Diabetes screening data processing method and management system

Abstract

The invention relates to the technical field of medical data processing, in particular to a diabetes screening data processing method and management system. The method comprises the following steps: acquiring a blood glucose sequence of a person to be screened, and obtaining reference data according to an abnormal coefficient of the data; acquiring a reference blood glucose sequence of the to-be-screened person from the blood glucose database, and analyzing to acquire a data set and to-be-selected data in the data set; calculating the preference rate of the to-be-selected data, taking the to-be-selected data with the maximum preference rate as a reference value of the data set, and further calculating a first weight of non-reference data in the blood glucose sequence; calculating a second weight of non-reference data according to the position difference between the non-reference data and the reference data in the blood glucose sequence and the variance of the data in the window taking the non-reference data as the center; and correcting the reference data by using the first weight and the second weight, and finally judging the blood glucose condition of the to-be-screened person according to the corrected reference data. The accuracy of blood glucose data can be improved.

Description

Technical Field

[0001] The present invention relates to the technical field of medical data processing, and in particular to a method and a management system for processing diabetes screening data. Background Art

[0002] As a chronic disease that affects the health of hundreds of millions of people worldwide, the prevention, control and early screening of diabetes have always attracted much attention. Early screening and intervention are key strategies for controlling the development of diabetes. The conventional method of diabetes screening is to determine whether one has diabetes by measuring fasting blood sugar levels. Patients are required to fast for at least 8 hours, and blood is drawn and blood sugar levels are measured to determine whether there are symptoms of hyperglycemia. The normal fasting blood sugar range is usually 7099 mg / dL (or 3.96.1 mmol / L). If the fasting blood sugar level exceeds 100 mg / dL (or 7.0 mmol / L), further examination may be required.

[0003] In the process of collecting blood sugar in the glucose tolerance test, the collector may not follow the standard collection process, such as improper selection of blood collection sites, improper use of blood collection needles, etc., or the subjects may not be fully explained and guided during the collection, which may cause the subjects to be nervous or improperly cooperate, resulting in the deviation of blood sugar data from normal blood sugar data, affecting the results of diabetes screening. Blood sugar at a single moment will be affected by the above factors, resulting in abnormal data, causing the blood sugar screening results to show high blood sugar data and diabetes risk, so the abnormal data needs to be corrected. Traditionally, the replacement method is used to correct outliers. The replacement method generally uses the mean of the data sequence for replacement, but the mean is sensitive to outliers and is easily affected by extreme values. In addition, the single-moment data of a single person is unreliable, resulting in inaccurate results of the outlier correction method, making the final collected blood sugar results inaccurate, affecting the screening results. Therefore, it is necessary to refer to the blood sugar conditions of people of the same type for analysis. Summary of the invention

[0004] In order to solve the above technical problems, the purpose of the present invention is to provide a method and management system for processing diabetes screening data. The technical solutions adopted are as follows: In a first aspect, an embodiment of the present invention provides a method for processing diabetes screening data, the method comprising: Collect blood sugar data of the person to be screened and form a blood sugar sequence; calculate the abnormal coefficient of the data according to the change of a data in the blood sugar sequence and the data before and after the data; use the abnormal coefficient of each data to obtain the reference data in the blood sugar sequence; Setting screening conditions, obtaining blood glucose sequences of persons who meet the screening conditions in the blood glucose database as reference blood glucose sequences of persons to be screened; forming a data set using the data at the same position in the blood glucose sequence and each reference blood glucose sequence, and obtaining the data to be selected in the data set; Calculate the preference rate of the candidate data according to the frequency of the candidate data in the data set and the variance and quantity of the data within a fixed range centered on the candidate data; take the candidate data with the largest preference rate as the reference value of the data set; calculate the first weight of the non-reference data according to the non-reference data in the blood glucose sequence and the reference value of the corresponding data set; Calculate a second weight of the non-reference data according to a position difference between a non-reference data and reference data in the blood glucose sequence and a variance of data in a window centered on the non-reference data; The modified weight of the non-reference data is calculated according to the first weight and the second weight of the non-reference data, and the modified reference data is calculated using the modified weight of each non-reference data.

