WO2023130599A1 - 标签生成方法和装置、监控血糖水平的系统 - Google Patents

标签生成方法和装置、监控血糖水平的系统 Download PDF

Info

Publication number
WO2023130599A1
WO2023130599A1 PCT/CN2022/085452 CN2022085452W WO2023130599A1 WO 2023130599 A1 WO2023130599 A1 WO 2023130599A1 CN 2022085452 W CN2022085452 W CN 2022085452W WO 2023130599 A1 WO2023130599 A1 WO 2023130599A1
Authority
WO
WIPO (PCT)
Prior art keywords
label
tag
group
case
category
Prior art date
Application number
PCT/CN2022/085452
Other languages
English (en)
French (fr)
Inventor
韩洋
蒋娟
雷大鹏
Original Assignee
苏州百孝医疗科技有限公司
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 苏州百孝医疗科技有限公司 filed Critical 苏州百孝医疗科技有限公司
Publication of WO2023130599A1 publication Critical patent/WO2023130599A1/zh

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F18/00—Pattern recognition
    • G06F18/20—Analysing
    • G06F18/24—Classification techniques
    • G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F18/00—Pattern recognition
    • G06F18/20—Analysing
    • G06F18/25—Fusion techniques
    • G06F18/254—Fusion techniques of classification results, e.g. of results related to same input data
    • G—PHYSICS
    • G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
    • G16H—HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
    • G16H50/00—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics
    • G16H50/30—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics for calculating health indices; for individual health risk assessment

