A health management integrated data collection optimization method and system

Abstract

The present invention relates to the field of data processing technology, and specifically to a health management integrated data collection optimization method and system, the method comprising: obtaining user health data of several months of user history; determining the data continuous variation index of various data sequences based on the degree of deviation between various data sequences in the blood pressure sequence and their normal data range, the data changes of physiological diseases and non-physiological diseases expressed by various data sequences, and the differences between local subsequences in various data sequences; determining the noise index in combination with the external factor sequence and the maximum data continuous variation index; and filtering the heart rate sequence, systolic pressure sequence, and diastolic pressure sequence by improving the Kalman filter algorithm based on the noise index. The present invention aims to eliminate the interference of the user's own behavior and external factors on the collected data, so as to obtain more accurate user health data and facilitate the user to evaluate the health status.

Description

Technical Field

[0001] The present invention relates to the technical field of data processing, and in particular to a health management integrated data collection optimization method and system. Background Art

[0002] With the continuous improvement of global health awareness and the rapid development of science and technology, health management has become an important part of people's daily life. With the widespread application of smart wearable devices, mobile Internet and big data technology, more and more individual physiological and living status data can be collected and recorded in real time. However, how to effectively integrate and utilize this data, accurately capture the changing trends of health conditions, and timely discover and prevent potential health problems has become a major issue that needs to be solved in modern health management. Although the existing data collection and processing solutions have made significant progress, there are still many challenges in data quality optimization, anomaly detection, and the formulation of personalized health management strategies.

[0003] Currently, although most health management platforms can collect a wealth of user physiological data, their ability to control data quality, remove noise, and identify abnormal data is limited. For example, physiological indicators such as heart rate, systolic blood pressure, and diastolic blood pressure collected by smart wearable devices are easily interfered by various factors such as individual activities and emotions, resulting in abnormal data fluctuations. In addition, existing data analysis methods often ignore the impact of user living habits on physiological indicators, such as diet, sleep, mental state, etc., and it is difficult to distinguish data anomalies caused by physiological lesions or changes in living habits. These problems make the assessment of user health status less effective. Summary of the invention

[0004] In order to solve the above technical problems, the present invention provides a health management integrated data collection optimization method and system, and the technical solutions adopted are as follows:

[0005] In a first aspect, an embodiment of the present invention provides a method for optimizing integrated health management data collection, the method comprising the following steps:

[0006] Obtaining the user's health data for several months in history; the user health data includes a blood pressure sequence and an external factor sequence;

[0007] Each data sequence in the blood pressure sequence is evenly divided to obtain local subsequences of each data sequence every day; based on the degree of deviation between each data sequence and its normal data range, the data changes of physiological diseases and non-physiological diseases represented by each data sequence, and the differences between local subsequences in each data sequence, the data continuous variation index of each data sequence in the user's blood pressure sequence for each month is determined;

[0008] The noise index of the user each month is determined by combining the variation of the data in the external factor sequence, the correlation between the data sequences in the blood pressure sequence, and the maximum continuous variation index of the data in the data sequences in the blood pressure sequence.

[0009] The noise matrix in the original Kalman gain calculation process is corrected based on the noise index, and the heart rate sequence, systolic blood pressure sequence and diastolic blood pressure sequence are filtered using the improved Kalman filter algorithm.

[0010] Preferably, the blood pressure sequence includes a systolic pressure sequence, a diastolic pressure sequence and a heart rate sequence; the external factor sequence includes a sleep state score sequence, a dietary habit score sequence, a mental state score sequence and a psychological state score sequence.

[0011] Preferably, the method for determining the data continuous variation index of each type of data sequence in the user's monthly blood pressure sequence includes:

[0012] Determine the data variation coefficient of each data series based on the degree of deviation between each data in each data series and the boundary value of its normal data range;

[0013] Based on the data changes of physiological diseases and non-physiological diseases in each data series in a short period of time, the data self-slowdown coefficient of each data series is determined;

[0014] Calculate the distances and values ​​between all local subsequences in various data sequences;

[0015] Based on the data variation coefficient, data self-relaxation coefficient, and distance and value, the data continuous variation index of each type of data series is determined; among them, the data variation coefficient and distance and value are both positively correlated with the data continuous variation index, and the data self-relaxation coefficient is negatively correlated with the data continuous variation index.

