Preprocessing method of biological component measurement data for predicting growth

Abstract

The present invention relates to a method for preprocessing input data for predicting growth of a child or a teenager, an apparatus and a computer program, the method for preprocessing biological component measurement data for predicting growth according to an exemplary embodiment of the present invention comprising: a step of inputting body data for obtaining a subject; a step of generating a first variable on the basis of the body data of the subject; and a step of comparing the first variable with a preset value to determine an error in the body data.

Description

Technical Field

[0001] The present invention relates to a method, a device and a computer program for preprocessing input data for predicting the growth of children or adolescents. Background Art

[0002] Recently, with the development of artificial intelligence technology, artificial intelligence technology is applied to various fields, and methods of generating additional information by extracting features from data through neural network models are being developed and used to replace existing data processing methods.

[0003] Recently, AI techniques have gone beyond simple tracking and detection of objects and have been applied to learn from past experiences to predict the future or to derive current features that reflect information about time series changes.

[0004] In addition, in AI that learns from data, the quality of data is considered one of the key factors in AI outcomes. For example, leading companies with the most advanced AI-based autonomous driving technology all emphasize that ensuring high-quality data is key.

[0005] Among them, predictive analysis is a technology in the field of statistics and data mining that extracts information from data and uses it to predict trends, behavior patterns, etc. This predictive analysis can be applied to all fields where decisions need to be made based on information obtained from data. The core of predictive analysis is to predict unknown variables after understanding the relationship between variables.

[0006] To this end, various approaches are adopted depending on the data characteristics and predicted subjects.

[0007] Among the various fields that require predictive analysis is the physical growth of children or adolescents. Parents and adolescents are very interested in questions such as when and how much height they will grow.

[0008] In the past, a method has been proposed for predicting height or analyzing the relationship with genetic / environmental factors by photographing the growth plate through X-ray (Korean Patent Registration No. 10-2075743, Korean Patent Registration No. 10-1866208). In addition, a method has been proposed for preparing the physical data of sample subjects with different measurement periods or measurement times into a form suitable for learning a growth prediction model (Korean Patent Registration No. 10-2198302).

[0009] Since children or adolescents have their own unique growth stages, this needs to be taken into account to increase the reliability of solutions provided through predictive data and analysis.

[0010] Especially for children or adolescents, the body undergoes a period of rapid changes according to the growth stage. Therefore, it is necessary to first perform a process to determine the errors in the input (or measured) data (biological component data) so as to improve the accuracy of growth prediction and the results of artificial intelligence learning. Summary of the invention

[0011] Technical issues

[0012] One of the various objects of the present invention is to provide a method, device and computer program for preprocessing input data to predict the growth of children or adolescents and provide customized solutions according to the growth stage.

[0013] Means of solving problems

[0014] In order to solve various embodiments of the subject of the present invention, a preprocessing method for biological component measurement data for predicting growth can be provided, which is a preprocessing method for biological component measurement data for predicting growth executed by a computing device, and includes: a step of inputting physical data of a subject; a step of generating a first variable based on the physical data of the subject; and a step of comparing the first variable with a preset value to determine errors in the physical data.

[0015] The present invention may also be characterized as: the step of inputting the body data of the subject, including: the step of inputting the biological component data of the subject; and the step of inputting the identification data of the subject.

[0016] The present invention may also be characterized in that, after the physical data of the subject is input, the method further includes connecting the biological component data of the subject and the identification data of the subject to generate the first data.

[0017] The present invention may also be characterized in that: after generating the first data, if a preset value is measured in the biological component data, the step of deleting the biological component data is included.

[0018] The present invention may also be characterized in that: the first variable is generated based on biological component data of the subject.

[0019] The present invention may also be characterized in that the first variable is generated as the sum of muscle mass (Soft Lean Mass), body fat mass (Body Fat Mass), and bone mineral mass (Osseous mineral) in the biological composition data of the subject.

[0020] The present invention may also be characterized in that the step of determining an error includes a step of comparing the first variable with a value in the biological component data.

[0021] The present invention may also be characterized in that the step of judging an error also includes a step of judging whether the age of the subject in the identification data meets a preset standard before comparing the first variable with a value in the biological component data.

[0022] The present invention may also be characterized in that: it also includes a step of deleting the biological component data when a result value of comparing the first variable with a value in the biological component data falls within a preset range.

[0023] The present invention may also be characterized in that: it also includes a step of generating second data when a result value of comparing the first variable with a value in the biological component data does not fall within a preset range.

[0024] An exemplary embodiment of the present invention may provide a program stored in a computer-readable recording medium including a program code for executing the above-mentioned method of preprocessing biological component measurement data for predicting growth.

[0025] An exemplary embodiment of the present invention may provide a computer-readable recording medium having recorded therein a program for executing the above-mentioned method for preprocessing biological component measurement data for predicting growth.