Definitions

  • the present application relates to the field of physiological signal processing, for example, to a method and device for generating a label, and a system for monitoring blood sugar levels.
  • the continuous glucose monitoring system Compared with the traditional blood glucose measurement equipment that needs to frequently measure fingertip blood, the continuous glucose monitoring system based on the development of biosensor technology (continuously providing real-time glucose concentration data at a certain frequency) provides people with a better understanding of the changes in their glucose concentration data. Level opportunities, the system provides a data basis for monitoring glucose concentration, which is very useful for personal health management.
  • the continuous glucose monitoring system can provide users with real-time glucose concentration data, and these glucose concentration data change as the data of each user using the system changes.
  • Glucose predictions are calculated based on multiple sources of input data that are indicative of the user's current state and future state to some extent.
  • predicting the future state is realized by calculating the prediction of future glucose on the basis of big data based on various inputs.
  • the above solution is to predict the future state of the user based on current user data information (such as various inputs, measurement data, or estimated carbohydrate values, etc.).
  • the above scheme does not take into account the user's personalization and the real effect brought by some actual user data, and only provides a relatively simple glucose level prediction, and the prediction accuracy is far from the purpose of giving users reliable advice , resulting in poor user experience.
  • the present application provides a label generation method and device, and a system for monitoring blood sugar levels.
  • This application provides a label generation method, including:
  • the first case includes glucose level data in a first time period and user data at the current moment;
  • the second tag group is determined based on the glucose level data of the second time period; the tag set formed by all historical tag groups and the second tag group contains multiple tag categories, the Each tag category in the tag set includes tags with at least two content attributes;
  • each tag category For each tag category, comparing the content attributes of tags corresponding to each tag category in the first tag group and the second tag group to obtain a comparison result, wherein the comparison result includes each tag category Whether the content attributes of the corresponding tags are the same;
  • a third tag group is determined; wherein, the third tag group is selected from the second tag group or a fusion tag group, and the fusion tag group is the first tag group Labels corresponding to the same label category in the second label group are obtained after fusion based on a first preset rule;
  • the present application also provides a label generation device, including:
  • the first case acquisition module is configured to acquire the first case, wherein the first case includes glucose level data in the first time period and user data at the current moment;
  • the comparison module is configured to use at least one model for data comparison to obtain a comparison result, wherein the comparison result includes similarity; the data of one side of the data comparison is the first case, and the data of the other side is Historical cases in the user knowledge base; the user knowledge base contains a natural number of historical cases, and each historical case has a corresponding historical label group;
  • the first label group output module is set to select the maximum value in the similarity, if the maximum value is greater than the similarity threshold, the historical case corresponding to the maximum value is used as the second case, and the second case corresponds to The historical label group of is used as the first label group and output;
  • the second tag group acquisition module is configured to acquire a second tag group, wherein the second tag group is determined based on the glucose level data in the second time period; all historical tag groups and the tags formed by the second tag group
  • the set contains multiple label categories, and each label category in the label set includes labels with at least two content attributes;
  • the update module is configured to compare the content attributes of the labels corresponding to each label category in the first label group and the second label group for each label category, and obtain a comparison result, wherein the comparison result Including whether the content attributes of the tags corresponding to each tag category are the same; based on the comparison result, determine a third tag group; wherein, the third tag group is selected from the second tag group or the fusion tag group , the fused tag group is obtained after merging tags corresponding to the same tag category in the first tag group and the second tag group based on a first preset rule; combining the first case and the The third tag group corresponding to the first case is updated to the user knowledge base.
  • the present application also provides a system for monitoring blood sugar levels, comprising:
  • a sensor configured to acquire glucose level data
  • a wireless transmitter configured to transmit said glucose level data
  • a mobile computing device comprising:
  • a wireless receiver configured to receive said glucose level data
  • a memory configured to store data comprising said glucose level data
  • a processor configured to process said data, and a software application comprising instructions stored in said memory, said instructions implementing said label generation method as described above when executed.
  • the present application also provides an electronic device, which includes a memory, a processor, and a computer program stored on the memory and operable on the processor.
  • the processor executes the program, the label generation method as described above is realized.
  • the present application also provides a non-transitory computer-readable storage medium, on which a computer program is stored, and when the computer program is executed by a processor, the label generation method as described above is realized.
  • FIG. 1 is a schematic structural diagram of an implementation environment provided by an embodiment of the present application.
  • FIG. 2 is a schematic flow diagram of a label generation method provided by an embodiment of the present application.
  • Fig. 3 is a schematic diagram of the use effect of a label generation method provided by the embodiment of the present application.
  • Fig. 4 is a schematic structural diagram of a label generation device provided by an embodiment of the present application.
  • Fig. 5 is a schematic structural diagram of an electronic device provided by an embodiment of the present application.
  • FIG. 1 is a schematic structural diagram of an implementation environment provided by an embodiment of the present application.
  • the implementation environment includes: a terminal 100 and/or a server 200 .
  • the terminal 100 may be an electronic device with data processing capabilities, such as a mobile phone, a tablet computer, an e-book reader, a Moving Picture Experts Group Audio Layer III (MP3) player, and a Moving Picture Experts Group Audio Layer III (MP3) player. Audio layer 4 (Moving Picture Experts Group Audio Layer IV, MP4) player, laptop portable computer and desktop computer, etc.
  • MP3 Moving Picture Experts Group Audio Layer III
  • MP3 Moving Picture Experts Group Audio Layer III
  • MP3 Moving Picture Experts Group Audio Layer III
  • MP3 Moving Picture Experts Group Audio Layer III
  • MP3 Moving Picture Experts Group Audio Layer III
  • MP3 Moving Picture Experts Group Audio Layer III
  • MP3 Moving Picture Experts Group Audio Layer III
  • MP3 Moving Picture Experts Group Audio Layer III
  • MP3 Moving Picture Experts Group Audio Layer III
  • MP3 Moving Picture Experts Group Audio Layer III
  • MP3 Moving Picture Experts Group Audio Layer III
  • An application program client or a browser may be installed in the terminal 100, and the web page client of the application program may be accessed through the browser.
  • the application program client and the web page client are collectively referred to as clients, which will not be specifically stated below.
  • the server 200 may be a local or remote server, or a server cluster composed of multiple servers, or a cloud computing service center. When the terminal 100 and the server 200 are simultaneously processing services related to the application, the server 200 may be configured to interact with the terminal 100 to provide services related to the application.
  • the server 200 is a server corresponding to the client.
  • the server 200 and the terminal 100 can be combined to implement multiple functions provided by the client, and are usually set up by an Internet service provider.
  • the terminal 100 and the server 200 may be connected through a wireless network or a wired network.
  • the glucose level data for the first time period includes, but is not limited to, time-stamped glucose concentration (i.e., blood glucose concentration, blood glucose measurement, or blood glucose value) data, data associated with the glucose concentration, the glucose level data for the first time period Acquisition methods include but are not limited to acquisition through a glucose concentration sensor, collection through network delivery, or other collection methods. For example, the user's glucose concentration data of the first time period with a time stamp continuously output by the glucose concentration sensor in a certain period may be acquired. The first time period extends from the first moment to the current moment, and the first moment is before the current moment.
  • time-stamped glucose concentration i.e., blood glucose concentration, blood glucose measurement, or blood glucose value
  • Acquisition methods include but are not limited to acquisition through a glucose concentration sensor, collection through network delivery, or other collection methods.
  • the user's glucose concentration data of the first time period with a time stamp continuously output by the glucose concentration sensor in a certain period may be acquired.
  • the first time period extends from the first moment to the current moment, and
  • the first time period may be 10 minutes to 2 hours before the current moment, and the first moment continues
  • the user's glucose concentration data up to the current moment can be represented as a section of data waveform with the past first time period as the horizontal axis and the collected data as the vertical axis.
  • User data is data at the current moment, and is generally data associated with glucose concentration input by the user or in other ways, for example, data related to life events such as carbohydrate intake and exercise.
  • Carbohydrate data can be obtained by software that automatically recognizes calories in food pictures.
  • At least one model may be a formula for calculating similarity, which is used to compare the similarity between the first case and the historical cases in the user knowledge base.
  • the user knowledge base contains a natural number of historical cases, that is, there may be no historical cases in the user knowledge base.
  • the comparison result does not include the similarity.
  • the comparison result is a comparison failure.
  • S3 can be suspended, and the first label group can not be output, and directly jump to S4.
  • Each historical label group can contain multiple labels.
  • the multiple labels can be labels of different categories, or labels of the same category.
  • the categories can be classified according to the change rate category, target blood sugar time ratio category, etc.
  • Each label group contains at least one label. For example, each historical case has a label for each category in the historical label group.
  • a higher similarity means that the history label group of the historical case corresponding to the similarity is more in line with the combination of the user's current blood sugar status and the current user data. Sort the similarity, select the maximum value in the sorting results, and then compare the maximum value with the similarity threshold. If the maximum value is not greater than the similarity threshold, it means that no similar historical cases have been compared. In this case, no output label, and continue to execute S4 to S5; if the maximum value is greater than the similarity threshold, then use the historical case corresponding to the maximum value as the second case (also known as the old case), and output the first label group on the display interface.
  • the second tag group is determined based on the glucose level data in the second time period; wherein, the tag set formed by all the historical tag groups and the second tag group contains a plurality of tag categories, the Each label category in the label set includes labels with at least two content attributes.
  • the tag set covers all historical tag groups and the second tag group, which may be an initial small tag set, and gradually expand, optimize and personalize a tag set as the number of executions of the method of this application increases.
  • the reason why the concept of label set is proposed is to illustrate that each label category in the label set includes at least two labels of content attributes, that is, the labels corresponding to each label category in the first label group and the second label group.
  • the content attributes may be the same, or they may be different.
  • the same content attributes indicate that the first output label group is predicted more accurately, which is highly consistent with the second label group.
  • the content attributes are different, indicating that the output first label group is not predicted accurately.
  • the matching degree of the two label groups is low, or even opposite.
  • the acquisition of the second tag group is for updating the user knowledge base based on the first case in S5, the second time period extends from the current moment to the second moment, and the second moment is after the current moment.
  • the second tag group is determined based on the glucose level data of the second time period (the current moment extends to the second moment, which may include the current moment or not). Therefore, the second tag group is very close and can reflect A set of tags for the user's true glucose level for the second time period.
  • each tag category For each tag category, compare the content attributes of the tags corresponding to each tag category in the first tag group and the second tag group to obtain a comparison result, the comparison result including each tag category Whether the content attributes of the tags are the same; based on the comparison result, a third tag group is determined; the third tag group is selected from the second tag group or the fusion tag group, and the fusion tag group is the The tags corresponding to the same tag category in the first tag group and the second tag group are fused based on the first preset rule; the first case and its corresponding third tag group are updated to the user knowledge library.
  • the third tag group corresponding to the updated first case needs to be determined based on the comparison result.
  • the determination method can adopt various methods in the following embodiments.
  • the third tag group comes from the second tag group, the first tag group and the second tag group.
  • the fused tag group after the tag group is fused.
  • the fusion label group is to fully utilize the combination of the first label group and the second label group to optimize the third label group under certain circumstances. In this specific case, the effect of the fusion label group is better than the independent first label group, or the first Two tab groups.
  • the label generation scheme in the user knowledge base becomes more and more accurate and personalized, and better labels can be realized.
  • the user knowledge base is updated once. If the data in the user knowledge base exceeds a certain range, expired data can be regularly deleted and new data within a period of time can be retained.
  • the method of the present application acquires the first case, compares the first case with the historical cases in the user knowledge base, obtains the similarity, selects the historical case whose maximum value of similarity is greater than the similarity threshold as the second case, and compares the second
  • the historical label group corresponding to the case is output as the first label group; and based on the user's real glucose level in the second time period, the second label group is determined, and the first label group and the second label group are jointly used to update the user knowledge base.
  • the second label group is very close to and can reflect the user's real glucose level; for each label category, comparing the content attributes of the labels corresponding to the same label category in the first label group and the second label group, the actual effect will be better
  • the label group of the user is updated to the user knowledge base; as the number of times the user uses the method of this application increases, the more precise label groups to choose from in the historical cases in the user knowledge base, and the output will be more accurate and reliable.
  • This application Fully consider the user's personalized characteristic data and glucose level data for a period of time, so that users can enjoy richer and more reasonable labels, the label output is accurate, the effect is better, and the user experience is improved.
  • the output label group can be given to the user Control glucose levels for better, more reliable advice.
  • the method further includes: acquiring the user's glucose level data from the blood glucose measuring device associated with the user through the network.
  • the network is a wired network or a wireless network
  • the blood glucose measurement equipment associated with the user includes, but is not limited to, a blood glucose meter with a blood glucose sensor, blood collection measurement equipment, and other physiological data collection equipment that can collect glucose concentration data.
  • the blood glucose measuring device is a continuous blood glucose monitoring device capable of continuously collecting glucose concentration data in real time.
  • the glucose level data includes the glucose level data of the first time period and the glucose level data of the second time period; the glucose level data of the first time period includes the first glucose concentration data and the first glucose concentration change rate, the The glucose level data for the second time period includes second glucose concentration data and a second rate of change of glucose concentration.
  • the first glucose concentration data includes: the first blood glucose measurement value at the current moment and its corresponding first time stamp, the historical blood glucose collection data between the first moment and the current moment, and the historical blood glucose collection data includes Continuously distributed multiple historical blood glucose measurement values and corresponding multiple historical time stamps.
  • the preset time interval is the interval at which the continuous blood glucose monitoring device generates blood glucose, such as 3 minutes.
  • the historical blood glucose collection data includes a plurality of historical blood glucose measurement values continuously distributed in a period of 3 minutes and a plurality of corresponding historical time stamps in the first time period from the current moment.
  • the first rate of change of glucose concentration is obtained based on the rate of change of positive and negative values based on the first glucose concentration data; the second rate of change of glucose concentration is based on the change of positive and negative values based on the second data of glucose concentration obtained at a rate.
  • the glucose concentration change may be determined based on the first blood glucose measurement value and its corresponding first time stamp, and a second value selected from the plurality of historical blood glucose measurement values and its corresponding plurality of historical time stamps rate; the second value includes a second blood glucose measurement value and its corresponding second time stamp, the second time stamp being associated with the first time stamp.
  • the first glucose concentration change rate is obtained based on the first glucose concentration data in the form of positive and negative value change rates.
  • the calculation method of the first glucose concentration change rate is: (first blood glucose measurement value - second Blood glucose measurement value)/(first time stamp-second time stamp), the selection of the second value can be selected within the third time period from the current moment, the third time period can be 1 minute to 30 minutes, for example, The data of 3 minutes before the current time can be selected. If there is any data missing or abnormal, other data in the third time period from the current time can be selected.
  • the glucose level data of the second time period includes the second glucose concentration data and the second glucose concentration change rate; similarly, the second glucose concentration change rate is based on the second glucose concentration data with positive and negative value change rate
  • the rate of change of positive and negative values can reflect the rise, fall or steady level of glucose concentration.
  • the rate of change of positive and negative values can be converted into a value between 0-100 through a function before use.
  • the second glucose concentration data in the second time period may overlap with the first glucose concentration data in the first time period at the current moment, and the second glucose concentration data in the second time period may not include the data at the current moment.
  • This application fully considers the glucose level data of the first time period to generate and output the label group; and based on the glucose level data of the second time period and the generated first label group, update the user knowledge base, the second time period Glucose level data is mainly used to combine the first tag group to form the third tag group for updating, so that after the user knowledge base is updated many times, the generated scheme is more suitable for the user.
  • the acquisition of the first case includes:
  • the user data includes one or more events associated with glucose concentration, and one or more user feature data.
  • User data includes data associated with glucose concentration, such as carbohydrate intake and its intake, exercise amount, etc., which are manually input by the user, identified by using a picture obtained by a mobile phone application (Application, APP), or obtained by other devices. User data at the current moment is also timestamped. When generating the tag group, the current user data associated with the glucose concentration can be fully considered, ensuring the accuracy of the tag group.
  • glucose concentration such as carbohydrate intake and its intake, exercise amount, etc.
  • the one or more events are associated with one or more of carbohydrate consumption, exercise, sleep, and substance administration; the substance administration includes medication type, medication dosage, carbohydrate Administration amount; the type of medication includes at least one of long-acting insulin, short-acting insulin and fast-acting insulin.
  • the main effect time period of fast-acting insulin is within 30 minutes, and the main time period of long-acting insulin is within 2 hours , so events at the current moment need to be considered.
  • the one or more user characteristic data are associated with at least one of the user's basic physiological information and personal information;
  • the basic physiological information includes insulin sensitivity coefficient, insulin-carbohydrate ratio At least one;
  • the personal information includes at least one of gender, region, type of diabetes, age, weight, and insulin history.
  • the region, diabetes type, age, etc. are all personalized based on the user's own settings. Considering the above factors, a more personalized user knowledge base can be obtained, and the generated tag group is more in line with the real situation of the user.
  • the type of diabetes includes type I diabetes, type II diabetes, and gestational diabetes, because users of each type of diabetes may have similar glucose concentration patterns, as well as similar food and drug sensitivities, etc.
  • the change of has a certain regularity, so when the type of diabetes is taken into account when generating the label group, a personalized user knowledge base can be obtained that is distinguished according to the type of diabetes the user belongs to, and the output results are more suitable for users with this type of diabetes.
  • the data comparison is performed using at least one model to obtain a comparison result, and the comparison result includes similarity, including:
  • Regularization includes converting non-numerical key feature data into numbers using preset normalization rules, such as "0" for type I diabetes and "1” for type II diabetes, such as male diabetic patients is “0", and the value of female diabetic patients is "1", using the preset normalization rules such as this to perform regularization processing on some data in the first case that does not have quantitative values.
  • Glucose concentration data that is, glucose concentration value
  • blood glucose change rate blood glucose change rate
  • carbohydrate intake and exercise amount, which already have quantitative values, do not need to be regularized in this step.
  • the at least two compared feature values include feature values obtained from at least two of the glucose level data in the first time period and the user data at the current moment.
  • At least two absolute distances are obtained based on at least two comparison feature values and at least two historical feature values corresponding to the at least two comparison feature values respectively in the historical feature value group included in the historical case.
  • At least two comparison feature values can form a new case of the user, that is, the first case, and case matching is performed in the user knowledge base based on the new case.
  • the historical eigenvalue group contained in each historical case contains multiple historical eigenvalues.
  • each eigenvalue in the new case is compared with the corresponding historical eigenvalue in the historical eigenvalue group.
