Blood spectral analysis method for preventing and controlling obesity and metabolic diseases in infants and young children

Abstract

The present invention relates to the technical field of spectral data analysis, and specifically to a blood spectral analysis method for preventing and controlling obesity metabolic diseases in infants and young children. The initial blood glucose absorption rate time series data is decomposed into multiple component time series data and residual time series data, so as to have a deeper understanding of the changes in blood glucose absorption rate and the influence of different factors. By analyzing the change trend, short-term fluctuation and comprehensive influence index of each component time series data, its state is determined. For the component time series data belonging to the normal influence state, it is regarded as the standard component; for the component belonging to the abnormal state, the frequency domain information of the component time series data and the residual time series data is used for correction. Finally, all standard components and corrected components are synthesized to obtain the corrected blood glucose absorption rate time series data. It can more accurately reflect the real changes in blood glucose in the blood of infants and young children, and provide a reliable basis for the subsequent monitoring of obesity metabolic diseases.

Description

Technical Field

[0001] The present invention relates to the technical field of spectral data analysis, and in particular to a blood spectral analysis method for preventing and controlling obesity metabolic diseases in infants and young children. Background Art

[0002] Infancy is a critical stage of growth and development, and the nutritional status during this stage has a profound impact on the long-term health of children. With the change of modern lifestyle, the incidence of obesity and metabolic diseases (such as diabetes, cardiovascular disease, etc.) in infants and young children has gradually increased, becoming an important issue in the field of public health. Early prevention and control of obesity and metabolic diseases is of great significance to reduce the risk of chronic diseases in the future.

[0003] When analyzing obesity and metabolic diseases in infants and young children, the existing technology generally uses near-infrared spectroscopy to monitor blood sugar levels, and then monitors sugar metabolism in infants and young children based on the temporal changes in blood sugar. However, physiological behaviors of infants and young children, such as diet and exercise, can cause fluctuations in blood sugar data, and changes in temperature can also significantly affect the absorption characteristics of human tissues, thereby affecting the absorption and reflection of near-infrared light; therefore, the metabolism of sugar in infants and young children may be affected by a variety of factors, so the blood sugar data obtained by near-infrared spectroscopy cannot accurately represent the actual metabolism of sugar in infants and young children. If it is not processed and directly analyzed, the accuracy of the monitoring results may be low. Summary of the invention

[0004] In order to solve the technical problem that the metabolism of sugar in infants and young children may be affected by multiple factors, the blood sugar data obtained by near-infrared spectroscopy technology cannot accurately represent the actual metabolism of sugar in infants and young children. If it is not processed and directly analyzed, it may lead to low accuracy of the monitoring results. The purpose of the present invention is to provide a blood spectral analysis method for preventing and controlling obesity metabolic diseases in infants and young children. The technical scheme adopted is as follows:

[0005] Acquire initial blood glucose absorption rate time series data of infants and young children; decompose the initial blood glucose absorption rate time series data to obtain residual time series data and multiple component time series data;

[0006] For any component time series data, based on the change trend of the data value, determine the long-term trend change index of the component time series data; based on the local change amplitude and fluctuation of the data value, determine the short-term fluctuation change index of the component time series data; based on the overall fluctuation of the data value and the local difference, determine the comprehensive impact index of the component time series data;

[0007] Based on the long-term trend change index, the short-term fluctuation change index and the comprehensive impact index, the state of the component time series data is judged; if the component time series data belongs to the normal impact state, the component time series data is used as the standard component; if the component time series data does not belong to the normal impact state, it is corrected based on the frequency domain information in the component time series data and the residual time series data, so as to obtain the corrected component;

[0008] All standard components and correction components are synthesized to obtain the corrected blood glucose absorption rate time series data.

[0009] Furthermore, the method for obtaining the long-term trend change indicator includes:

[0010] For any component time series data, obtain the time series data curve corresponding to the component time series data;

[0011] On the time series data curve, the slope value at each data point is obtained, and the absolute value of the difference between the slope value of each data point and the next adjacent data point is calculated as the trend change factor corresponding to each data point;

[0012] The mean of the trend change factors of all data points is negatively correlated and normalized to obtain the value as the local stability factor;

[0013] Calculate the mean of all data values ​​in the component time series data as the mean eigenvalue;

[0014] Calculate the absolute value of the difference between each data value and the mean eigenvalue as a deviation factor, and take the mean of all deviation factors as the overall change factor;

[0015] The normalized value of the sum of the local stability factor and the overall change factor is used as the long-term trend change indicator of the component time series data.

[0016] Furthermore, the method for obtaining the short-term fluctuation change index includes:

[0017] Determine the local neighborhood corresponding to each data point;

[0018] In the local neighborhood of each data point, the difference between the maximum data value and the minimum data value is calculated as the local fluctuation amplitude value corresponding to each data point;

[0019] In the local neighborhood of the next data point adjacent to each data point, the variance of all data values ​​is calculated as the lag fluctuation factor corresponding to each data point;

[0020] Determine the local fluctuation factor corresponding to each data point according to the local fluctuation amplitude value and the lag fluctuation factor corresponding to each data point, wherein the local fluctuation amplitude value is positively correlated with the local fluctuation factor, and the lag fluctuation factor is negatively correlated with the local fluctuation factor;

[0021] The normalized sum of the local fluctuation factors of all data points in the component time series data is used as the short-term fluctuation change indicator of the component time series data.