[0026] An exemplary embodiment of the present invention may provide a preprocessing device for biological component measurement data for predicting growth, which is a preprocessing device for biological component measurement data for predicting growth, comprising: an input unit for inputting physical data of a subject; a variable generating unit for generating a first variable based on the physical data of the subject; and an error judging unit for judging an error in the physical data by comparing the first variable with a preset value.

[0027] The present invention may also be characterized in that: the input unit includes: a first input unit for inputting the biological component data of the subject; and a second input unit for inputting the identification data of the subject;

[0028] The present invention may also be characterized in that: it further includes a connection unit for connecting data inputted through the first input unit and the second input unit.

[0029] Unless the features of the above-mentioned embodiments are inconsistent with or exclusive of other embodiments, they may be combined and embodied in other embodiments.

[0030] Effects of the Invention

[0031] According to various embodiments of the present invention, when predicting the growth of a child or adolescent, the accuracy of the growth prediction can be improved by determining errors in the biological component data measured to predict the growth of the child or adolescent.

[0032] According to various embodiments of the present invention, when providing solutions to the growth needs of children or adolescents, the accuracy of growth prediction can be improved by eliminating errors in input data.

[0033] According to various embodiments of the present invention, the accuracy of growth prediction for children or adolescents may be improved by determining errors in biological component data measured for predicting the growth of children or adolescents and generating refined input data.

[0034] According to various embodiments of the present invention, the accuracy of the solution required for the growth of a child or adolescent can be improved by determining errors in the biological component data measured to generate the solution required for the growth of the child or adolescent and generating refined input data.

[0035] The effects of the present invention are not limited to the above-mentioned effects, and those skilled in the art will clearly recognize other effects not mentioned from the following description. BRIEF DESCRIPTION OF THE DRAWINGS

[0036] Figure 1 is a diagram illustrating a structure of a growth prediction system according to an exemplary embodiment of the present invention.

[0037] Figure 2 is a diagram illustrating a structure of a growth prediction and solution generating apparatus according to an exemplary embodiment of the present invention.

[0038] Figure 3 It is shown Figure 2 Figure 2. Data preprocessing section of the .

[0039] Figure 4 is a graph illustrating predicted height and target height by growth stage according to an exemplary embodiment of the present invention.

[0040] Figure 5 is a graph illustrating obesity and growth at normal weight by growth stage of boys according to an exemplary embodiment of the present invention.

[0041] Figure 6 is a graph illustrating obesity and growth at normal weight according to growth stages of girls according to an exemplary embodiment of the present invention.

[0042] Figure 7 is a diagram illustrating a structure of a neural network for growth prediction and solution provision according to an exemplary embodiment of the present invention.

[0043] Figure 8 is a diagram illustrating a first neural network model according to an exemplary embodiment of the present invention.

[0044] Fig. 9is a diagram illustrating a second neural network model according to an exemplary embodiment of the present invention.

[0045] Figures 10 to 12 is a diagram illustrating a preprocessing method of biological component measurement data for predicting growth according to an exemplary embodiment of the present invention.

[0046] Fig.13 is a graph showing biological component data of a subject according to an exemplary embodiment of the present invention.

[0047] Fig.14 is a diagram showing data connecting biological component data and identification data of a subject according to an exemplary embodiment of the present invention. DETAILED DESCRIPTION

[0048] The following is a description of the specific implementation of the present invention with reference to the accompanying drawings. The following specific description is provided to help a comprehensive understanding of the methods, devices and / or systems described in this specification. However, these are only examples, and the present invention is not limited thereto.

[0049] When describing the embodiments of the present invention, if it is judged that the specific description of the known technology related to the present invention will make the gist of the present invention unclear, the detailed description thereof will be omitted. In addition, the terms described below are defined in consideration of the functions in the present invention, and can be changed according to the intention or habit of the user or operator. Therefore, the terms should be defined based on the contents in the entire specification.

[0050] The terms used in the detailed description are only for explaining the embodiments of the present invention and are not limiting. Unless clearly used otherwise, expressions in the singular form include the meaning of the plural form.

[0051] In this specification, expressions such as "comprise" or "have" should be interpreted as referring to certain characteristics, numbers, steps, actions, elements, parts of these, or combinations thereof, and do not exclude the existence or possibility of one or more other characteristics, numbers, steps, actions, elements, parts of these, or combinations thereof other than those mentioned.

[0052] In addition, when describing the components of the embodiments of the present invention, terms such as first, second, A, B, (a), (b), etc. may be used. These terms are only used to distinguish the components from other components, and the nature, order, sequence, etc. of the components are not limited by these terms.

[0053] Figure 1 is a diagram illustrating a structure of a growth prediction system according to an exemplary embodiment of the present invention.

[0054] refer to Figure 1The growth prediction system of this embodiment can receive the time series physical data 1 of the subject and distinguish the gender of the subject based on the input physical data 1 2 .