  • the data under each category for example, the glucose concentration data belongs to the same category, and the carbohydrate intake data belongs to the same category, and the feature values under each category are compared one by one; at least two absolute distances can be obtained.
  • the similarity is obtained, and the feature weight is based on the correlation between the comparison feature value corresponding to the feature weight and the glucose level. definite.
  • F represents the eigenvalue
  • F new represents a eigenvalue of a type of data of a new case
  • F old represents a eigenvalue of a historical case corresponding to F new
  • d' represents the absolute distance between two eigenvalues before normalization (difference degree); then, utilize preset rules to normalize the absolute distance not in the range of 0-1, and convert it into a numerical value of 0-1, such as the absolute distance normalization process of glucose concentration value can be expressed as:
  • d Glu represents the absolute distance of the blood glucose concentration value after normalization
  • d' Glu represents the absolute distance of the blood glucose concentration value before normalization
  • F NewGlu represents the blood glucose concentration value of the new case.
  • the absolute distances of other values can also be normalized one by one with the absolute distance normalization processing method of the glucose concentration value; the difference score between the new case and the historical case is:
  • d 1 , d 2 ,..., d n represent the absolute distance from the first eigenvalue to the nth eigenvalue
  • a 1 , a 2 ,..., a n represent the first eigenvalue to the nth eigenvalue respectively.
  • the characteristic weight (factor) of the value the greater the influence of the characteristic value on the glucose level, the greater the value of the corresponding characteristic weight.
  • S the greater the similarity between the two cases is.
  • the similarity threshold is preset and can be user-defined or user-input. The higher the similarity threshold is set, the smaller the probability of successful comparison and the more accurate the result. In order to take into account both the probability of successful comparison and the accuracy of comparison, the similarity threshold is set based on the comprehensive effect. For example, the similarity threshold can be set to 0.9, that is, when the similarity is greater than 90%, it is judged that the comparison is successful.
  • the comparison of data is performed using at least one model to obtain a comparison result, and the comparison result includes similarity; the data of one side of the data comparison is the first case, the other side
  • the data are historical cases in the user knowledge base, including:
  • pre-screening is performed in the user knowledge base to obtain a pre-selected library; historical cases in the pre-selected library contain historical feature values that are identical to the at least one comparison feature value part; at least one model is used for data comparison to obtain a comparison result, and the comparison result includes similarity; one side of the data comparison is the first case, and the other side is the data in the pre-selected library historical case.
  • pre-screening can also be performed based on some comparison feature values.
  • pre-screening is performed in the user knowledge base to obtain a pre-selected library; historical cases in the pre-selected library contain historical feature values that are identical to the at least one comparison feature value part; at least one model is used for data comparison to obtain a comparison result, and the comparison result includes similarity; one side of the data comparison is the first case, and the other side is the data in the pre-selected library historical case.
  • At least one comparison feature value can be one or two or more.
  • the type of medication screen out the historical cases consistent with the type of medication used by the user, and then further compare the new cases with the historical cases with the same type of medication; that is, first exclude the historical cases with inconsistent medication types.
  • the duration of medication historical cases that are consistent with the duration of medication of the user can be screened out, and then new cases are compared with these historical cases with the same duration of medication; that is, historical cases with inconsistent duration of medication can be excluded first.
  • multiple comparison feature values can also be combined to perform pre-screening, and only historical cases that are consistent with the user's medication period and medication type can be screened out, and then new cases can be further compared with these historical cases that are consistent with these medication years and medication types. Comparison; that is, first exclude historical cases with inconsistent drug use years or drug types. It is also possible not to perform pre-screening, and directly compare one by one based on the comparison feature value, and then compare the similarity with the similarity threshold after obtaining the similarity.
  • the label generation method in the selection of the maximum value in the similarity, if the maximum value is greater than the similarity threshold, the historical case corresponding to the maximum value is used as the second case, and the second case corresponds to After the history label group is used as the first label group and output, it also includes:
  • the user's execution of the first case may be evaluated based on the user's operation after the output of the first label group, or the glucose level data in the second time period.
  • the second tag group cannot be associated with the first case, so there is no need to obtain the second tag group, and there is no need to update the first case to the user knowledge base.
  • the execution of S4 is suspended to S5.
  • the first case may be updated to the user knowledge base, and at this time, the execution of S4 to S5 is continued. It is ensured that the cases entered into the user knowledge base have actually occurred, and the label group corresponding to the case can also reflect the real situation after the case occurred.
  • the data comparison is performed using at least one model and the comparison result is obtained, it also includes:
  • the process is suspended at this time.
  • the acquiring the second tag group, the second tag group is determined based on the glucose level data in the second time period, including:
  • the second tag group is obtained;
  • the second preset rule includes: dividing the glucose level data in the second time period based on each content attribute to obtain the first attribute group; based on the first attribute group and a predetermined association relationship to obtain the second label group, the association relationship is the association relationship between multiple content attribute groups and multiple pre-stored label groups, and the second label group is obtained from the multiple The tab set selected from the pre-stored tab sets.
  • the association relationship between multiple content attribute groups and multiple pre-stored tag groups is shown in Table 1 below.
  • the pre-stored tag groups include tag group 1, tag group 2, . . . , tag group M.
  • the second label group is obtained, for example, if the first attribute group is the first alarm interval, the first change rate interval, and the first target blood glucose time ratio interval , select the tag group 1 corresponding to the first attribute group as the second tag group, if the first attribute group is the second alarm interval, the second change rate interval, and the second target blood glucose time ratio interval, select The label group 2 is used as the second label group.
  • the target blood glucose time ratio interval is determined based on the expected blood glucose level and the glucose level data of the second time period, and the expected blood glucose level can be system-defined or user-defined.
  • the target blood glucose time ratio can be converted into a value between 0-100 by a function firstly and divided into the first target blood sugar time ratio interval, the second target blood glucose time ratio interval to the Mth target blood glucose time ratio interval.
  • the first target blood glucose time proportion interval is 0-20
  • the second target blood glucose time proportion interval is 20-40
  • the Mth target blood glucose time proportion interval is 80-100, and so on.
  • the first change rate interval is 0-20
  • the second change rate interval is 20-40
  • the Mth change rate interval is 80-100, and so on.
  • the first alarm interval is low blood sugar alarm
  • the second alarm interval is no alarm
  • the Mth alarm interval is high blood sugar alarm and so on.
  • the label set composed of all the historical label groups and the second label group contains multiple label categories, including:
  • the tag set formed by all the historical tag groups and the second tag group contains at least a plurality of tag categories, and the plurality of tag categories include at least two of alarm, rate of change, and target blood glucose time ratio.
  • the tags corresponding to the alarm category can include tags such as hyperglycemia alarm after 45 minutes, hyperglycemia alarm after 20 minutes, and ultra-low blood sugar alarm; tags corresponding to rate of change can include rapid rise in blood sugar, stable blood sugar change, Tags such as blood sugar steady drop; tags corresponding to the target blood sugar time ratio category can include tags such as target blood sugar time ratio higher than the stable value and target blood sugar time ratio lower than the stable value.
  • each label category in the label set includes labels of at least two content attributes, including at least one of the following situations:
  • tags corresponding to the alarm class there are at least tags whose content attribute is the first alarm interval and tags whose content attribute is the second alarm interval; among the tags corresponding to the change rate class, at least the content attribute is the first change rate interval and the label whose content attribute is the second change rate interval; among the tags corresponding to the target blood glucose time ratio class, there are at least tags whose content attribute is the first target blood glucose time ratio interval, and whose content attribute is the second target The label of the blood glucose time ratio interval.
  • each label category can also be distinguished.
  • the above-mentioned hyperglycemia alarm occurs after 45 minutes, and the hyperglycemia alarm occurs after 20 minutes.
  • the first alarm interval; the rapid rise of blood sugar is the Mth change rate interval; the steady decline of blood sugar is the first change rate interval; the target blood sugar time ratio is higher than the ratio threshold is the Mth target blood sugar time ratio interval; the target blood sugar time ratio Below the proportion threshold is the first target blood glucose time proportion interval.
  • the label generating method before the historical label group corresponding to the second case is used as the first label group and output, it also includes:
  • the presence of abnormal data in the first case refers to the existence of abnormalities in glucose level data or user data.
  • the abnormality of glucose level data and user data is generally judged based on big data, past user data history and experience.
  • Abnormal glucose level data may be caused by sensor abnormality, network abnormality, data exceeding a certain range, or data missing.
  • User data exceptions may be due to user-entered values outside a certain range of normal applicability.
  • the label generation method before the second label group is obtained, it also includes:
  • the determination and reason of the abnormality of the glucose level data in the second time period are the same as those of the abnormal data in the first case, and will not be repeated here. If there is abnormal data in the glucose level data in the second time period, the update of the second label group with erroneous data to the user knowledge base is suspended, so as to avoid applying the wrong second label group to the next label generation method.
  • the third label group is determined based on the comparison result, including:
  • the comparison result is the first comparison result, it is determined that the fused label group is the third label group; the first comparison result is: the number of label categories corresponding to the same label category with the same content attribute accounts for the number of all label categories The ratio is greater than the ratio threshold; when the comparison result is the second comparison result, it is determined that the second tag group is the third tag group; the second comparison result is: tags corresponding to the same tag category have the same content attributes The ratio of the number of label categories to the number of all label categories is not greater than the ratio threshold.
  • the above ratio threshold can be 80%-90%.
  • the first tag group and the second tag group There are many same content attributes, it can be determined that the fusion label group is the third label group; when the proportion of the number of label categories with the same content attributes of the labels corresponding to the same label category to the number of all label categories is not greater than the ratio threshold, then The same content attributes in the first label group and the second label group are few, and the second label group may be determined to be the third label group, and the second label group closer to the real level may be used for updating.
  • the above group-by-group update method can save system resources and improve update efficiency and response speed.
  • the third tag group is determined based on the comparison result, including:
  • the label corresponding to the first label category in the fusion label group is used as the label corresponding to the first label category in the third label group;
  • the multiple The label category includes at least the first label category, and the third comparison result is that the content attributes of the labels corresponding to the first label category are the same;
  • the comparison result includes the first label category corresponding to the fourth comparison result, the The label corresponding to the first label category in the first label group is used as the label corresponding to the first label category in the third label group;
  • the fourth comparison result is that the content attributes of the labels corresponding to the first label category are different.
  • Both the first label group and the second label group are classified based on label categories, and when comparing, the labels of each label category in the third label group are determined by comparing each label category one by one. If the content attributes of the tags corresponding to the first tag category in the first tag group and the second tag group are the same, then determine that the tag corresponding to the first tag category in the fusion tag group is the first tag in the third tag group The label corresponding to the category; if the content attributes of the label corresponding to the first label category in the first label group and the second label group are different, then determine that the label corresponding to the first label category in the first label group is the first The tags corresponding to the first tag category in the three-tag group, that is, each tag category is compared one by one, and each tag category is updated one by one based on the comparison result of each tag category. The method of updating each label category one by one can update each new case and its label group more accurately into the user knowledge base, so as to improve the accuracy and user satisfaction of the next generated label.
  • the fused tag group is obtained by fusing tags corresponding to the same tag category in the first tag group and the second tag group based on a first preset rule, including:
  • the fusion label group is based on the first sub-attribute of the label corresponding to each label category in the first label group and the second sub-attribute of the label corresponding to each label category in the second label group.
  • a preset rule is obtained after fusion; wherein, the content attribute of the label corresponding to each label category in the first label group has the first sub-attribute, and the content attribute of the label corresponding to each label category in the second label group has the first sub-attribute Two sub-attributes; the first preset rule is: assign a first weight to the first sub-attribute, and assign a second weight to the second sub-attribute; calculate the first sub-attribute and the first a first product of weights and a second product of the second subattribute and the second weight; adding the first product to the second product.
  • the sub-attributes of the same content attribute may be the same or similar, or may be different. Based on different sub-attributes, each content attribute can be distinguished.
  • the sub-attributes can be numerical values, or subdivided sub-intervals of each content attribute.
  • Each content attribute may have multiple sub-attributes. The definition of sub-attributes is to illustrate how to get the fusion tag group.
  • the first sub-attribute of the high blood sugar alarm after 45 minutes in the above-mentioned first alarm interval is 45 minutes
  • the second sub-attribute of the high blood sugar alarm after 20 minutes is 20 minutes
  • the first weight is 0.5
  • the second weight is 0.5
  • the fusion sub-attribute calculated at this time is 32.5 minutes
  • the fusion label is 32.5 minutes after which a hyperglycemia alarm occurs.
  • the content attribute of the label corresponding to the first label category of the first label group and the second label group is the same, then determine the label corresponding to the first label category in the fusion label group (including label: hyperglycemia alarm occurs after 32.5 minutes ) is the tag corresponding to the first tag category in the third tag group.
  • the sum of the first weight and the second weight is 1; the settings of the first weight and the second weight can also be defined or customized in other ways.
  • the method also includes:
  • the output is visualized using at least one display module.
  • the display module may be configured to display a first set of tabs, a user data entry box, and the like.
  • At least one collection module is used to obtain user data.
  • the acquisition module can be configured as an input module for acquiring user data, especially user data of the first time period, such as one or more events, one or more user feature data.
  • the user is currently recording a diet of 300g of carbohydrates when the blood sugar is stable and the blood sugar value is 5mmol/L.
  • the system of this application searches from the user knowledge base, finds a case A with the highest similarity, and the maximum similarity is 92%, and it is confirmed that it is greater than the similarity threshold of 90%.
  • Case A represents the diet case that the user has ever had, and the rate of change of blood sugar, blood sugar level and the amount of diet are the most similar to this case.
  • the labels of case A are "hyperglycemia alarm after 30 minutes" and "rapid rise in blood sugar". At this time, the above two labels are output to the patient as the first label group through the user interface.
  • the abscissa represents the time
  • the ordinate represents the blood sugar level, where the blood sugar level is divided into high blood sugar interval, target blood sugar interval and low blood sugar interval by two parallel dotted lines in the figure, and t1 represents the diet
  • the event is logged, at which point 2 tabs appear.
  • Labels are based on glucose levels after similar cases in the past (historical cases represented by dotted curves). When the confirmation of this diet occurs, the glucose level change from t1 to t2 is monitored, that is, this case represented by the solid line curve in FIG. 3 , so as to obtain the second label group.
  • a new second label group is generated by utilizing the change of the glucose level 3 hours after the meal, such as "high blood sugar alarm occurs after 50 minutes” and "blood sugar rises slowly".
  • the first label with the second label in the alarm category, "hyperglycemia alarm after 30 minutes” and “hyperglycemia alarm after 50 minutes” belong to the same label category and have different sub-attributes; in the rate of change category, "blood sugar alarm "A rapid rise” and "a slow rise in blood sugar” are opposite labels. Therefore, the third label group stored in the user knowledge base after the fusion of this case and historical cases should be “a high blood sugar alarm occurred in about 40 minutes” and "slow blood sugar rise”.
  • the label generation device includes:
  • the first case acquiring module 10 is configured to acquire the first case, the first case includes the glucose level data in the first time period and the user data at the current moment.
  • the comparison module 20 is configured to use at least one model for data comparison to obtain a comparison result, and the comparison result includes similarity; the data of one side of the data comparison is the first case, and the data of the other side is the user Historical cases in the knowledge base; the user knowledge base contains a natural number of historical cases, and each historical case has a corresponding historical label group.
  • the first label group output module 30 is configured to select the maximum value in the similarity, if the maximum value is greater than the similarity threshold, then use the historical case corresponding to the maximum value as the second case, and use the second case The corresponding historical tag group is used as the first tag group and output.
  • the second tag group acquisition module 40 is configured to acquire a second tag group, which is determined based on the glucose level data in the second time period; wherein, all historical tag groups and the second tag group form the tag set It contains multiple tag categories, and each tag category in the tag set includes tags of at least two content attributes.
  • the update module 50 is configured to, for each label category, compare the content attributes of the labels corresponding to each label category in the first label group and the second label group, and obtain a comparison result, the comparison result including each Whether the content attributes of the tags corresponding to the tag category are the same; based on the comparison result, a third tag group is determined; the third tag group is selected from the second tag group or the fusion tag group, and the fusion tag The group is obtained after merging the tags corresponding to the same tag category in the first tag group and the second tag group based on the first preset rule; updating the first case and its corresponding third tag group to The user knowledge base.
  • any embodiment of the above label generation method is also applicable to the label generation device, which will not be repeated here.
  • the present application also provides a system for monitoring blood sugar levels, comprising:
  • a sensor configured to obtain glucose level data
  • a wireless transmitter configured to transmit said glucose level data
  • a mobile computing device comprising: a wireless receiver configured to receive said glucose level data; a memory configured to store said data of glucose level data; a processor configured to process the data stored in said memory, and a software application comprising instructions stored in said memory which, when executed, perform various embodiments described above label generation method.
  • FIG. 5 is a schematic diagram of an electronic device provided by an embodiment of the present application.
  • the electronic device may include: a processor (processor) 510, a communication interface (Communications Interface) 520, a memory (memory) 530, and a communication bus 540, wherein the processing The device 510, the communication interface 520, and the memory 530 communicate with each other through the communication bus 540.
  • the processor 510 may invoke logic instructions in the memory 530 to execute the tag generation method provided in the above-mentioned embodiments.
  • the above logic instructions in the memory 530 may be implemented in the form of software function units and be stored in a computer-readable storage medium when sold or used as an independent product.
  • the technical solution of the present application can be embodied in the form of a software product in essence.
  • the computer software product is stored in a storage medium and includes a plurality of instructions to make a computer device (which can be a personal computer, a server, or a network device, etc.) ) Execute all or part of the steps of the method described in the embodiments of the present application.
  • the aforementioned storage media include: U disk, mobile hard disk, read-only memory (Read-Only Memory, ROM), random access memory (Random Access Memory, RAM), magnetic disk or optical disc and other media that can store program codes. .
  • the present application also provides a computer program product
  • the computer program product includes a computer program stored on a non-transitory computer-readable storage medium
  • the computer program includes program instructions, and when the program instructions are executed by a computer During execution, the computer can execute the label generation method provided by the above embodiments.
  • the present application also provides a non-transitory computer-readable storage medium, on which a computer program is stored, and when the computer program is executed by a processor, it is implemented to execute the tag generation method provided by the above-mentioned embodiments.
  • the device embodiments described above are only illustrative, and the units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in One place, or it can be distributed to multiple network elements. Part or all of the modules can be selected according to actual needs to achieve the purpose of the solution of this embodiment.
  • the embodiments can be implemented by means of software plus a necessary general-purpose hardware platform, and of course also by hardware.
  • the above-mentioned technical solutions can be embodied in the form of software products in essence, and the computer software products can be stored in computer-readable storage media, such as ROM/RAM, magnetic disks, optical disks, etc., and include multiple instructions to make a computer device (It may be a personal computer, a server, or a network device, etc.) executes the methods described in the embodiments or some parts of the embodiments.