[0016] Preferably, the method for determining the data variation coefficient includes:

[0017] Based on the difference between each data in each data sequence and the boundary value of its normal data range, a deviation function is constructed;

[0018] The average value of the deviation function of all data in each data sequence is used as the first evaluation coefficient of each data sequence;

[0019] The value of the deviation function from the average value of all data in each data sequence is used as the second evaluation coefficient of each data sequence;

[0020] The data variation coefficient is determined based on the first evaluation coefficient and the second evaluation coefficient; wherein both the first evaluation coefficient and the second evaluation coefficient are positively correlated with the data variation coefficient.

[0021] Preferably, the method for constructing the deviation function includes:

[0022] Get the upper and lower boundaries of the normal data range for various data series;

[0023] When each data in each data sequence belongs to the normal data range, the value of the deviation function of each data is set to 0;

[0024] On the contrary, the value of each data exceeding the upper boundary or the lower boundary is taken as the value of the deviation function of each data.

[0025] Preferably, the method for determining the data self-restraint coefficient includes:

[0026] Set the length of the local interval of self-ease;

[0027] For each data in each data sequence, the accumulation of the difference between each data and all data within the self-slowing local interval length interval is taken as the first interval difference of each data;

[0028] Calculate the average data value of all data in each data sequence; take the result of the preset multiple of the difference between each data and the average data value as the second interval difference of each data; the preset multiple is the length of the self-slowing local interval;

[0029] The ratio of the first interval difference to the second interval difference of each data is calculated, the ratio is forward fused with the value of the deviation function of each data, and the sum of the forward fusion results of all data in each data sequence is taken as the data self-relaxation coefficient of each data sequence.

[0030] Preferably, the method for determining the noise index includes:

[0031] Merge all types of score sequences in the external factor sequence to obtain a comprehensive score sequence of external factors; obtain the coefficient of variation of the comprehensive score sequence of external factors;

[0032] The data mean of all local subsequences in each data sequence is used to form the average sequence of each data sequence;

[0033] Calculate the correlation between the average sequence of various data sequences and the comprehensive score sequence of external factors to obtain the maximum correlation;

[0034] The noise index is determined based on the coefficient of variation, the maximum correlation and the maximum data continuous variation index; wherein the coefficient of variation and the maximum correlation are both positively correlated with the noise index, and the maximum data continuous variation index is negatively correlated with the noise index.

[0035] Preferably, the method for correcting the noise matrix of the original Kalman gain by the noise index is: multiplying the noise matrix of the original Kalman gain by the noise index to obtain the noise matrix of the corrected Kalman gain.

[0036] Preferably, after filtering the heart rate sequence, systolic blood pressure sequence and diastolic blood pressure sequence, the method further includes: fusing the filtering results of various data sequences of all months, using the fusion results of various data sequences as inputs of the moving average algorithm, and outputting the corresponding moving average sequence; marking the data in the moving average sequence that exceeds its normal data range as abnormal data, and reminding the user of health conditions based on the abnormal data.

[0037] In the second aspect, an embodiment of the present invention also provides an integrated health management data collection and optimization system, comprising a memory, a processor, and a computer program stored in the memory and running on the processor, wherein the processor implements the steps of any one of the above-mentioned integrated health management data collection and optimization methods when executing the computer program.

[0038] In the above scheme, the present invention realizes the collection optimization of user health data. Compared with the existing technology, most of them directly use the collected user health data to analyze the user's health status. However, the analysis process is easily interfered by the user's other individual activities, emotions, etc. on the blood pressure data, and the distinction between changes in living habits and physiological lesions in external factors is not considered, resulting in poor evaluation of the user's health status. The present invention analyzes the heart rate, systolic pressure and diastolic pressure in the blood pressure data, compares the degree of deviation between them and the normal data range, and accurately reflects the degree to which the data deviates from the normal range; at the same time, the data changes of the user's physiological diseases and non-physiological diseases are analyzed in the data, which greatly eliminates the abnormal data generated by non-physiological diseases for the evaluation of data abnormalities. The influence of the normal; then analyze the difference between the data sequence itself in multiple local subsequences, and analyze the possibility of abnormality based on the user's own data; combined with the above three aspects, the situation of continuous abnormal changes in user data can be jointly evaluated; the present invention combines the comprehensive score of external factors and the correlation between various data sequences and external factors, calculates the noise index, and judges whether the abnormality of heart rate, systolic pressure and diastolic pressure is caused by the change of these external factors, thereby minimizing the impact of these external factors on user health data and improving the accuracy of data evaluation; finally, the noise index is applied to the noise matrix in the calculation process of the modified Kalman gain, and the abnormal data is corrected according to the size of the abnormal fluctuation of the data affected by external factors. In this way, the problem that the interference of other individual activities, emotions, etc. of the user on the blood pressure data in the process of evaluating the user's health status is not considered in the prior art, and the changes in living habits and physiological lesions in the external factors are accurately distinguished. It not only considers whether the data has continuous abnormal changes, but also can minimize the impact of external factors on the user's health data. The corrected user health data is used to obtain more accurate data evaluation results, which is conducive to accurate evaluation of the user's health status. BRIEF DESCRIPTION OF THE DRAWINGS

[0039] 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.

