Early warning method and device for chronic complications of diabetes mellitus, medium and computer equipment

Abstract

The invention provides a diabetes chronic complication early warning method and device, a medium and computer equipment. The early warning method comprises the following steps: acquiring body data of a diabetic patient; screening the body data by adopting an LASSO logistic regression model to generate a data training set; training the multi-layer perception model by using the data training set to obtain a diabetic complication early warning model; obtaining to-be-detected body data, inputting the body data into the diabetic complication early warning model for prediction after screening, and generating an early warning result; by the adoption of the method, dimension reduction is conducted on the collected body data through the LASSO logistic regression model, the training data of the model can be effectively reduced, the computing power requirement for the training and reasoning process is lowered, the model can be conveniently deployed to a local computing end, and therefore the diabetic patient can conveniently achieve home detection. Compared with conventional blood detection, the non-invasive blood detection device has the advantages of being non-invasive and small in operation difficulty.

Description

Technical Field

[0001] The present invention relates to the field of intelligent medical technology, and more specifically, to a method, device, medium and computer equipment for early warning of chronic complications of diabetes. Background Art

[0002] Currently, the early diagnosis and prediction of chronic complications of diabetes rely heavily on modern medical theory, using various biomarkers and imaging tests. Traditional prediction models for diabetes complications primarily focus on intervention and prevention through invasive blood tests, blood glucose monitoring, and blood pressure monitoring. However, while blood glucose monitoring and HbA1c testing are crucial for diabetes management, these methods typically rely on frequent blood tests, which can be invasive, costly, and inconvenient for long-term self-monitoring at home.

[0003] In view of the above shortcomings, it is very important for diabetic patients to provide a non-invasive and easy-to-operate diabetes complication early warning method so that diabetic patients can detect it at home. Summary of the Invention

[0004] In view of the shortcomings of the existing technology, the purpose of the present invention is to provide a method, device, medium and computer equipment for early warning of chronic complications of diabetes, so as to overcome the shortcomings of the existing diabetes complication detection, which requires blood tests, is invasive, has high testing costs and cannot be detected at home.

[0005] The above technical objectives of the present invention are achieved through the following technical solutions:

[0006] In a first aspect, a method for early warning of chronic complications of diabetes mellitus comprises:

[0007] S1. Obtaining body data of a diabetic patient; the body data at least including: meridian conductance data of twelve pairs of Jing acupoints, and meridian conductance data of twelve pairs of Yuan acupoints;

[0008] S2. Using the LASSO logistic regression model to screen the physical data to generate a data training set; using the data training set to train the multi-layer perception model to obtain a diabetes complication early warning model;

[0009] S3. Obtaining the physical data of the person to be tested, and inputting the data into the diabetes complications early warning model for prediction after screening;

[0010] S4. Generate an early warning result, which specifically includes: the probability of occurrence of chronic complications of diabetes.

[0011] In one embodiment, obtaining the physical data of the diabetic patient specifically includes:

[0012] Obtain the first meridian conductance data of diabetic patients without complications;

[0013] Obtain the second meridian conductance data of patients with diabetic peripheral neuropathy.

[0014] In one embodiment, the first meridian conductance data and the second meridian conductance data respectively include:

[0015] Conductivity data of twelve pairs of well points;

[0016] Conductivity data of twelve pairs of Yuan acupoints;

[0017] The ratio of the sum of the conductance data of the twelve Yuan acupoints on the left to the sum of the conductance data of the twelve Yuan acupoints on the right;

[0018] The ratio of the sum of the conductance data of the twelve well-points on the left to the sum of the conductance data of the twelve well-points on the right;

[0019] The ratio of the sum of the conductance data of the six pairs of Yuan acupoints on the hand to the sum of the conductance data of the six pairs of Yuan acupoints on the foot;

[0020] The ratio of the sum of the conductance data of the six pairs of Jing acupoints on the hand to the sum of the conductance data of the six pairs of Jing acupoints on the foot;

[0021] The ratio of the sum of the conductance data of the six pairs of original acupoints of the Yin meridians to the sum of the conductance data of the six pairs of original acupoints of the Yang meridians;

[0022] The ratio of the sum of the conductance data of the six pairs of well-points of the yin meridians to the sum of the conductance data of the six pairs of well-points of the yang meridians.

[0023] In one embodiment, the LASSO logistic regression model is used to screen the body data to generate a data training set, specifically including:

[0024] Constructing feature matrix

[0025] Where n is the number of patients, p is the number of features; x np represents the value of the pth feature of the nth patient;

[0026] The LASSO logistic regression model is used to determine the importance of each feature in the feature matrix X, and the characteristic coefficient of each feature is adjusted according to the importance of each feature, wherein the characteristic coefficient includes a zero coefficient and a non-zero coefficient;

[0027] A data training set is constructed using feature items whose feature coefficients are non-zero coefficients.

[0028] In one embodiment, the LASSO logistic regression model is used to determine the importance of each feature in the feature matrix X, and the characteristic coefficient of each feature is adjusted according to the importance of each feature. The characteristic coefficient includes a zero coefficient and a non-zero coefficient, specifically including:

[0029] Construct a logistic regression model with LASSO regularization:

[0030]

[0031] Determine the objective function of the LASSO logistic regression model:

[0032]

[0033] Define a candidate parameter set Λ, which contains m candidate parameters λ l , Λ={λ1,λ2,λ3,…,λ m};

[0034] For each candidate parameter λ l , respectively use the body data to perform ten-fold cross validation to obtain each candidate parameter λ l The average cross-validation performance index CV(λ l );

[0035] Select the optimal parameter λ opt ,include:

[0036] λ opt =arg maxCV(λ),λ∈Λ;

[0037] Using the optimal parameter λ opt The LASSO logistic regression model is retrained based on all the body data, and the coefficients of each feature are adjusted and compressed into zero coefficients and non-zero coefficients based on the training results, including:

[0038] In one embodiment, for each candidate parameter λ l , respectively use the body data to perform ten-fold cross validation to obtain each candidate parameter λ l The average cross-validation performance index CV(λ l ), specifically including:

[0039] The body data D is randomly divided into 10 subsets D = {D1, D2, ... D 10}, the difference between the number of samples contained in any two subsets is less than a predetermined threshold;

