Intelligent noninvasive blood glucose detection method suitable for diabetics

Abstract

The invention relates to the technical field of blood glucose detection, in particular to an intelligent noninvasive blood glucose detection method suitable for diabetics. The method comprises the following steps: acquiring physiological data and NIR wavelengths of different dimensions of a plurality of diabetic reference patients; obtaining decision-making correlation values according to the aggregation distribution characteristics and the numerical value distribution characteristics of the physiological data of each dimension of the reference patient, and screening an optimal dimension by combining the difference of the physiological data of each two dimensions of different reference patients and the difference of the decision-making correlation values; according to the difference of the wavelengths corresponding to every two categories and the number distribution of similar patients of the reference patient, combining the different categories to obtain a plurality of groups; and determining a target wavelength by combining all wavelengths corresponding to each group, the physiological data of the preferable dimension of the reference patient in each group and the physiological data of the preferable dimension of the diabetic patient to be detected, and detecting the blood sugar of the diabetic patient to be detected by using the target wavelength. The accuracy of the blood glucose detection result of the diabetic patient is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of blood sugar detection, and in particular to an intelligent non-invasive blood sugar detection method suitable for diabetic patients. Background Art

[0002] Diabetes is a chronic metabolic disease characterized by high blood sugar. The continuous fluctuation of blood sugar levels in patients can cause damage to the patient's organs and tissues, and may cause a variety of complications such as cardiovascular disease, retinopathy, neuropathy and nephropathy. Near-infrared (NIR) spectroscopy technology has become the focus of research on non-invasive blood sugar detection due to its non-invasive and rapid characteristics. It uses the light absorption or scattering characteristics at different wavelengths to reflect blood sugar concentration. After near-infrared light contacts the skin surface, part of the light is absorbed by glucose in the blood. The received light of the photoelectric sensor will present a periodic band-like and carry glucose information. The appropriate NIR wavelength is obtained by invasively detecting the absorption spectrum of blood sugar, and then a unified blood sugar prediction model is established based on the linear relationship between blood sugar and NIR absorption rate, realizing the detection of blood sugar using spectral information.

[0003] Since there are significant individual differences in the physiological characteristics, skin tissue composition, blood sugar fluctuation range and dynamic regulation mechanism of diabetic patients, there are differences in the appropriate NIR wavelengths for patients. Therefore, inappropriate wavelength settings will lead to inaccurate blood sugar test results for patients. Summary of the invention

[0004] In order to solve the problem that the inappropriate wavelength setting of the existing near-infrared spectroscopy technology when performing blood glucose testing on diabetic patients may lead to inaccurate blood glucose testing results, the purpose of the present invention is to provide an intelligent non-invasive blood glucose testing method suitable for diabetic patients. The technical scheme adopted is as follows:

[0005] The present invention provides an intelligent non-invasive blood sugar detection method suitable for diabetic patients, the method comprising the following steps:

[0006] Obtain physiological data of different dimensions and NIR wavelengths of a preset number of diabetic reference patients;

[0007] Obtaining the decision-related value of the physiological data of each dimension according to the clustered distribution characteristics and numerical distribution characteristics of the physiological data of each dimension of all reference patients; evaluating the interchangeability of the physiological data of each two dimensions according to the differences in the physiological data of different reference patients in each dimension and the differences in the decision-related values; and selecting the preferred dimension in combination with all the decision-related values ​​and all the interchangeability degrees;

[0008] Based on the differences between the physiological data of the preferred dimensions of the reference patients in different categories, similar patients in other categories of the reference patients in each category are screened, wherein the wavelengths of all the reference patients in the same category are the same; according to the differences in the wavelengths corresponding to each two categories and the number distribution of the similar patients of the reference patients in each two categories, different categories are merged to obtain several groups;

[0009] The target wavelength is determined by combining all wavelengths corresponding to each group, the physiological data of the preferred dimension of the reference patient in each group, and the physiological data of the preferred dimension of the diabetic patient to be detected, and the blood glucose of the diabetic patient to be detected is detected using the target wavelength.

[0010] Preferably, obtaining the decision-related value of the physiological data of each dimension according to the aggregate distribution characteristics and the numerical distribution characteristics of the physiological data of each dimension of all reference patients includes:

[0011] Calculate the difference in the number of all reference patients in each two categories and the difference in the mean values ​​of the physiological data of the candidate dimensions of the reference patients in each two categories;

[0012] Obtaining a decision-related value of the physiological data of the candidate dimension according to all of the quantity differences and all of the average value differences;

[0013] The candidate dimension is any dimension.

[0014] Preferably, the step of comprehensively evaluating the difference in physiological data of each two dimensions between different reference patients and the difference in decision-related values ​​to evaluate the interchangeability of physiological data of each two dimensions includes:

[0015] respectively calculating the first difference between the physiological data of each dimension of each two reference patients;

[0016] For any two dimensions:

[0017] Obtaining first characteristic values ​​of each two reference patients in the arbitrary two dimensions according to the first differences corresponding to each two reference patients in the arbitrary two dimensions;

[0018] According to the differences between all the first eigenvalues ​​of all the reference patients in the any two dimensions and the differences between the decision-related values ​​of the physiological data of the any two dimensions, the degree of interchangeability of the physiological data of the any two dimensions is obtained, and the differences between all the first eigenvalues ​​and the differences between the decision-related values ​​are negatively correlated with the degree of interchangeability.

[0019] Preferably, obtaining the first characteristic values ​​of every two reference patients in the arbitrary two dimensions according to the first differences corresponding to every two reference patients in the arbitrary two dimensions includes:

[0020] Calculating a first sum value between the first difference corresponding to one dimension of the arbitrary two dimensions for each two reference patients and a preset first adjustment parameter, and a second sum value between the first difference corresponding to the other dimension of the arbitrary two dimensions for each two reference patients and a preset first adjustment parameter; wherein the preset first adjustment parameter is greater than 0;

[0021] The ratio between the first sum value and the second sum value is determined as the first feature values ​​corresponding to the two reference patients in the arbitrary two dimensions.