[0022] Furthermore, the method for obtaining the comprehensive impact index includes:

[0023] Calculate the absolute value of the difference between each data point and the next adjacent data point as the local difference value corresponding to each data point;

[0024] The mean of all local difference values ​​is taken as the local difference feature value;

[0025] The value obtained by normalizing the sum of the local difference characteristic value and the overall change factor is used as the comprehensive impact indicator of the component time series data.

[0026] Further, judging the state of the component time series data based on the long-term trend change index, the short-term fluctuation change index and the comprehensive impact index includes:

[0027] For any component time series data, the maximum value among the long-term trend change index, short-term fluctuation change index and comprehensive impact index of the component time series data is taken as the state value;

[0028] If the state value of the component time series data is a long-term trend change indicator, the state of the component time series data is a normal impact state;

[0029] If the state value of the component time series data is a short-period fluctuation change indicator, the state of the component time series data is a temperature-affected state;

[0030] If the status value of the component time series data is a comprehensive impact index, then the status of the component time series data is a comprehensive impact status.

[0031] Furthermore, when the component time series data is in a temperature-affected state, the method for acquiring the correction component includes:

[0032] For any component time series data, if the component time series data is in a temperature-affected state, Fourier transform is performed on the component time series data to obtain first frequency domain data;

[0033] A first low-pass filter is preset, and the first frequency domain data is low-pass filtered based on the first low-pass filter, and the obtained first filtered data is inversely Fourier transformed to obtain a corrected component of the component time series data.

[0034] Furthermore, when the component time series data is in a comprehensive influence state, the method for obtaining the correction component includes:

[0035] For any component time series data, if the component time series data is in a comprehensive influence state, Fourier transform is performed on the component time series data to obtain second frequency domain data;

[0036] Performing Fourier transform on the residual time series data to obtain comparison frequency domain data, and taking the mean of all frequencies in the comparison frequency domain data as the comparison frequency value;

[0037] A second low-pass filter is set according to the comparison frequency value, and the second frequency domain data is low-pass filtered based on the second low-pass filter, and the obtained second filtered data is inversely Fourier transformed to obtain a corrected component of the component time series data.

[0038] Furthermore, the initial blood glucose absorption rate time series data is decomposed to obtain residual time series data and multiple component time series data, including:

[0039] The initial blood glucose absorption rate time series data is decomposed based on an empirical mode decomposition algorithm to obtain multiple component time series data and residual time series data.

[0040] Furthermore, the method for obtaining the time series data of the corrected blood glucose absorption rate includes:

[0041] All standard components and correction components are synthesized based on the empirical mode decomposition algorithm to obtain the corrected blood glucose absorption rate time series data.

[0042] Furthermore, the method for acquiring the initial blood glucose absorption rate time series data includes:

[0043] Obtain blood spectrum data of infants and young children at different times;

[0044] The characteristic wavelength is preset, the absorbance of the characteristic wavelength in the blood spectrum data at all times is extracted and sorted in time series, so as to obtain the initial blood glucose absorbance time series data.

[0045] The present invention has the following beneficial effects:

[0046] First, the initial blood glucose absorption rate time series data of infants and young children is obtained, and the initial blood glucose absorption rate time series data is decomposed into residual time series data and multiple component time series data, which can analyze the change of blood glucose absorption rate more carefully, and help to identify the influence of different factors on blood glucose metabolism. The influence of normal physiological behavior on blood glucose in the blood is an obvious trend change, and it is a long-term influence behavior. Therefore, for each component time series data, the change trend of the data value is analyzed, so as to determine the long-term trend change index of each component time series data, which is used to characterize the situation that the change of the component time series data is affected by normal physiological activities; the influence of temperature change on blood glucose in infants and young children is a short-term behavior, because temperature change is often a short-term change process, so the present invention is based on the local change amplitude and fluctuation of the data value in the component time series data, and determines the short-term fluctuation change index, which is used to characterize the situation that the change of the component time series data is affected by temperature; and when it is affected by physiological activities and temperature, the component time series data will have both data fluctuation characteristics caused by normal physiological behavior and fluctuation characteristics affected by temperature, so based on the overall fluctuation of the data value and the local difference, the comprehensive influence index is determined. Then, based on the above three indicators, the state of each component time series data is determined. If it belongs to the normal influence state, then the component time series data is a standard component, which means that it can characterize the actual blood sugar changes of infants and young children. If it does not belong to the normal influence state, it means that it is affected by temperature, so it needs to be corrected. Since frequency domain analysis can reveal the frequency components of the data and identify the dominant frequency in the data, the present invention uses the frequency domain information in the component time series data and the residual time series data to correct the component time series data to obtain the corrected component. The corrected component at this time can eliminate the influence of temperature. Therefore, finally, by synthesizing all the standard components and the corrected components, the corrected blood sugar absorption rate time series data can be obtained, thereby more accurately reflecting the actual changes in blood sugar in the blood of infants and young children, and ensuring the accuracy of subsequent obesity metabolic disease monitoring. BRIEF DESCRIPTION OF THE DRAWINGS

[0047] 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.