[0055] Furthermore, according to gender, in the case of a boy, growth prediction 5-1 can be performed using growth prediction model 3-1, and in the case of a girl, growth prediction 5-2 can be performed using growth prediction model 3-1 used for boys and other growth prediction models 3-2.

[0056] The growth rate by growth stage may be different for boys and girls. For example, as described below, the period when boys and girls enter the rapid growth stage may be different, and therefore, the growth rate or the generation of a solution considering the growth rate may be different. Therefore, preferably, it is necessary to perform growth prediction using a model that learns different learning data based on the subject's gender.

[0057] Figure 2 is a diagram showing a structure of a growth prediction and solution generation apparatus according to an exemplary embodiment of the present invention, Figure 3 It is shown Figure 2 Figure 1. Data preprocessing part of the diagram. Fig.13 is a diagram showing biological component data of a subject according to an exemplary embodiment of the present invention, Fig.14 is a diagram showing data connecting biological component data and identification data of a subject according to an exemplary embodiment of the present invention.

[0058] Below, refer to Figure 2 , Figure 3 , Fig.13 and Fig.14 Provide explanation.

[0059] According to an exemplary embodiment of the present invention, the growth prediction and solution generation device may include a data preprocessing unit 100, an input unit 10, a gender determination unit 20, a growth stage determination unit 30, a prediction unit 50, a solution generation unit 70, and a display unit 90.

[0060] The growth prediction and solution generation device can obtain the time series physical data of the subject by inputting it through the input unit 10.

[0061] The physical data of the subject may include data capable of identifying the subject (identification data), and biological component data of the subject.

[0062] Exemplarily, the identification data may include data for identifying the subject, such as grade (or age in months), gender and height, and the biological component data may include data on weight, protein, mineral content, body fat, body water, muscle mass, body fat mass, bone tissue, skeletal muscle mass, body mass index, basal metabolic rate, neck circumference, chest circumference, abdominal circumference, thigh circumference, arm circumference and hip circumference.

[0063] More specifically, refer to Fig.13 and Fig.14 The subject's identification data 1201 may include date of birth 1201-1, gender 1201-2, age in months 1201-3 and measurement date 1201-4.

[0064] In addition, the subject's biological component data 1101 may include data number 1101-1, height 1101-2, weight 1101-3, protein content 1101-4, mineral content 1101-5, body fat mass 1101-6 (Body Fat Mass), muscle mass 1101-7 (soft lean mass), bone mineral mass 1101-8 (Osseous mineral), and skeletal muscle mass 1101-9.

[0065] These body data are merely examples to help understand the present invention, and the present embodiment is not limited thereto. Of course, the types of data constituting the body data may be changed in various ways according to the embodiment.

[0066] In addition, exemplary embodiments of the present invention may include a data preprocessing unit 100 for determining errors in the subject's physical data and generating refined data.

[0067] Since the biological component data 1101 may change rapidly according to the growth stage due to the physical characteristics of children or adolescents, errors (errors in the measurement process) may occur in the measurement process of other biological component data, such as measurement method, measurement period, measurement environment, etc.

[0068] In addition, during the transmission or storage of the measured biological component data, errors (data transmission errors) may occur.

[0069] Therefore, the data preprocessor 100 of this embodiment, after generating a virtual variable (hereinafter referred to as the first variable) based on the measured biological component data 1101, compares the first variable with a value in the biological component data, thereby generating refined data, which is obtained through a process of judging errors occurring during the measurement of the measured biological component data 1101 and data transmission errors.

[0070] Reference Figures 10 to 12A more detailed description is given.

[0071] The data input to the input section 10 may include time-series body data of the subject.

[0072] The time series body data may be continuous data or discontinuous data, but may be data included in at least one of the periods corresponding to the growth stages of children or adolescents.

[0073] More specifically, the collection period and the collection frequency of the subject's time-series physical data may vary.

[0074] For example, the first subject may have physical data measured from 8 to 12 years old, which is part of the childhood or adolescence period, and the second subject may have physical data measured irregularly, such as at 8 years old, 10 to 12 years old, 15 years old, etc.

[0075] In addition, the third subject may have physical data measured multiple times within a certain period (during one of the growth stages of a child or an adolescent), while the fourth subject may have physical data measured only once within a certain period of time (during one of the growth stages of a child or an adolescent).

[0076] As described above, depending on the collection period and the number of collections, the subject's physical data may include two or more growth stages of children or adolescents (first subject, second subject), but may not be the case (third subject, fourth subject).

[0077] As for the third subject as described, when there are multiple measured physical data during one growth stage among multiple growth stages, the growth stage judgment unit 30 classifies the growth stage including the physical data of the third subject through the growth stage classification unit 31, and can extract the physical data through the physical data extraction unit 33. The physical data extracted here can generally include the physical data of the third subject input through the input unit 10.