[0022] Preferably, the combining of all the decision-related values ​​and all the interchangeability degrees to select the preferred dimension comprises:

[0023] The product of the interchangeability between the candidate dimension and the physiological data of each of its related dimensions and the decision-related value of the physiological data of each of the related dimensions of the candidate dimension is recorded as the second eigenvalue of each of the related dimensions of the candidate dimension; wherein the related dimension of the candidate dimension is a dimension whose interchangeability with the candidate dimension is greater than a preset interchangeability threshold;

[0024] Combining the second eigenvalues ​​of all related dimensions of the candidate dimension and the interchangeability of the physiological data of the candidate dimension, obtaining the preference degree of the candidate dimension, wherein the second eigenvalue is negatively correlated with the preference degree, and the interchangeability of the physiological data of the candidate dimension is positively correlated with the preference degree;

[0025] Filter the preferred dimensions based on the relative preference of all dimensions.

[0026] Preferably, the preferred dimensions are screened according to the magnitude relationship of the preferred degrees of all dimensions, including:

[0027] Sort all dimensions in descending order of preference to obtain a dimension sequence;

[0028] The first preset number of dimensions in the dimension sequence are determined as preferred dimensions.

[0029] Preferably, the method of screening similar patients in other categories of the reference patients in each category based on the differences between the physiological data of the preferred dimensions of the reference patients in different categories comprises:

[0030] Obtaining a similarity value between the first patient and the second patient according to a difference between the physiological data of the first patient and the second patient in the same preferred dimension, wherein the difference between the physiological data of the same preferred dimension is negatively correlated with the similarity value;

[0031] If the similarity value is greater than a preset similarity threshold, it is determined that the first patient and the second patient are similar patients to each other;

[0032] The first patient and the second patient are two reference patients in different categories.

[0033] Preferably, the merging of different categories to obtain a plurality of groups according to the difference in wavelengths corresponding to each two categories and the number distribution of similar patients to the reference patients in each two categories comprises:

[0034] Calculate the third sum between the number of all similar patients in the second category of the reference patient in the first category and the number of all similar patients in the first category of the reference patient in the second category; obtain the fusion degree of the arbitrary two categories according to the difference between the wavelengths corresponding to the first category and the second category and the third sum, the difference between the wavelengths is negatively correlated with the fusion degree, and the third sum is positively correlated with the fusion degree; wherein the first category and the second category are any two categories among all categories;

[0035] Based on the fusion degree of every two categories, the corresponding categories are merged to obtain several groups.

[0036] Preferably, the corresponding categories are merged based on the fusion degree of each two categories to obtain several groups, including:

[0037] Sorting all categories in a preset order to obtain a category sequence, wherein the preset order is an order from small to large wavelength or an order from large to small wavelength;

[0038] The two categories with the largest fusion degree are merged, and each category is judged in turn according to the order of the categories in the category sequence whether the fusion degree with any category of all the merged categories is greater than the preset fusion degree threshold. If it is greater, the corresponding category is merged with the merged category; if it is less than or equal to, it is regarded as a new category; each category after all the mergers are completed is regarded as a group.

[0039] Preferably, the step of determining the target wavelength by combining all wavelengths corresponding to each group, the physiological data of the preferred dimension of the reference patient in each group, and the physiological data of the preferred dimension of the diabetic patient to be detected includes:

[0040] Training the wavelength selection model using all wavelengths corresponding to each group and the physiological data of the preferred dimensions of the reference patients in each group to obtain a trained wavelength selection model;

[0041] The physiological data of the preferred dimension of the diabetic patient to be detected is input into the trained wavelength selection model to obtain the target wavelength.

[0042] The present invention has at least the following beneficial effects:

[0043] The present invention firstly calculates the decision-related value of the physiological data of each dimension according to the clustered distribution characteristics and numerical distribution characteristics of the physiological data of each dimension of all reference patients, and then evaluates the interchangeability of the physiological data of different dimensions by comprehensively analyzing the differences in the physiological data of each two dimensions of different reference patients and the differences in the decision-related values, and obtains the corresponding interchangeability. Since the collected physiological data have many dimensions, the data of some dimensions have little effect on the test results. In order to reduce the amount of calculation, the present invention selects a plurality of preferred dimensions from all dimensions in combination with the decision-related value and the interchangeability, that is, the physiological data are screened in combination with the different characteristics of the influence of different dimensions on blood glucose detection, and the physiological data of the preferred dimensions of reference patients of different categories are selected. The differences between the two groups were obtained, and similar patients of the reference patients in each category in other categories were screened. According to the differences in wavelengths corresponding to different categories and the number distribution of similar patients of the reference patients, different categories were merged to obtain multiple groups. Further, all wavelengths corresponding to each group, the physiological data of the preferred dimensions of the reference patients in each group, and the physiological data of the preferred dimensions of the diabetic patients to be detected were combined to select a near-infrared wavelength for the diabetic patients to be detected, and the blood glucose of the diabetic patients to be detected was detected using this wavelength. The method provided by the present invention makes the wavelength selection of the diabetic patients to be detected more accurate and reliable, and uses the target wavelength as the wavelength for non-invasive blood glucose detection of the diabetic patients to be detected for blood glucose detection, thereby improving the accuracy of the blood glucose detection results of diabetic patients. BRIEF DESCRIPTION OF THE DRAWINGS

[0044] 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.

[0045] Figure 1 A flow chart of an intelligent non-invasive blood glucose detection method suitable for diabetic patients provided by an embodiment of the present invention;

[0046] Figure 2 This is a structural block diagram of an intelligent non-invasive blood glucose detection system suitable for diabetic patients provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0047] In order to further explain the technical means and effects adopted by the present invention to achieve the predetermined purpose of the invention, the following is a detailed description of an intelligent non-invasive blood glucose detection method suitable for diabetic patients proposed in accordance with the present invention in combination with the accompanying drawings and preferred embodiments.

