary analysis module, a deep analysis module and a prediction module;

[0134] The data acquisition module is used to obtain the physiological data of all preset physiological indicators of patients in the Department of Nephrology every day; according to the distribution characteristics of the daily physiological data compared with the normal range of the preset physiological indicators, the daily standard physiological vector is obtained;

[0135] The preliminary analysis module is used to cluster the standard vital sign vectors, segment the time domain according to the clustering results, and obtain time series sub-segments; obtain abnormal sub-segments according to the overall characteristics of the standard vital sign vectors in each time series sub-segment; obtain the abnormal factor of each abnormal sub-segment according to the number and abnormal characteristics of the standard vital sign vectors in each abnormal sub-segment;

[0136] A deep analysis module is used to merge adjacent similar time series sub-segments to obtain normal sequence segments and abnormal sequence segments; obtain the corrected abnormal factor of each abnormal sequence segment according to the number of abnormal sub-segments in each abnormal sequence segment, the fluctuation characteristics and statistical characteristics of the abnormal factor; obtain the deterioration factor of each abnormal sequence segment according to the difference characteristics between the standard eigenvector in each abnormal sequence segment and the standard eigenvector in the adjacent normal sequence segment, combined with the corrected abnormal factor; obtain the corrected deterioration factor of each abnormal sequence segment according to the difference characteristics between each abnormal sequence segment and the adjacent normal sequence segment, combined with the deterioration factor;

[0137] The prediction module is used to obtain the weight of daily vital sign physiological data according to the type of sequence segment at the end of the time domain, combined with the modified deterioration factor and the similarity and time difference between different abnormal sequence segments; based on the weight of daily vital sign physiological data, the ARIMA model is used to obtain predicted physiological data.

[0138] In summary, in order to solve the technical problem that the ARIMA model has low sensitivity to abnormal data of patients in the Department of Nephrology and the prediction results are not ideal, the present invention proposes a method, system and medical equipment for monitoring the vital signs and physiological data of patients in the Department of Nephrology. The present invention first obtains the standard vital signs vectors of patients in the Department of Nephrology every day; further clusters the standard vital signs vectors to obtain time series sub-segments, and screens out abnormal sub-segments; further obtains the abnormal factors of each abnormal sub-segment according to the number and abnormal characteristics of the standard vital signs vectors in each abnormal sub-segment; further merges adjacent similar time series sub-segments to obtain normal sequence segments and abnormal sequence segments; further obtains the corrected deterioration factor of each abnormal sequence segment according to the number of abnormal sub-segments in each abnormal sequence segment, the fluctuation characteristics and statistical characteristics of the abnormal factors, combined with the difference characteristics between each abnormal sequence segment and the adjacent normal sequence segment; finally obtains the weight of daily vital signs and physiological data, and uses the ARIMA model to obtain predicted physiological data. The present invention identifies the abnormal physical signs and physiological data of the patient's history and analyzes the degree of abnormality, analyzes the deterioration of the patient at different historical stages, and ultimately gives a greater weight to the abnormal data, thereby improving the sensitivity of the ARIMA model to abnormal data and providing more accurate predicted physiological data for relevant personnel.

[0139] It should be noted that the sequence of the above embodiments of the present invention is only for description and does not represent the advantages and disadvantages of the embodiments. The processes depicted in the accompanying drawings do not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0140] The various embodiments in this specification are described in a progressive manner, and the same or similar parts between the various embodiments can be referenced to each other, and each embodiment focuses on the differences from other embodiments.