[0040] Figure 1 A flowchart of a health management integrated data collection optimization method provided by the present invention;

[0041] Figure 2A flow chart of a method for determining the data continuous variation index of various data sequences in the user's monthly blood pressure sequence provided by the present invention;

[0042] Figure 3 A flow chart of a method for determining a data variation coefficient provided by the present invention;

[0043] Figure 4 A flow chart of a method for constructing a deviation function provided by the present invention;

[0044] Figure 5 A flow chart of a method for determining a data self-slowing coefficient provided by the present invention;

[0045] Figure 6 This is a flow chart of the method for determining the noise index provided by 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 the specific implementation method, structure, features and effects of a health management integrated data collection optimization method and system proposed by the present invention in combination 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 specific scheme of the health management integrated data collection optimization method and system provided by the present invention is described in detail below with reference to the accompanying drawings.

[0049] An embodiment of the present invention provides a health management integrated data collection optimization method and system.

[0050] Specifically, the following health management integrated data collection optimization method is provided, please refer to Figure 1 , the method comprises the following steps:

[0051] The first step is to obtain the user's historical health data for several months; the user health data includes a blood pressure sequence and an external factor sequence.

[0052] The user's blood pressure and sleep status are monitored through the smart wearable device carried by the personal user. The blood pressure data is collected once every hour, and the blood pressure data includes systolic pressure, diastolic pressure and heart rate data. The sleep status is collected once a day, and the sleep status is assigned corresponding sleep status scores of 4, 3, 2, and 1 according to excellent, good, average, and poor respectively.

[0053] Based on the information filled out by users on the mobile health application, the user's daily eating habits, mental state and psychological state in the past month are counted on a monthly basis. The eating habits are scored as 3, 2 and 1 according to heavy oil and salt, general, and light respectively, and each meal is scored. The daily eating habit score is the average of the eating habit scores within that day; the mental state is scored as 3, 2 and 1 according to mental tension, general, and mental relaxation respectively; the psychological state is scored as 3, 2 and 1 according to high psychological pressure, general, and low psychological pressure respectively.

[0054] According to the above method, the user's daily sleep state score, eating habit score, mental state score and psychological state score are counted, and according to the time sequence, a sleep state score sequence, eating habit score sequence, mental state score sequence and psychological state score sequence are respectively formed, and the sleep state score sequence, eating habit score sequence, mental state score sequence and psychological state score sequence are used as external factor sequences; all the systolic blood pressure data, diastolic blood pressure data and heart rate data of the user in one month are respectively formed into a systolic blood pressure sequence, diastolic blood pressure sequence and heart rate sequence for one month according to the collection time, and the systolic blood pressure sequence, diastolic blood pressure sequence and heart rate sequence are used as blood pressure sequences. All blood pressure sequences and external factor sequences collected in one month together constitute the user's health data for one month.

[0055] For example Indicates user Months of user health data, including , and They are user Months of systolic blood pressure series, diastolic blood pressure series and heart rate series, , , , They are user In this way, the user's health data for n months of history is collected. In this embodiment, the value of n is 12, and the implementer can set it according to the actual situation.

[0056] At this point, the user's health data for several months in history can be obtained.

[0057] In the second step, each type of data sequence in the blood pressure sequence is evenly divided to obtain local subsequences of each type of data sequence every day; based on the degree of deviation between each type of data sequence and its normal data range, the data changes of physiological diseases and non-physiological diseases expressed by each type of data sequence, and the differences between local subsequences in each type of data sequence, the data continuous variation index of each type of data sequence in the user's blood pressure sequence each month is determined.

[0058] Among the various diseases that affect the user's health, there are some common and more harmful physiological diseases that can be monitored at home using smart wearable devices or home medical instruments. For example, cardiovascular disease can monitor the user's heart rate, systolic blood pressure and diastolic blood pressure data at home. However, due to the complexity of the equipment use environment and the diversity of factors that affect heart rate, systolic blood pressure and diastolic blood pressure data, only analyzing these data and ignoring the impact of some external factors on these data will greatly reduce the accuracy of data analysis. In order to reduce the errors caused by user health analysis based on one-sided data, it is necessary to consider more external factors, remove the errors caused by external factors, calibrate the obtained data, and improve the accuracy of data evaluation results.