[0048] Figure 1 A method flow chart of a blood spectral analysis method for preventing and controlling obesity metabolic diseases in infants and young children provided by one embodiment of the present invention;

[0049] Figure 2A method flow chart of a method for obtaining a long-term trend change indicator provided by an embodiment of the present invention;

[0050] Figure 3 A flow chart of a method for obtaining a short-period fluctuation change index provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0051] 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 blood spectral analysis method for preventing and controlling obesity metabolic diseases in infants and young children proposed by the present invention, its specific implementation method, structure, characteristics and effects, 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.

[0052] 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.

[0053] The following is a detailed description of a specific scheme of a blood spectral analysis method for preventing and controlling obesity metabolic diseases in infants and young children provided by the present invention in conjunction with the accompanying drawings.

[0054] See also Figure 1 , which shows a method flow chart of a blood spectral analysis method for preventing and controlling obesity metabolic diseases in infants and young children provided by an embodiment of the present invention, the method comprising the following steps:

[0055] Step S1: obtaining the initial blood glucose absorption rate time series data of the infant; decomposing the initial blood glucose absorption rate time series data to obtain residual time series data and multiple component time series data.

[0056] Due to multiple reasons such as overeating, lack of exercise and genetic factors, more and more infants and young children are overweight and obese. Infant obesity is not only a problem of appearance, but more seriously, it may lead to a series of metabolic diseases. For example, abnormal high blood sugar metabolism may lead to the development of diabetes in childhood. What's more, the early occurrence of metabolic diseases may affect the growth and development and immune system of infants and young children, increase the risk of chronic diseases such as cardiovascular disease and diabetes in adulthood, and seriously affect their health throughout their lives.

[0057] For the prevention and control of obesity and metabolic diseases in infants and young children, near-infrared spectroscopy data is usually used to extract blood sugar data from the blood of infants and young children, so as to monitor the blood sugar data. However, near-infrared spectroscopy technology analyzes the chemical composition of blood by measuring the absorption characteristics of near-infrared light, and changes in temperature will significantly affect the absorption characteristics of tissues. The main reason is that changes in temperature will cause changes in the vibration energy of chemical bonds and changes in the probability of energy level transitions. Therefore, spectral data is easily affected by temperature, resulting in the blood sugar data extracted by near-infrared spectroscopy technology cannot accurately represent the actual blood sugar characteristics of infants and young children.

[0058] In the embodiment of the present invention, by analyzing and processing the blood sugar data obtained by near-infrared spectral data, unnecessary temperature influence can be effectively reduced, so that the final blood sugar data can more accurately reflect the actual blood sugar changes of infants and young children.

[0059] Therefore, it is first necessary to obtain the time series data of the initial blood glucose absorption rate of infants and young children.

[0060] Preferably, in one embodiment of the present invention, the method for acquiring the initial blood glucose absorption rate time series data includes:

[0061] Select a near-infrared spectrometer that can cover an appropriate wavelength range (approximately 800nm-2500nm), use a non-invasive near-infrared probe, and obtain the infant's blood spectrum data through the infant's skin reflection spectrum or transmission spectrum. Within a preset period of time, collect blood spectrum data at certain time intervals to obtain the infant's blood spectrum data at different times.

[0062] Since different substances have different absorption peaks in the blood spectrum data, and a more obvious characteristic absorption peak of glucose is about 1210nm, 1210nm is used as the characteristic wavelength, and the absorption rate of the characteristic wavelength in the blood spectrum data at all times is extracted and sorted in time series, so that the initial blood glucose absorption rate time series data can be obtained.

[0063] It should be noted that the time interval is set to 2 seconds, and the preset period is set to 30 minutes. The specific time interval and the length of the preset period can be adjusted according to the implementation scenario and are not limited here.

[0064] Time series decomposition is a deep analysis method that can decompose complex time series data into components that are easier to understand and analyze. Through decomposition, we can more clearly see the changing pattern of blood glucose absorption rate and possible fluctuations, so we can use the information decomposition algorithm to decompose the initial blood glucose absorption rate time series data to obtain the decomposition results.

[0065] Preferably, in one embodiment of the present invention, the method for obtaining the decomposition result includes:

[0066] The empirical mode decomposition algorithm is an adaptive data processing or mining method, which is very suitable for processing nonlinear and non-stationary time series. Because blood sugar levels are affected by many factors, the empirical mode decomposition algorithm can adapt well to this data characteristic and effectively decompose it.

[0067] Therefore, the initial blood glucose absorption rate time series data is decomposed based on the empirical mode decomposition algorithm to obtain multiple component time series data and residual time series data. Among them, the component time series data can represent the changing characteristics of blood glucose absorption rate at different time scales, such as short-term fluctuations, medium-term trends, and long-term laws. This decomposition helps to understand the changing mechanism of blood glucose absorption rate more deeply, and the residual time series data is the remaining part after empirical mode decomposition, which represents the part that cannot be explained by components. In the decomposition of blood glucose absorption rate time series data, the residual time series data can be regarded as the long-term stable blood glucose trend of infants and young children, that is, it is regarded as component data that is not or rarely affected by temperature.