Landscapes

  • Engineering & Computer Science (AREA)
  • Data Mining & Analysis (AREA)
  • Theoretical Computer Science (AREA)
  • General Engineering & Computer Science (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Bioinformatics & Computational Biology (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Evolutionary Biology (AREA)
  • Evolutionary Computation (AREA)
  • Physics & Mathematics (AREA)
  • Artificial Intelligence (AREA)
  • General Physics & Mathematics (AREA)
  • Bioinformatics & Cheminformatics (AREA)
  • Health & Medical Sciences (AREA)
  • Public Health (AREA)
  • Medical Informatics (AREA)
  • Databases & Information Systems (AREA)
  • Pathology (AREA)
  • Biomedical Technology (AREA)
  • Epidemiology (AREA)
  • General Health & Medical Sciences (AREA)
  • Primary Health Care (AREA)
  • Measurement Of The Respiration, Hearing Ability, Form, And Blood Characteristics Of Living Organisms (AREA)

Abstract

本文公开一种标签生成方法和装置、监控血糖水平的系统。所述标签生成方法包括：获取第一案例；使用至少一个模型进行数据比对，得到比对结果，比对结果包括相似度；选取相似度中的最大值，若最大值大于相似阈值，则将最大值对应的历史案例作为第二案例，将第二案例对应的历史标签组作为第一标签组并输出；获取第二标签组，第二标签组是基于第二时间段的葡萄糖水平数据确定的；针对每一个标签类别，比较第一标签组和第二标签组中的每一个标签类别对应的标签的内容属性，得到比较结果；基于比较结果，确定出第三标签组；将第一案例及第一案例对应的第三标签组更新至用户知识库。