[0059] Take the heart rate data in the blood pressure data as an example for analysis: Since the user's heart rate is usually in the range of 60 to 100 beats per minute under normal circumstances, although the heart rate will fluctuate depending on the individual's physical condition, it will not be too far from the upper and lower limits of the range of 60 to 100. By analyzing the changes in the user's heart rate data, systolic blood pressure data, and diastolic blood pressure data relative to their normal data range, the data changes of various data sequences in the user's blood pressure sequence can be mined. Among them, the normal data ranges for systolic and diastolic blood pressure in blood pressure data are 90 to 140 mmHg and 60 to 90 mmHg, respectively.

[0060] In order to reduce the local fluctuation or trend change of the heart rate data and more clearly reveal the inherent fluctuation characteristics and change trend of the heart rate data itself, this embodiment uses the DFA detrending algorithm for various data sequences in the blood pressure sequence to obtain various data sequences after detrending. Among them, the DFA detrending algorithm is a well-known technology and will not be described in detail.

[0061] In order to analyze the changes of various data sequences in the blood pressure sequence in a short period of time, the present invention evenly divides various data sequences in the blood pressure sequence to obtain local subsequences of each data sequence every day. In this embodiment, the length of the divided local subsequence is set to 24, and the implementer can set it according to the actual situation.

[0062] Based on the above analysis, the present invention determines the data variation coefficient of each data sequence based on the degree of deviation between each data in each data sequence and the boundary value of its normal data range, specifically:

[0063] Based on the difference between each data in each data sequence and the boundary value of its normal data range, a deviation function is constructed; the average value of the deviation function of all data in each data sequence is used as the first evaluation coefficient of each data sequence; the value of the deviation function of the average value of all data in each data sequence is used as the second evaluation coefficient of each data sequence; based on the first evaluation coefficient and the second evaluation coefficient, the data variation coefficient is determined; wherein the first evaluation coefficient and the second evaluation coefficient are both positively correlated with the data variation coefficient.

[0064] It can be understood that a positive correlation means that the dependent variable will increase as the independent variable increases, and the dependent variable will decrease as the independent variable decreases. The specific relationship can be a multiplication relationship, an addition relationship, or the idempotence of an exponential function, which is determined by actual application and is not particularly limited in the present invention.

[0065] Among them, the method for constructing the deviation function is: obtain the upper and lower boundaries of the normal data range of each data sequence; when each data in each data sequence belongs to the normal data range, set the value of the deviation function of each data to 0; otherwise, use the value of the part of each data that exceeds the upper or lower boundary as the value of the deviation function of each data.

[0066] In one embodiment of the present invention, the sum of the first evaluation coefficient and the second evaluation coefficient is used as the data variation coefficient.

[0067] In other embodiments of the present invention, other forward fusion methods such as multiplying the first evaluation coefficient by the second evaluation coefficient may be used to obtain the data variation coefficient.

[0068] It can be understood that the forward fusion is a fusion method such as addition and multiplication between data. The specific forward fusion method is determined by the implementer according to the actual situation, and the present invention does not impose any special limitation.

[0069] It should be understood that, taking heart rate data as an example, when a user has an abnormal heart rate due to an obvious physiological disease, the heart rate will be continuously too high or too low compared to the normal heart rate range, so that there will be a large difference between the upper and lower limits of the normal heart rate range, and it will continue to exist, that is, the value of the deviation function will continue to be large, that is, the first evaluation coefficient is large; and because the deviation function is continuously too high or too low, there will be a large difference between the mean heart rate and the upper and lower limits of the normal heart rate range, that is, it will cause the second evaluation coefficient to be large, thereby making the value of the data variation coefficient larger.

[0070] In the present invention, the flow chart of the method for determining the data variation coefficient is shown in the attached figure. Figure 3 The flowchart of the method for constructing the deviation function is shown in the attached Figure 4 shown.

[0071] If the user has an obvious physiological disease that causes an abnormal heart rate, the heart rate will be abnormally high or low for a long time; if the abnormal heart rate is caused by a physiological disease that does not show obvious physical symptoms in the early stage or by the user's personal behavior or state, such as exercise, short-term emotional excitement, or a sudden outburst of emotions, the data is often close to the normal data or automatically alleviates within a short period of time and drops to a normal level, which will not cause a persistent and obvious abnormal heart rate data. At the same time, systolic and diastolic blood pressure data also have automatic relief characteristics similar to heart rate data.