[0078] However, as in the case of the fourth subject, when the subject's body data corresponds to only one of multiple growth stages and there is body data measured only once during the period, body data corresponding to any period before the growth stage judgment unit 30 inputs the body data of the fourth subject can be further generated.

[0079] More specifically, the inputted physical data of the subject may be used to generate time-series physical data corresponding to the subject in any period based on the time-series physical growth information of a plurality of sample subjects stored in advance.

[0080] As an example, physical data can be generated based on the input subject's physical data and a distribution model (similarity) of time series physical growth data of multiple sample subjects stored in advance, and physical data can also be generated based on a Bayesian inference model (conditional probability).

[0081] The growth stage determination unit 30 can classify the subject's body data input through the input unit 10 into any one of a plurality of growth stages in the growth stage classification unit 31 , and can extract body data corresponding to the classified growth stage in the body data extraction unit 33 .

[0082] Figure 4 is a graph illustrating predicted height and target height by growth stage according to an exemplary embodiment of the present invention.

[0083] Figure 4 The x-axis in the figure represents age in months, and the y-axis represents height (cm). In contrast, the dotted line P represents the predicted growth rate of the evaluation subject, and the solid line G represents the target growth rate of the evaluation subject.

[0084] Reference Figure 4 The growth stages of the children or adolescents described above may include a general growth stage (301), a rapid growth stage (303), a slow growth stage (305), and a growth plate-free stage (307).

[0085] Each growth stage can be distinguished according to the growth rate. The height growth each year is different in each growth stage. Even in the same growth stage, the actual height growth will vary depending on the growth type.

[0086] The general growth stage (301) usually refers to the period before puberty when secondary sexual characteristics appear. During this period, the growth plates of children or adolescents are generally open, so depending on the growth environment, the shorter growth type will generally grow 4 to 5 cm per year, while the taller growth type will grow 6 to 7 cm per year.

[0087] The rapid growth phase (303) is when secondary sexual characteristics begin to appear. Girls' breasts swell and lumps appear, boys' testicles enlarge, pubic hair begins to grow, and voice changes. The rapid growth phase (303) lasts about 2 to 3 years after the normal growth phase (301), with an average annual growth of 7 to 10 cm.

[0088] The slow growth stage (305) refers to the period when secondary sexual characteristics are completed. During this period, girls can be distinguished by the menarche, while boys can clearly see their changes through pubic hair, changes in voice, and armpit hair. If the growth rate drops rapidly in the slow growth stage (305) compared to the rapid growth stage (303), it generally lasts for about 2 to 3 years, with an average annual growth of about 5 to 6 cm, and then naturally stops growing. The growth plates begin to gradually close after the rapid growth stage (304), and about 6 months after entering the slow growth stage (305), the growth plates are closed by about 50%.

[0089] The no growth plate stage (307) refers to the period when the growth plates have closed. Although the growth period has not completely ended, it has become difficult to grow taller naturally. Generally, for girls, the no growth plate stage (307) is entered within about 1 year and 6 months to 2 years from the time of menarche, and for boys, the no growth plate stage (307) is entered within about 1 year and 6 months to 2 years from the time of armpit hair growth. In the no growth plate stage (307), although the growth plates have closed and natural growth has stopped, by changing incorrect living habits and improving body functions through customized exercises, posture correction and nutritional intake, the height can also be increased by about 1 to 3 cm.

[0090] In addition, the prediction unit 50 is a prediction model that can be implemented using artificial intelligence with a recursive neural network (RNN) structure, so that not only current values ​​but also time series values ​​can be used. For example, the prediction model can be implemented using a recursive neural network, a long short-term memory (LSTM), or a gated recurrent unit (GRU). Of course, in addition to this, various previous artificial intelligence architectures can also be applied to the prediction model of this embodiment, which will be referred to in the following. Figures 7 to 9 Provide detailed explanation.

[0091] The solution generating section 70 may generate a growth management solution based on the body data of the subject corresponding to the classified growth stage.

[0092] More specifically, when the subject is in the general growth stage (301), a solution for increasing the growth prediction value of the subject can be provided. The growth prediction value corresponds to Figure 4 The value of the y-axis in the figure can provide the subject with solutions for increasing the predicted growth value through various solution display units 90 for increasing the predicted target value of the y-axis.

[0093] Examples of solutions provided by the display unit 90 may include current height, predicted height, obesity level, body fat mass, skeletal muscle mass, protein mass, mineral content, sleep volume, exercise volume, nutritional information, living habits, posture, etc. Each indicator may be gradually expressed as attention, normal, good, etc. according to a preset range, or may also be expressed as a level.