Description

标签生成方法和装置、监控血糖水平的系统
本申请要求在2022年01月06日提交中国专利局、申请号为202210007243.2的中国专利申请的优先权，该申请的全部内容通过引用结合在本申请中。
技术领域
本申请涉及生理信号处理领域，例如涉及一种标签生成方法和装置、监控血糖水平的系统。
背景技术
随着社会的发展和技术的进步，一些不健康的生活习惯，导致越来越多的人出现多种健康问题，特别是患有慢性病的人群逐渐增多。人体出现健康问题会有多方面的反应，其中葡萄糖浓度数据的异常是较为常见且可以进行测量的健康指标。
与传统需要频繁测量指尖血的血糖测量设备相比，基于生物传感器的技术发展的持续葡萄糖监测系统(按一定频率连续提供实时葡萄糖浓度数据)给人们提供了更好地了解其葡萄糖浓度数据变化水平的机会，该系统提供了监测葡萄糖浓度的数据基础，这对于个人的健康管理而言是非常有用的。持续葡萄糖监测系统可提供用户实时的葡萄糖浓度数据，且这些葡萄糖浓度数据是随着每个使用该系统的用户的数据变化而变化的。比如，在有饮食摄入后一定时间内会导致葡萄糖浓度的持续上升，或在饥饿的情况下会导致葡萄糖浓度的持续下降，或糖尿病患者在注射胰岛素后一定时间内将原本上升或具有上升趋势的葡萄糖浓度改变为下降或具有下降趋势。这些葡萄糖浓度的波动对于健康人群来讲可能无需担心，但是对于糖尿病患者而言，有时可能是致命的。基于此，相关技术中公开了一种基于智能预测的葡萄糖报警设备，其可以基于葡萄糖预测来生成与葡萄糖相关的警报。葡萄糖预测是基于多个输入数据源来计算的，该输入数据源在一定程度上可以指示用户的当前状态和未来状态。上述方案中，预测未来状态是通过基于多种输入在大数据基础上计算对未来葡萄糖的预测来实现的。上述方案是基于目前用户数据信息(如多种输入、测量数据或碳水化合物估计值等)对用户未来状态进行预测。
上述方案未考虑到用户的个性化、以及一些实际的用户数据所带来的真实效果，提供的也仅仅是比较单一化的葡萄糖水平预测，预测精度远远达不到给予用户可靠性建议的目的，而导致用户体验感差。
发明内容
本申请提供一种标签生成方法和装置、监控血糖水平的系统。
本申请提供一种标签生成方法，包括：
获取第一案例，其中，所述第一案例包括第一时间段的葡萄糖水平数据和当前时刻的用户数据；
使用至少一个模型进行数据比对，得到比对结果，其中，所述比对结果包括相似度；所述数据比对的一方数据为所述第一案例、另一方数据为用户知识库中的历史案例；所述用户知识库中包含自然数个历史案例，每个历史案例具有对应的历史标签组；
选取所述相似度中的最大值，若所述最大值大于相似阈值，则将所述最大值对应的历史案例作为第二案例，将所述第二案例对应的历史标签组作为第一标签组并输出；
获取第二标签组，其中，所述第二标签组是基于第二时间段的葡萄糖水平数据确定的；所有历史标签组和所述第二标签组构成的标签集中含有多个标签类别，所述标签集中每一个标签类别中包括至少两种内容属性的标签；
针对每一个标签类别，比较所述第一标签组和所述第二标签组中的所述每一个标签类别对应的标签的内容属性，得到比较结果，其中，所述比较结果包括每一个标签类别对应的标签的内容属性是否相同；
基于所述比较结果，确定出第三标签组；其中，所述第三标签组是在所述第二标签组或融合标签组中选取的，所述融合标签组是将所述第一标签组与所述第二标签组中的相同标签类别对应的标签基于第一预设规则进行融合后得到的；
将所述第一案例及所述第一案例对应的第三标签组更新至所述用户知识库。
本申请还提供了一种标签生成装置，包括：
第一案例获取模块，设置为获取第一案例，其中，所述第一案例包括第一时间段的葡萄糖水平数据和当前时刻的用户数据；
比对模块，设置为使用至少一个模型进行数据比对，得到比对结果，其中，所述比对结果包括相似度；所述数据比对的一方数据为所述第一案例、另一方数据为用户知识库中的历史案例；所述用户知识库中包含自然数个历史案例，每个历史案例具有对应的历史标签组；
第一标签组输出模块，设置为选取所述相似度中的最大值，若所述最大值 大于相似阈值，则将所述最大值对应的历史案例作为第二案例，将所述第二案例对应的历史标签组作为第一标签组并输出；
第二标签组获取模块，设置为获取第二标签组，其中，所述第二标签组是基于第二时间段的葡萄糖水平数据确定的；所有历史标签组和所述第二标签组构成的标签集中含有多个标签类别，所述标签集中每一个标签类别中包括至少两种内容属性的标签；
更新模块，设置为针对每一个标签类别，比较所述第一标签组和所述第二标签组中的所述每一个标签类别对应的标签的内容属性，得到比较结果，其中，所述比较结果包括每一个标签类别对应的标签的内容属性是否相同；基于所述比较结果，确定出第三标签组；其中，所述第三标签组是在所述第二标签组或融合标签组中选取的，所述融合标签组是将所述第一标签组与所述第二标签组中的相同标签类别对应的标签基于第一预设规则进行融合后得到的；将所述第一案例及所述第一案例对应的第三标签组更新至所述用户知识库。
本申请还提供了一种监控血糖水平的系统，包括：
传感器，设置为获取葡萄糖水平数据；
无线发射器，设置为发射所述葡萄糖水平数据；
以及
移动计算装置，其包括：
无线接收器，设置为接收所述葡萄糖水平数据；
存储器，设置为存储包含所述葡萄糖水平数据的数据；
处理器，设置为处理所述数据，以及软件应用程序，其包含存储于所述存储器中的指令，所述指令执行所述时实现如上述所述的标签生成方法。
本申请还提供一种电子设备，包括存储器、处理器及存储在存储器上并可在处理器上运行的计算机程序，所述处理器执行所述程序时实现如上述所述的标签生成方法。
本申请还提供一种非暂态计算机可读存储介质，其上存储有计算机程序，该计算机程序被处理器执行时实现如上述所述的标签生成方法。
附图说明
图1是本申请实施例提供的一种实施环境的结构示意图；
图2是本申请实施例提供的一种标签生成方法的流程示意图；
图3是本申请实施例提供的一种标签生成方法的使用效果示意图；
图4是本申请实施例提供的一种标签生成装置的结构示意图；
图5是本申请实施例提供的一种电子设备的结构示意图。
具体实施方式
下面将结合本申请中的附图，对本申请中的技术方案进行描述，所描述的实施例是本申请一部分实施例。
请参考图1，图1是本申请实施例提供的一种实施环境的结构示意图。该实施环境包括：终端100和/或服务器200。
终端100可以是具备数据处理能力的电子设备，如手机、平板电脑、电子书阅读器、动态影像专家压缩标准音频层面3(Moving Picture Experts Group Audio Layer III，MP3)播放器、动态影像专家压缩标准音频层面4(Moving Picture Experts Group Audio Layer IV，MP4)播放器、膝上型便携计算机和台式计算机等等。
终端100中可以安装有应用程序客户端，或者安装有浏览器，通过浏览器访问应用程序的网页客户端。本申请实施例将应用程序客户端和网页客户端统称为客户端，下文不再特别声明。
服务器200可以是一台近端或远端服务器，或者由多个台服务器组成的服务器集群，或者是一个云计算服务中心。当终端100和服务器200同时处理本申请相关业务时，服务器200可设置为与终端100交互提供本申请相关业务。服务器200是与客户端对应的服务器，服务器200与终端100可以结合实现客户端提供的多项功能，通常由互联网服务商来设立。
终端100与服务器200之间可以通过无线网络或者有线网络相连。
下面结合图2描述本申请的一种标签生成方法，该方法包括：
S1、获取第一案例，所述第一案例包括第一时间段的葡萄糖水平数据和当前时刻的用户数据。
第一时间段的葡萄糖水平数据包括但不限于带有时间戳的葡萄糖浓度(即血糖浓度、血糖测量值或血糖值)数据、与葡萄糖浓度相关联的数据，第一时间段的葡萄糖水平数据的获取方式包括但不限于通过葡萄糖浓度传感器获取、由网络传递采集或其他采集方式。例如可以获取由葡萄糖浓度传感器按一定周期连续输出的带有时间戳的第一时间段的用户葡萄糖浓度数据。所述第一时间段从第一时刻延续至当前时刻，所述第一时刻在所述当前时刻之前，例如，第 一时间段可以是当前时刻往前的10分钟至2小时，第一时刻延续至当前时刻(包含当前时刻)的用户葡萄糖浓度数据可以体现为以过去第一时间段的时间为横轴、以采集数据为纵轴的一段数据波形。
用户数据为当前时刻的数据，一般是由用户输入或其他方式输入的与葡萄糖浓度相关联的数据，例如，碳水摄入量、运动量等生活事件相关数据。碳水数据可以由自动识别食物图片中的热量的软件获得。
S2、使用至少一个模型进行数据比对，得到比对结果，所述比对结果包括相似度；所述数据比对的一方数据为所述第一案例、另一方数据为用户知识库中的历史案例；所述用户知识库中包含自然数个历史案例，每个历史案例具有对应的历史标签组。
至少一个模型可以是计算相似度的公式，用于比对第一案例与用户知识库中的历史案例的相似度。用户知识库中包含自然数个历史案例，也就是用户知识库中可以没有历史案例。当没有历史案例时，比对结果中不包含相似度，此时比对结果为比对失败，在比对失败的情况下，可以暂停S3，不输出第一标签组，直接跳转至S4。每个历史标签组中可以包含多个标签，多个标签可以是不同类别的标签，也可以是同类别标签，类别可以是按照变化率类、目标血糖时间占比类等进行分类。每个标签组中至少包含一个标签。例如，每个历史案例对应的历史标签组中的每个类别具有一个标签。
S3、选取所述相似度中的最大值，若所述最大值大于相似阈值，则将所述最大值对应的历史案例作为第二案例，将所述第二案例对应的历史标签组作为第一标签组并输出。
相似度越高代表该相似度对应的历史案例的历史标签组越符合用户当前血糖状况与当前用户数据相结合的情况。对相似度进行排序，选取排序结果中的最大值，再将最大值与相似阈值进行比较，若最大值不大于相似阈值，则说明未比对到比较相似的历史案例，该情况下不输出任何标签，并继续执行S4至S5；若最大值大于相似阈值，则将所述最大值对应的历史案例作为第二案例(也称旧案例)，并在显示界面上输出第一标签组。
S4、获取第二标签组，所述第二标签组是基于第二时间段的葡萄糖水平数据确定的；其中，所有历史标签组和第二标签组构成的标签集中含有多个标签类别，所述标签集中每一个标签类别中包括至少两种内容属性的标签。
标签集涵盖所有的历史标签组和第二标签组，其可以是初始的一个较小的标签集合，随着本申请方法执行次数的增多而逐渐扩大并优化、个性化的一个标签集合。之所以提出标签集的概念，是为了说明，在标签集中每一个标签类 别中包括至少两种内容属性的标签，也就是说，第一标签组和第二标签组中每一个标签类别对应的标签可能内容属性相同，也可能不同，内容属性相同表示输出的第一标签组预测的较为精准，与第二标签组相符度高，内容属性不同表示输出的第一标签组预测的不够精准，与第二标签组相符度低、甚至相反。第二标签组的获取是为了S5中基于第一案例对用户知识库进行更新，所述第二时间段从当前时刻延续至第二时刻，所述第二时刻在所述当前时刻之后。第二标签组是基于第二时间段(当前时刻延续至第二时刻，可以包含当前时刻，也可以不含当前时刻)的葡萄糖水平数据确定的，因此，第二标签组是非常接近并能够反映用户在第二时间段的真实葡萄糖水平的标签组。
S5、针对每一个标签类别，比较所述第一标签组和第二标签组中的所述每一个标签类别对应的标签的内容属性，得到比较结果，所述比较结果包括每一个标签类别对应的标签的内容属性是否相同；基于所述比较结果，确定出第三标签组；所述第三标签组是在所述第二标签组或融合标签组中选取的，所述融合标签组是将所述第一标签组与第二标签组中的相同标签类别对应的标签基于第一预设规则进行融合后得到的；将所述第一案例及其对应的第三标签组更新至所述用户知识库。
除了第二时间段的葡萄糖水平数据出错的情况以外，每一次执行标签生成方法时，需要基于第二标签组将第一案例更新至用户知识库中。更新的第一案例所对应的第三标签组需要基于比较结果来确定，确定方式可以采用下面实施例中的多种方式，第三标签组来源于第二标签组、第一标签组与第二标签组相融合后的融合标签组。融合标签组是在特定情况下，充分利结合第一标签组和第二标签组对第三标签组进行优化，该特定情况下，融合标签组的效果优于独立的第一标签组、或第二标签组。随着用户使用本申请方法的次数的增多，用户知识库中的标签生成方案也越来越精准、个性化，能够实现更佳的标签。每一次执行本申请的方法，用户知识库就更新一次，如果用户知识库中的数据超过一定的范围，可以定期删除掉过期的数据，保留一段时间内的新数据。
本申请的方法通过获取第一案例，并将第一案例与用户知识库中的历史案例进行比对，得到相似度，选取相似度最大值大于相似阈值的历史案例作为第二案例，将第二案例对应的历史标签组作为第一标签组输出；并基于用户在第二时间段的真实葡萄糖水平，确定出第二标签组，第一标签组和第二标签组共同用于更新用户知识库，第二标签组非常接近并能够反映用户真实葡萄糖水平；针对每一个标签类别，将第一标签组、以及第二标签组中的相同标签类别对应的标签的内容属性进行比较，将实际效果更好的标签组更新到用户知识库中；随着用户使用本申请方法的次数的增加，用户知识库中的历史案例中可供选择的精准标签组就越多，输出也会更加精准可靠，本申请充分考虑了用户个性化 特征数据以及一段时间内的葡萄糖水平数据，使得用户享受到更丰富、更合理的标签，标签输出精准、效果更好，提升了用户体验感，输出的标签组可以给予用户控制葡萄糖水平以更好、更可靠的建议。
所述的标签生成方法中，所述方法还包括：通过网络从关联所述用户的血糖测量设备处获取用户的葡萄糖水平数据。
网络为有线网络或无线网络，关联所述用户的血糖测量设备包括但不限于含血糖传感器的血糖仪、采血测量设备、其他可以采集葡萄糖浓度数据的生理数据采集设备。例如，所述血糖测量设备是能够连续实时采集葡萄糖浓度数据的连续血糖监测设备。
所述葡萄糖水平数据包括第一时间段的葡萄糖水平数据及第二时间段的葡萄糖水平数据；所述第一时间段的葡萄糖水平数据包括第一葡萄糖浓度数据和第一葡萄糖浓度变化率，所述第二时间段的葡萄糖水平数据包括第二葡萄糖浓度数据和第二葡萄糖浓度变化率。
第一葡萄糖浓度数据包括：当前时刻的第一血糖测量值及其对应的第一时间戳、第一时刻至当前时刻之间的历史血糖采集数据，所述历史血糖采集数据包括按预设时间间隔连续分布的多个历史血糖测量值及其对应的多个历史时间戳。预设时间间隔为连续血糖监测设备产生血糖的间隔，如3分钟。历史血糖采集数据包括距离当前时刻第一时间段的、以3分钟为周期连续分布的多个历史血糖测量值及其对应的多个历史时间戳。
所述第一葡萄糖浓度变化率是基于所述第一葡萄糖浓度数据以正负值变化率的方式得到的；所述第二葡萄糖浓度变化率是基于所述第二葡萄糖浓度数据以正负值变化率的方式得到的。
可以基于所述第一血糖测量值及其对应的第一时间戳、以及在所述多个历史血糖测量值及其对应的多个历史时间戳中选取的第二值，确定所述葡萄糖浓度变化率；所述第二值包括第二血糖测量值及其对应的第二时间戳，所述第二时间戳与所述第一时间戳相关联。
所述第一葡萄糖浓度变化率是基于所述第一葡萄糖浓度数据以正负值变化率的方式得到的，例如，第一葡萄糖浓度变化率的计算方式为：(第一血糖测量值-第二血糖测量值)/(第一时间戳-第二时间戳)，第二值的选择可在距离当前时刻的第三时间段内进行选择，第三时间段可以是1分钟至30分钟，例如，可选择当前时刻之前的3分钟时的数据，若有数据缺失或异常，可选择其他距离当前时刻的第三时间段内的数据。
所述第二时间段的葡萄糖水平数据包括第二葡萄糖浓度数据和第二葡萄糖 浓度变化率；同理，所述第二葡萄糖浓度变化率是基于所述第二葡萄糖浓度数据以正负值变化率的方式得到的，正负值变化率能够体现葡萄糖浓度的升、降或平稳水平，正负值变化率可以先通过函数转化为0-100之间的数值再使用。第二时间段的第二葡萄糖浓度数据与第一时间段的第一葡萄糖浓度数据可以在当前时刻有重合，第二时间段的第二葡萄糖浓度数据也可以不包含当前时刻的数据。
本申请充分考虑了第一时间段的葡萄糖水平数据，进行标签组生成和输出；并基于第二时间段的葡萄糖水平数据和已生成的第一标签组、更新用户知识库，第二时间段的葡萄糖水平数据主要用于结合第一标签组、构成用于更新的第三标签组，使得用户知识库多次更新后，生成的方案更适合用户。
所述的标签生成方法中，所述获取第一案例，包括：
获取当前时刻的用户数据；所述用户数据包括与葡萄糖浓度相关联的一个或多个事件、一个或多个用户特征数据。
用户数据包括用户手动输入、利用手机应用程序(Application，APP)获取的图片识别出的、或其他设备获取的与葡萄糖浓度相关联的数据，例如碳水摄入及其摄入量、运动量等。当前时刻的用户数据也带有时间戳。在生成标签组时，可以充分考虑到与葡萄糖浓度相关联的当前时刻的用户数据，确保了标签组的准确性。
所述的标签生成方法中，所述一个或多个事件与碳水消耗、锻炼、睡眠以及物质的施予中的一个或多者相关联；所述物质的施予包括用药类型、用药剂量、碳水施予量；所述用药类型包括长效胰岛素、短效胰岛素、速效胰岛素中的至少一种。
考虑到不同的事件，例如用药类型、用药剂量、碳水施予量均对葡萄糖浓度有较大影响，例如，速效胰岛素主要影响时间段为30分钟内，长效胰岛素主要影响时间段为2小时内，因此需要考虑当前时刻的事件。
所述的标签生成方法中，所述一个或多个用户特征数据与用户的基本生理信息、以及个人信息中的至少一个相关联；所述基本生理信息包括胰岛素敏感系数、胰岛素-碳水比中的至少一种；所述个人信息包括性别、所处地域、糖尿病类型、年龄、体重、胰岛素历史使用年限中的至少一种。
所处地域、糖尿病类型、年龄等都是基于用户自身进行个性化设定的，考虑以上因素，可以得到更加个性化的用户知识库，生成的标签组更符合用户的真实情况。例如，糖尿病类型包括I型糖尿病、Ⅱ型糖尿病、妊娠糖尿病，因为每个糖尿病类型的用户可能存在相似的葡萄糖浓度规律、以及相似的食物、药 物敏感性等，每个糖尿病类型的用户对于葡萄糖浓度的变化具有一定的规律性，因此在生成标签组时考虑到糖尿病类型，可以得到具有按照用户所属糖尿病类型进行区分的个性化用户知识库，输出的结果更适用于患有该糖尿病类型的用户。
所述的标签生成方法中，所述使用至少一个模型进行数据比对，得到比对结果，所述比对结果包括相似度，包括：
对所述第一案例进行规则化，得到至少两个比对特征值。
规则化包括对于非数值类的关键特征数据使用预设的归一化规则转化为数字，如I型糖尿病取值为“0”，Ⅱ型糖尿病取值为“1”，如男性糖尿病患者取值为“0”，女性糖尿病患者取值为“1”，利用诸如此类的预设的归一化规则对部分不具有量化值的第一案例中的数据进行规则化处理。葡萄糖浓度数据(即为葡萄糖浓度值)、血糖变化率、碳水摄入量、运动量这些已经具有量化值的数据则在该步骤中不必进行规则化处理。至少两个比对特征值包括根据第一时间段的葡萄糖水平数据和当前时刻的用户数据中的至少两个数据得到的特征值。
基于至少两个比对特征值与所述历史案例包含的历史特征值组中与所述至少两个比对特征值分别对应的至少两个历史特征值，得到至少两个绝对距离。至少两个比对特征值可以形成用户的一个新案例，也就是第一案例，基于新案例在用户知识库中进行案例的匹配。每个历史案例包含的历史特征值组中包含多个历史特征值，匹配过程中，将新案例中的每一个特征值与历史特征值组中相对应的历史特征值进行对比，相对应是指每一个类别下的数据，例如葡萄糖浓度数据为同一类，碳水摄入数据为同一类，对每一个类别下的特征值进行一一比对；可以得到至少两个绝对距离。
基于所述至少两个绝对距离和每个比对特征值对应的特征权值，得到相似度，所述特征权值是基于所述特征权值对应的比对特征值与葡萄糖水平的相关性大小确定的。
每种比对特征值采用以下公式计算绝对距离：
d'＝|F new-F old|
其中，F表示特征值，F new表示新案例的一类数据的一个特征值，F old表示历史案例对应于F new的一个特征值；d'表示归一化之前的两个特征值的绝对距离(差异度)；然后，利用预设的规则对不在0-1范围内的绝对距离进行归一化，转化为0-1的数值，如葡萄糖浓度值的绝对距离归一化处理可表示为：