[0005] Preferably, the calculation formula of the abnormal coefficient is: , Among them, α represents the abnormal coefficient of a data in the blood glucose sequence; norm represents the normalization operation; n represents the number of data before the data in the blood glucose sequence; Represents the i-th data in the data before this data in the blood glucose sequence; represents the average value of the data before this data in the blood glucose sequence; m represents the number of data after this data in the blood glucose sequence; Indicates the jth data after the data in the blood glucose sequence; Indicates the average value of the data after this data in the blood glucose series; Indicates the slope of the data and the adjacent previous data; It represents the slope of the data and the adjacent next data.

[0006] Preferably, obtaining reference data in the blood glucose sequence using the abnormal coefficient of each data includes: A first threshold is set, and if the abnormal coefficient of a data in the blood glucose sequence is greater than the first threshold, the data is recorded as reference data.

[0007] Preferably, the screening conditions include: Screening criteria include blood glucose collection time range, height range, weight range, age range, gender restrictions and medical conditions.

[0008] Preferably, obtaining the data to be selected in the data set includes: A statistical histogram is established based on the data in a data set, with the horizontal axis representing the blood glucose value and the vertical axis representing the number. The frequency of each blood glucose value is obtained based on the statistical histogram, and the inverse of the number of blood glucose value types is used as the frequency threshold. The blood glucose data corresponding to the blood glucose values ​​with a frequency greater than or equal to the frequency threshold are recorded as candidate data.

[0009] Preferably, the calculation formula for the preference rate of the selected data is: , in, represents the preference rate of the data to be selected; P represents the frequency of the data to be selected; represents the variance of the data within a fixed range centered on the data to be selected; C represents the number of data within a fixed range centered on the data to be selected.

[0010] Preferably, calculating the first weight of the non-reference data according to the non-reference data in the blood glucose sequence and the reference value of the corresponding data set includes: An absolute value of a difference between the non-reference data and a reference value of a data set where the non-reference data is located is obtained, and the reciprocal of the sum of the absolute value of the difference and a preset value is a first weight of the non-reference data.

[0011] Preferably, calculating the second weight of the non-reference data includes: Calculate the absolute value of the position difference between a non-reference data and the reference data, and take the inverse to obtain the position difference; obtain the inverse of the sum of the variance of the data in the window centered on the non-reference data and a preset value, and record it as the fluctuation effect; average and normalize the position difference and the fluctuation effect to obtain the second weight of the non-reference data; if the variance of the data in the window centered on the non-reference data is greater than or equal to the second threshold, the result of the normalized position difference is the second weight of the non-reference data.

[0012] Preferably, calculating the modified weight of the non-reference data according to the first weight and the second weight of the non-reference data, and calculating the modified reference data using the modified weight of each non-reference data, comprises: Set a weight coefficient corresponding to the first weight and a weight coefficient corresponding to the second weight; perform weighted summation of the first and second weights according to the weight coefficient corresponding to the first weight and the weight coefficient corresponding to the second weight to obtain a corrected weight for the non-reference data; perform weighted summation of the non-reference data in the blood glucose sequence except the reference data using the corresponding normalized corrected weights to obtain the corrected reference data.

[0013] In a second aspect, the present invention also provides a management system for diabetes screening data, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the computer program is executed by the processor, the steps of a method for processing diabetes screening data are implemented.

[0014] The embodiments of the present invention have at least the following beneficial effects: the present invention obtains the blood glucose sequence of the person to be screened, calculates the abnormal coefficient of each data therein, and then obtains the reference data in the blood glucose sequence, and can obtain the data that needs to be corrected in the blood glucose sequence; further, the screening conditions are set, and the blood glucose sequences of the persons who meet the screening conditions are obtained in the blood glucose database as the reference blood glucose sequence of the person to be screened, and the data at the same position in the blood glucose sequence and each reference blood glucose sequence are used to form a data set, and the selected data in the data set are obtained, and the data set is used for subsequent analysis. The correction of the reference data not only refers to the blood glucose data of the person to be screened, but also refers to the blood glucose data of other persons in similar situations, thereby increasing the amount of data and making the subsequent analysis more accurate; then, the first weight of the non-reference data is calculated according to the non-reference data in the blood glucose sequence and the benchmark value of the corresponding data set, the second weight of the non-reference data is calculated according to the position difference between a non-reference data in the blood glucose sequence and the reference data and the variance of the data in the window centered on the non-reference data, and then the correction weight is calculated, and the corrected reference data is calculated using the correction weight of each non-reference data, so as to ensure that the corrected data value is more accurate, so that the screening result of diabetes is more accurate and reliable. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] 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.