[0072] Based on this, according to the above analysis, the present invention determines the data self-slowing coefficient of each data sequence based on the data changes of physiological diseases and non-physiological diseases in each data sequence in a short period of time, specifically:

[0073] Set the length of the local self-slowing interval; for each data in each data sequence, take the accumulation of the difference between each data and all data within the length interval of the local self-slowing interval as the first interval difference of each data; calculate the average data value of all data in each data sequence; take the result of the preset multiple of the difference between each data and the average data value as the second interval difference of each data; the preset multiple is the length of the local self-slowing interval; calculate the ratio of the first interval difference to the second interval difference of each data, forward fuse the ratio with the value of the deviation function of each data, and take the cumulative sum of the forward fusion results of all data in each data sequence as the data self-slowing coefficient of each data sequence. In order to avoid the situation where the second interval difference is 0 when used as the denominator, a preset parameter adjustment coefficient is added to the second interval difference, and the value is 0.1 in this embodiment.

[0074] In this embodiment, the length of the local self-slowing interval is set to 3, and the implementer can set the value according to the actual situation.

[0075] Preferably, in one embodiment of the present invention, the difference between each data and the average data value is set as the difference between each data and the average data value. In other embodiments of the present invention, other negative fusion methods may be used for calculation.

[0076] It can be understood that negative fusion is a fusion method such as subtraction and division between data. The specific negative fusion method is determined by the implementer according to the actual situation, and the present invention does not impose any special limitation.

[0077] Meanwhile, in this embodiment, the method for forward fusion of the ratio and the value of the deviation function of each data is set to: calculate the product of the ratio and the value of the deviation function of each data.

[0078] Because the body's early response to physiological diseases is not obvious, or the abnormal heart rate data is caused by the user's personal behavior or state, it is often close to the normal data and can automatically drop to the normal level within a short time, so that the heart rate gradually decreases over time, that is, when the difference of the first interval difference is calculated and accumulated, the difference should be greater than 0. Secondly, the heart rate is getting closer and closer to the mean heart rate, and the positive and negative first interval difference and the second interval difference in the data self-suspension coefficient are consistent, so the value is a positive number, that is, the data self-suspension coefficient is positive; on the contrary, when a physiological disease causes an abnormal heart rate, the ratio of the first interval difference to the second interval difference in the data self-suspension coefficient is a negative number, and due to the frequency of abnormal heart rate caused by physiological diseases, the data self-suspension coefficient is negative.

[0079] In the present invention, the flow chart of the method for determining the data self-restraint coefficient is shown in the attached figure. Figure 5 shown.

[0080] With respect to the changes of various data sequences in the blood pressure data relative to their normal data ranges, and whether the data in a short period of time has an automatic relief condition, the present invention combines the change characteristics of these two types of blood pressure data, based on the degree of deviation between various data sequences and their normal data ranges, the data changes of physiological diseases and non-physiological diseases expressed by various data sequences, and the differences between local subsequences in various data sequences, to determine the data continuous variation index of various data sequences in the user's blood pressure sequence for each month, specifically:

[0081] Calculate the distances and values ​​between all local subsequences in each type of data sequence; determine the data continuous mutation index of each type of data sequence based on the data variation coefficient, data self-relaxation coefficient, and distance and value; among them, the data variation coefficient and distance and value are both positively correlated with the data continuous mutation index, and the data self-relaxation coefficient is negatively correlated with the data continuous mutation index.

[0082] It can be understood that a negative correlation relationship indicates that the dependent variable will decrease as the independent variable increases, and the dependent variable will increase as the independent variable decreases. It can be a subtraction relationship, a division relationship, etc., which is determined by actual application and the present invention does not impose any special limitations.

[0083] In one embodiment of the present invention, the data variation coefficient is added to the distance and value, and the data self-relaxation coefficient is subtracted from the result of the addition to obtain the data continuous variation index. In other embodiments of the present invention, other suitable positive correlation and negative correlation relationships can also be combined to calculate the data continuous variation index.

[0084] In this embodiment, the distances between all local subsequences in each type of data sequence and the distances in the values ​​are calculated using the DTW distance. Other embodiments of the present invention may also use the absolute value of the difference between the values ​​to calculate the distance.