[0094] In addition, the current status, customized solutions, and precautions for each indicator can be displayed. The current status can be displayed by stage or level according to the target value. If it is a customized solution, it can include information such as protein, mineral content, body fat, body water, muscle mass (soft lean mass), body fat mass (fat free mass), bone tissue, skeletal muscle mass, body mass index (BMI, Body Mass Index), basal metabolic rate, etc. that need to be adjusted to achieve the current target value based on the input body data.

[0095] Based on the input body data, precautions may include information on the current deficiencies that need to be adjusted, such as protein, mineral content, body fat, body water, muscle mass (soft lean mass), body fat mass (fat free mass), bone tissue, skeletal muscle mass, body mass index (BMI, Body Mass Index), basal metabolic rate, etc.

[0096] In addition, when the subject is in a rapid growth stage (303), a solution for increasing the growth prediction value of the subject in the rapid growth stage (303) can be provided. The growth prediction value corresponds to Figure 2 The value of the y-axis in the figure can provide the subject with solutions for increasing the predicted growth value through various solution display units 90 for increasing the predicted target value of the y-axis.

[0097] Figure 5 is a graph showing obesity and growth at normal weight according to growth stages of boys according to an exemplary embodiment of the present invention, Figure 6 is a graph illustrating obesity and growth at normal weight according to growth stages of girls according to an exemplary embodiment of the present invention.

[0098] Figure 5 The x-axis represents age in months, and the y-axis represents growth rate in cm. The solid line G1 represents the growth rate of each growth stage when boys are normal weight, and the dotted line P1 represents the growth rate of each growth stage when boys are obese.

[0099] Figure 6The x-axis represents age in months, and the y-axis represents growth rate in cm. The solid line G2 represents the growth rate of each growth stage when a girl is of normal weight, and the dotted line P2 represents the growth rate of each growth stage when a girl is obese.

[0100] Next, we will refer to Figure 5 and Figure 6 Provide explanation.

[0101] Obesity is not simply weight gain, but a disease accompanied by excessive accumulation of fat tissue in the body, accompanied by overweight or metabolic disorders caused by it. Medically, childhood or adolescent obesity is defined as a situation in which the weight of a child from infancy to adolescence exceeds the standard weight for height by more than 20%.

[0102] Most cases of obesity in infancy disappear after the baby's first birthday as the child becomes more active and physically active. However, in some cases, the obesity persists, and in many cases, the obesity temporarily returns to normal only to relapse during school age.

[0103] 75% to 80% of childhood and adolescent obesity will turn into adult obesity. In addition, it will inhibit the secretion of growth hormone, especially for girls, which will advance puberty and shorten growth potential, thus leading to growth stunting or causing precocious puberty.

[0104] Therefore, when obesity is predicted in school-age children or adolescents who are prone to obesity and obesity is predicted, it is necessary to provide a solution that can systematically prevent obesity.

[0105] In addition, childhood or adolescent obesity can be divided into simple obesity of unknown cause and symptomatic obesity caused by special causes, among which more than 99% of childhood obesity is simple obesity.

[0106] The general characteristics of simple obese children are that during the normal growth stage (301), both boys and girls are average or slightly taller than their peers (multiple sample subjects), but after the rapid growth stage (303), they will be shorter than their peers (multiple sample subjects) or have a lower growth rate.

[0107] In summary, childhood or adolescent obesity can be understood as a group of diseases caused by various reasons accompanied by excessive weight or metabolic disorders. Figure 5 and Figure 6 In general, boys and girls together have an average or slightly higher growth rate than multiple sample subjects of the same age during the general growth stage (301), but after the rapid growth stage (303), they show a growth rate that is often lower than multiple sample subjects of the same age.

[0108] Therefore, in this embodiment, in order to more accurately predict obesity and generate a solution based on the growth stage, the gender can be classified by the gender judgment unit 20 based on the physical data of the subject input through the input unit 10, and then the growth stage can be classified in the growth stage judgment unit 30 according to the classified gender, and the physical data can be extracted. Then, the solution generation unit 70 generates a solution in consideration of the gender and growth stage of the subject.

[0109] More specifically, in the case where both boys and girls are obese, it can be confirmed that the growth prediction value in the rapid growth stage (303) is lower than normal. Therefore, when the subject is in the rapid growth stage (303), a solution for increasing the growth prediction value can be provided, and when the subject is in the normal growth stage (301), the considered physical data can be included, especially the content related to the adjustment of indicators that can alleviate the abnormal increase of sex hormones.

[0110] In addition, when the subject belongs to the slow growth stage (305), a solution for adjusting the period of the slow growth stage (305) of the subject can be provided. Adjusting the period of the growth stage can distinguish between a case where the subject's physical data is in the initial stage of the slow growth stage (305) and a case where it is in the middle or late stage of the slow growth stage (305) in the growth stage classified based on the input physical data of the subject.