[0016] Figure 1 A method flow chart of a method for processing diabetes screening data provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0017] In order to further explain the technical means and effects adopted by the present invention to achieve the predetermined invention purpose, the following is a detailed description of a diabetes screening data processing method and management system proposed by the present invention, its specific implementation, structure, features and effects, in conjunction with the accompanying drawings and preferred embodiments. In the following description, different "one embodiment" or "another embodiment" does not necessarily refer to the same embodiment. In addition, specific features, structures, or characteristics in one or more embodiments may be combined in any suitable form.

[0018] 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.

[0019] The following is a detailed description of a method for processing diabetes screening data and a specific solution of a management system provided by the present invention in conjunction with the accompanying drawings. Example

[0020] The main application scenarios of the present invention are: establishing a street blood glucose database in a hospital with a diabetes screening program based on the street range, regularly organizing street residents to undergo diabetes screening in designated hospitals and saving the screening data in the database. The screening method is a conventional method for diabetes screening (glucose tolerance test), that is, determining whether diabetes is present by measuring fasting blood glucose levels. Patients are required to fast for at least 8 hours, and blood is drawn and blood glucose levels are measured to determine whether there are symptoms of hyperglycemia. The normal fasting blood glucose range is usually 7099 mg / dL (or 3.96.1 mmol / L). If the fasting blood glucose level exceeds 100 mg / dL (or 7.0 mmol / L), further examination may be required.

[0021] See also Figure 1 , which shows a method flow chart of a method for processing diabetes screening data provided by an embodiment of the present invention, the method comprising the following steps: Step S1, collecting blood sugar data of the person to be screened and forming a blood sugar sequence; calculating the abnormal coefficient of the data according to the change of a data in the blood sugar sequence and the data before and after the data; and obtaining the reference data in the blood sugar sequence using the abnormal coefficient of each data.

[0022] First, since conventional blood sugar data is only collected once, but once is easily subject to data anomalies caused by the above-mentioned operational specification problems, the glucose tolerance test is performed to collect blood sugar data multiple times. The fasting blood sugar data of the person to be screened is collected by drawing blood, etc., and the blood sugar data is collected once every preset time after the person to be screened has not eaten for a specified time, such as 8 hours. The number of collections can be set to a preset number of times to obtain the blood sugar collection data at the corresponding time. The blood sugar data of the person to be screened constitutes a blood sugar sequence. Preferably, the preset time in the present invention is 20 minutes, and the preset number of times is 5 times. The implementer can adjust it according to the actual situation.

[0023] Then, determine whether there is abnormal data in the blood glucose collection data sequence that needs to be corrected, and correct it to obtain a corrected blood glucose sequence; then, perform a preliminary screening for diabetes based on the corrected blood glucose sequence, and determine whether there is a risk of diabetes by checking whether there is blood glucose data greater than or equal to 7.0mmol / L in the corrected blood glucose sequence, obtain the screening results, and save the relevant data in the blood glucose database; finally, determine whether the person to be screened has a risk of diabetes and whether further examination is needed based on the screening results. This is the entire screening process.

[0024] After obtaining the blood sugar sequence, the abnormal data in it needs to be corrected. During the initial screening of diabetes, if the fasting blood sugar concentration exceeds 7.0mmol / L, it is generally believed that the person participating in the screening may be at risk of diabetes and needs further examination. However, during the blood drawing process, the blood collector may not follow the standard collection process, such as improper selection of blood collection sites, improper use of blood collection needles, etc., or the subjects are not fully explained and guided during the collection, which leads to the subjects being nervous or not cooperating properly, resulting in the blood sugar data deviating from the normal blood sugar data, which can easily cause the collector to mistakenly believe that the subject's blood sugar is high and may have diabetes, affecting the accuracy of the screening results.