[0068] It should be noted that empirical mode decomposition is a well-known technology, and the specific process will not be described here.

[0069] Step S2: For any component time series data, based on the changing trend of the data value, determine the long-term trend change index of the component time series data; based on the local change amplitude and fluctuation of the data value, determine the short-term fluctuation change index of the component time series data; based on the overall fluctuation of the data value and the local difference, determine the comprehensive impact index of the component time series data.

[0070] The influence of normal physiological behavior on blood glucose in the blood is analyzed in the overall long term, and there is an obvious trend of change. For example, common diet and exercise, the former is that the blood glucose in the blood gradually rises, and then decreases after reaching a certain amplitude, and the behavior lasts for a long time. The latter is based on normal blood glucose, and gradually decreases with the change of exercise amount (increase in exercise intensity or duration of exercise); however, the influence of temperature on infant blood spectrum data is a short-term behavior. The specific influence process is that when the temperature changes, the infant blood spectrum data is affected by the temperature change, and because the temperature change often has a short change process, the infant blood glucose data will also have a short-term change behavior; and the blood glucose data under the joint influence of temperature and physiological behavior will have both a long-term obvious change trend and a short-term change in the short period. Therefore, in an embodiment of the present invention, the change of data values ​​in each component time series data can be analyzed based on the above-mentioned characteristics, so as to help determine the state of each component time series data in the subsequent process.

[0071] When the blood sugar data of infants and young children changes due to physiological behaviors such as diet and exercise, it is generally a relatively stable and continuous trend change. This change is basically difficult to distinguish in the near future, but it must be more obvious in the long term. Therefore, for any component time series data, the changing trend of the data value can be analyzed to determine the long-term trend change indicator of the component time series data.

[0072] Preferably, in one embodiment of the present invention, the method for obtaining the long-term trend change indicator includes:

[0073] See also Figure 2 , which shows a method flow chart of a method for obtaining a long-term trend change indicator in one embodiment of the present invention, the method comprising the following steps:

[0074] Step S201: for any component time series data, analyze the local change trend in the component time series data and determine the local stability factor.

[0075] For any component time series data, obtain the time series data curve corresponding to the component time series data, and obtain the slope value at each data point on the time series data curve. The slope value can reflect the changing trend between each two adjacent data values. Calculate the absolute value of the difference in slope value between each data point and the next adjacent data point as the trend change factor corresponding to each data point. The larger the trend change factor, the worse the consistency of the changing trend between the two adjacent data points, that is, the larger the local fluctuation of the data. Based on the above analysis, if the component time series data is affected by normal physiological activities, the local fluctuation will be weaker, so the mean value of the trend change factor of all data points is negatively correlated and normalized as the local stability factor. The larger the local stability factor, the more consistent the adjacent data values ​​in the component time series data, so the component time series data is more likely to be affected by normal physiological activities. The negative correlation mapping and normalization processing here can be used. Function, where It represents an exponential function with the natural constant e as the base, and x represents the independent variable.

[0076] It should be noted that the method for obtaining the time series data curve can use the least squares method to fit the component time series data, and the least squares method is a well-known technology, and the specific process will not be described here; the slope value at the last data point is set to the mean of the slope values ​​at all data points before its time series, and the trend change factor corresponding to the last data point is set to be consistent with the trend change factor corresponding to the previous adjacent data point.

[0077] Step S202: Analyze the overall change trend of the data values ​​in the component time series data to determine the overall change factor.

[0078] To analyze the overall change trend, we can calculate the mean of all data values ​​in the component time series data as the mean eigenvalue, which is used to reflect the average level of all data values ​​in the component time series data. The absolute value of the difference between each data value and the mean eigenvalue is calculated as the deviation factor. The larger the deviation factor, the greater the difference between each data value and the overall average level. Therefore, the mean of all deviation factors is used as the overall change factor. At this time, the larger the overall change factor, the greater the overall fluctuation of the component time series data. Therefore, the longer the change trend, the more obvious it is. Therefore, it is more likely to be affected by normal sorting activities.

[0079] Step S203: Integrate the local stability factor and the overall change factor of the component time series data to obtain the long-term trend change index of the component time series data.

[0080] The normalized value of the sum of the local stability factor and the overall change factor of the component time series data is used as the long-term trend change index of the component time series data. The larger the long-term trend change index is, the more likely the data change in the component time series data is caused by normal physiological activities. Normalization is a technical means well known to those skilled in the art, and the normalization function can be linear normalization or standard normalization, etc. The specific normalization method is not limited here.

[0081] The influence of temperature on infant blood spectrum data is a short-term behavior. The specific influence process is: when the temperature changes, the infant blood spectrum data is affected by the temperature change, because the temperature can affect the body's metabolic rate, thereby affecting the consumption and level of blood sugar. For example, in a colder environment, the human body will accelerate metabolism in order to maintain body temperature, which may lead to increased blood sugar consumption and decreased blood sugar levels; on the contrary, in a hotter environment, the human body's metabolic rate may slow down, resulting in reduced blood sugar consumption and increased blood sugar levels; and because the change in temperature often has a short-term change process, in the embodiment of the present invention, the local change amplitude and fluctuation of the data value in each component time series data are analyzed to determine the short-term fluctuation change index of the component time series data.