[0111] The criteria for distinguishing the initial stage and the middle and late stages of the above-mentioned slow growth stage (305) can be as follows: Figures 4 to 6 The x-axis in the figure is used as the basis, and the distinction is made based on the predetermined range from the rapid growth stage (303) to the slow growth stage (305), or based on whether the secondary sexual characteristics have ended according to the input physical data of the subject. If it has not ended, it can be distinguished as the initial stage of the slow growth stage (305), and if it has ended, it can be distinguished as the middle and late stages of the slow growth stage (305).

[0112] Preferably, it is possible to determine whether the secondary sexual characteristics have ended based on the input physical data of the subject, so as to determine whether the subject's current physical data is in the initial stage or the middle and late stage of the slow growth stage (305). When it is impossible to determine whether the secondary sexual characteristics have ended based on the input physical data of the subject, it is possible to determine whether the subject's physical data is in the initial stage or the middle and late stage of the slow growth stage (305) based on the predetermined range from the rapid growth stage (303) to the slow growth stage (305).

[0113] In addition, when the input physical data of the subject is in the initial stage of the slow growth stage (305), a period adjustment plan can be provided to delay entering the slow growth stage (305).

[0114] As described above, the period when the secondary sexual characteristics transition from the rapid growth stage (303) to the slow growth stage (305) gradually ends. Therefore, a solution for delaying the completion time of the secondary sexual characteristics can be provided, and the display unit 90 can be used to provide the subject with a display of the secondary sexual characteristics. Figures 4 to 6 Various solutions for moving the range of the x-axis corresponding to the slow growth stage (305) to the right. In this case, the range of the slow growth stage (305) can be increased according to the physical data of the subject, or the range can be reduced as the fast growth stage (303) increases.

[0115] In addition, if the input physical data of the subject is in the middle and late stages of the slow growth stage (305), a period adjustment plan can be provided to increase the period of the slow growth stage (305).

[0116] As described above, the slow growth stage (305) refers to the period when the growth plate of the subject is closed. Generally, 6 months after entering the slow growth stage (305), the growth plate is closed by about 50%. When the growth plate is closed and natural growth stops, the subject enters the no growth plate stage (307). Therefore, in this case, a solution for increasing the period of the slow growth stage (305) can be provided. In other words, the display unit 90 can be used to provide the subject with a solution for increasing the period of the slow growth stage (305). Figures 4 to 6 Various solutions for the range of the x-axis corresponding to the slow growth phase (305).

[0117] When the above-mentioned subject belongs to the general growth stage (301), the physical data that needs to be considered may include, in particular, content related to adjusting indicators that can alleviate the degree of growth plate closure.

[0118] In addition, when the subject is in the growth plate-free stage (307), solutions such as improving physical functions through lifestyle habits, customized exercises, posture correction, nutritional intake, etc. can be provided based on the input physical data of the subject.

[0119] In the growth plate-free stage (307), the growth plates are closed and natural growth stops. Therefore, solutions to improve physical functions such as lifestyle habits, customized exercises, and posture correction can be provided based on the subject's weight, body fat, body water, muscle mass, skeletal muscle mass, body mass index (BMI), basal metabolic rate, neck circumference, chest circumference, abdominal circumference, thigh circumference, arm circumference, hip circumference, etc., or solutions to improve physical functions through nutritional intake can be provided based on protein, mineral content, bone tissue (bone density), etc.

[0120] In addition, in order to more accurately predict obesity and provide a solution in this embodiment, the gender of the subject is classified by the gender determination unit 20, and the growth stage classification unit 31 sets the period of the rapid growth stage differently according to the classified gender.

[0121] This is because, as described above, boys and girls enter the rapid growth stage at different times. Also, the body data of the subject corresponding to the rapid growth stage considering gender output from the body data extraction unit 33 is input to the prediction unit 50 to output a predicted value of obesity.

[0122] When the subject is predicted to be obese, and the gender is a boy, a solution for increasing the growth prediction value of the subject in the rapid growth stage (303) may be provided, which is as described above.

[0123] Also, when the subject is predicted to be obese and the gender is a girl, a solution for increasing the duration of the rapid growth phase (303) may be provided.

[0124] For more details, refer to Figure 5 and Figure 6 For girls, the difference in predicted growth values ​​during the rapid growth phase (303) is smaller than that for boys, but as the rapid growth phase (303) transitions to the slow growth phase (305), the speed of height increase (growth rate) at each age group will drop significantly.

[0125] For example, in Figure 5 It can be seen that the difference in the growth prediction value of a normal-weight boy in the rapid growth stage (303) is significantly different from the growth prediction values ​​in the normal-weight case (G1) and the obese case (P1). In addition, a solution for increasing the growth prediction value in the rapid growth stage (303) is required for obese boys.

[0126] Also, as an example, Figure 6 In the study, normal-weight girls grew 7 cm from age 9 to 10, and about 6.5 cm from age 10 to 11.