[0025] It takes time for blood sugar concentration to drop and rise, and this process may occur during the interval between blood draws. Therefore, the reliability of blood sugar data at a single moment is low, and it is impossible to directly identify interference anomalies according to the threshold. It is necessary to analyze it in combination with the characteristics of blood sugar data. Under normal circumstances, the fasting blood sugar data of people without blood sugar is relatively stable, generally maintained at 3.96mmol / L, while the fasting blood sugar data of diabetic patients is volatile, and the data is usually high, which may exceed 7.0mmol / L or even higher. Therefore, if the data is abnormal data, it may be reflected in the low overall volatility of the data before and after the abnormal data, and the high local volatility of the abnormal data. Abnormal data can be identified based on the above characteristics.

[0026] Based on the above analysis, the process of obtaining abnormal data is as follows: The abnormal coefficient of a data in the blood glucose sequence and the changes of the data before and after the data are calculated, and the overall volatility of the data before and after a data and the local volatility of the data and the surrounding blood glucose data are calculated. Among them, when the data is the first and last digit in the blood glucose sequence, the overall volatility is calculated as the overall volatility of the data after or before the data. The abnormal coefficient of the data is obtained according to the fluctuation characteristics of the data.

[0027] The calculation formula of the abnormal coefficient of the data in the blood glucose series is as follows: , Among them, α represents the abnormal coefficient of a data in the blood glucose sequence; norm represents the normalization operation; n represents the number of data before the data in the blood glucose sequence; Represents the i-th data in the data before this data in the blood glucose sequence; represents the average value of the data before this data in the blood glucose sequence; m represents the number of data after this data in the blood glucose sequence; Indicates the jth data after the data in the blood glucose sequence; Indicates the average value of the data after this data in the blood glucose series; Indicates the slope of the data and the adjacent previous data; It represents the slope of the data and the adjacent next data.

[0028] It represents the overall volatility of the n blood sugar data preceding a data. The smaller the value, the smaller the difference between the n blood sugar data preceding the data. The blood sugar data are closer and more stable. It represents the overall volatility of the m blood sugar data behind a data. The smaller the value, the smaller the difference between the m blood sugar data behind the data, the closer the blood sugar data are, and the more stable the blood sugar data are. It indicates the overall volatility of the data before and after a data. The smaller the value, the more stable the blood sugar data before and after the data. That is, the probability that the blood sugar data of the screened person is consistent with the normal blood sugar data is higher. If the data is too different from the surrounding data, it is likely to be abnormal. The larger the value is, the more mutations there are between the data and the previous and next data, and the greater the local volatility. The larger the value, the more stable the blood sugar data before and after the data is. If there is a large gap between the data and the previous and subsequent data, that is, the local volatility is large and the stability is low, then the probability that the data is abnormal blood sugar data is greater.

[0029] A first threshold is set. If the abnormal coefficient of one data in the blood glucose sequence is greater than the first threshold, the data is recorded as reference data and needs to be further corrected before being used. Preferably, the value of the first threshold is 0.6. In the blood glucose sequence, when one data is reference data, the other data are non-reference data. For example, if the first data is reference data, the other data in the blood glucose sequence except the first data are non-reference data. When the second data is reference data, the other data in the blood glucose sequence except the second data are non-reference data.

[0030] Step S2, set screening conditions, obtain blood glucose sequences of persons who meet the screening conditions in the blood glucose database as reference blood glucose sequences for persons to be screened; use the blood glucose sequences and data at the same position in each reference blood glucose sequence to form a data set, and obtain the data to be selected in the data set.

[0031] After finding the reference data of the person to be screened, the reference data needs to be corrected according to the data of the person to be screened. Since the blood sugar data around the reference data of the person to be screened is the single-time data of a single person, the reliability is difficult to determine. Because the blood sugar changes have a certain cycle, these values ​​will be more or less affected by improper operation, but the impact will not make the blood sugar data become reference data (abnormal data). Therefore, the data in the blood sugar database of the same street area can be analyzed to expand the data range, so that the reference of the correction is more, and the accuracy of the correction is guaranteed.