[0040] For each candidate parameter λl , select a subset D in turn k As validation set, the remaining nine subsets were used as training sets;

[0041] After the LASSO logistic regression model is trained using the training set, the classification performance index of the trained LASSO logistic regression model is calculated using the validation set corresponding to the training set to obtain the candidate parameter λ. l The corresponding ten classification performance indicators are:

[0042] Performance(D k ,β (k) (λ l )),k=1,2,3,…,10,λ l ∈Λ;

[0043] Calculate each of the candidate parameters λ l The average cross-validation performance indicators include:

[0044]

[0045] In one embodiment, the features with non-zero coefficients specifically include: the conductivity data of the left Lidui point, the conductivity data of the right Zhongchong point, the conductivity data of the left Chongyang point, the ratio of the sum of the conductivity data of the twelve left Yuan points to the sum of the conductivity data of the twelve right Yuan points, the conductivity data of the right Yongquan point, the conductivity data of the left Hegu point, the conductivity data of the left Taichong point, the conductivity data of the left Daling point, the ratio of the sum of the conductivity data of the six pairs of Yin meridian Yuan points to the sum of the conductivity data of the six pairs of Yang meridian Yuan points, the conductivity data of the left Dadun point, and the conductivity data of the left Yangchi point.

[0046] A diabetic chronic complication early warning device, comprising:

[0047] An acquisition unit is used to acquire body data of a diabetic patient; the body data at least includes: meridian conductance data of twelve pairs of Jing acupoints, and meridian conductance data of twelve pairs of Yuan acupoints;

[0048] A training unit, configured to screen the body data using a LASSO logistic regression model to generate a data training set; and train the multi-layer perception model using the data training set to obtain a diabetes complication early warning model;

[0049] A prediction unit, configured to obtain the physical data of the person to be tested, and input the data into the diabetes complications early warning model for prediction after screening;

[0050] The early warning unit is used to generate an early warning result, wherein the early warning result is specifically: the probability of occurrence of chronic complications of diabetes.

[0051] A computer-readable storage medium stores a computer program, which implements the steps of the above method when executed by a processor.

[0052] A computer device includes a memory and a processor, wherein the memory stores a computer program and the processor implements the steps of the above method when executing the computer program.

[0053] In summary, the present invention has the following beneficial effects: a method, device, medium and computer equipment for early warning of chronic complications of diabetes, the early warning method comprising: obtaining body data of a diabetic patient; the body data comprising at least: meridian conductance data of twelve pairs of Jing acupoints and meridian conductance data of twelve pairs of Yuan acupoints; using the LASSO logistic regression model to screen the body data to generate a data training set; using the data training set to train the multi-layer perception model to obtain a diabetes complication early warning model; obtaining body data of a person to be tested, and inputting the screened body data into the diabetes complication early warning model for prediction; generating an early warning result, the prediction result The warning results are specifically: the probability of occurrence of chronic complications of diabetes; by adopting the method of the present invention, the dimension of the collected body data is reduced through the LASSO logistic regression model, which can effectively reduce the training data of the model and reduce the computing power requirements for the training and reasoning processes, so that the model can be conveniently deployed to the local computing end, so that diabetic patients can achieve home testing, and the prediction method provided in this application can enable diabetic patients to achieve non-invasive testing. It only needs to measure the meridian conductance values of the twelve well points and twelve original points of diabetic patients to determine the risk of complications of diabetic patients. Compared with conventional blood tests, it has the advantages of being non-invasive and less difficult to operate. BRIEF DESCRIPTION OF THE DRAWINGS

[0054] Figure 1 This is a flow chart of a method for early warning of chronic complications of diabetes mellitus according to the present invention;

[0055] Figure 2 1 is a structural diagram of a device for early warning of chronic complications of diabetes in an embodiment of the present invention;

[0056] Figure 3 This is a diagram of the internal structure of a computer device according to an embodiment of the present invention;

[0057] Figure 4 Schematic diagram of the cross-validation results of LASSO regression of the present invention;

[0058] Figure 5 It is the LASSO regression coefficient path diagram of the present invention;

[0059] In the figure: 1. Acquisition unit; 2. Training unit; 3. Prediction unit; 4. Early warning unit. DETAILED DESCRIPTION

[0060] To make the objectives, features, and advantages of the present invention more readily apparent, the following detailed description of the present invention is provided with reference to the accompanying drawings. The accompanying drawings illustrate several embodiments of the present invention. However, the present invention may be implemented in many different forms and is not limited to the embodiments described herein.

[0061] In the embodiments of the present application, "at least one" refers to one or more, and "more" refers to two or more. "And / or" describes the association relationship of associated objects, indicating that three relationships may exist. For example, A and / or B can represent the existence of A alone, the existence of A and B at the same time, and the existence of B alone. Among them, A and B can be singular or plural. The character " / " generally indicates that the previous and next associated objects are in an "or" relationship. "At least one of the following" and similar expressions refer to any combination of these items, including any combination of single or plural items. For example, at least one of a, b and c can be represented by: a, b, c, ab, ac, bc, or abc, where a, b, c can be single or multiple.

[0062] Those skilled in the art will appreciate that the various units and algorithm steps described in the embodiments disclosed herein can be implemented using a combination of electronic hardware, computer software, and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professionals and technicians can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0063] Those skilled in the art will clearly understand that, for the convenience and brevity of description, the specific working processes of the systems, devices and units described above can refer to the corresponding processes in the aforementioned method embodiments and will not be repeated here.

[0064] In the several embodiments provided in this application, if any function is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, or the part that contributes to the prior art, or the part of the technical solution, can be embodied in the form of a software product. The computer software product is stored in a storage medium and includes several instructions for enabling a device (which can be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the method described in each embodiment of this application. The aforementioned storage medium includes: various media that can store program codes, such as a USB flash drive, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk.

[0065] The above description is merely a specific embodiment of the present application. Any person skilled in the art may easily conceive of variations or substitutions within the technical scope disclosed in this application, and such variations or substitutions shall be within the scope of protection of this application. The scope of protection of this application shall be subject to the scope of protection of the claims.

[0066] The present invention is described in detail below with reference to the accompanying drawings and embodiments.

[0067] Example 1

[0068] In order to facilitate understanding of the technical solution of this application, the relevant technologies in this application are first explained.