[0048] 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.

[0049] The following is a detailed description of a specific solution of an intelligent non-invasive blood glucose detection method suitable for diabetic patients provided by the present invention in conjunction with the accompanying drawings.

[0050] An embodiment of an intelligent non-invasive blood sugar detection method suitable for diabetic patients:

[0051] The specific scenario targeted by this embodiment is: when performing non-invasive blood glucose testing on diabetic patients, since the physiological characteristics of different diabetic patients may vary greatly, the traditional fixed wavelength or unified blood glucose prediction model often has certain limitations. The method provided in this embodiment can provide different wavelength options for diabetic patients with different physiological characteristics, thereby improving the accuracy and reliability of blood glucose test results.

[0052] This embodiment proposes an intelligent non-invasive blood sugar detection method suitable for diabetic patients. Figure 1 As shown, an intelligent non-invasive blood glucose detection method suitable for diabetic patients in this embodiment includes the following steps:

[0053] Step S1, obtaining physiological data of different dimensions and NIR wavelengths of a preset number of diabetic reference patients.

[0054] First, the electronic medical records, ages, several symptom levels, several disease levels, smoking indexes, drinking indexes, etc. of a preset number of diabetic patients are obtained, wherein the symptom level is obtained by the doctor scoring the level of each symptom, and the symptoms include polyuria, weight loss, blurred vision, etc.; the disease level is also obtained by the doctor scoring the level of each disease, and the diseases include hypertension, coronary heart disease, and common complications of diabetic patients; in this embodiment, smoking index = number of cigarettes smoked per day × number of years of smoking, drinking index = weekly alcohol consumption × number of years of drinking. Then, the physiological indicators of these diabetic patients are collected, including skin reflectivity (index finger pulp), tissue thickness, BMI index, heart rate, blood pressure, and several components of blood; each type of data collected is used as physiological data of a dimension, that is, the physiological data of different dimensions of these patients are collected, and the physiological data of these patients will be used to train the wavelength selection model. Therefore, these patients are recorded as reference patients, and all the collected physiological data are normalized to eliminate the influence of dimension. The normalization of data is a prior art and will not be described in detail here. The physiological data mentioned later are all normalized data, that is, the physiological data of multiple dimensions of multiple reference patients are collected. It should be noted that in this embodiment, the electronic medical records and physiological data of all diabetic patients are collected with permission and are non-privacy data. In this embodiment, the preset number is 500, that is, the number of reference patients is 500. In specific applications, the implementer can set it according to the specific situation.

[0055] The absorption spectrum of blood glucose is measured for each reference patient by drawing venous blood, and the glucose absorption peak is obtained as the NIR wavelength corresponding to the patient.

[0056] So far, this embodiment has collected physiological data of different dimensions of each diabetic reference patient and the NIR wavelength corresponding to each diabetic reference patient.

[0057] Step S2, obtaining the decision-related value of the physiological data of each dimension according to the clustered distribution characteristics and numerical distribution characteristics of the physiological data of each dimension of all reference patients; evaluating the degree of interchangeability of the physiological data of each two dimensions by comprehensively analyzing the differences in the physiological data of different reference patients in each dimension and the differences in the decision-related values; and screening the preferred dimension by combining all the decision-related values ​​and all the interchangeability degrees.

[0058] This embodiment is to measure the blood sugar of the diabetic patient in real time in a non-invasive way, so it is extremely important to select the best NIR wavelength to show the patient's blood sugar. The method of drawing venous blood is not only harmful, but also the best NIR wavelength of the diabetic patient may fluctuate. Therefore, this embodiment considers that by analyzing the relationship between the patient's physiological data and the NIR wavelength, the purpose of predicting the patient's best NIR wavelength only through the patient's physiological data can be achieved, and when the patient has symptom changes or physical changes, the NIR wavelength for measuring blood sugar can be adjusted in time to ensure the accuracy of blood sugar measurement.

[0059] First, it is necessary to determine the relationship between the patient's physiological data and the NIR wavelength. The random forest model predicts the category of the sample by constructing multiple decision trees, which is adapted to the purpose. Therefore, this method uses the random forest algorithm for training and learning.

[0060] Since not all dimensions of physiological data collected in step S1 will affect the optimal NIR wavelength, in order to optimize the model, the dimensions should be screened and the dimensions with different impacts should be weighted to improve the model classification effect. In addition, since the optimal NIR wavelength may have a large range, the solution of using each wavelength as a method has certain problems in implementation. By using the differences between patients corresponding to different NIR wavelengths, some similar available NIR wavelengths can be merged, which can improve the generalization ability of the model and provide patients with multiple choices.

[0061] This embodiment will first calculate the decision-related value of the physiological data of each dimension based on the clustered distribution characteristics and numerical distribution characteristics of the physiological data of each dimension of all reference patients, and then combine the differences in the physiological data of every two dimensions of different reference patients to screen the preferred dimension from all dimensions. The physiological data of the preferred dimension will be used for wavelength selection.

[0062] This embodiment is explained by taking one dimension as an example, and the method provided in this embodiment can be used to process other dimensions. Specifically, any dimension is recorded as a candidate dimension, and all reference patients corresponding to the same wavelength are taken as a category, that is, all reference patients are divided into multiple categories according to the wavelength, and the wavelengths corresponding to all patients in the same category are equal, and the wavelengths corresponding to patients in different categories are not equal. According to the physiological data of the candidate dimensions of all reference patients in each category, the average value of the physiological data of the candidate dimensions of the reference patients in each category is calculated. It should be noted that: each category corresponds to an average value under the candidate dimension. The difference in the number of all reference patients in each two categories and the difference in the average values ​​of the physiological data of the candidate dimensions of the reference patients in each two categories are calculated respectively. According to the differences in all the quantitative differences and all the average values, the decision-related values ​​of the physiological data of the candidate dimensions are obtained.