[0082] Preferably, in one embodiment of the present invention, the method for obtaining the short-period fluctuation change index includes:

[0083] See also Figure 3 , which shows a method flow chart of a method for obtaining a short-period fluctuation change index in one embodiment of the present invention, the method comprising the following steps:

[0084] Step S211: For any component time series data, analyze the local variation amplitude of the data value in the component time series data, and determine the local fluctuation amplitude value corresponding to each data point.

[0085] Since local features need to be analyzed, for any component time series data, the local neighborhood corresponding to each data point is first determined.

[0086] In the local neighborhood of each data point, the difference between the maximum data value and the minimum data value is calculated as the local fluctuation amplitude value corresponding to each data point. The larger the local fluctuation amplitude value, the greater the data difference in the local range of the data values ​​in the component time series data, which can be regarded as being affected by temperature, resulting in fluctuations in blood glucose data within a short period.

[0087] Step S212: Based on the local fluctuation of the data value in the component time series data, determine the lag fluctuation factor corresponding to each data point in the component time series data.

[0088] Given that temperature change is a short process, blood sugar change is also a short process, and because the human body has a regulatory function, after a short change in blood sugar, it will be in a stable state for a short period of time. Therefore, in the local neighborhood of the next data point adjacent to each data point, the variance value of all data values ​​is calculated as the hysteresis fluctuation factor corresponding to each data point. The smaller the hysteresis fluctuation factor, the more stable the change of the data value, and the blood sugar data is regarded as being temporarily affected by temperature. Among them, the hysteresis fluctuation factor corresponding to the last data point is set to the variance value of all data values ​​in the local neighborhood corresponding to the last data point.

[0089] Step S213: In the component time series data, the local fluctuation amplitude values ​​and the lagged fluctuation factors corresponding to all data points are integrated to obtain the short-term fluctuation change index of the component time series data.

[0090] According to the local fluctuation amplitude value and the lag fluctuation factor corresponding to each data point, the local fluctuation factor corresponding to each data point is determined. The formula model of the local fluctuation factor includes:

[0091] ;

[0092] in, represents the local fluctuation factor; Indicates the local fluctuation amplitude value; represents the lagged volatility factor; Indicates the preset first parameter.

[0093] In the formula model of the local fluctuation factor, based on the above analysis, it can be seen that the larger the local fluctuation amplitude value, the greater the data difference in the data values ​​in the component time series data within the local range, which can be regarded as being affected by temperature; the smaller the lag fluctuation factor, the more stable the change in the data value, which is regarded as the blood glucose data being temporarily affected by temperature; therefore, the local fluctuation amplitude value is positively correlated with the local fluctuation factor, and the lag fluctuation factor is negatively correlated with the local fluctuation factor, so the lag fluctuation factor is used as the denominator and the local fluctuation amplitude value as the numerator to obtain the local fluctuation factor. The larger the local fluctuation factor, the more the change in the data values ​​in the component time series data conforms to the characteristics of temperature influence.

[0094] Finally, the sum of the local fluctuation factors of all data points in the component time series data is normalized and used as the short-period fluctuation change index of the component time series data. At this time, the larger the short-period fluctuation change index is, the more likely the data change in the component time series data is caused by temperature change. Normalization is a technical means well known to those skilled in the art, and the normalization function can be linear normalization or standard normalization, etc. The specific normalization method is not limited here.

[0095] It should be noted that the first parameter is preset The role of is to prevent the denominator from being 0, and the value here can be 0.001. The specific value can be adjusted according to the implementation scenario and is not limited here. The data points in the local neighborhood of each data point include each data point and the two data points closest to each data point in time series. The size of the local neighborhood can be adjusted according to the implementation scenario and is not limited here.

[0096] Because the blood sugar data of infants and young children may be affected by both physiological activities and temperature, in an embodiment of the present invention, for any component time series data, a comprehensive impact index of the component time series data is calculated based on the overall fluctuation of the data value and the local difference, which is used to reflect the possibility that the component time series data is affected by both factors.

[0097] Preferably, in one embodiment of the present invention, the method for obtaining the comprehensive impact index includes:

[0098] The comprehensive impact state indicates that the component time series data is affected by both the long-term impact caused by physiological activities and the short-term impact caused by temperature changes. Therefore, the local differences of the data values ​​can be analyzed first to quantify the degree of short-term impact in the comprehensive impact state: the absolute value of the difference between the data value of each data point and the next adjacent data point is calculated as the local difference value corresponding to each data point. The larger the local difference value, the greater the fluctuation between adjacent data values, which can be regarded as an immediate impact. The mean of all local difference values ​​is taken as the local difference eigenvalue. The larger the local difference eigenvalue, the more it indicates that the data value in the component time series data has undergone a local mutation, and the possibility of being affected by the short-term impact caused by temperature changes is higher. Among them, the local difference value corresponding to the last data point is set to be consistent with the local difference value corresponding to the previous data point adjacent to it.