Figure PCTCN2022085452-appb-000001
d Glu表示归一化之后的血糖浓度值的绝对距离，d' Glu表示归一化之前的血糖浓度值的绝对距离，F NewGlu表示新案例的血糖浓度值。
其他值的绝对距离也可采用同葡萄糖浓度值的绝对距离归一化处理方式逐一进行归一化处理；新案例与历史案例的差异得分为：
D＝d 1*a 1+d 2*a 2+…+d n*a n
其中，d 1、d 2、…、d n表示第1个特征值至第n个特征值的绝对距离，a 1、a 2、…、a n分别表示第1个特征值至第n个特征值的特征权值(因子)，比对特征值对葡萄糖水平的影响越大，对应的特征权值的值越大。且满足：
Figure PCTCN2022085452-appb-000002
新案例与历史案例的相似度为：S＝1-D。S越大表示两个案例的相似度越大。
当所述相似度大于相似阈值，判断为比对成功；当所述相似度不大于相似阈值，判断为比对失败。相似阈值是预先设置的，可以是用户自定义或用户输入。相似阈值设置的越高，则比对成功的概率越小，结果越精准，为了兼顾比对成功概率和比对精准度，综合效果来设置相似阈值。例如，相似阈值可以设置为0.9，也就是当相似度大于90％时，判断为比对成功。
所述的标签生成方法中，所述使用至少一个模型进行数据比对，得到比对结果，所述比对结果包括相似度；所述数据比对的一方数据为所述第一案例、另一方数据为用户知识库中的历史案例，包括：
基于至少一个比对特征值，在所述用户知识库中进行预先筛选，得到预选库；所述预选库中的历史案例包含的历史特征值组中存在与所述至少一个比对特征值相同的部分；使用至少一个模型进行数据比对，得到比对结果，所述比对结果包括相似度；所述数据比对的一方数据为所述第一案例、另一方数据为所述预选库中的历史案例。
比对时，为了加快响应速度和响应精准度，也可以基于部分比对特征值进行预先筛选。基于至少一个比对特征值，在所述用户知识库中进行预先筛选，得到预选库；所述预选库中的历史案例包含的历史特征值组中存在与所述至少一个比对特征值相同的部分；使用至少一个模型进行数据比对，得到比对结果，所述比对结果包括相似度；所述数据比对的一方数据为所述第一案例、另一方数据为所述预选库中的历史案例。至少一个比对特征值可以是一个或两个及以 上均可。例如用药类型，筛选出与用户用药类型一致的历史案例，再进一步将新案例与这些用药类型一致的历史案例进行比对；即，先排除用药类型不一致的历史案例。例如，用药年限，可以筛选出与用户用药年限一致的历史案例，再将新案例与这些用药年限一致的历史案例进行比对；即，先排除用药年限不一致的历史案例。例如，也可以结合多个比对特征值进行预先筛选，可以仅筛选出与用户用药年限、用药类型均一致的历史案例，再进一步将新案例与这些用药年限、用药类型均一致的历史案例进行比对；即，先排除用药年限或用药类型不一致的历史案例。也可以不进行预先筛选，直接基于比对特征值逐一比较，得到相似度后再将相似度和相似阈值进行比较。
所述的标签生成方法中，在所述选取相似度中的最大值，若所述最大值大于相似阈值，则将所述最大值对应的历史案例作为第二案例，将所述第二案例对应的历史标签组作为第一标签组并输出之后，还包括：
获取所述用户对所述第一案例的执行情况；当所述执行情况为所述用户未执行所述第一案例时，则暂停执行所述S4至S5；当所述执行情况为所述用户执行所述第一案例时，则继续执行所述S4至S5。
用户对所述第一案例的执行情况可以基于用户在第一标签组输出之后的操作，或是第二时间段的葡萄糖水平数据进行评估。当用户不执行第一案例时，第二标签组是无法和第一案例相互关联的，因此无需获取第二标签组，也无需将第一案例更新至用户知识库，此时暂停执行所述S4至S5。当用户执行第一案例时，可以将第一案例更新至用户知识库，此时继续执行所述S4至S5。确保了进入用户知识库的案例都是实际发生过的，该案例对应的标签组也是能够反映在案例发生后的真实情况的。
所述的标签生成方法中，在所述使用至少一个模型进行数据比对，得到比对结果之后，还包括：
若所述最大值不大于相似阈值或所述比对结果为用户知识库为空，则暂停输出所述第一标签组，并继续执行所述S4至S5。
若所述最大值不大于相似阈值或所述比对结果为用户知识库为空，说明比对失败，也就是用户知识库中不存在与当前时刻的第一案例较相似的案例，此时暂停输出第一标签组，并利用第二标签组继续执行所述S4至S5、更新用户知识库。避免输出相似度不够的标签组，确保输出的标签组都是能够精准反映用户真实情况的标签组。
所述的标签生成方法中，所述获取第二标签组，所述第二标签组是基于第二时间段的葡萄糖水平数据确定的，包括：
基于第二预设规则，获取第二标签组；所述第二预设规则包括：对第二时间段的葡萄糖水平数据基于每种内容属性进行划分，得到第一属性组；基于所述第一属性组和预先确定的关联关系，得到所述第二标签组，所述关联关系为多个内容属性组与多个预存标签组之间的关联关系，所述第二标签组是从所述多个预存标签组中选取的标签组。其中，多个内容属性组与多个预存标签组之间的关联关系如下表1所示。
表1 预先确定的关联关系
Figure PCTCN2022085452-appb-000003
如表1所示，预存标签组包括标签组1、标签组2、…、标签组M。基于所述第一属性组和预先确定的关联关系，得到所述第二标签组，例如，若第一属性组为第一报警区间、第一变化率区间、第一目标血糖时间占比区间时，选择第一属性组对应的标签组1作为第二标签组，若第一属性组为第二报警区间、第二变化率区间、第二目标血糖时间占比区间时，选择第一属性组对应的标签组2作为第二标签组。其中，目标血糖时间占比区间是基于期望血糖水平和第二时间段的葡萄糖水平数据确定的，期望血糖水平可以是系统定义或用户自定 义的。目标血糖时间占比可以先通过函数转化为0-100之间的数值划分为第一目标血糖时间占比区间、第二目标血糖时间占比区间至第M目标血糖时间占比区间。例如，第一目标血糖时间占比区间为0-20，第二目标血糖时间占比区间为20-40，第M目标血糖时间占比区间为80-100等等。第一变化率区间为0-20、第二变化率区间为20-40，第M变化率区间为80-100等等。第一报警区间为低血糖报警、第二报警区间为不报警、第M报警区间为高血糖报警等等。
所述的标签生成方法中，所述所有历史标签组和第二标签组构成的标签集中含有多个标签类别，包括：
所述所有历史标签组和第二标签组构成的所述标签集中至少含有多个标签类别，所述多个标签类别包括报警类、变化率类、目标血糖时间占比类中的至少两种。
例如，报警类对应的标签可以包括45分钟后出现高血糖报警、20分钟后出现高血糖报警、即将出现超低血糖报警等标签；变化率类对应的标签可以包括血糖快速上升、血糖变化平稳、血糖平稳下降等标签；目标血糖时间占比类对应的标签可以包括目标血糖时间占比高于稳定值、目标血糖时间占比低于稳定值等标签。
所述的标签生成方法中，所述标签集中每一个标签类别中包括至少两种内容属性的标签，包括以下情况中的至少一种：
所述报警类对应的标签中至少存在内容属性为第一报警区间的标签、以及内容属性为第二报警区间的标签；所述变化率类对应的标签中至少存在内容属性为第一变化率区间的标签、以及内容属性为第二变化率区间的标签；所述目标血糖时间占比类对应的标签中至少存在内容属性为第一目标血糖时间占比区间的标签、以及内容属性为第二目标血糖时间占比区间的标签。
基于至少两种内容属性的不同，对每一个标签类别还可以进行区分，例如，上述45分钟后出现高血糖报警、20分钟后出现高血糖报警为第M报警区间，即将出现超低血糖报警为第一报警区间；血糖快速上升为第M变化率区间；血糖平稳下降为第一变化率区间；目标血糖时间占比高于占比阈值为第M目标血糖时间占比区间；目标血糖时间占比低于占比阈值为第一目标血糖时间占比区间。
所述的标签生成方法中，在所述将所述第二案例对应的历史标签组作为第一标签组并输出之前，还包括：
判断所述第一案例中是否存在异常数据，若所述第一案例中存在异常数据，则暂停输出所述第一标签组、并暂停执行所述S4至S5。
第一案例中存在异常数据是指葡萄糖水平数据中存在异常或用户数据中存在异常，葡萄糖水平数据和用户数据异常一般是根据大数据、以往的用户的数据历史和经验进行判断。葡萄糖水平数据异常可能是由于传感器异常、网络异常导致的数据超出一定范围、或者数据缺失等原因造成的。用户数据异常可能是由于用户输入的值超出一定的正常适用范围。在将所述第二案例对应的历史标签组作为第一标签组并输出之前，若第一案例中存在异常数据，则该第一标签不适用于输出，此时需要暂停输出所述第一标签组，避免将有误差的标签输出给用户而导致的安全隐患问题。且暂停将存在错误数据的第一案例更新至用户知识库，避免将有错误的第一案例应用于下一次的标签生成方法中。
所述的标签生成方法中，在所述获取第二标签组之前，还包括：
判断所述第二时间段的葡萄糖水平数据中是否存在异常数据，若所述第二时间段的葡萄糖水平数据中存在异常数据，则暂停执行所述S4至S5。
第二时间段的葡萄糖水平数据异常的判断和原因与第一案例中异常数据的判断和原因均相同，此处不再一一赘述。若第二时间段的葡萄糖水平数据中存在异常数据，暂停将存在错误数据的第二标签组更新至用户知识库，避免将有错误的第二标签组应用于下一次的标签生成方法中。
所述的标签生成方法中，所述基于所述比较结果，确定出第三标签组，包括：
当所述比较结果为第一比较结果时，确定所述融合标签组为第三标签组；所述第一比较结果为：相同标签类别对应的的内容属性相同的标签类别数目占所有标签类别数目的比例大于比例阈值；当所述比较结果为第二比较结果时，确定所述第二标签组为第三标签组；所述第二比较结果为：相同标签类别对应的标签的内容属性相同的标签类别数目占所有标签类别数目的比例不大于比例阈值。
上述比例阈值可以是80％-90％，当相同标签类别对应的标签的内容属性相同的标签类别数目占所有标签类别数目的比例大于比例阈值时，此时第一标签组和第二标签组中的相同内容属性较多，可以确定所述融合标签组为第三标签组；当相同标签类别对应的标签的内容属性相同的标签类别数目占所有标签类别数目的比例不大于比例阈值时，此时第一标签组和第二标签组中的相同内容属性较少，可以确定所述第二标签组为第三标签组，采用更接近真实水平的第二标签组来更新。上述按组更新的方式，可以节省系统资源，提高更新效率和响应速度。
所述的标签生成方法中，所述基于所述比较结果，确定出第三标签组，包 括：
当所述比较结果包括第一标签类别对应第三比较结果时，将所述融合标签组中的第一标签类别对应的标签作为第三标签组中第一标签类别对应的标签；所述多个标签类别至少包括第一标签类别，所述第三比较结果为所述第一标签类别对应的标签的内容属性相同；当所述比较结果包括第一标签类别对应第四比较结果时，将所述第一标签组中的第一标签类别对应的标签作为第三标签组中第一标签类别对应的标签；所述第四比较结果为所述第一标签类别对应的标签的内容属性不同。
第一标签组、以及第二标签组均是基于标签类别进行分类，在比较时，是按照每一个标签类别进行逐一比较确定第三标签组中的每个标签类别的标签的。若第一标签组以及第二标签组中的第一标签类别对应的标签的内容属性相同，则确定所述融合标签组中的第一标签类别对应的标签为第三标签组中的第一标签类别对应的标签；若第一标签组以及第二标签组中的所述第一标签类别对应的标签的内容属性不同，则确定所述第一标签组中的第一标签类别对应的标签为第三标签组中的第一标签类别对应的标签，也就是说，逐一对每一个标签类别比较、且基于每一个标签类别的比较结果逐一对每一个标签类别更新。逐一对每一个标签类别的更新方式能够将每一次的新案例及其标签组更加精准的更新到用户知识库中，以提高下一次生成标签的精度和用户满意度。
所述的标签生成方法中，所述融合标签组是将所述第一标签组与所述第二标签组中的相同标签类别对应的标签基于第一预设规则进行融合后得到的，包括：
所述融合标签组是将所述第一标签组中每个标签类别对应的标签的第一子属性与所述第二标签组中所述每个标签类别对应的标签的第二子属性基于第一预设规则进行融合后得到的；其中，第一标签组中每个标签类别对应的标签的内容属性具有第一子属性，第二标签组中每个标签类别对应的标签的内容属性具有第二子属性；所述第一预设规则为：对所述第一子属性赋予第一权重，并对所述第二子属性赋予第二权重；计算所述第一子属性与所述第一权重的第一乘积以及所述第二子属性与所述第二权重的第二乘积；将所述第一乘积与所述第二乘积相加。
在第一标签组和第二标签组的第一标签类别对应的标签的内容属性相同的情况下，相同内容属性的子属性可能相同或相似，也可能不同。基于子属性的不同，对每一种内容属性还可以进行区分，子属性可以是数值，也可以是每一个内容属性再进行细分后的子区间，每一个内容属性可能具有多个子属性。子属性的定义是为了说明如何得到融合标签组。例如，上述第一报警区间中的45 分钟后出现高血糖报警的第一子属性为45分钟，20分钟后出现高血糖报警的第二子属性为20分钟，第一权重为0.5，第二权重为0.5，此时计算出的融合子属性为32.5分钟，融合标签为32.5分钟后出现高血糖报警。当第一标签组和第二标签组的第一标签类别对应的标签的内容属性相同，则确定所述融合标签组中的第一标签类别对应的标签(包含标签：32.5分钟后出现高血糖报警)为第三标签组中的第一标签类别对应的标签。第一权重与第二权重之和为1；第一权重、第二权重的设置也可以基于其它方式定义或自定义。
所述的标签生成方法中，所述方法还包括：
利用至少一显示模块实现所述输出的可视化。
显示模块可以配置为显示第一标签组、用户数据输入框等。
和/或，利用至少一采集模块获取用户数据。
采集模块可以配置为获取用户数据的输入模块，尤其是第一时间段的用户数据，例如一个或多个事件、一个或多个用户特征数据。
为了说明本申请的标签生成方法，结合不同的当前场景，提供以下具体实施例。
用户当前在血糖平稳阶段且血糖值为5mmol/L的情况下记录饮食300g碳水。本申请的系统从用户知识库中进行查找，查找到一个相似度最高的案例A，且相似度最大值为92％，确认大于相似阈值90％。案例A表示该使用者曾经发生的饮食案例，且血糖变化率，血糖值和饮食的量与本次案例的相似度最高。案例A的标签为“30分钟后出现高血糖报警”和“血糖出现快速上升”。此时将以上两个标签作为第一标签组通过用户界面输出给患者。
当患者收到第一标签组后患者可以选择执行此次饮食或由于重新调整饮食并重新记录其饮食情况。如图3所示，图3中，横坐标表示时间，纵坐标表示血糖值，其中血糖值由图示两条平行的虚线分为高血糖区间、目标血糖区间、低血糖区间，t1表示饮食的记录事件，此时出现2个标签。标签是基于过往的相似案例发生后的葡萄糖水平(虚线曲线表示的历史案例)得出的。当本次饮食确认发生后，监测t1到t2的葡萄糖水平变化，即图3中的实线曲线部分表示的本次案例，从而得到第二标签组。
根据患者的记录，利用饮食后3小时的葡萄糖水平的变化生成新的第二标签组，如“50分钟后出现高血糖报警”和“血糖缓慢上升”。将第一标签与第二标签进行比较，报警类别中，“30分钟后出现高血糖报警”与“50分钟后出现高血糖报警”属于相同标签类别且子属性不同；变化率类别中，“血糖出现快速上升”与血糖缓慢上升”属于相反标签。因此本次案例与历史案例共同融 合后存储在用户知识库中的第三标签组应为“约40分钟左右出现高血糖报警”和“血糖缓慢上升”。
参见图4，下面对本申请提供的标签生成装置进行描述，下文描述的标签生成装置与上文描述的标签生成方法可相互对应参照，所述标签生成装置包括：
第一案例获取模块10，设置为获取第一案例，所述第一案例包括第一时间段的葡萄糖水平数据和当前时刻的用户数据。
比对模块20，设置为使用至少一个模型进行数据比对，得到比对结果，所述比对结果包括相似度；所述数据比对的一方数据为所述第一案例、另一方数据为用户知识库中的历史案例；所述用户知识库中包含自然数个历史案例，每个历史案例具有对应的历史标签组。
第一标签组输出模块30，设置为选取所述相似度中的最大值，若所述最大值大于相似阈值，则将所述最大值对应的历史案例作为第二案例，将所述第二案例对应的历史标签组作为第一标签组并输出。
第二标签组获取模块40，设置为获取第二标签组，所述第二标签组是基于第二时间段的葡萄糖水平数据确定的；其中，所有历史标签组和第二标签组构成的标签集中含有多个标签类别，所述标签集中每一个标签类别中包括至少两种内容属性的标签。
更新模块50，设置为针对每一个标签类别，比较所述第一标签组和第二标签组中的所述每一个标签类别对应的标签的内容属性，得到比较结果，所述比较结果包括每一个标签类别对应的标签的内容属性是否相同；基于所述比较结果，确定出第三标签组；所述第三标签组是在所述第二标签组或融合标签组中选取的，所述融合标签组是将所述第一标签组与第二标签组中的相同标签类别对应的标签基于第一预设规则进行融合后得到的；将所述第一案例及其对应的第三标签组更新至所述用户知识库。
由于标签生成装置与上述实施例中的标签生成方法是一一对应的，上述标签生成方法中的任一实施例同样适用于该标签生成装置，此处不再一一赘述。
本申请还提供了一种监控血糖水平的系统，包括：
传感器，设置为获取葡萄糖水平数据；无线发射器，设置为发射所述葡萄糖水平数据；以及移动计算装置，包括：无线接收器，设置为接收所述葡萄糖水平数据；存储器，设置为存储包含所述葡萄糖水平数据的数据；处理器，设置为处理所述存储器存储的数据，以及软件应用程序，所述软件应用程序包含存储于所述存储器中的指令，所述指令执行时执行上述多个实施例的标签生成方法。
图5是本申请实施例提供的一种电子设备的示意图，该电子设备可以包括：处理器(processor)510、通信接口(Communications Interface)520、存储器(memory)530和通信总线540，其中，处理器510，通信接口520，存储器530通过通信总线540完成相互间的通信。处理器510可以调用存储器530中的逻辑指令，以执行上述实施例所提供的标签生成方法。
此外，上述的存储器530中的逻辑指令可以通过软件功能单元的形式实现并作为独立的产品销售或使用时，可以存储在一个计算机可读取存储介质中。本申请的技术方案本质上可以以软件产品的形式体现出来，该计算机软件产品存储在一个存储介质中，包括多个指令用以使得一台计算机设备(可以是个人计算机，服务器，或者网络设备等)执行本申请实施例所述方法的全部或部分步骤。而前述的存储介质包括：U盘、移动硬盘、只读存储器(Read-Only Memory，ROM)、随机存取存储器(Random Access Memory，RAM)、磁碟或者光盘等多种可以存储程序代码的介质。
另一方面，本申请还提供一种计算机程序产品，所述计算机程序产品包括存储在非暂态计算机可读存储介质上的计算机程序，所述计算机程序包括程序指令，当所述程序指令被计算机执行时，计算机能够执行上述实施例所提供的标签生成方法。
又一方面，本申请还提供一种非暂态计算机可读存储介质，其上存储有计算机程序，该计算机程序被处理器执行时实现以执行上述实施例所提供的标签生成方法。
以上所描述的装置实施例仅仅是示意性的，其中所述作为分离部件说明的单元可以是或者也可以不是物理上分开的，作为单元显示的部件可以是或者也可以不是物理单元，即可以位于一个地方，或者也可以分布到多个网络单元上。可以根据实际的需要选择其中的部分或者全部模块来实现本实施例方案的目的。
通过以上的实施方式的描述，实施方式可借助软件加必需的通用硬件平台的方式来实现，当然也可以通过硬件。上述技术方案本质上可以以软件产品的形式体现出来，该计算机软件产品可以存储在计算机可读存储介质中，如ROM/RAM、磁碟、光盘等，包括多个指令用以使得一台计算机设备(可以是个人计算机，服务器，或者网络设备等)执行实施例或者实施例的一些部分所述的方法。