[0032] Screening conditions are set according to the characteristics of the person to be screened, including the blood glucose collection time range, height range, weight range, age range, gender restrictions and medical conditions, and blood glucose sequences that meet the screening conditions of the person to be screened are searched in the blood glucose database. The person corresponding to the blood glucose sequence screened out from the blood glucose database should meet the screening conditions, the time for collecting blood glucose data should be within the blood glucose collection time range of the person to be screened, the height, weight and age should be within the height, weight and age range set for the person to be screened, the gender should be the same, and the medical conditions should be the same. It should be noted that the setting of the screening conditions requires professional medical staff to set them.

[0033] For example, for a person to be screened, when screening is only performed based on age, height, weight and gender, the screening conditions are set as follows: the data of the remaining screened persons who are in the same age range (such as 20 to 29, or 30 to 39) and the age difference is less than 3 years, the same gender, the same height range (such as 160cm to 169cm, or 170cm to 179cm) and the height difference is less than 3cm, and the same weight range (such as 50kg to 55kg) and the weight difference is less than 3kg are statistically analyzed. For example, if the screened person is male, weighs 70kg, is 177cm tall and is 33 years old, the historical blood glucose data of the remaining screened persons in the same street area who are male, 30-36 years old, weigh 70-73kg and are 174-179cm tall should be selected for analysis.

[0034] Because the number of data in the blood glucose sequence of the person to be screened is the same as that in the reference blood glucose sequence, each position in the sequence corresponds to multiple blood glucose data, where the blood glucose sequence of the person to be screened is recorded as , the reference blood glucose sequence is recorded as ,in represents the vth reference blood glucose sequence, then the first reference blood glucose sequence is , and the rest of the reference blood glucose sequences are the same. Since the reference value of the same type of data is higher and less affected by interference, the data values ​​at the same position in the blood glucose sequence and each reference blood glucose sequence are extracted as a data set, recorded as , for example, the data sets consisting of the data at the first position and the data at the second position are and The same is true for data sets composed of data at other locations.

[0035] In order to ensure the accuracy of statistical blood sugar data, it is necessary to obtain an optimal benchmark value in each data set. Statistical data should be based on the data of most people. Although there may be abnormal data in the blood sugar data of other screened people, the abnormal data are in the minority. Therefore, the frequency of the numerical value of each blood sugar data will be different and it is impossible to be evenly distributed. The frequency of abnormal data is lower and the frequency of normal data is higher. Therefore, for any data set, the candidate data can be obtained according to the frequency of blood sugar data. Specifically: first, a statistical histogram is established based on the data in a data set, the horizontal axis represents the blood sugar value, and the vertical axis represents the number. The frequency of each blood sugar value is obtained according to the statistical histogram, and the inverse of the number of blood sugar value types is used as the frequency threshold. The blood sugar data corresponding to the blood sugar value with a frequency greater than or equal to the frequency threshold is recorded as the candidate data.

[0036] Thereby, a reference blood glucose sequence of the blood glucose sequence and candidate data of each data set can be obtained.

[0037] Step S3, calculates the preference rate of the candidate data according to the frequency of the candidate data in the data set and the variance and quantity of the data within a fixed range centered on the candidate data; takes the candidate data with the largest preference rate as the benchmark value of the data set; calculates the first weight of the non-reference data according to the non-reference data in the blood glucose sequence and the benchmark value of the corresponding data set.

[0038] Since the candidate data are not necessarily adjacent, there is an indefinite distance between the candidate data, that is, the difference in blood glucose values. Since normal blood glucose data fluctuates within a fixed value range, the more concentrated blood glucose values ​​in the candidate data are more reliable. Since the normal fasting blood glucose value is generally maintained at 3.96mmol / L, the normal range of fasting blood glucose for patients without diabetes is 3.9-6.1mmol / L, so the fixed fluctuation range can be set to a minimum of 0.5mmol / L and a maximum of no more than (6.1-3.9) / 3. All data within the fixed fluctuation range centered on each candidate data are obtained, and the distribution uniformity of these blood glucose values ​​and the number of candidate data within the fixed fluctuation range except the selected blood glucose value are calculated. The preference rate of each candidate data is obtained according to the frequency of the candidate data, the distribution uniformity of the surrounding blood glucose values, and the number of candidate data, and the candidate data with the largest preference rate is selected as the benchmark value.