[0069] Diabetic complications are multiple organ damages caused by long-term high blood sugar, which are mainly divided into two categories: acute complications and chronic complications. For chronic complications, since their symptoms are hidden in the early stages, patients ignore mild symptoms and cause irreversible damage. The chronic complications of diabetes mainly include diabetic retinopathy, diabetic nephropathy and diabetic neuropathy. The prevention and treatment of chronic complications of diabetes must be achieved through blood sugar control + multi-factor management + regular screening. During regular screening, it is usually recommended that diabetic patients go to the hospital for follow-up screening according to a predetermined cycle (such as half a year or three months) to ensure that complications can be detected and intervened early to avoid delayed treatment and more serious consequences. However, the above-mentioned regular screening and follow-up process still has the following defects: in the period between the two screenings, it is difficult for patients to clearly understand their own status, and they cannot understand their own situation through home testing. Multiple blood draws will also cause trauma to the patient's body and affect the quality of life of diabetic patients.

[0070] To address this situation, some studies have utilized non-invasive bioelectrical signal measurement methods to predict chronic complications of diabetes. The DMS (Digital Meridian System) measures the resistance of twelve meridian acupoints on each side of the human body, converting this resistance into a meridian energy index, reflecting the state of deficiency or excess in the meridians and assisting in Traditional Chinese Medicine (TCM) syndrome differentiation and treatment. However, existing DMS meridian instruments generally suffer from insufficient sensitivity, significant susceptibility to environmental influences, and a lack of specificity for early diagnosis of diabetic complications, making it difficult to provide direct early warning and assessment of diabetic complications.

[0071] In Traditional Chinese Medicine (TCM), studies have shown that the "Network Disease Theory" holds important guiding significance for the diagnosis and treatment of diabetes. It posits that chronic complications of diabetes are primarily caused by internal obstruction of the collaterals. Based on this theory, TCM can mitigate the risk of diabetic complications to a certain extent by regulating and unblocking the body's meridians. However, traditional TCM meridian detection methods are limited by the complexity and time-consuming nature of manual procedures, as well as the lack of precise quantitative measurement methods, making rapid, objective, and continuous monitoring difficult.

[0072] In order to solve the above problems, the present invention provides a method for early warning of chronic complications of diabetes, such as Figure 1 Shown, including:

[0073] S1. Obtaining body data of a diabetic patient; the body data at least including: meridian conductance data of twelve pairs of Jing acupoints, and meridian conductance data of twelve pairs of Yuan acupoints;

[0074] The physical analysis of diabetic patients comes from the diabetes chronic disease management clinic and the endocrinology inpatient department. In order to ensure the accuracy of the test data, diabetic patients need to be screened, including the following conditions: 1. Diabetic patients are over 18 years old; 2. Signed informed consent; 3. Diabetic patients meet the diagnostic criteria for type 2 diabetes, and the diagnostic criteria for type 2 diabetes are one of the following four conditions: ① Glycated hemoglobin ≥6.5%. ② Fasting blood glucose ≥7.0mmol / L. ③ 2-hour blood glucose ≥11.1mmol / L during oral glucose tolerance test. ④ In patients with typical symptoms of hyperglycemia or hyperglycemic crisis, random blood glucose ≥11.1mmol / L; 4. Diabetic patients meet the criteria for chronic complications. Specifically, chronic complications mainly include: at least one of the three complications of diabetic nephropathy, diabetic retinopathy and diabetic peripheral neuropathy.

[0075] To ensure the accuracy of the measurement data, it is necessary to exclude patients who cannot be measured. This is primarily due to the exclusion of patients who cannot be evaluated using the conductivity method. Specific exclusions include: ① Patients with implanted pacemakers; ② Patients with metal implants; ③ Extremity amputations; ④ Local skin damage at the measurement site; ⑤ Scars at the measurement site; ⑥ Patients with comprehension difficulties and inability to communicate; ⑦ Patients preparing for pregnancy, currently pregnant, or breastfeeding.

[0076] To measure patient conductance data, Sichuan Chongzhou Jinhai Medical Instrument Co., Ltd. manufactures and sells the JH / DMS-K100 Traditional Chinese Medicine Meridian Detector, Medical Device Certificate No.: Sichuan Xiezhuzhun 20172270231. The instrument's testing environment and operation comply with the September 2018 Chinese Association of Traditional Chinese Medicine standard, "Technical Operation Specifications for Preventive Treatment with Traditional Chinese Medicine - Conductivity Method Acupoint Assessment" (T / CACM 1089-2018). The meridian values for each well point are calculated using the meridian calculation method outlined in the standard.

[0077] In the human body, Jing point is one of the five Shu points (Jing, Ying, Shu, Jing, He), located at the ends of the limbs (fingers or toes). There are six Jing points on the hands, namely Shaoshang (Lung Meridian), Shangyang (Large Intestine Meridian), Zhongchong (Pericardium Meridian), Guanchong (Triple Burner Meridian), Shaochong (Heart Meridian), and Shaoze (Small Intestine Meridian); there are six Jing points on the feet, namely Yinbai (Spleen Meridian), Lidui (Stomach Meridian), Dadun (Liver Meridian), Zuqiaoyin (Gallbladder Meridian), Zhiyin (Urinary Bladder Meridian), and Yongquan (Kidney Meridian). The distribution pattern of Jing points corresponds to the internal organs, and they are widely used in clinical practice. They are an important means of "preventive treatment" and first aid in Traditional Chinese Medicine. Yuan points are acupoints on the extremities where the vital energy of the internal organs is infused, passes through, and resides along the twelve meridians. There are six Yuan points on the hand: Taiyuan (Hand Taiyin Lung Meridian), Hegu (Hand Yangming Large Intestine Meridian), Shenmen (Hand Shaoyin Heart Meridian), Daling (Hand Jueyin Pericardium Meridian), Wangu (Hand Taiyang Small Intestine Meridian), and Yangchi (Hand Shaoyang Triple Burner Meridian). There are six Yuan points on the foot: Chongyang (Hand Yangming Stomach Meridian), Taibai (Hand Taiyin Spleen Meridian), Taixi (Hand Shaoyin Kidney Meridian), Jinggu (Hand Taiyang Bladder Meridian), Qiuxu (Hand Shaoyang Gallbladder Meridian), and Taichong (Hand Jueyin Liver Meridian). By measuring the Jing and Yuan points on the left hand, right hand, left foot, and right foot, we can obtain meridian conductance data for twelve pairs of Jing points and twelve pairs of Yuan points.