[0085] It should be understood that since the heart rate abnormalities caused by physiological diseases are more irregular, the heart rate change curve after detrending is different, and thus the changes between the local subsequences of the heart rate after detrending for any two days are different, that is, the distances and values ​​between all local subsequences are larger; therefore, when the user has an abnormal heart rate due to an obvious physiological disease, the data variation coefficient obtained by the heart rate deviating from its normal data range is larger, and the data can be automatically relieved. The data self-relaxation coefficient is negative and has a large value, which will make the data continuous variation index larger, indicating that the continuous variation of the heart rate is more serious.

[0086] In the present invention, the flow chart of the method for determining the data continuous variation index of each data sequence in the user's monthly blood pressure sequence is shown in the attached figure. Figure 2 shown.

[0087] The third step is to determine the noise index of the user each month by combining the variation of the data in the external factor sequence, the correlation between the various data sequences in the blood pressure sequence, and the maximum data continuous variation index in the various data sequences in the blood pressure sequence.

[0088] Since the user's heart rate, systolic blood pressure, and diastolic blood pressure may also be affected by external factors, causing data fluctuations, the external factors collected in this embodiment are the user's recent sleep status, eating habits, mental state, and psychological state. At this time, it is necessary to further analyze whether the abnormalities in heart rate, systolic blood pressure, and diastolic blood pressure are caused by changes in these external factors, so as to minimize the impact of these external factors on the user's health data and improve the accuracy of data evaluation.

[0089] The present invention sums the sleep state score, eating habit score, mental state score and psychological state score of each day in a month to obtain a comprehensive score of external factors for one day, and the comprehensive scores of external factors of all days in a month are combined in chronological order to form a comprehensive score sequence of external factors for one month. At the same time, the coefficient of variation of the comprehensive score sequence of external factors is obtained. Among them, the coefficient of variation is used to reflect the degree of discreteness of the data. When the data is more discrete, it means that there may be noise caused by external factors in the data, and the coefficient of variation is larger. The calculation method of the coefficient of variation is a known technology and will not be repeated.

[0090] At the same time, the data means of all local subsequences in each type of data sequence are used to form the average sequence of each type of data sequence; the correlation between the average sequence of each type of data sequence and the comprehensive score sequence of external factors is calculated to obtain the maximum correlation; the maximum correlation is used to reflect the changes in external factors that have a large correlation with the abnormal blood pressure data, which will also cause abnormal fluctuations in blood pressure data.

[0091] In this embodiment, the calculation method of the correlation is set to Spearman rank correlation coefficient. The calculation method of Spearman rank correlation coefficient is a well-known technology and will not be described in detail.

[0092] In order to measure the impact of external factors on blood pressure data, the present invention combines the variation of data in the external factor sequence, the correlation between various data sequences in the blood pressure sequence, and the maximum data continuous variation index in various data sequences in the blood pressure sequence to determine the noise index of the user each month, specifically:

[0093] The noise index is determined based on the coefficient of variation, the maximum correlation and the maximum data continuous variation index; wherein the coefficient of variation and the maximum correlation are both positively correlated with the noise index, and the maximum data continuous variation index is negatively correlated with the noise index.

[0094] In one embodiment of the present invention, the specific method for calculating the noise index is: calculate the product of the coefficient of variation and the maximum correlation as the first product; use the maximum data continuous variation index as the exponent of an exponential function with the natural constant e as the base, and use the ratio of the first product to the calculation result of the exponential function as the noise index.

[0095] It should be understood that there is a certain mutual influence relationship between heart rate, systolic pressure and diastolic pressure. When at least one of the continuous variation indexes of heart rate, systolic pressure and diastolic pressure is large, it means that the main reason for the abnormal fluctuation of the data observed at this time is that the user has obvious physiological diseases; in addition, if the coefficient of variation of the comprehensive score sequence of external factors is large, and the abnormality of the data is highly correlated with these external factors, and the change of external factors with a large correlation with the abnormality of the data will also cause abnormal fluctuation of the data, thereby increasing the noise, but because the exponential function grows faster, the overall noise index is small. When the continuous variation index of heart rate, systolic pressure and diastolic pressure is small, that is, the maximum continuous variation index of data is small, it means that the main reason for the abnormal fluctuation of the data observed at this time is that the physical response of the physiological disease is not obvious in the early stage or it is caused by non-physiological diseases. At this time, if the coefficient of variation of the comprehensive score sequence of external factors is large, and the abnormality of the data is highly correlated with the external factors, that is, the coefficient of variation and the maximum correlation are both large, so the noise index is relatively large.

[0096] In the present invention, the flow chart of the method for determining the noise index is shown in the attached figure. Figure 6 shown.