Claims

1. A method for monitoring vital signs and physiological data of patients in nephrology, characterized in that: The method comprises: Obtaining the physical sign and physiological data of all preset physiological index dimensions of patients in the Department of Nephrology every day; obtaining the standard physical sign vector every day according to the distribution characteristics of the physical sign and physiological data compared with the normal range of the preset physiological index; Clustering the standard vital sign vectors, segmenting the time domain according to the clustering results to obtain time series subsegments; obtaining abnormal subsegments according to the overall characteristics of the standard vital sign vectors in each of the time series subsegments; obtaining the abnormal factor of each of the abnormal subsegments according to the number and abnormal characteristics of the standard vital sign vectors in each of the abnormal subsegments; Merging adjacent time sequence sub-segments of the same type to obtain normal sequence segments and abnormal sequence segments; obtaining a corrected abnormal factor for each abnormal sequence segment according to the number of abnormal sub-segments in each abnormal sequence segment, the fluctuation characteristics and statistical characteristics of the abnormal factor; obtaining a deterioration factor for each abnormal sequence segment according to the difference characteristics between the standard eigenvector in each abnormal sequence segment and the standard eigenvector in the adjacent normal sequence segment, combined with the corrected abnormal factor; obtaining a corrected deterioration factor for each abnormal sequence segment according to the difference characteristics between each abnormal sequence segment and the adjacent normal sequence segment, combined with the deterioration factor; According to the type of the sequence segment at the end of the time domain, combined with the modified deterioration factor and the similarity and time difference between different abnormal sequence segments, the weight of the physical sign physiological data of each day is obtained; based on the weight of the physical sign physiological data of each day, the predicted physiological data is obtained using the ARIMA model; The method for obtaining the standard sign vector includes: Select any vital sign physiological data of any physiological indicator dimension as the data to be processed; based on the preset normal range of the physiological indicator, obtain the maximum value and the median value of the range; use the absolute value of the difference between the data to be processed and the corresponding median value of the range as the numerator, use the difference between the maximum value and the median value of the range corresponding to the data to be processed as the denominator, and use the fraction as the processed vital sign data of the data to be processed; The processed vital sign data of all dimensions of each day are used to form a standard vital sign vector of each day; the order of the physiological index dimensions in all the standard vital sign vectors is the same; The method for obtaining the abnormal sub-segment includes: Acquire the average value vector of each of the time series sub-segments as the comprehensive sub-vector of each of the time series sub-segments; when there is data greater than a first preset threshold value in the comprehensive sub-vector, determine the corresponding time series sub-segment as an abnormal sub-segment; The method for obtaining the abnormal factor includes: According to the number of data greater than a first preset threshold in the comprehensive sub-vector of the abnormal sub-segment, combined with the maximum data value in the comprehensive sub-vector, an abnormal characteristic coefficient corresponding to the abnormal sub-segment is obtained; the number of data greater than the first preset threshold and the maximum data value are both positively correlated with the abnormal characteristic coefficient; The product of the number of the standard sign vectors in the abnormal sub-segment and the abnormal characteristic coefficient is used as the abnormal factor corresponding to the abnormal sub-segment; The method for obtaining the modified abnormal factor includes: Select any of the abnormal sequence segments as a target sequence segment; obtain a fluctuation intensity parameter according to the fluctuation intensity characteristics of the abnormal factors of the abnormal sub-segment in the target sequence segment; obtain the sum of all the abnormal factors of the target sequence segment as a statistical sum value; According to the number of the abnormal sub-segments in the target sequence segment, the statistical sum value and the violent fluctuation parameter, a corrected abnormal factor of the target sequence segment is obtained; the number of the abnormal sub-segments, the statistical sum value and the violent fluctuation parameter are all positively correlated with the corrected abnormal factor; The calculation formula of the abnormal factor includes: ； in, Indicates the sequence number of the abnormal sub-segment; Indicates The abnormal factor of the abnormal sub-segment; represents the linear normalization function; Indicates The number of data in the comprehensive sub-vector of the abnormal sub-segments is greater than a first preset threshold; Indicates The maximum data value in the integrated sub-vector of the abnormal sub-segments; Indicates The number of standard sign vectors of abnormal sub-segments; Indicates The abnormal characteristic coefficient of the abnormal sub-segment; The method for obtaining the deterioration factor includes: Obtaining the average value vector of each of the normal sequence segments as the comprehensive vector of each of the normal sequence segments; obtaining the average value vector of each of the abnormal sequence segments as the comprehensive vector of each of the abnormal sequence segments; According to the difference characteristics of the comprehensive vector of each abnormal sequence segment and the comprehensive vector of the adjacent normal sequence segment, combined with the corresponding corrected abnormal factor, the deterioration