[0063] In this embodiment, a specific calculation formula for the decision-related value is given, and the decision-related value of the physiological data of the i-th dimension can be expressed as:

[0064]

[0065] Among them, L i represents the decision-related value of the physiological data of the i-th dimension, A represents the number of categories, represents the number of all reference patients in the sth category, represents the number of all reference patients in the t-th category, C s represents the average value of the physiological data of the candidate dimension of the reference patient in the sth category, C t represents the average value of the physiological data of the candidate dimension of the reference patient in the tth category, norm() represents the normalization function, and || represents the absolute value sign.

[0066] represents the cumulative sum of the differences in the number of reference patients in different categories, Characterizes the difference in the average values ​​of the physiological data of the i-th dimension of reference patients in different categories, both of which reflect the preliminary decision-making relevance of the physiological data of the i-th dimension in terms of quantity and density.

[0067] For any two dimensions, if the changes in physiological data are correlated between any two patients, that is, they are larger for one patient and smaller for another patient, or one dimension is larger for one patient and smaller for the other, and the opposite is true for another patient, then the correlation between the physiological data of these two dimensions is high. At the same time, if the decision correlation is also similar, then these two features are interchangeable for wavelength selection decisions.

[0068] Based on this, the difference between the physiological data of each dimension of each two reference patients is calculated respectively, and the difference is recorded as the first difference. For any two dimensions: calculate the first sum value between the first difference corresponding to one dimension of the two reference patients and the preset first adjustment parameter, and the second sum value between the first difference corresponding to the other dimension of the two reference patients and the preset first adjustment parameter; determine the ratio between the first sum value and the second sum value as the first eigenvalue corresponding to the two reference patients in the two dimensions. According to the difference between all the first eigenvalues ​​of all reference patients in the two dimensions and the difference between the decision-related values ​​of the physiological data of the two dimensions, the degree of interchangeability of the physiological data of the two dimensions is obtained, and the difference between all the first eigenvalues ​​and the difference between the decision-related values ​​are both negatively correlated with the degree of interchangeability.

[0069] Among them, the negative correlation relationship means 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 practical applications.

[0070] In this embodiment, a specific calculation formula for the first eigenvalue and the degree of interchangeability is given, which can be specifically expressed as:

[0071]

[0072] Among them, M i,j represents the degree of interchangeability between the physiological data of the i-th dimension and the physiological data of the j-th dimension, L i represents the decision-related value of the physiological data of the i-th dimension, L j represents the decision-related value of the physiological data of the jth dimension, Di x represents the physiological data of the ith dimension of the xth reference patient, Di y represents the physiological data of the ith dimension of the yth reference patient, Dj x represents the physiological data of the jth dimension of the xth reference patient, Dj y represents the physiological data of the jth dimension of the yth reference patient, represents the first eigenvalue of the x-th reference patient and the y-th reference patient in the i-th dimension and the j-th dimension, represents the nth first eigenvalue of all reference patients in the i-th dimension and the j-th dimension, represents the rth first eigenvalue of all reference patients in the i-th dimension and the j-th dimension, and λ1 is the preset first adjustment parameter.

[0073] The purpose of introducing the preset first adjustment parameter into the calculation formula of the first eigenvalue in this embodiment is to prevent the denominator from being 0. In this embodiment, the preset first adjustment parameter is 0.01. In specific applications, the implementer can set it according to specific circumstances. norm(|Di x -Di y |) represents the first difference between the physiological data of the xth reference patient and the yth reference patient in the i-th dimension, norm(|Dj x -Dj y |) represents the first difference between the physiological data of the xth reference patient and the yth reference patient in the jth dimension, norm(|Di x -Di y |)+λ1 represents the first sum value, norm(|Dj x -Dj y |)+λ1 represents the second sum value.

[0074] The greater the difference in decision-related values ​​of physiological data between the i-th dimension and the j-th dimension, and the greater the difference between different first eigenvalues, the more inconsistent the changes between the i-th dimension and the j-th dimension among different patients, and the smaller the interchangeability between the i-th dimension and the j-th dimension, that is, the smaller the interchangeability between the physiological data of the i-th dimension and the physiological data of the j-th dimension.

[0075] In multiple dimensions with a high degree of interchangeability, the higher the decision relevance, the greater the possibility that it will be selected as a preferred feature to participate in model building, and the degree of preference of each dimension is calculated accordingly.

[0076] Next, the candidate dimension is still used as an example for explanation. Specifically, the dimension whose degree of interchangeability with the candidate dimension is greater than the preset interchangeability threshold among all dimensions is taken as the related dimension of the candidate dimension. In this embodiment, the preset interchangeability threshold is 0.5. In specific applications, the implementer can set it according to the specific situation. The product of the degree of interchangeability of the physiological data of the candidate dimension and each of its related dimensions and the decision-related value of the physiological data of each related dimension of the candidate dimension is recorded as the second eigenvalue of each related dimension of the candidate dimension. It should be noted that each related dimension of the candidate dimension corresponds to a second eigenvalue. Combining the second eigenvalues ​​of all related dimensions of the candidate dimension and the degree of interchangeability of the physiological data of the candidate dimension, the preferred degree of the candidate dimension is obtained. The second eigenvalue is negatively correlated with the preferred degree, and the degree of interchangeability of the physiological data of the candidate dimension is positively correlated with the preferred degree.

[0077] Among them, the 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; the positive correlation relationship indicates that the dependent variable will increase as the independent variable increases, and the dependent variable will decrease as the independent variable decreases. It can be an addition relationship, a multiplication relationship, etc., which is determined by actual application.