[0039] , in, represents the preference rate of the data to be selected; P represents the frequency of the data to be selected; represents the variance of the data within a fixed range centered on the data to be selected; C represents the number of data within a fixed range centered on the data to be selected.

[0040] In this way, the preference rate of each candidate data in the data set can be obtained, and the candidate data with the largest preference rate is selected as the benchmark value of the data set.

[0041] If the non-reference blood glucose data of the person to be screened is closer to the baseline value of the data set, it means that the reliability of the non-reference blood glucose data is higher. When correcting the reference blood glucose data, a higher weight can be given, which is recorded as the first weight of the non-reference blood glucose data.

[0042] Furthermore, the absolute value of the difference between the non-reference data and the reference value of the data set where the non-reference data is located is obtained, and the reciprocal of the sum of the absolute value of the difference and the preset value is the first weight of the non-reference data. The specific calculation formula is: , Among them, β represents the first weight of non-reference data; Represents the tth data in the blood glucose sequence; Indicates the reference value of the tth data set, where the data in the tth data set is composed of the tth data in the blood glucose sequence and the reference blood glucose sequence. The larger β is, the more reliable the value of the non-reference blood glucose data is, and the larger the first weight is. The preset value is 1.

[0043] It should be noted that the purpose of the present invention is to correct the reference data. There may be one, multiple, or no reference data in the blood glucose sequence. When there are multiple reference data, one is selected as the reference data, and the other data are non-reference data.

[0044] Step S4, calculating a second weight of the non-reference data according to the position difference between the non-reference data and the reference data in the blood glucose sequence and the variance of the data in a window centered on the non-reference data.

[0045] Since statistical characteristics will lose local change trends and there are differences between individuals, it is necessary to obtain the second weight based on the blood sugar sequence of the person to be screened. Since the change of blood sugar caused by abnormal operation has a trend, that is, blood sugar will gradually tend to normal levels from a large value, and the real high blood sugar data has a fluctuating characteristic, the second weight can be obtained based on whether the blood sugar curve of the person to be screened has the corresponding characteristics.

[0046] The second weight of the non-reference data is calculated based on the position difference between a non-reference data and the reference data in the blood glucose sequence and the variance of the data in the window centered on the non-reference data. Specifically, the absolute value of the position difference between the non-reference data and the reference data is calculated, and the position difference is obtained by taking the inverse; the inverse of the sum of the variance of the data in the window centered on the non-reference data and the preset value is obtained, which is recorded as the fluctuation effect; the position difference and the fluctuation effect are averaged and normalized to obtain the second weight of the non-reference data; if the variance of the data in the window centered on the non-reference data is greater than or equal to the second threshold, the result of the normalized position difference is the second weight of the non-reference data. The specific calculation formula is: , Where γ represents the second weight of non-reference data; Indicates the location of the reference data, Indicates the location of non-reference data; Represents the variance of the data in the window centered on the non-reference data. The size of the window is 1*3, and 1 is the preset value. If the volatility is greater, the local blood glucose data may change unstably. If there is volatility, the blood glucose data of the person to be screened may have the precursor characteristics of diabetes. At this time, the credibility of each data is similar. At this time, the weight is only assigned based on the distance. That is, when the variance of the data in the window centered on the non-reference data is greater than the second threshold 0.6, the calculation formula of the second weight of the non-reference data is .

[0047] Through step S3 and step S4, a first weight and a second weight of non-reference data can be obtained.

[0048] Step S5, calculating the modified weight of the non-reference data according to the first weight and the second weight of the non-reference data, and calculating the modified reference data using the modified weight of each non-reference data.