[0078] Based on the above data acquisition method, a total of 120 cases were collected in this example, including 91 cases of diabetes without complications and 29 cases of diabetes with complications. In terms of gender distribution, there were 57 females in the group without complications, accounting for 62.64%, and 16 females in the group with complications, accounting for 55.17%. There was no statistical difference in the gender composition between the two groups (χ 2=0.51, P=0.47). The median age was 59.00 and 62.00 years, respectively (P=0.41), indicating that the age distribution of the two groups was similar, eliminating the influence of age on the results. The median BMI was 23.44 and 23.31, respectively (P=0.54), indicating that the obesity level of the two groups was similar and BMI did not significantly affect the risk of complications. No significant differences were found between the two groups. In the analysis of blood glucose indicators, there were no statistically significant differences between the two groups in the lowest (P=0.19) and highest (P=0.09) values of fasting blood glucose and the lowest (P=0.95) and highest (P=0.46) values of postprandial blood glucose. The results suggest that the complication group and the non-complication group were homogeneous in terms of demographic characteristics and baseline blood glucose indicators.

[0079] Table 1 Comparison of gender composition between the two groups of patients

[0080]

[0081] Table 2 Comparison of basic conditions of patients in the two groups

[0082]

[0083] In Table 2, N represents the sample size; = represents the mean and standard deviation; M represents the median, which is applicable to non-normally distributed data and represents the middle position; Q75-25 represents the interquartile range, which shows the range and dispersion of the data distribution. The Z-score is a standardized indicator used in statistical tests. The P-value represents the probability of observing the current data or a more extreme data under the assumption that the null hypothesis (no difference between the two groups of data) is true.

[0084] By measuring the conductance values of the Jing acupoints of patients with known diabetic complications and labeling each conductance value, it can be used as training data for model prediction to train the model. According to the clinical diagnostic criteria for chronic complications of diabetes, the subjects were labeled into different categories: 1. Non-complication group; 2. Diabetic retinopathy group; 3. Diabetic nephropathy group; 4. Diabetic peripheral neuropathy group. The patient's physical data includes multiple contents, such as the patient's basic information, the subject's ID, age, gender, diabetes course, diagnosis date and other basic demographic information. Secondly, based on the research content of this application, it is also necessary to include 48 conductance values of the twelve pairs of Jing acupoints and twelve pairs of Yuan acupoints corresponding to each patient, as well as the ratio between the conductance values of the acupoints. Since the data collected from patients with diabetic complications are mostly patients with diabetic peripheral neuropathy, and the number of patients in the other two groups of diabetic retinopathy and diabetic nephropathy is relatively small, this embodiment only analyzes the conductance data of patients with diabetic peripheral neuropathy. In other words, the acquisition of the body data of diabetic patients specifically includes: acquiring the first meridian conductance data of diabetic patients without complications; acquiring the second meridian conductance data of patients with diabetic peripheral neuropathy. Specifically, the first meridian conductance data and the second meridian conductance data in this embodiment respectively include the following data:

[0085] The conductivity data of the twelve pairs of well points, a total of 24 items, are represented as x in the matrix. i,1 ~x i,24 ;

[0086] The conductivity data of the twelve pairs of Yuan acupoints, a total of 24 items of data, are represented as x in the matrix. i,25 ~x i,48 ;

[0087] The ratio of the sum of the conductance data of the twelve original acupoints on the left to the sum of the conductance data of the twelve original acupoints on the right is represented in the matrix as x i,49 =∑left side Yuan point / ∑right side Yuan point;

[0088] The ratio of the sum of the conductance data of the twelve wells on the left to the sum of the conductance data of the twelve wells on the right is represented in the matrix as x i,50 =∑ left side Jing point / ∑ right side Jing point;

[0089] The ratio of the sum of the conductance data of the six pairs of Yuan acupoints on the hand to the sum of the conductance data of the six pairs of Yuan acupoints on the foot is expressed as x in the matrix. i,51 =∑ Yuan acupoints on the hand / ∑ Yuan acupoints on the foot;

[0090] The ratio of the sum of the conductance data of the six pairs of Jing acupoints on the hand to the sum of the conductance data of the six pairs of Jing acupoints on the foot is expressed as x in the matrix. i,52 =∑ Jing points on the hand / ∑ Jing points on the foot;

[0091] The ratio of the sum of the conductance data of the six pairs of original acupoints of the Yin meridians to the sum of the conductance data of the six pairs of original acupoints of the Yang meridians is expressed as x in the matrix. i,53 =∑Yin meridian original acupoints / ∑Yang meridian original acupoints;

[0092] The ratio of the sum of the conductance data of the six pairs of yin meridians to the sum of the conductance data of the six pairs of yang meridians is represented in the matrix as x i,54 =∑Yin meridian well point / ∑Yang meridian well point;

[0093] For the specific data collection standards, please refer to the standard of the Chinese Association of Traditional Chinese Medicine "Technical Operation Specifications for Preventive Treatment of Diseases with Traditional Chinese Medicine - Conductivity Method Acupoint Evaluation" (T / CACM 1089-2018) issued in September 2018.

[0094] Based on the above data, construct the feature matrix

[0095] The horizontal feature is the number of patients. In this embodiment, the number of patients n=120. The vertical feature is the feature category. In this embodiment, the number of feature categories p=54.

[0096] In the actual measurement process, it is also necessary to classify the patient's data according to the patient's clinical standards and add labels to the patient's data according to each category to facilitate model training. The data mainly includes: obtaining the first meridian conductance data of diabetic patients without complications; obtaining the second meridian conductance data of patients with diabetic peripheral neuropathy.