Claims (23)

  1. 一种标签生成方法，包括：
    获取第一案例，其中，所述第一案例包括第一时间段的葡萄糖水平数据和当前时刻的用户数据；
    使用至少一个模型进行数据比对，得到比对结果，其中，所述比对结果包括相似度；所述数据比对的一方数据为所述第一案例、另一方数据为用户知识库中的历史案例；所述用户知识库中包含自然数个历史案例，每个历史案例具有对应的历史标签组；
    选取所述相似度中的最大值，在所述最大值大于相似阈值的情况下，将所述最大值对应的历史案例作为第二案例，将所述第二案例对应的历史标签组作为第一标签组并输出；
    获取第二标签组，其中，所述第二标签组是基于第二时间段的葡萄糖水平数据确定的；所有历史标签组和所述第二标签组构成的标签集中含有多个标签类别，所述标签集中每一个标签类别中包括至少两种内容属性的标签；
    针对每一个标签类别，比较所述第一标签组和所述第二标签组中的所述每一个标签类别对应的标签的内容属性，得到比较结果，其中，所述比较结果包括每一个标签类别对应的标签的内容属性是否相同；
    基于所述比较结果，确定出第三标签组；其中，所述第三标签组是在所述第二标签组或融合标签组中选取的，所述融合标签组是将所述第一标签组与所述第二标签组中的相同标签类别对应的标签基于第一预设规则进行融合后得到的；
    将所述第一案例及所述第一案例对应的第三标签组更新至所述用户知识库。
  2. 根据权利要求1所述的标签生成方法，还包括：
    通过网络从关联所述用户的血糖测量设备处获取用户的葡萄糖水平数据；
    其中，所述葡萄糖水平数据包括第一时间段的葡萄糖水平数据及第二时间段的葡萄糖水平数据；所述第一时间段的葡萄糖水平数据包括第一葡萄糖浓度数据和第一葡萄糖浓度变化率，所述第二时间段的葡萄糖水平数据包括第二葡萄糖浓度数据和第二葡萄糖浓度变化率；
    所述第一葡萄糖浓度变化率是基于所述第一葡萄糖浓度数据以正负值变化率的方式得到的；所述第二葡萄糖浓度变化率是基于所述第二葡萄糖浓度数据以正负值变化率的方式得到的。
  3. 根据权利要求1所述的标签生成方法，其中，所述获取第一案例，包括：
    获取当前时刻的用户数据；其中，所述用户数据包括与葡萄糖浓度相关联的至少一个事件、以及至少一个用户特征数据。
  4. 根据权利要求3所述的标签生成方法，其中，所述至少一个事件与碳水消耗、锻炼、睡眠以及物质的施予中的至少一者相关联；所述物质的施予包括用药类型、用药剂量、碳水施予量；所述用药类型包括长效胰岛素、短效胰岛素、速效胰岛素中的至少一种。
  5. 根据权利要求4所述的标签生成方法，其中，所述至少一个用户特征数据与用户的基本生理信息、以及个人信息中的至少一个相关联；所述基本生理信息包括胰岛素敏感系数、胰岛素-碳水比中的至少一种；所述个人信息包括性别、所处地域、糖尿病类型、年龄、体重、胰岛素历史使用年限中的至少一种。
  6. 根据权利要求1所述的标签生成方法，其中，所述使用至少一个模型进行数据比对，得到比对结果，包括：
    对所述第一案例进行规则化，得到至少两个比对特征值；
    基于所述至少两个比对特征值与所述历史案例包含的历史特征值组中与所述至少两个比对特征值分别对应的至少两个历史特征值，得到至少两个绝对距离；
    基于所述至少两个绝对距离和每个比对特征值对应的特征权值，得到相似度，所述特征权值是基于所述特征权值对应的比对特征值与葡萄糖水平的相关性大小确定的。
  7. 根据权利要求6所述的标签生成方法，其中，所述使用至少一个模型进行数据比对，得到比对结果，包括：
    基于至少一个比对特征值，在所述用户知识库中进行预先筛选，得到预选库；所述预选库中的历史案例包含的历史特征值组中存在与所述至少一个比对特征值相同的部分；
    使用至少一个模型进行数据比对，得到比对结果，所述比对结果包括相似度；所述数据比对的一方数据为所述第一案例、另一方数据为所述预选库中的历史案例。
  8. 根据权利要求1所述的标签生成方法，在所述选取相似度中的最大值，在所述最大值大于相似阈值的情况下，将所述最大值对应的历史案例作为第二案例，将所述第二案例对应的历史标签组作为第一标签组并输出之后，还包括：
    获取所述用户对所述第一案例的执行情况；
    在所述执行情况为所述用户未执行所述第一案例的情况下，暂停执行所述 获取第二标签组以及所述针对每一个标签类别，比较所述第一标签组和所述第二标签组中的所述每一个标签类别对应的标签的内容属性，得到比较结果的操作；
    在所述执行情况为所述用户执行所述第一案例的情况下，继续执行所述获取第二标签组以及所述针对每一个标签类别，比较所述第一标签组和所述第二标签组中的所述每一个标签类别对应的标签的内容属性，得到比较结果的操作。
  9. 根据权利要求1所述的标签生成方法，在所述使用至少一个模型进行数据比对，得到比对结果之后，还包括：
    在所述最大值不大于相似阈值或所述比对结果为用户知识库为空的情况下，暂停输出所述第一标签组，并继续执行所述获取第二标签组以及所述针对每一个标签类别，比较所述第一标签组和所述第二标签组中的所述每一个标签类别对应的标签的内容属性，得到比较结果的操作。
  10. 根据权利要求1所述的标签生成方法，其中，所述获取第二标签组，所述第二标签组是基于第二时间段的葡萄糖水平数据确定的，包括：
    基于第二预设规则，获取所述第二标签组；
    所述第二预设规则包括：
    对所述第二时间段的葡萄糖水平数据基于每种内容属性进行划分，得到第一属性组；
    基于所述第一属性组和预先确定的关联关系，得到所述第二标签组，其中，所述关联关系为多个内容属性组与多个预存标签组之间的关联关系，所述第二标签组是从所述多个预存标签组中选取的标签组。
  11. 根据权利要求1所述的标签生成方法，其中，
    所述多个标签类别包括报警类、变化率类、目标血糖时间占比类中的至少两种。
  12. 根据权利要求11所述的标签生成方法，其中，所述标签集中每一个标签类别中包括至少两种内容属性的标签，包括以下情况中的至少一种：
    所述报警类对应的标签中至少存在内容属性为第一报警区间的标签、以及内容属性为第二报警区间的标签；
    所述变化率类对应的标签中至少存在内容属性为第一变化率区间的标签、以及内容属性为第二变化率区间的标签；
    所述目标血糖时间占比类对应的标签中至少存在内容属性为第一目标血糖 时间占比区间的标签、以及内容属性为第二目标血糖时间占比区间的标签。
  13. 根据权利要求1所述的标签生成方法，在所述将所述第二案例对应的历史标签组作为第一标签组并输出之前，还包括：
    判断所述第一案例中是否存在异常数据，响应于所述第一案例中存在异常数据，暂停输出所述第一标签组、并暂停执行所述获取第二标签组以及所述针对每一个标签类别，比较所述第一标签组和所述第二标签组中的所述每一个标签类别对应的标签的内容属性，得到比较结果的操作。
  14. 根据权利要求1所述的标签生成方法，在所述获取第二标签组之前，还包括：
    判断所述第二时间段的葡萄糖水平数据中是否存在异常数据，响应于所述第二时间段的葡萄糖水平数据中存在异常数据，暂停执行所述获取第二标签组以及所述针对每一个标签类别，比较所述第一标签组和所述第二标签组中的所述每一个标签类别对应的标签的内容属性，得到比较结果的操作。
  15. 根据权利要求1所述的标签生成方法，其中，所述基于所述比较结果，确定出第三标签组，包括：
    在所述比较结果为第一比较结果的情况下，确定所述融合标签组为所述第三标签组；所述第一比较结果为相同标签类别对应的标签的内容属性相同的标签类别数目占所有标签类别数目的比例大于比例阈值；
    在所述比较结果为第二比较结果的情况下，确定所述第二标签组为所述第三标签组；所述第二比较结果为相同标签类别对应的标签的内容属性相同的标签类别数目占所有标签类别数目的比例不大于比例阈值。
  16. 根据权利要求1所述的标签生成方法，其中，所述基于所述比较结果，确定出第三标签组，包括：
    在所述比较结果包括第一标签类别对应第三比较结果的情况下，将所述融合标签组中的所述第一标签类别对应的标签作为所述第三标签组中所述第一标签类别对应的标签；所述多个标签类别至少包括所述第一标签类别，所述第三比较结果为所述第一标签类别对应的标签的内容属性相同；
    在所述比较结果包括第一标签类别对应第四比较结果的情况下，将所述第一标签组中的所述第一标签类别对应的标签作为所述第三标签组中所述第一标签类别对应的标签；所述多个标签类别至少包括所述第一标签类别，所述第四比较结果为所述第一标签类别对应的标签的内容属性不同。
  17. 根据权利要求1所述的标签生成方法，其中，
    所述融合标签组是将所述第一标签组中每个标签类别对应的标签的第一子属性与所述第二标签组中所述每个标签类别对应的标签的第二子属性基于所述第一预设规则进行融合后得到的；其中，所述第一标签组中每个标签类别对应的标签的内容属性具有第一子属性，所述第二标签组中每个标签类别对应的标签的内容属性具有第二子属性；
    所述第一预设规则为：
    对所述第一子属性赋予第一权重，并对所述第二子属性赋予第二权重；
    计算所述第一子属性与所述第一权重的第一乘积以及所述第二子属性与所述第二权重的第二乘积；
    将所述第一乘积与所述第二乘积相加。
  18. 根据权利要求1所述的标签生成方法，其中，所述第一时间段从第一时刻延续至当前时刻，所述第一时刻在当前时刻之前，所述第二时间段从当前时刻延续至第二时刻，所述第二时刻在当前时刻之后。
  19. 根据权利要求1所述的标签生成方法，其中，所述方法还包括以下至少之一：
    利用至少一显示模块实现所述输出的可视化；
    利用至少一采集模块获取用户数据。
  20. 一种标签生成装置，包括：
    第一案例获取模块，设置为获取第一案例，其中，所述第一案例包括第一时间段的葡萄糖水平数据和当前时刻的用户数据；
    比对模块，设置为使用至少一个模型进行数据比对，得到比对结果，其中，所述比对结果包括相似度；所述数据比对的一方数据为所述第一案例、另一方数据为用户知识库中的历史案例；所述用户知识库中包含自然数个历史案例，每个历史案例具有对应的历史标签组；
    第一标签组输出模块，设置为选取所述相似度中的最大值，在所述最大值大于相似阈值的情况下，将所述最大值对应的历史案例作为第二案例，将所述第二案例对应的历史标签组作为第一标签组并输出；
    第二标签组获取模块，设置为获取第二标签组，其中，所述第二标签组是基于第二时间段的葡萄糖水平数据确定的；所有历史标签组和所述第二标签组构成的标签集中含有多个标签类别，所述标签集中每一个标签类别中包括至少两种内容属性的标签；
    更新模块，设置为针对每一个标签类别，比较所述第一标签组和所述第二标签组中的所述每一个标签类别对应的标签的内容属性，得到比较结果，其中，所述比较结果包括每一个标签类别对应的标签的内容属性是否相同；基于所述比较结果，确定出第三标签组；其中，所述第三标签组是在所述第二标签组或融合标签组中选取的，所述融合标签组是将所述第一标签组与所述第二标签组中的相同标签类别对应的标签基于第一预设规则进行融合后得到的；将所述第一案例及所述第一案例对应的第三标签组更新至所述用户知识库。
  21. 一种监控血糖水平的系统，包括：
    传感器，设置为获取葡萄糖水平数据；
    无线发射器，设置为发射所述葡萄糖水平数据；
    以及
    移动计算装置，包括：
    无线接收器，设置为接收所述葡萄糖水平数据；
    存储器，设置为存储包含所述葡萄糖水平数据的数据；
    处理器，设置为处理所述存储器存储的数据，以及软件应用程序，所述软件应用程序包含存储于所述存储器中的指令，所述指令执行时实现如权利要求1至19任一项所述的标签生成方法。
  22. 一种电子设备，包括存储器、处理器及存储在所述存储器上并可所述在处理器上运行的计算机程序，其中，所述处理器执行所述程序时实现如权利要求1至19任一项所述的标签生成方法。
  23. 一种非暂态计算机可读存储介质，存储有计算机程序，其中，所述计算机程序被处理器执行时实现如权利要求1至19任一项所述的标签生成方法。
PCT/CN2022/085452 2022-01-06 2022-04-07 标签生成方法和装置、监控血糖水平的系统 WO2023130599A1 (zh)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN202210007243.2 2022-01-06
CN202210007243.2A CN114021672B (zh) 2022-01-06 2022-01-06 标签生成方法和装置、监控血糖水平的系统