[0127] In contrast, obese girls grew about 6.7 cm taller between the ages of 9 and 10, and about 5.7 cm taller between the ages of 10 and 11. Figure 5 As can be seen, the difference in growth predictions is relatively small compared to the case of boys.

[0128] Therefore, for obese girls, it is necessary to provide a solution for increasing the duration of the rapid growth phase (303) to reduce the amount of growth rate reduction that occurs when transitioning from the rapid growth phase to the slow growth phase.

[0129] Figure 7 is a diagram showing the structure of a neural network for growth prediction and solution provision according to an exemplary embodiment of the present invention, Figure 8 is a diagram showing a first neural network model according to an exemplary embodiment of the present invention, Fig. 9 is a diagram illustrating a second neural network model according to an exemplary embodiment of the present invention.

[0130] Below, refer to Figures 7 to 9 Provide explanation.

[0131] An exemplary embodiment of the present invention may include a first model 50 and a second model 13 , and may construct a pipeline in which at least a portion of an output of the second model 13 is input to the first model 50 .

[0132] More specifically, the first model 50 is a model that learns body data corresponding to at least one of a plurality of growth stages based on time-series body data of a plurality of sample subjects as learning data.

[0133] The growth stage may be a rapid growth stage (303) which is a period when growth begins to slow down due to obesity, but as described above, any one or more of a plurality of growth stages may be used in order to more accurately predict obesity.

[0134] The first model 50 includes LSTM neural networks 50-1, 50-2, 50-3, and 50-4 for learning time series data, and uses the past physical data of multiple sample subjects to train the LSTM neural networks 50-1, 50-2, 50-3, and 50-4. Then, the physical data of the current subject is input into the trained LSTM neural networks 50-1, 50-2, 50-3, and 50-4, and the predicted growth rate and the solution considering the growth rate are output.

[0135] The subject's body data may refer to data refined by the data preprocessing unit 100 after the error judgment process, which will be processed by the following Fig.10 and Fig.11 Provide detailed explanation.

[0136] LSTM neural networks 50-1, 50-2, 50-3, and 50-4 are learned using at least one of the body data of multiple sample subjects as a default value. For example, for height, training is performed using height data of each year in an arbitrary period or a specific growth stage, and predictions for the next year are made and compared with the actual data. By comparing in this way, the training set is moved to the future in units of an arbitrary period or a specific growth stage and learned.

[0137] In addition, the LSTM neural networks 50-1, 50-2, 50-3, and 50-4 can be learned for each growth stage. Therefore, each of the general growth stage (301), the rapid growth stage (303), the slow growth stage (305), and the growth plate-free stage (307) can be learned as past body data of the corresponding growth stage.

[0138] The body data of the subject used for the study may refer to the data refined by the data preprocessing unit 100 after the error judgment process, which will be processed by the following method. Fig.10 and Fig.11 Provide detailed explanation.

[0139] In addition, exemplarily, in this embodiment, time series physical data of multiple sample subjects can be input in sequence according to age in months or any period as learning data, and the calculation results of the predicted values ​​of the past time points or growth rates can be transmitted to the growth rate prediction of the next age in months or any period.

[0140] Therefore, the LSTM neural networks 50-1, 50-2, 50-3, and 50-4 can not only predict the growth rate based on the current physical data, but can also learn the extent to which the prediction results of various indicators at past time points affect the current growth rate prediction. Thus, the items that have the greatest impact on the growth rate change can be extracted based on the age in months or any period in the indicator and reflected in the growth rate prediction.

[0141] In addition, for time series learning, it is necessary to obtain the physical data of multiple sample subjects regularly. However, as mentioned above, it may be difficult to obtain the physical data of multiple sample subjects regularly in months or arbitrary periods. Therefore, it is possible to remove the abnormal value physical data or non-continuous physical data of each unit period and normalize them in time for use.

[0142] In addition, the second model 13 can derive bone maturity (age in months) from the hand bone image using a convolutional neural network (Convolution Neural Network) learned using the subject's bone maturity data as learning data.

[0143] More specifically, the convolutional neural network includes multiple convolution layers and pooling layers, wherein the multiple convolution layers extract feature maps of features in the image of the subject to be analyzed in the hand bone image, and the pooling layer performs sub-sampling between the multiple convolution layers, so that features of different levels of the subject area can be extracted and analyzed, and features can be probabilistically inferred through activation functions, or regression analysis can be performed to derive bone maturity through weight learning between nodes.

[0144] The bone maturity extracted by the second model 13 is input into the LSTM neural networks 50-1, 50-2, 50-3, and 50-4 together with at least a portion of the subject's physical data, which can improve the accuracy of the subject's growth rate prediction and thus further improve the accuracy of obesity prediction.

[0145] Figures 10 to 12 is a diagram illustrating a preprocessing method of biological component measurement data for growth prediction according to an exemplary embodiment of the present invention.