[0097] S2. Using the LASSO logistic regression model to screen the physical data to generate a data training set; using the data training set to train the multi-layer perception model to obtain a diabetes complication early warning model;

[0098] Specifically, the LASSO logistic regression model can adjust the coefficients before each feature based on the importance of each feature, adjust the coefficients before unimportant features to zero, and adjust the coefficients before important features to non-zero and retain them. The specific process is as follows:

[0099] First, build a logistic regression model with LASSO regularization:

[0100]

[0101] Among them, P(y i =1|x i ) represents the feature vector x of a given sample i When , the conditional probability that the sample belongs to category 1, β0 represents the intercept term, that is, when all x i,j The base value of the logarithmic probability of the event when both are 0. jis the regression coefficient, which represents the influence weight of the jth feature on the target variable. The larger the absolute value, the higher the feature importance. Through the regularization effect of the LASSO logistic regression model, the coefficients of unimportant features can be compressed to 0 to achieve feature selection.

[0102] Then determine the objective function of the LASSO logistic regression model:

[0103]

[0104] Define a candidate parameter set Λ, which contains m candidate parameters λ l , Λ={λ1,λ2,λ3,…,λ m The candidate parameters are predefined and usually take values with uniform logarithmic distribution. In this embodiment, they are: Λ={10 -3 ,10 -2 ,10 -2 ,10 0 ,10 1 ,10 2 ,10 3};

[0105] For each candidate parameter λ l , respectively use the body data to perform ten-fold cross validation to obtain each candidate parameter λ l The average cross-validation performance index CV(λ l );

[0106] The ten-fold cross validation mainly includes the following steps:

[0107] All body data D are randomly divided into ten subsets of similar size, D = {D1, D2, ... D 10}, that is, the difference in the number of samples contained in any two subsets is smaller than a predetermined threshold, for example, the predetermined threshold can be 1 or 2;

[0108] For each candidate parameter λ l , select a subset D in turn k As the validation set, the remaining nine subsets are used as training sets; that is, the LASSO logistic regression model uses a candidate parameter λ for each l In each case, ten training sessions will be conducted, and ten validation sessions will be conducted after training to determine the classification accuracy of the model.

[0109] After the LASSO logistic regression model is trained using the training set, the classification performance index of the trained LASSO logistic regression model is calculated using the validation set corresponding to the training set to obtain the candidate parameter λ. l The corresponding ten classification performance indicators are:

[0110] Performance(D k ,β (k) (λ l )),k=1,2,3,…,10,λ l ∈Λ;

[0111] Calculate each of the candidate parameters λ l The average cross-validation performance indicators include:

[0112]

[0113] Select the optimal parameter λ opt ,include:

[0114] λ opt =arg maxCV(λ),λ∈Λ;

[0115] Using the optimal parameter λ opt The LASSO logistic regression model is retrained based on all the body data, and the coefficients of each feature are adjusted and compressed into zero coefficients and non-zero coefficients based on the training results, including:

[0116]

[0117] Constructing a data training set using the feature item whose feature coefficient is a non-zero coefficient specifically includes:

[0118] The feature items with non-zero coefficients screened out based on the LASSO logistic regression model are shown in Table 3 below. There are 11 features in total, namely: the conductance data of the left Lidui point, the conductance data of the right Zhongchong point, the conductance data of the left Chongyang point, the ratio of the sum of the conductance data of the twelve original acupoints on the left to the sum of the conductance data of the twelve original acupoints on the right, the conductance data of the right Yongquan point, the conductance data of the left Hegu point, the conductance data of the left Taichong point, the conductance data of the left Daling point, the ratio of the sum of the conductance data of the six pairs of original acupoints of the yin meridians to the sum of the conductance data of the six pairs of original acupoints of the yang meridians, the conductance data of the left Dadun point, and the conductance data of the left Yangchi point.

[0119] Table 3 Number of variables selected for LASSO regression

[0120] variable Confidence Li Dui Zuo -0.309376167 center right -0.263052763 Chongyang Left 0.236398802 Yuan acupoint left-right ratio 0.226022918 Yongquan Right -0.162602622 Hegu Left 0.126664577 Taichong Left 0.115733904 Dalingzuo 0.074213521 Yuan acupoint Yin / Yang ratio 0.04265653 Dadunzuo 0.022318924 Yangchi Zuo 0.0116041

[0121] like Figure 4 As shown, this embodiment also provides cross-validation results based on LASSO regression, where Figure 4The figure shows the results of ten-fold cross-validation, where the X-axis variable is Log(λ), which represents the logarithm of the regularization parameter λ; the Y-axis variable is Binomial Deviance, which measures the goodness of fit of the binary classification model; the red dotted line represents the average cross-validation error under different λ values; the gray error bar represents the standard error of the cross-validation error; the top number (53, 53, 53, 52…0) represents the number of non-zero coefficients (i.e., selected variables) retained in the model under the corresponding λ value.

[0122] exist Figure 4 In the figure, as λ increases from left to right, the selected variables gradually decrease, the model complexity decreases, and the error first increases slowly and then rises rapidly. The two vertical dotted lines in the figure, the left dotted line represents λ min , the value of λ that produces the minimum cross-validation error; the dotted line on the right represents λ 1se , the lambda value of the most parsimonious model within one standard error of the minimum error. Each gray line in the figure represents the trajectory of the coefficient of a variable as lambda changes. Most variable coefficients are compressed to zero as lambda increases. Several key variables still maintain non-zero coefficients at larger lambda values. The coefficients can be positive (positive correlation) or negative (negative correlation). Figure 5 Choose the optimal value of λ, usually λ min or λ 1se ,pass Figure 5 Identify the important variables to retain at the optimal λ and determine the number of variables to include in the final model.

[0123] Through the above process, the data can be reduced in dimensionality to obtain a training dataset. The number of features in the training dataset is reduced from 55 to 11, effectively reducing the computational resource requirements during subsequent model training. After the data is reduced in dimensionality, it is input into a pre-built multi-layer perceptron model for training, enabling the model to learn and perceive the implicit features of the data. Specifically, the multi-layer perceptron model includes: an input layer, the number of nodes in the input layer equals the number of nodes in the standard normal distribution of the training data; the input layer's main task is to pass data samples (feature vectors) to the model for processing by subsequent layers. The number of nodes in the input layer equals the number of nodes in the standard normal distribution of the training data, i.e., the data's feature dimension. Due to the data's dimensionality reduction, the number of nodes in the input layer is relatively small, reducing computational complexity. There are two hidden layers, each of which is used to extract features from the input data and pass the extracted data to the output layer. The hidden layers extract and transform the input data, learning the implicit features and complex relationships in the data through a combination of neurons and nonlinear activation functions. In this application, the two hidden layers have 128 and 64 nodes, respectively. Each neuron in the hidden layer learns through weight and bias parameters to adjust the feature representation. The output layer, which has 1 node, is used to output a binary classification result for the early warning of chronic complications of diabetes. The output layer specifically uses a sigmoid activation function. The output layer is used to generate the model's final prediction results.