[0099] Then we analyze the long-term impact of physiological activities. Based on the analysis of the aforementioned long-term trend change indicators, we know that the larger the overall change factor is, the greater the overall fluctuation of the component time series data is, and thus there is a long-term change trend, so it is more likely to be affected by normal sorting activities.

[0100] Therefore, the overall change factor is finally integrated with the local difference characteristic value, and the sum of the local difference characteristic value and the overall change factor is normalized as the comprehensive impact index of the component time series data. At this time, the larger the comprehensive impact index is, the more it indicates that the component time series data has both a long-term change trend caused by physiological activities and a short-term fluctuation characteristic caused by temperature changes, that is, the component time series data is affected by a comprehensive impact. Normalization is a technical means well known to those skilled in the art, and the normalization function can be linear normalization or standard normalization, etc. The specific normalization method is not limited here.

[0101] Step S3: Based on the long-term trend change index, the short-term fluctuation change index and the comprehensive impact index, the state of the component time series data is judged; if the component time series data belongs to the normal impact state, the component time series data is used as the standard component; if the component time series data does not belong to the normal impact state, it is corrected based on the frequency domain information in the component time series data and the residual time series data to obtain the corrected component.

[0102] Based on the above steps, three indicators corresponding to each component time series data can be obtained, namely, the long-term trend change indicator, which indicates the possibility that the fluctuation of each component time series data is affected by the long period of physiological activities; the short-term fluctuation change indicator, which indicates the possibility that the fluctuation of each component time series data is affected by the short period of temperature changes; the comprehensive impact indicator, which indicates the possibility that the fluctuation of each component time series data is affected by the combined effect of physiological activities and temperature changes; for different influencing factors, the processing method of component time series data will also be different, so in an embodiment of the present invention, firstly, based on the three indicators of each component time series data, its state is determined, and then different processing methods are determined according to the different states, so as to obtain data that can more accurately reflect the actual blood sugar changes of infants and young children.

[0103] Preferably, in one embodiment of the present invention, judging the state of the component time series data based on the long-term trend change index, the short-term fluctuation change index and the comprehensive impact index includes:

[0104] For any component time series data, the maximum value among the long-term trend change index, short-term fluctuation change index and comprehensive impact index of the component time series data is taken as the state value.

[0105] If the status value of the component time series data is a long-term trend change indicator, then the status of the component time series data is a normal influence status, that is, the data fluctuations in the component time series data are all caused by normal physiological activities; if the status value of the component time series data is a short-term fluctuation change indicator, then the status of the component time series data is a temperature influence status, that is, the data fluctuations in the component time series data are more likely to be caused by temperature changes; if the status value of the component time series data is a comprehensive influence indicator, then the status of the component time series data is caused by both physiological activities and temperature changes, so it is in a comprehensive influence state.

[0106] After determining the state to which the component time series data belongs, the data can be processed differently according to different states. When the component time series data belongs to the normal influence state, there is no need to correct or adjust the data, and it can characterize the actual blood sugar changes of infants and young children. Therefore, the component time series data belonging to the normal influence state is directly used as the standard component.

[0107] For the component time series data that is not in a normal impact state, because its main part is still the blood sugar data with normal changes, and the normal blood sugar data is in a long-term relatively stable change trend, it can be regarded as a low-frequency process, and the frequency domain analysis can identify the dominant frequency components in the data. Therefore, for the component time series data that is not in a normal impact state, in an embodiment of the present invention, it is corrected based on the frequency domain information in the component time series data and the residual time series data to obtain the corrected component.

[0108] Preferably, in one embodiment of the present invention, the method for obtaining the correction component includes:

[0109] For any component time series data, if the component time series data is in a temperature-affected state, Fourier transform is performed on the component time series data to obtain first frequency domain data.

[0110] A low-pass filter is a circuit that allows low-frequency signals to pass through but blocks high-frequency signals from passing through. It can be used to remove high-frequency components in the signal and retain low-frequency useful information. Therefore, a first low-pass filter is preset here, and the first frequency domain data is low-pass filtered based on the first low-pass filter, so as to filter out the high-frequency data and obtain the first filtered data. At this time, the first filtered data is the data from which the high-frequency information generated by the temperature effect is filtered out. Then, the first filtered data is inversely Fourier transformed to obtain the corrected component of the component time series data. It should be noted that the frequency threshold in the first low-pass filter here is set to the frequency value in the first quarter of the descending sequence of all frequencies in the first frequency domain data. The specific value can be adjusted according to the implementation scenario and is not limited here.

[0111] If the component time series data is in a comprehensive influence state, the factors affecting the component time series data are relatively complex. In the embodiment of the present invention, the residual time series data is considered to be the long-term stable blood sugar data change trend of infants and young children, that is, the component of blood sugar data that is not affected by temperature (or is rarely affected by temperature). All frequencies of the corresponding frequency domain data are probably normal frequencies. Therefore, the component time series data is filtered with the help of the frequency domain information in the residual time series data:

[0112] Perform Fourier transform on the residual time series data to obtain comparison frequency domain data, and take the mean of all frequencies in the comparison frequency domain data as the comparison frequency value; and perform Fourier transform on the component time series data to obtain second frequency domain data.