factor of each abnormal sequence segment is obtained; the difference characteristics of the comprehensive vector and the corrected abnormal factor are both positively correlated with the deterioration factor; The method for obtaining the modified deterioration factor includes: Marking the normal sequence segments adjacent to the target sequence segment as a first sequence segment and a second sequence segment respectively; In the comprehensive vector of the target sequence segment and the comprehensive vector of the first sequence segment, the absolute value of the difference of each preset physiological indicator dimension is calculated to obtain a first difference vector; in the first difference vector, the physiological indicators with the largest values ​​of the second preset number are selected as the first target physiological indicators, and the others are selected as the first reference physiological indicators; according to the difference between the numerical mean of the first target physiological indicator in the first difference vector and the numerical mean of the first reference physiological indicator, the numerical mean of the first target physiological indicator in the first difference vector and the data mean of the first target physiological indicator in the comprehensive vector of the target sequence segment, the first deterioration correction coefficient of the target sequence segment is obtained; the difference between the numerical mean of the first target physiological indicator in the first difference vector and the numerical mean of the first reference physiological indicator, the numerical mean of the first target physiological indicator in the first difference vector and the data mean of the first target physiological indicator in the comprehensive vector of the target sequence segment are all positively correlated with the first deterioration correction coefficient; acquiring a second degradation correction coefficient of the target sequence segment according to the comprehensive vector of the second sequence segment and the target sequence segment; acquiring a modified degradation factor of the target sequence segment according to the first degradation correction coefficient, the second degradation correction coefficient and the degradation factor of the target sequence segment; the first degradation correction coefficient, the second degradation correction coefficient and the degradation factor are all positively correlated with the modified degradation factor; Wherein, the second preset number is less than the total number of preset physiological indicator dimensions; The calculation formula of the first deterioration correction factor includes: ； in, Indicates the sequence number of the abnormal sequence segment; Indicates The first deterioration correction factor of the abnormal sequence segment; Indicates The numerical mean of the first target physiological index in the first difference vector of the abnormal sequence segments; Indicates The numerical mean of the first reference physiological index in the first difference vector of the abnormal sequence segments; Indicates The data mean of the first target physiological index in the comprehensive vector of the abnormal sequence segments; The calculation formula of the second deterioration correction factor includes: ； in, Indicates the sequence number of the abnormal sequence segment; Indicates The second deterioration correction factor of the abnormal sequence segment; Indicates The numerical mean of the second target physiological index in the second difference vector of the abnormal sequence segments; Indicates The numerical mean of the second reference physiological index in the second difference vector of the abnormal sequence segments; Indicates The data mean of the second target physiological index in the comprehensive vector of the abnormal sequence segments; The method for obtaining the weight of the daily physiological data includes: Setting the weight of the physical sign physiological data of each day in the normal sequence segment to a third preset positive value; When the end of the time domain is a normal sequence segment, the sum of the normalized value of the modified deterioration factor of the abnormal sequence segment and the third preset positive value is used as the weight of the vital sign physiological data of each day in the abnormal sequence segment; When the end of the time domain is an abnormal sequence segment, the abnormal sequence segment at the end of the time domain is used as the current state sequence segment; according to the similar characteristics of the vital sign physiological data of each abnormal sequence segment and the current state sequence segment in all the same physiological indicator dimensions, combined with the time difference between the abnormal sequence segment and the current state sequence segment, and the corrected deterioration factor, a weight factor of each abnormal sequence segment is obtained; the normalized value of the weight factor and the sum of the third preset positive value are used as the weight of the vital sign physiological data of each day in the abnormal sequence segment; The similar characteristics of the vital sign physiological data in all the same physiological indicator dimensions, the time difference and the corrected deterioration factor are positively correlated with the weight factor; The calculation formula of the weight factor includes: ； in, Indicates the sequence number of the abnormal sequence segment; Indicates The weight factor of each abnormal sequence segment; represents the linear normalization function; Indicated by natural constant The exponential function with base ; represents the number of preset physiological indicator dimensions; n represents the sequence number of preset physiological indicator dimensions; Indicates The first abnormal sequence segment Physiological data of preset physiological indicators; Indicates the current sequence segment Physiological data of preset physiological indicators; represents the DTW distance function; Indicates The time difference between the abnormal sequence segment and the current sequence segment; Indicates Corrected deterioration factor for each abnormal sequence segment.