[0124] In this embodiment, the model training process adopts a gradient descent optimization algorithm, and a binary cross entropy loss function is used to measure the difference between the probability distribution predicted by the model and the true label. The cross entropy loss can directly measure the degree of closeness between the predicted probability of the model and the true label, and when the ratio of positive and negative samples is unbalanced, the cross entropy loss can still provide an accurate error metric. The optimization direction is clear. In the binary classification problem of this application, the cross entropy avoids the gradient vanishing problem and makes the neural network learning more stable. The gradient descent method is used to minimize the loss function, and by adaptively adjusting the learning rate, fast and stable convergence is provided, training time is reduced, and it has strong robustness to noise and gradient fluctuations.

[0125] S3. Obtaining the physical data of the person to be tested and inputting it into the diabetes complications early warning model for prediction;

[0126] After the model training is completed, in real life, diabetic patients can achieve non-invasive, home testing by measuring the meridian conductance data of the twelve well points and twelve original points on the hands and feet. Between each regular monitoring, patients can predict their own conditions through this method and the model trained by this method, providing diabetic patients with an effective means of detecting chronic complications.

[0127] S4. Generate an early warning result, which specifically includes: the probability of occurrence of chronic complications of diabetes.

[0128] Specifically, in this application, the output layer does not directly output the binary classification results, but outputs the probability of the classification results, that is, the probability of the patient developing chronic complications. When the probability of chronic complications in the patient is high, the patient will be advised to go to the hospital for examination to provide more accurate test results for diabetic patients.

[0129] Table 4. Confusion matrix results of neural network model

[0130]

[0131]

[0132] Based on the above training process, as shown in Table 4 above, the overall accuracy (Accuracy) of the diabetes complications warning model on the training data is 0.97, the sensitivity (Sensitivity) is 0.86, and the specificity (Specificity) is 1.00. The confusion matrix results show that the model correctly identified 91 negative samples and 25 positive samples, while making 4 errors in mispredicting positive as negative. It is worth noting that the model showed good accuracy in predicting negative samples. Although the prediction accuracy of the positive class is slightly lower (86.21%), its overall balance (F1-Score = 89.29%) still shows that the model has high clinical practical value.

[0133] Table 5 Confusion matrix results of external validation

[0134]

[0135] As shown in Table 5 above, to evaluate the applicability of the constructed model in real-world clinical settings, this study collected 30 blinded external validation samples. These 30 patients did not participate in model training or initial validation during data collection, ensuring the independence of the external data. After unblinding, it was revealed that 8 of these 30 diabetic patients had chronic complications. The confusion matrix results showed that of the predictions of "no complications," 20 actually had no complications and 2 actually had complications; of the predictions of "complications," 6 actually had complications and 2 actually did not. The sensitivity was 75%, the specificity was 90.9%, the accuracy was 86.7%, and the F1 value was approximately 75%.

[0136] Furthermore, by using the aforementioned LASSO logistic regression model to reduce the data dimension and then using the reduced data to train the multi-layer perception model, training speed can be increased and resource requirements reduced. Furthermore, during the home testing process, user data can be continuously collected, and the LASSO logistic regression model training process can be repeated to further adjust the coefficients of each feature to improve the model's classification accuracy.

[0137] Example 2

[0138] See also Figure 2 , a diabetic chronic complication early warning device, the diabetic chronic complication early warning device comprising:

[0139] An acquisition unit 1 is configured to acquire body data of a diabetic patient; the body data at least includes: meridian conductance data of twelve pairs of Jing acupoints and meridian conductance data of twelve pairs of Yuan acupoints;

[0140] Training unit 2 is used to screen the physical data using a LASSO logistic regression model to generate a data training set; and use the data training set to train the multi-layer perception model to obtain a diabetes complication early warning model;

[0141] Prediction unit 3, used to obtain the physical data of the person to be tested, and input it into the diabetes complications early warning model for prediction after screening;

[0142] The early warning unit 4 is used to generate an early warning result, which is specifically: the probability of occurrence of chronic complications of diabetes.

[0143] For the specific definition of the diabetes chronic complication early warning device, please refer to the definition of the diabetes chronic complication early warning method above, and will not be repeated here. The various modules in the above-mentioned diabetes chronic complication early warning device can be implemented in whole or in part through software, hardware, or a combination thereof. The above-mentioned modules can be embedded in or independent of the processor in the computer device in hardware form, or can be stored in the memory of the computer device in software form, so that the processor can call and execute the operations corresponding to the above modules.

[0144] Those skilled in the art will understand that Figure 2 The structure shown in the figure is only a block diagram of a part of the structure related to the present application scheme, and does not constitute a limitation of the present application scheme. The specific diabetes chronic complications early warning device may include more or fewer components than shown in the figure, or combine certain components, or have a different component arrangement.

[0145] Example 3

[0146] A computer-readable storage medium stores a computer program, which, when executed by a processor, implements the method for early warning of chronic complications of diabetes as described in Example 1.

[0147] Example 4

[0148] In one embodiment, a computer device is provided. The computer device may be a server, and its internal structure diagram may be as follows: Figure 3 As shown. The computer device includes a processor, memory, network interface, and database connected via a system bus. The processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program, and a database. The internal memory provides an environment for the operation of the operating system and computer program in the non-volatile storage medium. When executed by the processor, the computer program implements a method for early warning of chronic complications of diabetes.

[0149] Those skilled in the art will understand that Figure 3 The structure shown in the figure is only a block diagram of a part of the structure related to the solution of the present application, and does not constitute a limitation on the computer device to which the solution of the present application is applied. The specific computer device may include more or fewer components than shown in the figure, or combine certain components, or have a different component arrangement.