[0097] The fourth step is to modify the noise matrix in the original Kalman gain calculation process based on the noise index, and use the improved Kalman filter algorithm to filter the heart rate sequence, systolic blood pressure sequence and diastolic blood pressure sequence.

[0098] The user's monthly heart rate sequence, systolic blood pressure sequence and diastolic blood pressure sequence are respectively used as input, and the Kalman filter algorithm is used to output the corrected heart rate sequence, systolic blood pressure sequence and diastolic blood pressure sequence for each month. The Kalman filter algorithm is a well-known technology, and the specific process is not repeated here.

[0099] When the Kalman filter algorithm is used to filter the heart rate sequence, systolic pressure sequence and diastolic pressure sequence, the present invention corrects the noise matrix in the Kalman gain calculation process, thereby removing the influence of noise from the collected data and more accurately reflecting the user's health status.

[0100] The method for correcting the noise matrix of the original Kalman gain by the noise index is as follows: multiplying the noise matrix of the original Kalman gain by the noise index to obtain the noise matrix of the corrected Kalman gain.

[0101] It should be understood that when the noise index is small, it means that the main reason for abnormal data fluctuations is that the user has obvious physiological diseases. At this time, the corrected noise matrix is ​​also small, so that the Kalman gain will be relatively large, and the corrected data will be more inclined to the measured value; when the noise index is large, it means that the main reason for abnormal data fluctuations is that the early physical response of physiological diseases is not obvious or caused by non-physiological diseases. The abnormal data fluctuations are affected by external factors. At this time, the corrected noise matrix is ​​also large, so the corresponding Kalman gain will be relatively small, and the corrected data will be more inclined to the predicted value. Using the Kalman filter algorithm, by correcting the noise matrix, the health data of the collected users can be corrected according to the extent to which the abnormal data fluctuations are affected by external factors.

[0102] The above steps can obtain the corrected heart rate sequence, systolic blood pressure sequence and diastolic blood pressure sequence for n months, and the corrected heart rate sequence, systolic blood pressure sequence and diastolic blood pressure sequence for n months are fused respectively to obtain the fused heart rate sequence, systolic blood pressure sequence and diastolic blood pressure sequence for n months.

[0103] Taking the fused heart rate series, systolic blood pressure series and diastolic blood pressure series of n months as input, the moving average method in time series analysis is used to output the corresponding moving average value series. The data of the most recent year is used to analyze whether there is any abnormal fluctuation in the data.

[0104] The data in the moving average sequence that exceeds the normal data range is marked as abnormal data, and the user is reminded of health conditions based on the abnormal data. The moving average method is a well-known technology, and the specific calculation process is not repeated here.

[0105] Based on the same inventive concept as the above method, an embodiment of the present invention also provides an integrated health management data collection and optimization system, including a memory, a processor, and a computer program stored in the memory and running on the processor, and when the processor executes the computer program, it implements the steps of any one of the above methods.

[0106] The various embodiments of the present invention are described in a progressive manner, and the same or similar parts between the various embodiments can be referred to each other, and each embodiment focuses on the differences from other embodiments.

[0107] It should be noted that, unless otherwise specified and limited, terms such as "include", "comprises" or any other variants thereof are intended to cover non-exclusive inclusion, so that a circuit structure, article or device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such article or device. In the absence of further restrictions, an element defined by the sentence "including one..." does not exclude the existence of other identical elements in the article or device including the element. In addition, the term "and\or" used herein includes any and all combinations of one or more related listed items.

[0108] Those skilled in the art will readily appreciate other embodiments of the invention after considering the specification and practicing the invention herein. The present invention is intended to cover any variations, uses or adaptations of the present invention that follow the general principles of the present invention and include common knowledge or customary techniques in the art that are not invented by the present invention.

[0109] It will be appreciated that the present invention is not limited to the exact construction that has been described above and shown in the drawings, and that various modifications and changes may be made without departing from the scope thereof.