2. A system for monitoring vital signs and physiological data of patients in the Department of Nephrology, the system comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the computer program, the steps of the method for monitoring vital signs and physiological data of a nephrology patient as claimed in claim 1 are implemented.

3. A medical device for monitoring vital signs and physiological data of patients in the Department of Nephrology, characterized in that: The medical device includes a data acquisition module, a preliminary analysis module, a deep analysis module and a prediction module; The data acquisition module is used to obtain the physiological data of all preset physiological index dimensions of the patients in the Department of Nephrology every day; according to the distribution characteristics of the physiological data of the daily signs compared with the normal range of the preset physiological indexes, the standard physiological vector of each day is obtained; A preliminary analysis module is used to cluster the standard vital sign vectors, segment the time domain according to the clustering results, and obtain time series subsegments; obtain abnormal subsegments according to the overall characteristics of the standard vital sign vectors in each of the time series subsegments; and obtain the abnormal factor of each abnormal subsegment according to the number and abnormal characteristics of the standard vital sign vectors in each of the abnormal subsegments; A depth analysis module is used to merge adjacent time sequence sub-segments of the same type to obtain normal sequence segments and abnormal sequence segments; obtain a corrected abnormal factor for each abnormal sequence segment according to the number of abnormal sub-segments in each abnormal sequence segment, the fluctuation characteristics and statistical characteristics of the abnormal factor; obtain a deterioration factor for each abnormal sequence segment according to the difference characteristics between the standard eigenvector in each abnormal sequence segment and the standard eigenvector in the adjacent normal sequence segment, combined with the corrected abnormal factor; obtain a corrected deterioration factor for each abnormal sequence segment according to the difference characteristics between each abnormal sequence segment and the adjacent normal sequence segment, combined with the deterioration factor; A prediction module, used to obtain the weight of the physical sign physiological data of each day according to the type of the sequence segment at the end of the time domain, combined with the modified deterioration factor and the similarity and time difference between different abnormal sequence segments; Based on the weights of the physiological data of the physical signs mentioned every day, the ARIMA model is used to obtain the predicted physiological data; The method for obtaining the modified abnormal factor includes: Select any of the abnormal sequence segments as a target sequence segment; obtain a fluctuation intensity parameter according to the fluctuation intensity characteristics of the abnormal factors of the abnormal sub-segment in the target sequence segment; obtain the sum of all the abnormal factors of the target sequence segment as a statistical sum value; According to the number of the abnormal sub-segments in the target sequence segment, the statistical sum value and the violent fluctuation parameter, a corrected abnormal factor of the target sequence segment is obtained; the number of the abnormal sub-segments, the statistical sum value and the violent fluctuation parameter are all positively correlated with the corrected abnormal factor; The calculation formula of the abnormal factor includes: ； in, Indicates the sequence number of the abnormal sub-segment; Indicates The abnormal factor of the abnormal sub-segment; represents the linear normalization function; Indicates The number of data in the comprehensive sub-vector of the abnormal sub-segments is greater than a first preset threshold; Indicates The maximum data value in the integrated sub-vector of the abnormal sub-segments; Indicates The number of standard sign vectors of abnormal sub-segments; Indicates The abnormal characteristic coefficient of the abnormal sub-segment; The method for obtaining the deterioration factor includes: Obtaining the average value vector of each of the normal sequence segments as the comprehensive vector of each of the normal sequence segments; obtaining the average value vector of each of the abnormal sequence segments as the comprehensive vector of each of the abnormal sequence segments; According to the difference characteristics of the comprehensive vector of each abnormal sequence segment and the comprehensive vector of the adjacent normal sequence segment, combined with the corresponding corrected abnormal factor, the deterioration factor of each abnormal sequence segment is obtained; the difference characteristics of the comprehensive vector and the corrected abnormal factor are both positively correlated with the deterioration factor; The method for obtaining the modified deterioration factor includes: Marking the normal sequence segments adjacent to the target sequence segment as a first sequence segment and a second sequence segment respectively; In the comprehensive vector of the target sequence segment and the comprehensive vector of the first sequence segment, the absolute value of the difference of each preset physiological indicator dimension is calculated to obtain a first difference vector; in the first