[0113] The second low-pass filter is set with the comparison frequency value as a threshold, and the second frequency domain data is low-pass filtered based on the second low-pass filter. The obtained second filtered data is inversely Fourier transformed to obtain the corrected component of the component time series data.

[0114] So far, after the above steps, the standard component and the correction component corresponding to the component time series data of the initial blood glucose absorption rate time series data of the infant can be obtained.

[0115] Step S4: synthesize all standard components and correction components to obtain corrected blood glucose absorption rate time series data.

[0116] The standard component is the component time series data that is not affected by temperature, and the corrected component is the component time series data that eliminates the influence of temperature. Therefore, by synthesizing all the standard components and corrected components, we can obtain the corrected blood glucose absorption rate time series data, which is used to accurately characterize the actual blood glucose changes of infants and young children.

[0117] Preferably, in one embodiment of the present invention, the method for obtaining the time series data of the corrected blood glucose absorption rate includes:

[0118] All standard components and correction components are synthesized based on the empirical mode decomposition algorithm to obtain the corrected blood glucose absorption rate time series data. It should be noted that the empirical mode decomposition algorithm is a well-known technology and will not be described in detail here.

[0119] In other embodiments of the present invention, all standard components, correction components and residual time series data may be synthesized based on the empirical mode decomposition algorithm to obtain the corrected blood glucose absorption rate time series data.

[0120] After obtaining the corrected blood glucose absorption rate time series data of infants and young children, the corrected blood glucose absorption rate time series data can be monitored and judged by the general threshold method. That is, when the blood glucose absorption rate exceeds the normal range, an early warning is needed to remind the doctor to make further judgments and analyses. Finally, a health record can be established for the infant based on the doctor's analysis results to prevent obesity and metabolic diseases or to prevent and control early metabolic diseases.

[0121] In summary, firstly, the initial blood glucose absorption rate time series data of infants and young children is obtained, and the initial blood glucose absorption rate time series data is decomposed into residual time series data and multiple component time series data, which can analyze the changes in blood glucose absorption rate in more detail, and help to identify the influence of different factors on blood glucose metabolism. The influence of normal physiological behavior on blood glucose in blood is an obvious trend change, and it is a long-term influence behavior. Therefore, for each component time series data, the change trend of the data value is analyzed, so as to determine the long-term trend change index of each component time series data, which is used to characterize the situation that the change of the component time series data is affected by normal physiological activities; the influence of temperature change on blood glucose in infants and young children is a short-term behavior, because temperature change is often a short-term change process, so the embodiment of the present invention is based on the local change amplitude and fluctuation of the data value in the component time series data, and determines the short-term fluctuation change index, which is used to characterize the situation that the change of the component time series data is affected by temperature; and when it is affected by physiological activities and temperature, the component time series data will have both data fluctuation characteristics caused by normal physiological behavior and fluctuation characteristics affected by temperature, so based on the overall fluctuation of the data value and the local difference, the comprehensive influence index is determined. Then, based on the above three indicators, the state of each component time series data is determined. If it belongs to the normal influence state, then the component time series data is a standard component, which means that it can characterize the actual blood sugar changes of infants and young children. If it does not belong to the normal influence state, it means that it is affected by temperature, so it needs to be corrected. Since frequency domain analysis can reveal the frequency components of the data and identify the dominant frequencies in the data, the frequency domain information in the component time series data and the residual time series data is used in the embodiment of the present invention to correct the component time series data to obtain a corrected component. The corrected component at this time can eliminate the influence of temperature. Therefore, finally, by synthesizing all the standard components and the corrected components, the corrected blood sugar absorption rate time series data can be obtained, thereby more accurately reflecting the actual changes in blood sugar in the blood of infants and young children, and ensuring the accuracy of subsequent obesity metabolic disease monitoring.

[0122] 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.

[0123] 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 blood spectral analysis method for preventing and controlling obesity metabolic diseases in infants and young children, characterized in that: The method comprises: Acquire initial blood glucose absorption rate time series data of infants and young children; decompose the initial blood glucose absorption rate time series data to obtain residual time series data and multiple component time series data; For any component time series data, based on the change trend of the data value, determine the long-term trend change index of the component time series data; based on the local change amplitude and fluctuation of the data value, determine the short-term fluctuation change index of the component time series data; based on the overall fluctuation of the data value and the local difference, determine the comprehensive impact index of the component time series data; Based on the long-term trend change index, the short-term fluctuation change index and the comprehensive impact index, the state of the component time series data is judged; if the component time series data belongs to the normal impact state, the component time series data is used as the standard component; if the component time series data does not belong to the normal impact state, it is corrected based on the frequency domain information in the component time series data and the residual time series data, so as to obtain the corrected component; All standard components and correction components are synthesized to obtain the corrected blood glucose absorption rate time series data.