[0150] In one embodiment, a computer device is provided, including a memory and a processor. The memory stores a computer program, and when the processor executes the computer program, the following steps are implemented:

[0151] S1. Obtaining body data of a diabetic patient; the body data at least including: meridian conductance data of twelve pairs of Jing acupoints, and meridian conductance data of twelve pairs of Yuan acupoints;

[0152] S2. Using the LASSO logistic regression model to screen the physical data to generate a data training set; using the data training set to train the multi-layer perception model to obtain a diabetes complication early warning model;

[0153] S3. Obtaining the physical data of the person to be tested, and inputting the data into the diabetes complications early warning model for prediction after screening;

[0154] S4. Generate an early warning result, which specifically includes: the probability of occurrence of chronic complications of diabetes.

[0155] In one embodiment, obtaining the physical data of the diabetic patient specifically includes:

[0156] Obtain the first meridian conductance data of diabetic patients without complications;

[0157] Obtain the second meridian conductance data of patients with diabetic peripheral neuropathy.

[0158] In one embodiment, the first meridian conductance data and the second meridian conductance data respectively include:

[0159] Conductivity data of twelve pairs of well points;

[0160] Conductivity data of twelve pairs of Yuan acupoints;

[0161] The ratio of the sum of the conductance data of the twelve Yuan acupoints on the left to the sum of the conductance data of the twelve Yuan acupoints on the right;

[0162] The ratio of the sum of the conductance data of the twelve well-points on the left to the sum of the conductance data of the twelve well-points on the right;

[0163] The ratio of the sum of the conductance data of the six pairs of Yuan acupoints on the hand to the sum of the conductance data of the six pairs of Yuan acupoints on the foot;

[0164] The ratio of the sum of the conductance data of the six pairs of Jing acupoints on the hand to the sum of the conductance data of the six pairs of Jing acupoints on the foot;

[0165] The ratio of the sum of the conductance data of the six pairs of original acupoints of the Yin meridians to the sum of the conductance data of the six pairs of original acupoints of the Yang meridians;

[0166] The ratio of the sum of the conductance data of the six pairs of well-points on the yin meridians to the sum of the conductance data of the six pairs of well-points on the yang meridians;

[0167] The ratio of the sum of the conductance data of the twelve pairs of Yuan acupoints to the sum of the conductance data of the twelve pairs of Jing acupoints.

[0168] In one embodiment, the LASSO logistic regression model is used to screen the body data to generate a data training set, specifically including:

[0169] Constructing feature matrix

[0170] Where n is the number of patients, p is the number of features; x n,p represents the value of the pth feature of the nth patient;

[0171] The LASSO logistic regression model is used to determine the importance of each feature in the feature matrix X, and the characteristic coefficient of each feature is adjusted according to the importance of each feature, wherein the characteristic coefficient includes a zero coefficient and a non-zero coefficient;

[0172] A data training set is constructed using feature items whose feature coefficients are non-zero coefficients.

[0173] In one embodiment, the LASSO logistic regression model is used to determine the importance of each feature in the feature matrix X, and the characteristic coefficient of each feature is adjusted according to the importance of each feature. The characteristic coefficient includes a zero coefficient and a non-zero coefficient, specifically including:

[0174] Construct a logistic regression model with LASSO regularization:

[0175]

[0176] Determine the objective function of the LASSO logistic regression model:

[0177]

[0178] Define a candidate parameter set Λ, which contains M candidate parameters λ l , Λ={λ1,λ2,λ3,…,λ m};

[0179] For each candidate parameter λ l , respectively use the body data to perform ten-fold cross validation to obtain each candidate parameter λ l The average cross-validation performance index CV(λ l );

[0180] Select the optimal parameter λ opt ,include:

[0181] λ opt =arg maxCV(λ),λ∈Λ;

[0182] Using the optimal parameter λ optThe LASSO logistic regression model was retrained based on all the body data, including:

[0183]

[0184] The trained LASSO logistic regression model is used to adjust the coefficients of various features in the body data to obtain zero coefficients and non-zero coefficients.

[0185] In one embodiment, for each candidate parameter λ l , respectively use the body data to perform ten-fold cross validation to obtain each candidate parameter λ l The average cross-validation performance index CV(λ l ), specifically including:

[0186] The body data D is randomly divided into 10 subsets D = {D1, D2, ... D 10}, the difference between the number of samples contained in any two subsets is less than a predetermined threshold;

[0187] For each candidate parameter λ l , select a subset D in turn k As validation set, the remaining nine subsets were used as training sets;

[0188] After the LASSO logistic regression model is trained using the training set, the classification performance index of the trained LASSO logistic regression model is calculated using the validation set corresponding to the training set to obtain the candidate parameter λ. l The corresponding ten classification performance indicators are:

[0189] Performance(D k ,β (k) (λ l )),k=1,2,3,…,10,λ l ∈Λ;

[0190] Calculate each of the candidate parameters λ l The average cross-validation performance indicators include:

[0191]

[0192] In one embodiment, the features with non-zero coefficients specifically include: the conductance data of the Lidui point on the left side, the conductance data of the Zhongchong point on the right side, the conductance data of the Chongyang point on the left side, the ratio of the sum of the conductance data of the twelve original points on the left side to the sum of the conductance data of the twelve original points on the right side, the conductance data of the Yongquan point on the right side, the conductance data of the Hegu point on the left side, the conductance data of the Taichong point on the left side, the conductance data of the Daling point on the left side, the ratio of the sum of the conductance data of the six pairs of original points of the yin meridians to the sum of the conductance data of the six pairs of original points of the yang meridians, the conductance data of the Dadun point on the left side, and the conductance data of the Yangchi point on the left side.

[0193] Those skilled in the art will appreciate that all or part of the processes in the above-mentioned embodiments can be implemented by instructing the relevant hardware through a computer program. The computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to memory, storage, database or other media used in the embodiments provided in this application can include non-volatile and / or volatile memory. Non-volatile memory can include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM) or flash memory. Volatile memory can include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in various forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link (Synchlink) DRAM (SLDRAM), memory bus (Rambus) direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM).

[0194] The technical features of the above embodiments can be combined arbitrarily. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0195] The above description is merely a preferred embodiment of the present invention. The scope of protection of the present invention is not limited to the above embodiment. All technical solutions based on the concept of the present invention are within the scope of protection of the present invention. It should be noted that for those skilled in the art, various improvements and modifications that do not depart from the principles of the present invention should also be considered within the scope of protection of the present invention.