difference vector, the physiological indicators with the largest values ​​of the second preset number are selected as the first target physiological indicators, and the others are selected as the first reference physiological indicators; according to the difference between the numerical mean of the first target physiological indicator in the first difference vector and the numerical mean of the first reference physiological indicator, the numerical mean of the first target physiological indicator in the first difference vector and the data mean of the first target physiological indicator in the comprehensive vector of the target sequence segment, the first deterioration correction coefficient of the target sequence segment is obtained; the difference between the numerical mean of the first target physiological indicator in the first difference vector and the numerical mean of the first reference physiological indicator, the numerical mean of the first target physiological indicator in the first difference vector and the data mean of the first target physiological indicator in the comprehensive vector of the target sequence segment are all positively correlated with the first deterioration correction coefficient; acquiring a second degradation correction coefficient of the target sequence segment according to the comprehensive vector of the second sequence segment and the target sequence segment; acquiring a modified degradation factor of the target sequence segment according to the first degradation correction coefficient, the second degradation correction coefficient and the degradation factor of the target sequence segment; the first degradation correction coefficient, the second degradation correction coefficient and the degradation factor are all positively correlated with the modified degradation factor; Wherein, the second preset number is less than the total number of preset physiological indicator dimensions; The calculation formula of the first deterioration correction factor includes: ； in, Indicates the sequence number of the abnormal sequence segment; Indicates The first deterioration correction factor of the abnormal sequence segment; Indicates The numerical mean of the first target physiological index in the first difference vector of the abnormal sequence segments; Indicates The numerical mean of the first reference physiological index in the first difference vector of the abnormal sequence segments; Indicates The data mean of the first target physiological index in the comprehensive vector of the abnormal sequence segments; The calculation formula of the second deterioration correction factor includes: ； in, Indicates the sequence number of the abnormal sequence segment; Indicates The second deterioration correction factor of the abnormal sequence segment; Indicates The numerical mean of the second target physiological index in the second difference vector of the abnormal sequence segments; Indicates The numerical mean of the second reference physiological index in the second difference vector of the abnormal sequence segments; Indicates The data mean of the second target physiological index in the comprehensive vector of the abnormal sequence segments; The method for obtaining the weight of the daily physiological data includes: Setting the weight of the physical sign physiological data of each day in the normal sequence segment to a third preset positive value; When the end of the time domain is a normal sequence segment, the sum of the normalized value of the modified deterioration factor of the abnormal sequence segment and the third preset positive value is used as the weight of the vital sign physiological data of each day in the abnormal sequence segment; When the end of the time domain is an abnormal sequence segment, the abnormal sequence segment at the end of the time domain is used as the current state sequence segment; according to the similar characteristics of the vital sign physiological data of each abnormal sequence segment and the current state sequence segment in all the same physiological indicator dimensions, combined with the time difference between the abnormal sequence segment and the current state sequence segment, and the corrected deterioration factor, a weight factor of each abnormal sequence segment is obtained; the normalized value of the weight factor and the sum of the third preset positive value are used as the weight of the vital sign physiological data of each day in the abnormal sequence segment; The similar characteristics of the vital sign physiological data in all the same physiological indicator dimensions, the time difference and the corrected deterioration factor are positively correlated with the weight factor; The calculation formula of the weight factor includes: ； in, Indicates the sequence number of the abnormal sequence segment; Indicates The weight factor of each abnormal sequence segment; represents the linear normalization function; Indicated by natural constant The exponential function with base ; represents the number of preset physiological indicator dimensions; n represents the sequence number of preset physiological indicator dimensions; Indicates The first abnormal sequence segment Physiological data of preset physiological indicators; Indicates the current sequence segment Physiological data of preset physiological indicators; represents the DTW distance function; Indicates The time difference between the abnormal sequence segment and the current sequence segment; Indicates Corrected deterioration factor for each abnormal sequence segment.

Images