Publications (1)

Publication Number Publication Date
WO2023130599A1 true WO2023130599A1 (zh) 2023-07-13

Family

ID=80069599

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2022/085452 WO2023130599A1 (zh) 2022-01-06 2022-04-07 标签生成方法和装置、监控血糖水平的系统

Country Status (2)

Country Link
CN (1) CN114021672B (zh)
WO (1) WO2023130599A1 (zh)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN117494045A (zh) * 2023-11-06 2024-02-02 南京海汇装备科技有限公司 一种基于数据融合的数据集成智能管控系统及方法

Families Citing this family (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN114021672B (zh) * 2022-01-06 2022-04-22 苏州百孝医疗科技有限公司 标签生成方法和装置、监控血糖水平的系统

Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN110085318A (zh) * 2019-03-12 2019-08-02 平安科技（深圳）有限公司 预测未来血糖值的方法、装置及计算机设备
US20210265034A1 (en) * 2020-02-18 2021-08-26 Shaklee Corporation Systems and methods for refining a dietary treatment regimen using ranked based scoring
CN113743991A (zh) * 2021-09-03 2021-12-03 上海幻电信息科技有限公司 生命周期价值预测方法及装置
CN114021672A (zh) * 2022-01-06 2022-02-08 苏州百孝医疗科技有限公司 标签生成方法和装置、监控血糖水平的系统

Family Cites Families (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN110168655B (zh) * 2016-12-13 2024-05-07 赛诺菲 用于支持健康控制的数据管理单元
CN111883228B (zh) * 2020-07-28 2023-07-07 平安科技（深圳）有限公司 基于知识图谱的健康信息推荐方法、装置、设备及介质

Patent Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN110085318A (zh) * 2019-03-12 2019-08-02 平安科技（深圳）有限公司 预测未来血糖值的方法、装置及计算机设备
US20210265034A1 (en) * 2020-02-18 2021-08-26 Shaklee Corporation Systems and methods for refining a dietary treatment regimen using ranked based scoring
CN113743991A (zh) * 2021-09-03 2021-12-03 上海幻电信息科技有限公司 生命周期价值预测方法及装置
CN114021672A (zh) * 2022-01-06 2022-02-08 苏州百孝医疗科技有限公司 标签生成方法和装置、监控血糖水平的系统

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN117494045A (zh) * 2023-11-06 2024-02-02 南京海汇装备科技有限公司 一种基于数据融合的数据集成智能管控系统及方法
CN117494045B (zh) * 2023-11-06 2024-04-26 南京海汇装备科技有限公司 一种基于数据融合的数据集成智能管控系统及方法

Also Published As

Publication number Publication date
CN114021672A (zh) 2022-02-08
CN114021672B (zh) 2022-04-22

Similar Documents

Publication Publication Date Title
WO2023130599A1 (zh) 标签生成方法和装置、监控血糖水平的系统
US11574742B2 (en) Diabetes management therapy advisor
US11538587B2 (en) Dynamic data-driven biological state analysis
KR20220016487A (ko) 생체 모니터링 및 혈당 예상을 위한 시스템들, 및 연관된 방법들
Aldhyani et al. Soft clustering for enhancing the diagnosis of chronic diseases over machine learning algorithms
US20200077931A1 (en) Forecasting blood glucose concentration
WO2021208902A1 (zh) 一种睡眠报告的生成方法、装置、终端以及存储介质
US20170278017A1 (en) System and method for managing behavior change applications for mobile users
CN110349671B (zh) 体检数据处理方法、系统、电子设备及存储介质
US20210241916A1 (en) Forecasting and explaining user health metrics
US20130124440A1 (en) Data mining technique with maintenance of fitness history
US20090326981A1 (en) Universal health data collector and advisor for people
US20110184754A1 (en) System and method for remote health care management
TWI502537B (zh) 生理資料量測管理系統及其方法
CN114023418A (zh) 胰岛素推荐方法和装置、监控血糖水平的系统
JP2023507175A (ja) 連続血糖モニタリングシステムによる多状態エンゲージメント
WO2023115751A1 (zh) 血糖预测方法和装置、监测血糖水平的系统
Ding et al. Diabetic complication prediction using a similarity-enhanced latent Dirichlet allocation model
CN112528009A (zh) 生成用户慢病调理方案的方法、装置及计算机可读介质
WO2020005822A1 (en) Activity tracking and classification for diabetes management system, apparatus, and method
US20210383925A1 (en) Systems for adaptive healthcare support, behavioral intervention, and associated methods
CN108022653A (zh) 管控药物的特征获取方法、电子装置及计算机可读存储介质
CN115171840B (zh) 一种个性化体验套餐生成系统及生成方法
US12040096B2 (en) Diabetes management therapy advisor
US20240071623A1 (en) Patient health platform