Claims

1. A method for early warning of chronic complications of diabetes, characterized in that: include: Obtain physical data of diabetic patients; The body data at least includes: meridian conductance data of twelve pairs of Jing acupoints, and meridian conductance data of twelve pairs of Yuan acupoints; The body data is screened using a LASSO logistic regression model to generate a data training set; the multi-layer perception model is trained using the data training set to obtain a diabetes complication early warning model; Obtaining the physical data of the person to be tested, and inputting it into the diabetes complication early warning model for prediction after screening; Generate a warning result, which specifically is: the probability of occurrence of chronic complications of diabetes.

2. A method for early warning of chronic complications of diabetes according to claim 1, characterized in that: The obtaining of the physical data of the diabetic patient specifically includes: Obtain the first meridian conductance data of diabetic patients without complications; Obtain the second meridian conductance data of patients with diabetic peripheral neuropathy.

3. A method for early warning of chronic complications of diabetes according to claim 2, characterized in that: The first meridian conductance data and the second meridian conductance data respectively include: Conductivity data of twelve pairs of well points; Conductivity data of twelve pairs of Yuan acupoints; The ratio of the sum of the conductance data of the twelve Yuan acupoints on the left to the sum of the conductance data of the twelve Yuan acupoints on the right; The ratio of the sum of the conductance data of the twelve well-points on the left to the sum of the conductance data of the twelve well-points on the right; The ratio of the sum of the conductance data of the six pairs of Yuan acupoints on the hand to the sum of the conductance data of the six pairs of Yuan acupoints on the foot; The ratio of the sum of the conductance data of the six pairs of Jing acupoints on the hand to the sum of the conductance data of the six pairs of Jing acupoints on the foot; The ratio of the sum of the conductance data of the six pairs of original acupoints of the Yin meridians to the sum of the conductance data of the six pairs of original acupoints of the Yang meridians; The ratio of the sum of the conductance data of the six pairs of well-points of the yin meridians to the sum of the conductance data of the six pairs of well-points of the yang meridians.

4. A method for early warning of chronic complications of diabetes according to claim 3, characterized in that: The LASSO logistic regression model is used to screen the physical data to generate a data training set, specifically including: Constructing feature matrix Where n is the number of patients, p is the number of features; x n,p represents the value of the pth feature of the nth patient; The LASSO logistic regression model is used to determine the importance of each feature in the feature matrix X, and the characteristic coefficient of each feature is adjusted according to the importance of each feature, wherein the characteristic coefficient includes a zero coefficient and a non-zero coefficient; A data training set is constructed using feature items whose feature coefficients are non-zero coefficients.

5. A method for early warning of chronic complications of diabetes according to claim 4, characterized in that: The LASSO logistic regression model is used to determine the importance of each feature in the feature matrix X, and the characteristic coefficients of each feature are adjusted according to the importance of each feature. The characteristic coefficients include zero coefficients and non-zero coefficients, specifically including: Construct a logistic regression model with LASSO regularization: Determine the objective function of the LASSO logistic regression model: Define a candidate parameter set Λ, which contains m candidate parameters λ l , Λ={λ1,λ2,λ3,…,λ m }; For each candidate parameter λ l , respectively use the body data to perform ten-fold cross validation to obtain each candidate parameter λ l The average cross-validation performance index CV(λ l ); Select the optimal parameter λ opt ,include: l opt =arg max CV(λ),λ∈Λ; Using the optimal parameter λ opt The LASSO logistic regression model is retrained based on all the body data, and the coefficients of each feature are adjusted and compressed into zero coefficients and non-zero coefficients based on the training results, including:

6. A method for early warning of chronic complications of diabetes according to claim 5, characterized in that: For each of the candidate parameters λ l , respectively use the body data to perform ten-fold cross validation to obtain each candidate parameter λ l The average cross-validation performance index CV(λ l ), specifically including: The body data D is randomly divided into 10 subsets D = {D1, D2, ... D 10 }, the difference between the number of samples contained in any two subsets is less than a predetermined threshold; For each candidate parameter λ l , select a subset D in turn k As validation set, the remaining nine subsets were used as training sets; After the LASSO logistic regression model is trained using the training set, the classification performance index of the trained LASSO logistic regression model is calculated using the validation set corresponding to the training set to obtain the candidate parameter λ. l The corresponding ten classification performance indicators are: Performance(D k ,b (k) (l l )),k=1,2,3,…,10,λ l ∈Λ; Calculate each of the candidate parameters λ l The average cross-validation performance indicators include:

7. A method for early warning of chronic complications of diabetes according to claim 6, characterized in that: The characteristics of non-zero coefficients include: The conductance data of the left Lidui acupoint, the conductance data of the right Zhongchong acupoint, the conductance data of the left Chongyang acupoint, the ratio of the sum of the conductance data of the twelve original acupoints on the left to the sum of the conductance data of the twelve original acupoints on the right, the conductance data of the right Yongquan acupoint, the conductance data of the left Hegu acupoint, the conductance data of the left Taichong acupoint, the conductance data of the left Daling acupoint, the ratio of the sum of the conductance data of the six pairs of original acupoints of the yin meridians to the sum of the conductance data of the six pairs of original acupoints of the yang meridians, the conductance data of the left Dadun acupoint, and the conductance data of the left Yangchi acupoint.

8. A chronic diabetic complication early warning device, characterized in that: The device comprises: An acquisition unit is used to acquire body data of a diabetic patient; the body data at least includes: meridian conductance data of twelve pairs of Jing acupoints, and meridian conductance data of twelve pairs of Yuan acupoints; A training unit, configured to screen the body data using a LASSO logistic regression model to generate a data training set; and train the multi-layer perception model using the data training set to obtain a diabetes complication early warning model; A prediction unit, configured to obtain the physical data of the person to be tested, and input the data into the diabetes complications early warning model for prediction after screening; The early warning unit is used to generate an early warning result, wherein the early warning result is specifically: the probability of occurrence of chronic complications of diabetes.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the method for early warning of chronic complications of diabetes as claimed in any one of claims 1 to 7 is implemented.

10. A computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the method for early warning of chronic complications of diabetes as described in any one of claims 1 to 7 is implemented.

Images