[0078] In this embodiment, a specific calculation formula for the preference degree is given, and the preference degree of the i-th dimension can be expressed as:

[0079]

[0080] Among them, P i Indicates the preference of the i-th dimension, L i represents the decision-related value of the physiological data of the i-th dimension, G represents the number of relevant dimensions of the i-th dimension, and M i,g represents the interchangeability of physiological data of the g-th related dimension of the i-th dimension, L i,g Represents the decision-related value of the g-th related dimension of the i-th dimension.

[0081] M i,g ×L i,g Represents the second eigenvalue of the g-th related dimension of the i-th dimension. The proportion of decision-related values ​​of physiological data used to characterize the i-th dimension. The higher the proportion, the higher the decision-related relevance when its related dimensions can replace the i-th dimension. Then the i-th dimension should be selected as the preferred feature among all related dimensions, that is, the greater the preference of the i-th dimension.

[0082] By adopting the above method, the preference degree of each dimension can be obtained, and then the preferred dimensions are selected according to the size relationship of the preference degrees of all dimensions. Specifically, all dimensions are sorted in order from large to small in preference degree to obtain a dimension sequence, and the first preset number of dimensions in the dimension sequence are determined as preferred dimensions, that is, multiple preferred dimensions are selected from all dimensions. In this embodiment, the preset number is 50% of the total number of all dimensions. It should be noted that: if 50% of the total number of all dimensions is not an integer, it is rounded up and the result is used as the preset number.

[0083] Step S3, based on the differences between the physiological data of the preferred dimensions of the reference patients of different categories, screen similar patients of each category of reference patients in other categories, wherein the wavelengths of all reference patients in the same category are the same; according to the differences in the wavelengths corresponding to every two categories and the number distribution of similar patients of the reference patients in every two categories, merge the different categories to obtain several groups.

[0084] After obtaining the preferred dimensions, the NIR wavelengths of different patients are analyzed. The NIR wavelength is 780-2526nm, so the range of the optimal NIR wavelength is relatively large. However, the benefits of setting up such a number of blood glucose monitors with different wavelengths are relatively small. If the information sequences of two patients with different optimal NIR wavelengths are similar, then the optimal NIR wavelengths used by these two types of patients can also be replaced and fused into one band. This can reduce the possible types of classification results when establishing a random forest, and improve the accuracy and generalization ability of wavelength selection.

[0085] Since the data in the physiological data of the preferred dimension will have a certain influence on the selection of the NIR wavelength, the difference in the physiological data of the preferred dimension of different categories of patients can represent the difference in the NIR wavelength.

[0086] Physiological data has various dimensions, and patients also have different physiological data in preferred dimensions. Therefore, the relationship between physiological data in preferred dimensions and NIR wavelength is diverse. The following four are considered:

[0087] a) physiological data of different preferred dimensions have the same NIR wavelength;

[0088] b) Physiological data of different preferred dimensions have different NIR wavelengths;

[0089] c) physiological data of the same preferred dimension have the same NIR wavelength;

[0090] d) Physiological data of the same preferred dimension have different NIR wavelengths.

[0091] For a), patients belonging to the same category have physiological data of preferred dimensions with large differences, indicating that the effects of the preferred dimensions on the NIR wavelengths are offset; for b), it indicates that the effects are not completely offset; for c), it indicates that consistent physiological conditions ideally use consistent NIR wavelengths; for d), it indicates that different NIR wavelengths are fusible. Therefore, when fusing NIR wavelengths, only the d) condition is considered.

[0092] Among reference patients in different categories, the more patients have physiological data of the same preferred dimension, the more different categories should be merged.

[0093] Based on the above features, this embodiment first screens similar patients of each category of reference patients in other categories based on the differences between the physiological data of the preferred dimensions of reference patients of different categories. Specifically, two reference patients in different categories are respectively recorded as the first patient and the second patient, and the similarity value between the first patient and the second patient is obtained based on the difference between the physiological data of the same preferred dimension of the first patient and the second patient, and the difference between the physiological data of the same preferred dimension is negatively correlated with the similarity value.

[0094] Among them, the negative correlation relationship means 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 practical applications.

[0095] In this embodiment, a specific calculation formula for the similarity value is given, and the similarity value between the first patient and the second patient can be expressed as:

[0096]

[0097] Among them, P 1,2 represents the similarity value between the first patient and the second patient, H represents the number of preferred dimensions, and D 1,h represents the physiological data of the first patient in the hth preferred dimension, D 2,h represents the physiological data of the hth preferred dimension of the second patient, and exp() represents an exponential function with a natural constant as the base.

[0098] |D 1,h -D 2,h| represents the difference between the physiological data of the first patient and the second patient in the hth preferred dimension. The larger the value, the greater the difference between the two. When the difference between the physiological data of all preferred dimensions of the first patient and the second patient is larger, it means that the first patient and the second patient are less similar, that is, the similarity value between the first patient and the second patient is smaller.

[0099] The larger the similarity value, the more similar the conditions of the two patients are. Therefore, if the similarity value between the first patient and the second patient is greater than the preset similarity threshold, the first patient and the second patient are determined to be similar patients. In this embodiment, the preset similarity threshold is 0.5. In specific applications, the implementer can set it according to specific circumstances.

[0100] For any two categories, the higher the similarity between all patients, the stronger the correlation between the two categories, and the more likely the two categories are to be merged. In addition, the closer the NIR wavelengths corresponding to the two categories are, and the closer they belong to the same band, the more likely the two categories are to be merged. Therefore, next, based on the difference in wavelengths corresponding to each two categories and the number of similar patients in each two categories, the fusion between each two categories is evaluated.

[0101] The following description will be given by taking any two categories as an example, and the method provided in this embodiment can be used to process other categories.