2. A blood spectral analysis method for preventing and controlling obesity metabolic diseases in infants and young children according to claim 1, characterized in that: The method for obtaining the long-term trend change indicator includes: For any component time series data, obtain the time series data curve corresponding to the component time series data; On the time series data curve, the slope value at each data point is obtained, and the absolute value of the difference between the slope value of each data point and the next adjacent data point is calculated as the trend change factor corresponding to each data point; The mean of the trend change factors of all data points is negatively correlated and normalized to obtain the value as the local stability factor; Calculate the mean of all data values ​​in the component time series data as the mean eigenvalue; Calculate the absolute value of the difference between each data value and the mean eigenvalue as the deviation factor, and take the mean of all deviation factors as the overall change factor; The normalized value of the sum of the local stability factor and the overall change factor is used as the long-term trend change indicator of the component time series data.

3. A blood spectral analysis method for preventing and controlling obesity metabolic diseases in infants and young children according to claim 1, characterized in that: The method for obtaining the short-term fluctuation change index includes: Determine the local neighborhood corresponding to each data point; In the local neighborhood of each data point, the difference between the maximum data value and the minimum data value is calculated as the local fluctuation amplitude value corresponding to each data point; In the local neighborhood of the next data point adjacent to each data point, the variance of all data values ​​is calculated as the lag fluctuation factor corresponding to each data point; Determine the local fluctuation factor corresponding to each data point according to the local fluctuation amplitude value and the lag fluctuation factor corresponding to each data point, wherein the local fluctuation amplitude value is positively correlated with the local fluctuation factor, and the lag fluctuation factor is negatively correlated with the local fluctuation factor; The normalized sum of the local fluctuation factors of all data points in the component time series data is used as the short-term fluctuation change indicator of the component time series data.

4. A blood spectral analysis method for preventing and controlling obesity metabolic diseases in infants and young children according to claim 2, characterized in that: The method for obtaining the comprehensive impact index includes: Calculate the absolute value of the difference between each data point and the next adjacent data point as the local difference value corresponding to each data point; The mean of all local difference values ​​is taken as the local difference feature value; The value obtained by normalizing the sum of the local difference characteristic value and the overall change factor is used as the comprehensive impact indicator of the component time series data.

5. A blood spectral analysis method for preventing and controlling obesity metabolic diseases in infants and young children according to claim 1, characterized in that: The determining of the state of the component time series data based on the long-term trend change index, the short-term fluctuation change index and the comprehensive impact index includes: For any component time series data, the maximum value among the long-term trend change index, short-term fluctuation change index and comprehensive impact index of the component time series data is taken as the state value; If the state value of the component time series data is a long-term trend change indicator, the state of the component time series data is a normal impact state; If the state value of the component time series data is a short-period fluctuation change indicator, the state of the component time series data is a temperature-affected state; If the status value of the component time series data is a comprehensive impact index, then the status of the component time series data is a comprehensive impact status.

6. A blood spectral analysis method for preventing and controlling obesity metabolic diseases in infants and young children according to claim 5, characterized in that: When the component time series data is in a temperature-affected state, the method for obtaining the correction component includes: For any component time series data, if the component time series data is in a temperature-affected state, Fourier transform is performed on the component time series data to obtain first frequency domain data; A first low-pass filter is preset, and the first frequency domain data is low-pass filtered based on the first low-pass filter, and the obtained first filtered data is inversely Fourier transformed to obtain a corrected component of the component time series data.

7. A blood spectral analysis method for preventing and controlling obesity metabolic diseases in infants and young children according to claim 5, characterized in that: When the component time series data is in a comprehensive influence state, the method for obtaining the correction component includes: For any component time series data, if the component time series data is in a comprehensive influence state, Fourier transform is performed on the component time series data to obtain second frequency domain data; Performing Fourier transform on the residual time series data to obtain comparison frequency domain data, and taking the mean of all frequencies in the comparison frequency domain data as the comparison frequency value; A second low-pass filter is set according to the comparison frequency value, and the second frequency domain data is low-pass filtered based on the second low-pass filter, and the obtained second filtered data is inversely Fourier transformed to obtain a corrected component of the component time series data.

8. A blood spectral analysis method for preventing and controlling obesity metabolic diseases in infants and young children according to claim 1, characterized in that: The initial blood glucose absorption rate time series data is decomposed to obtain residual time series data and multiple component time series data, including: The initial blood glucose absorption rate time series data is decomposed based on an empirical mode decomposition algorithm to obtain multiple component time series data and residual time series data.

9. A blood spectral analysis method for preventing and controlling obesity metabolic diseases in infants and young children according to claim 1, characterized in that: The method for obtaining the time series data of the corrected blood glucose absorption rate comprises: All standard components and correction components are synthesized based on the empirical mode decomposition algorithm to obtain the corrected blood glucose absorption rate time series data.

10. A blood spectral analysis method for preventing and controlling obesity metabolic diseases in infants and young children according to claim 1, characterized in that: The method for acquiring the initial blood glucose absorption rate time series data comprises: Obtain blood spectrum data of infants and young children at different times; The characteristic wavelength is preset, the absorbance of the characteristic wavelength in the blood spectrum data at all times is extracted and sorted in time series, so as to obtain the initial blood glucose absorbance time series data.

Images