[0049] The first weight is obtained according to the reliability of the non-reference data, and the second weight is obtained according to the influence of the non-reference data on the reference data. The weight of each non-reference data when the reference data is corrected can be obtained according to the first and second weights, that is, the correction weight, and the blood glucose value of the target reference data can be corrected according to the correction weight.

[0050] Furthermore, a weight coefficient corresponding to the first weight and a weight coefficient corresponding to the second weight are set; and the first and second weights are weighted and summed according to the weight coefficient corresponding to the first weight and the weight coefficient corresponding to the second weight to obtain a correction weight of the non-reference data. The specific calculation formula is: , Where W represents the correction weight of the selected non-reference data, represents the first weight of the selected non-reference data, represents the second weight of the selected non-reference data, They respectively represent the weight coefficient corresponding to the first weight and the weight coefficient corresponding to the second weight. Since the reliability of blood glucose data is more important for the accuracy of the correction result, in the embodiment of the present invention, is 0.7, is 0.3. The larger the correction weight, the more important the non-reference data is. Using this blood glucose value to correct the blood glucose value can obtain a more accurate correction result. Similarly, the correction weights of the remaining non-reference data can be obtained, and the correction weights of all non-reference data are normalized. The non-reference data in the blood glucose sequence other than the reference data are weighted and summed using the corresponding normalized correction weights to obtain the corrected reference data. In this way, the corrected reference data of each reference data in the blood glucose sequence can be obtained; the corrected reference data and other uncorrected data constitute the corrected blood glucose sequence of the person to be screened.

[0051] Finally, according to the screening requirements, determine whether there is blood glucose data greater than or equal to 7.0mmol / L in the corrected blood glucose sequence; if there is no blood glucose data greater than or equal to 7.0mmol / L, the screening data and results will be saved in the blood glucose database of the street to which the screened personnel belong; if there is blood glucose data greater than or equal to 7.0mmol / L, notify the screening personnel to conduct a more detailed examination to determine whether the person has diabetes; based on the final examination results, the data will be saved in the corresponding blood glucose database, and if the person has diabetes, the data will be stored in the diabetic patient database. Example

[0052] This embodiment provides a management system for diabetes screening data, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the computer program is executed by the processor, the steps of a method for processing diabetes screening data are implemented. Since a method for processing diabetes screening data has been described in detail in Example 1, it will not be described in detail here.

[0053] In summary, the present invention determines the abnormal data (reference data) in the blood glucose data of the person to be screened according to the characteristics of the blood glucose data, obtains the correction weights of the remaining non-reference data when correcting the target reference data according to the blood glucose collection data of the screened persons with the same conditions in the blood glucose database, and obtains the data value of the corrected reference data. It ensures that the corrected data value is more accurate and the screening results are more accurate and reliable.

[0054] It should be noted that the sequence of the above embodiments of the present invention is only for description and does not represent the advantages and disadvantages of the embodiments. The above is a description of a specific embodiment of this specification. In addition, the processes depicted in the accompanying drawings do not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0055] The various embodiments in this specification are described in a progressive manner, and the same or similar parts between the various embodiments can be referenced to each other, and each embodiment focuses on the differences from other embodiments.

[0056] 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 spirit and principle of the present invention should be included in the protection scope of the present invention.

Claims

1. A method for processing diabetes screening data, characterized in that: The method includes: Collect blood sugar data of the person to be screened and form a blood sugar sequence; calculate the abnormal coefficient of the data according to the change of a data in the blood sugar sequence and the data before and after the data; use the abnormal coefficient of each data to obtain the reference data in the blood sugar sequence; Setting screening conditions, obtaining blood glucose sequences of persons who meet the screening conditions in the blood glucose database as reference blood glucose sequences of persons to be screened; forming a data set using the data at the same position in the blood glucose sequence and each reference blood glucose sequence, and obtaining the data to be selected in the data set; Calculate the preference rate of the candidate data according to the frequency of the candidate data in the data set and the variance and quantity of the data within a fixed range centered on the candidate data; take the candidate data with the largest preference rate as the reference value of the data set; calculate the first weight of the non-reference data according to the non-reference data in the blood glucose sequence and the reference value of the corresponding data set; Calculate a second weight of the non-reference data according to a position difference between a non-reference data and reference data in the blood glucose sequence and a variance of data in a window centered on the non-reference data; The modified weight of the non-reference data is calculated according to the first weight and the second weight of the non-reference data, and the modified reference data is calculated using the modified weight of each non-reference data.