[0102] Specifically, any two categories among all categories are recorded as the first category and the second category, and the third sum between the number of all similar patients in the second category with reference to the patient in the first category and the number of all similar patients in the first category with reference to the patient in the second category is calculated; the fusion degree of the any two categories is obtained according to the difference between the wavelengths corresponding to the first category and the second category and the third sum, the difference between the wavelengths is negatively correlated with the fusion degree, and the third sum is positively correlated with the fusion degree.

[0103] Among them, 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. It can be an additive relationship, a multiplicative relationship, etc., which is determined by actual application; a negative correlation means 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 subtractive relationship, a division relationship, etc., which is determined by actual application.

[0104] In this embodiment, a specific calculation formula for the fusion degree is given, and the fusion degree of the s-th category and the t-th category can be expressed as:

[0105]

[0106] Among them, Q s,trepresents the fusion degree of the s-th category and the t-th category, l s represents the wavelength corresponding to the sth category, l t represents the wavelength corresponding to the t-th category, X1 represents the number of reference patients in the s-th category, and R s,x1 (t) represents the number of all reference patients in the tth category for the X1th reference patient in the sth category, X2 represents the number of reference patients in the tth category, R t,x2 (s) represents the number of all reference patients in the s-th category for the X1-th reference patient in the t-th category, exp() represents an exponential function with a natural constant as the base, norm() represents a normalization function, and || represents the absolute value sign.

[0107] |l s -l t | represents the difference between the wavelengths corresponding to the sth category and the tth category. The larger the absolute value, the greater the difference between the two wavelengths. Represents the third sum value. When the wavelength difference between the sth category and the tth category is smaller and the third sum value is larger, it means that the sth category and the tth category are more suitable for fusion, that is, the fusion degree of the sth category and the tth category is higher.

[0108] By using the above method, the fusion degree of each two categories can be obtained. The greater the fusion degree, the more suitable the corresponding two categories are for merging. Therefore, the corresponding categories are then merged based on the fusion degree of each two categories to obtain several groups.

[0109] Specifically, all categories are sorted in a preset order to obtain a category sequence, wherein the preset order is the order of wavelength from small to large or the order of wavelength from large to small; in this embodiment, the preset order is the order of wavelength from small to large, that is, all categories are sorted in the order of wavelength from small to large to obtain a category sequence. Then, the two categories with the largest fusion degree are merged, and then, according to the order of categories in the category sequence, it is judged in turn whether the fusion degree of each category with any category in all merged categories is greater than the preset fusion degree threshold. If greater than, the corresponding category is merged with the merged category. If less than or equal to, it is used as a new category; all categories in the category sequence are judged in turn and the categories that need to be merged are merged, and each category after all the merges are taken as a group, that is, multiple groups are obtained. In this embodiment, the fusion degree threshold is 0.5. In specific applications, the implementer can set it according to the specific situation.

[0110] Step S4, determining a target wavelength by combining all wavelengths corresponding to each group, the physiological data of the preferred dimension of the reference patient in each group, and the physiological data of the preferred dimension of the diabetic patient to be detected, and detecting the blood glucose of the diabetic patient to be detected using the target wavelength.

[0111] Next, this embodiment will construct a wavelength selection model, and use all wavelengths corresponding to each group and the physiological data of the preferred dimensions of the reference patients in each group to train the wavelength selection model, use the trained wavelength selection model to obtain the optimal wavelength for the diabetic patients to be tested, and use their optimal wavelength to test their blood sugar.

[0112] Specifically, each reference patient was taken as a sample, and the Bootstrap method was used to resample from the sample. A subset of training samples with the same sample size was randomly selected with replacement to construct a single decision tree. This was repeated several times to form a random forest model, which was the wavelength selection model for the patient. The input of the model was the physiological data of all preferred dimensions corresponding to each sample, and the output of the model was the wavelength.

[0113] The physiological data of the preferred dimension of the diabetic patient to be detected is collected and normalized, and the normalized physiological data of the preferred dimension of the diabetic patient to be detected is input into the trained wavelength selection model to obtain the target wavelength, and the target wavelength is used to detect the blood glucose of the diabetic patient to be detected. It should be noted that: since there may be multiple wavelengths in a group, there may be more than one target wavelength obtained in the end. When the blood glucose of the diabetic patient to be detected is detected, one wavelength can be selected from all the target wavelengths for blood glucose detection, thereby realizing non-invasive detection of blood glucose of diabetic patients and improving the convenience and accuracy of blood glucose detection of patients.

[0114] This embodiment first calculates the decision-related value of the physiological data of each dimension according to the clustered distribution characteristics and numerical distribution characteristics of the physiological data of each dimension of all reference patients, and then evaluates the interchangeability of the physiological data of different dimensions by comprehensively analyzing the differences in the physiological data of each two dimensions of different reference patients and the differences in the decision-related values, and obtains the corresponding interchangeability. Since the collected physiological data has many dimensions, the data of some dimensions have little effect on the test results. In order to reduce the amount of calculation, this embodiment selects multiple preferred dimensions from all dimensions based on the decision-related values ​​and the interchangeability, that is, the physiological data is screened based on the different characteristics of the impact of different dimensions on blood glucose detection, and the physiological data of the preferred dimensions of reference patients of different categories are selected. The differences between the two groups were obtained, and similar patients of the reference patients in each category in other categories were screened. According to the differences in wavelengths corresponding to different categories and the number distribution of similar patients of the reference patients, different categories were merged to obtain multiple groups. Further, all wavelengths corresponding to each group, the physiological data of the preferred dimensions of the reference patients in each group, and the physiological data of the preferred dimensions of the diabetic patients to be detected were combined to select a near-infrared wavelength for the diabetic patients to be detected, and the blood glucose of the diabetic patients to be detected was detected using this wavelength. The method provided in this embodiment makes the wavelength selection of the diabetic patients to be detected more accurate and reliable, and uses the target wavelength as the wavelength for non-invasive blood glucose detection of the diabetic patients to be detected for blood glucose detection, thereby improving the accuracy of the blood glucose detection results of diabetic patients.