Claims

1. A health management integrated data collection optimization method, characterized in that: The method comprises the following steps: Obtaining the user's health data for several months in history; the user health data includes a blood pressure sequence and an external factor sequence; Each data sequence in the blood pressure sequence is evenly divided to obtain local subsequences of each data sequence every day; based on the degree of deviation between each data sequence and its normal data range, the data changes of physiological diseases and non-physiological diseases represented by each data sequence, and the differences between local subsequences in each data sequence, the data continuous variation index of each data sequence in the user's blood pressure sequence for each month is determined; The noise index of the user each month is determined by combining the variation of the data in the external factor sequence, the correlation between the data sequences in the blood pressure sequence, and the maximum continuous variation index of the data in the data sequences in the blood pressure sequence. The noise matrix in the original Kalman gain calculation process is corrected based on the noise index, and the heart rate sequence, systolic blood pressure sequence and diastolic blood pressure sequence are filtered using the improved Kalman filter algorithm. The blood pressure sequence includes a systolic pressure sequence, a diastolic pressure sequence and a heart rate sequence; the external factor sequence includes a sleep state score sequence, a dietary habit score sequence, a mental state score sequence and a psychological state score sequence; The method for determining the noise index includes: Merge all types of score sequences in the external factor sequence to obtain a comprehensive score sequence of external factors; obtain the coefficient of variation of the comprehensive score sequence of external factors; The data mean of all local subsequences in each data sequence is used to form the average sequence of each data sequence; Calculate the correlation between the average sequence of various data sequences and the comprehensive score sequence of external factors to obtain the maximum correlation; The noise index is determined based on the coefficient of variation, the maximum correlation and the maximum data continuous variation index; wherein the coefficient of variation and the maximum correlation are both positively correlated with the noise index, and the maximum data continuous variation index is negatively correlated with the noise index.

2. A health management integrated data collection and optimization method as claimed in claim 1, characterized in that: The method for determining the data continuous variation index of each data sequence in the user's monthly blood pressure sequence includes: Determine the data variation coefficient of each data series based on the degree of deviation between each data in each data series and the boundary value of its normal data range; Based on the data changes of physiological diseases and non-physiological diseases in each data series in a short period of time, the data self-slowdown coefficient of each data series is determined; Calculate the distances and values ​​between all local subsequences in various data sequences; Based on the data variation coefficient, data self-relaxation coefficient, and distance and value, the data continuous variation index of each type of data series is determined; among them, the data variation coefficient and distance and value are both positively correlated with the data continuous variation index, and the data self-relaxation coefficient is negatively correlated with the data continuous variation index.

3. A health management integrated data collection and optimization method as claimed in claim 2, characterized in that: The method for determining the data variation coefficient includes: Based on the difference between each data in each data sequence and the boundary value of its normal data range, a deviation function is constructed; The average value of the deviation function of all data in each data sequence is used as the first evaluation coefficient of each data sequence; The value of the deviation function from the average value of all data in each data sequence is used as the second evaluation coefficient of each data sequence; The data variation coefficient is determined based on the first evaluation coefficient and the second evaluation coefficient; wherein both the first evaluation coefficient and the second evaluation coefficient are positively correlated with the data variation coefficient.

4. A health management integrated data collection and optimization method as claimed in claim 3, characterized in that: The method for constructing the deviation function includes: Get the upper and lower boundaries of the normal data range for various data series; When each data in each data sequence belongs to the normal data range, the value of the deviation function of each data is set to 0; On the contrary, the value of each data exceeding the upper boundary or the lower boundary is taken as the value of the deviation function of each data.

5. A health management integrated data collection and optimization method as claimed in claim 4, characterized in that: The method for determining the data self-restraint coefficient includes: Set the length of the local interval of self-ease; For each data in each data sequence, the accumulation of the difference between each data and all data within the length interval of the self-slowing local interval is taken as the first interval difference of each data; Calculate the average data value of all data in each data sequence; take the result of the preset multiple of the difference between each data and the average data value as the second interval difference of each data; the preset multiple is the length of the self-slowing local interval; The ratio of the first interval difference to the second interval difference of each data is calculated, the ratio is forward fused with the value of the deviation function of each data, and the sum of the forward fusion results of all data in each data sequence is taken as the data self-relaxation coefficient of each data sequence.

6. The method for optimizing health management integrated data collection according to claim 1, characterized in that: The method for correcting the noise matrix of the original Kalman gain by the noise index is: multiplying the noise matrix of the original Kalman gain by the noise index to obtain the noise matrix of the corrected Kalman gain.

7. The health management integrated data collection and optimization method according to claim 1, characterized in that: After filtering the heart rate sequence, systolic blood pressure sequence and diastolic blood pressure sequence, the method also includes: fusing the filtering results of various data sequences of all months, using the fusion results of various data sequences as inputs of the moving average algorithm, and outputting the corresponding moving average sequence; marking the data in the moving average sequence that exceeds its normal data range as abnormal data, and reminding the user of health conditions based on the abnormal data.

8. A health management integrated data collection and optimization system, comprising a memory, a processor, and a computer program stored in the memory and running on the processor, characterized in that: When the processor executes the computer program, the steps of the health management integrated data acquisition optimization method as described in any one of claims 1-7 are implemented.

Images