2. A method for processing diabetes screening data according to claim 1, characterized in that: The calculation formula of the abnormal coefficient is: ， Among them, α represents the abnormal coefficient of a data in the blood glucose sequence; norm represents the normalization operation; n represents the number of data before the data in the blood glucose sequence; Represents the i-th data in the data before this data in the blood glucose sequence; represents the average value of the data before this data in the blood glucose sequence; m represents the number of data after this data in the blood glucose sequence; Indicates the jth data after the data in the blood glucose sequence; Indicates the average value of the data after this data in the blood glucose series; Indicates the slope of the data and the adjacent previous data; It represents the slope of the data and the adjacent next data.

3. A method for processing diabetes screening data according to claim 1, characterized in that: The method of obtaining reference data in the blood glucose sequence by using the abnormal coefficient of each data includes: A first threshold is set, and if the abnormal coefficient of a data in the blood glucose sequence is greater than the first threshold, the data is recorded as reference data.

4. The method for processing diabetes screening data according to claim 1, characterized in that: The screening conditions include: Screening criteria include blood glucose collection time range, height range, weight range, age range, gender restrictions and medical conditions.

5. The method for processing diabetes screening data according to claim 1, characterized in that: The step of obtaining the data to be selected from the data set includes: A statistical histogram is established based on the data in a data set, with the horizontal axis representing the blood glucose value and the vertical axis representing the number. The frequency of each blood glucose value is obtained based on the statistical histogram, and the inverse of the number of blood glucose value types is used as the frequency threshold. The blood glucose data corresponding to the blood glucose values ​​with a frequency greater than or equal to the frequency threshold are recorded as candidate data.

6. A method for processing diabetes screening data according to claim 1, characterized in that: The calculation formula of the preference rate of the selected data is: ， in, represents the preference rate of the data to be selected; P represents the frequency of the data to be selected; represents the variance of the data within a fixed range centered on the data to be selected; C represents the number of data within a fixed range centered on the data to be selected.

7. The method for processing diabetes screening data according to claim 1, characterized in that: The step of calculating the first weight of the non-reference data according to the non-reference data in the blood glucose sequence and the reference value of the corresponding data set includes: An absolute value of a difference between the non-reference data and a reference value of a data set where the non-reference data is located is obtained, and the reciprocal of the sum of the absolute value of the difference and a preset value is a first weight of the non-reference data.

8. The method for processing diabetes screening data according to claim 1, characterized in that: The calculating the second weight of the non-reference data includes: Calculate the absolute value of the position difference between a non-reference data and the reference data, and take the inverse to obtain the position difference; obtain the inverse of the sum of the variance of the data in the window centered on the non-reference data and a preset value, and record it as the fluctuation effect; average and normalize the position difference and the fluctuation effect to obtain the second weight of the non-reference data; if the variance of the data in the window centered on the non-reference data is greater than or equal to the second threshold, the result of the normalized position difference is the second weight of the non-reference data.

9. The method for processing diabetes screening data according to claim 1, characterized in that: The step of calculating the modified weight of the non-reference data according to the first weight and the second weight of the non-reference data, and calculating the modified reference data using the modified weight of each non-reference data comprises: Set a weight coefficient corresponding to the first weight and a weight coefficient corresponding to the second weight; perform weighted summation of the first and second weights according to the weight coefficient corresponding to the first weight and the weight coefficient corresponding to the second weight to obtain a corrected weight for the non-reference data; perform weighted summation of the non-reference data in the blood glucose sequence except the reference data using the corresponding normalized corrected weights to obtain the corrected reference data.

10. A management system for diabetes screening data, comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: When the computer program is executed by a processor, the steps of a method for processing diabetes screening data as described in any one of claims 1 to 9 are implemented.

Images