[0115] An embodiment of an intelligent non-invasive blood sugar detection system suitable for diabetic patients:

[0116] like Figure 2 As shown, the figure shows a structural block diagram of an intelligent non-invasive blood glucose detection system suitable for diabetic patients. The system includes a data acquisition module, a preferred dimension screening module, a category merging module and a blood glucose detection module.

[0117] The data acquisition module is used to obtain physiological data of different dimensions and NIR wavelengths of a preset number of diabetic reference patients;

[0118] The preferred dimension screening module is used to obtain the decision-related value of the physiological data of each dimension according to the cluster distribution characteristics and numerical distribution characteristics of the physiological data of each dimension of all reference patients; comprehensively evaluate the interchangeability of the physiological data of each two dimensions of different reference patients and the difference in decision-related values; and screen the preferred dimension in combination with all the decision-related values ​​and all the interchangeability degrees;

[0119] A category merging module is used to screen similar patients of each category of reference patients in other categories based on the differences between the physiological data of the preferred dimensions of reference patients of different categories, wherein the wavelengths of all reference patients in the same category are the same; and to merge different categories to obtain a plurality of groups according to the differences in the wavelengths corresponding to each two categories and the number distribution of the similar patients of reference patients in each two categories;

[0120] The blood glucose detection module is used to determine the target wavelength by combining all wavelengths corresponding to each group, the physiological data of the preferred dimensions of the reference patients in each group and the physiological data of the preferred dimensions of the diabetic patients to be detected, and use the target wavelength to detect the blood glucose of the diabetic patients to be detected.

[0121] It should be understood that Figure 2 The structural block diagram of an intelligent non-invasive blood glucose detection system suitable for diabetic patients and its modules can be implemented in various ways. For example, in some embodiments, the system and its modules can be implemented by hardware, software or a combination of software and hardware. Among them, the hardware part can be implemented using dedicated logic; the software part can be stored in a memory and executed by an appropriate instruction execution system, such as a microprocessor or a dedicated hardware design. Those skilled in the art can understand that the above-mentioned method and system can be implemented using computer executable instructions and / or included in a processor control code, for example, such as a carrier medium such as a disk, CD or DVD-ROM, a programmable memory such as a read-only memory (firmware), or a data carrier such as an optical or electronic signal carrier. Such code is provided on the carrier medium such as a disk, CD or DVD-ROM, a programmable memory such as a read-only memory (firmware), or a data carrier such as an optical or electronic signal carrier. The system and its modules of this specification can not only be implemented by hardware circuits such as ultra-large-scale integrated circuits or gate arrays, semiconductors such as logic chips, transistors, or programmable hardware devices such as field programmable gate arrays, programmable logic devices, etc., but can also be implemented by software executed by various types of processors, and can also be implemented by a combination of the above-mentioned hardware circuits and software (for example, firmware).

[0122] For more details about the above modules, please refer to other places in this manual and will not be repeated here.

[0123] In other embodiments, an intelligent non-invasive blood glucose detection device suitable for diabetic patients is also provided, including a memory and a processor. The memory is used to store executable program code, and the processor is used to call and run the executable program code from the memory, so that the device executes the above-mentioned intelligent non-invasive blood glucose detection method suitable for diabetic patients. The device can be specifically a chip, a component or a module, and the chip may include a connected processor and a memory; wherein the memory is used to store instructions, and when the processor calls and executes the instructions, the chip can execute the intelligent non-invasive blood glucose detection method suitable for diabetic patients provided in the above embodiment.

[0124] In other embodiments, a computer program product is also provided. When the computer program product is run on a computer, the computer executes the above-mentioned related steps to implement an intelligent non-invasive blood glucose detection method suitable for diabetic patients provided in the above embodiment.

[0125] In other embodiments, a computer-readable storage medium is also provided, in which a computer program code is stored. When the computer program code is executed on a computer, the computer executes the above-mentioned related method steps to implement an intelligent non-invasive blood glucose detection method suitable for diabetic patients provided in the above embodiment.

[0126] Among them, the provided system, electronic device, computer program product, and computer-readable storage medium are all used to execute the corresponding methods provided above. Therefore, the beneficial effects that can be achieved can refer to the beneficial effects in the corresponding methods provided above, and will not be repeated here.

[0127] It should be noted that the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the principles of the present invention should be included in the protection scope of the present invention.

Claims

1. An intelligent non-invasive blood sugar detection method suitable for diabetic patients, characterized in that: The method comprises the following steps: Obtain physiological data of different dimensions and NIR wavelengths of a preset number of diabetic reference patients; Obtaining the decision-related value of the physiological data of each dimension according to the clustered distribution characteristics and numerical distribution characteristics of the physiological data of each dimension of all reference patients; evaluating the interchangeability of the physiological data of each two dimensions according to the differences in the physiological data of different reference patients in each dimension and the differences in the decision-related values; and selecting the preferred dimension in combination with all the decision-related values ​​and all the interchangeability degrees; Based on the differences between the physiological data of the preferred dimensions of the reference patients in different categories, similar patients in other categories of the reference patients in each category are screened, wherein the wavelengths of all the reference patients in the same category are the same; according to the differences in the wavelengths corresponding to each two categories and the number distribution of the similar patients of the reference patients in each two categories, different categories are merged to obtain several groups; The target wavelength is determined by combining all wavelengths corresponding to each group, the physiological data of the preferred dimension of the reference patient in each group, and the physiological data of the preferred dimension of the diabetic patient to be detected, and the blood glucose of the diabetic patient to be detected is detected using the target wavelength.

2. The intelligent non-invasive blood sugar detection method suitable for diabetic patients according to claim 1, characterized in that: The step of obtaining the decision-related value of the physiological data of each dimension according to the aggregate distribution characteristics and the numerical distribution characteristics of the physiological data of each dimension of all reference patients includes: Calculate the difference in the number of all reference patients in each two categories and the difference in the mean values ​​of the physiological data of the candidate dimensions of the reference patients in each two categories; Obtaining a decision-related value of the physiological data of the candidate dimension according to all of the quantity differences and all of the average value differences; The candidate dimension is any dimension.

3. The intelligent non-invasive blood sugar detection method suitable for diabetic patients according to claim 1, characterized in that: The step of comprehensively evaluating the difference in physiological data of each two dimensions between different reference patients and the difference in decision-related values ​​to evaluate the interchangeability of physiological data of each two dimensions includes: respectively calculating the first difference between the physiological data of each dimension of each two reference patients; For any two dimensions: Obtaining first characteristic values ​​of each two reference patients in the arbitrary two dimensions according to the first differences corresponding to each two reference patients in the arbitrary two dimensions; According to the differences between all the first eigenvalues ​​of all the reference patients in the any two dimensions and the differences between the decision-related values ​​of the physiological data of the any two dimensions, the degree of interchangeability of the physiological data of the any two dimensions is obtained, and the differences between all the first eigenvalues ​​and the differences between the decision-related values ​​are negatively correlated with the degree of interchangeability.

4. The intelligent non-invasive blood sugar detection method suitable for diabetic patients according to claim 3, characterized in that: The obtaining, according to the first differences corresponding to the arbitrary two dimensions of the arbitrary two dimensions of the arbitrary two reference patients, the first characteristic values ​​of the arbitrary two dimensions of the arbitrary two reference patients comprises: Calculating a first sum value between the first difference corresponding to one dimension of the arbitrary two dimensions for each two reference patients and a preset first adjustment parameter, and a second sum value between the first difference corresponding to the other dimension of the arbitrary two dimensions for each two reference patients and a preset first adjustment parameter; wherein the preset first adjustment parameter is greater than 0; The ratio between the first sum value and the second sum value is determined as the first feature values ​​corresponding to the two reference patients in the arbitrary two dimensions.

5. The intelligent non-invasive blood sugar detection method suitable for diabetic patients according to claim 2, characterized in that: The combining of all the decision-related values ​​and all the interchangeability degrees to select the preferred dimension includes: The product of the interchangeability between the candidate dimension and the physiological data of each of its related dimensions and the decision-related value of the physiological data of each of the related dimensions of the candidate dimension is recorded as the second eigenvalue of each of the related dimensions of the candidate dimension; wherein the related dimension of the candidate dimension is a dimension whose interchangeability with the candidate dimension is greater than a preset interchangeability threshold; Combining the second eigenvalues ​​of all related dimensions of the candidate dimension and the interchangeability of the physiological data of the candidate dimension, obtaining the preference degree of the candidate dimension, wherein the second eigenvalue is negatively correlated with the preference degree, and the interchangeability of the physiological data of the candidate dimension is positively correlated with the preference degree; Filter the preferred dimensions based on the relative preference of all dimensions.

6. The intelligent non-invasive blood sugar detection method suitable for diabetic patients according to claim 5, characterized in that: Filter the preferred dimensions based on the relative preference of all dimensions, including: Sort all dimensions in descending order of preference to obtain a dimension sequence; The first preset number of dimensions in the dimension sequence are determined as preferred dimensions.

7. The intelligent non-invasive blood sugar detection method suitable for diabetic patients according to claim 1, characterized in that: The method of screening similar patients in other categories of the reference patients of each category based on the differences between the physiological data of the preferred dimensions of the reference patients of different categories comprises: Obtaining a similarity value between the first patient and the second patient according to a difference between the physiological data of the first patient and the second patient in the same preferred dimension, wherein the difference between the physiological data of the same preferred dimension is negatively correlated with the similarity value; If the similarity value is greater than a preset similarity threshold, it is determined that the first patient and the second patient are similar patients to each other; The first patient and the second patient are two reference patients in different categories.

8. The intelligent non-invasive blood sugar detection method suitable for diabetic patients according to claim 1, characterized in that: According to the difference in wavelengths corresponding to each two categories and the number distribution of similar patients to the reference patients in each two categories, different categories are merged to obtain several groups, including: Calculate the third sum between the number of all similar patients in the second category of the reference patient in the first category and the number of all similar patients in the first category of the reference patient in the second category; obtain the fusion degree of the arbitrary two categories according to the difference between the wavelengths corresponding to the first category and the second category and the third sum, the difference between the wavelengths is negatively correlated with the fusion degree, and the third sum is positively correlated with the fusion degree; wherein the first category and the second category are any two categories among all categories; Based on the fusion degree of every two categories, the corresponding categories are merged to obtain several groups.

9. The intelligent non-invasive blood sugar detection method suitable for diabetic patients according to claim 8, characterized in that: The corresponding categories are merged based on the fusion degree of each two categories to obtain several groups, including: Sorting all categories in a preset order to obtain a category sequence, wherein the preset order is an order from small to large wavelength or an order from large to small wavelength; The two categories with the largest fusion degree are merged, and each category is judged in turn according to the order of the categories in the category sequence whether the fusion degree with any category of all the merged categories is greater than the preset fusion degree threshold. If it is greater, the corresponding category is merged with the merged category; if it is less than or equal to, it is regarded as a new category; each category after all the mergers are completed is regarded as a group.

10. The intelligent non-invasive blood sugar detection method suitable for diabetic patients according to claim 1, characterized in that: The step of determining the target wavelength by combining all wavelengths corresponding to each group, the physiological data of the preferred dimension of the reference patient in each group, and the physiological data of the preferred dimension of the diabetic patient to be detected comprises: Training the wavelength selection model using all wavelengths corresponding to each group and the physiological data of the preferred dimensions of the reference patients in each group to obtain a trained wavelength selection model; The physiological data of the preferred dimension of the diabetic patient to be detected is input into the trained wavelength selection model to obtain the target wavelength.

Images