[0146] As described above, the physical data input when predicting the growth of a subject (child or adolescent) or providing a solution through growth prediction may have errors in the measurement process or data transmission.

[0147] Therefore, in order to more accurately predict the growth of the subject or provide a solution through growth prediction, the refined data that has undergone the above-mentioned process of detecting and judging erroneous data needs to be input into the input unit 10 .

[0148] In addition, the physical data of the plurality of sample subjects used to train the prediction model of the prediction unit 50 may also use refined data that has undergone the above-mentioned process of detecting and judging erroneous data.

[0149] Below, refer to Figure 3 , Figures 10 to 14 Describe the process of detecting and judging errors in physical data.

[0150] The exemplary embodiment of the present invention can generate a virtual first variable (S30) based on the input body data after receiving the body data (S10) in order to detect or determine errors in the body data, especially the biological component data of the subject. In addition, the generated virtual first variable can be compared with the body data input in the step (S10) (S50) to determine errors in the input body data (S70).

[0151] More specifically, if Fig.14 As shown, the subject's body data may include identification data 1201 and biological component data 1101, and the preprocessing unit 100 may obtain the subject's biological component data 1101 (S111) by inputting through the biological component data input unit 110, and may obtain the subject's identification data 1201 (S112) by inputting through the identification data input unit 120. In addition, the input biological component data 1101 and identification data 1201 may be generated as concatenated data 140d by the connection unit 140.

[0152] When the connected data 140d is determined to have detected an error in the biological component data 1101 input after the error detection unit 130 detects an error in the biological component data (S131) ​​(S131, yes), at least a portion of the biological component data constituting the connected data 140b may be deleted (S132).

[0153] Exemplarily, when at least one of the plurality of items 1101-1 to 1101-9 constituting the input biological component data 1101 includes a "Nan" value, the data preprocessing unit 100 of this embodiment may delete the biological component data 1101 through the data deleting unit 180. In this case, based on the measurement date 1201-4 of the identification data, all items (rows including the measurement date) measured on the measurement date including the above-mentioned "Nan" value may be deleted.

[0154] Alternatively, when the age in months 1201-3 of the identification data is less than or equal to 30 months, all items measured on the measurement date containing the 'Nan' value (rows containing the measurement date) can be deleted. In more detail, when the age in months is less than or equal to 30 months, this is a period of rapid growth for infants, and there may be many errors in the measurement. Therefore, the reliability of the biological component data is lower than that of the biological component data with a relatively high age. Therefore, as described above, if the biological component data contains the "Nan" value when the age in months is less than or equal to 30 months, all items measured on the measurement date containing the "Nan" value can be deleted.

[0155] The above-mentioned deletion of the biological component data is exemplary, and when an error (Nan value) is included in the items constituting the biological component data according to various standards, the error may be detected and deleted.

[0156] If no error is detected in the biological component data 1101 (S131: No), a dummy first variable may be generated by the variable generating unit 150 (S310).

[0157] The first variable may be generated as the sum of body fat mass 1101-6 (Body Fat Mass), muscle mass 1101-7 (Soft Lean Mass) and bone mineral mass 1101-8 (Osseous mineral). For reference, muscle mass 1101-7 may be defined as the sum of non-osseous mineral except for body water, protein mass 1101-4 and bone.

[0158] As described above, errors in the input biological component measurement data 1101 can be determined by comparing (S50) the first variable with the data input in the steps (S111, S112) after generating the virtual first variable (S310).

[0159] More specifically, after first determining whether the subject's age in months 1201-3 meets a preset standard in the subject's identification data 1201, it is determined whether to delete the biological component data based on the difference between the first variable and the weight 1101-3 in the biological component data 1101 according to the preset standard, thereby generating refined data (second data).

[0160] The biological composition data 1101 input in the step (S111) is usually measured by bioelectrical impedance analysis (BIA), which is a method of measuring body composition based on the speed of electric current passing through the body.

[0161] Since the biological component data measured in this way have a greater possibility of error in the measurement process, the younger the subject's age is, the greater the range of variation in body composition is, therefore, the biological component measurement data preprocessing method of this embodiment can determine whether the subject's age is included in multiple preset standards, and then determine the error of the biological component measurement data under different conditions according to each preset standard.

[0162] For example, in this embodiment, if the difference between the generated first variable and the weight 1101-3 is greater than or equal to the preset range, it is determined that an error has occurred in the measurement value generated for the first variable. At this time, refined data can be generated by deleting the row containing the measurement value generated for the first variable from the connected data 140d.

[0163] As described above, the virtual first variable is compared with the weight data because theoretically the weight data 1101-3 is the same as the first variable, but due to various variables, a difference between the weight data 1101-3 and the first variable may occur. Therefore, when the difference between the weight data 1101-3 and the first variable exceeds a preset range, the measurement value in the row containing the measurement value generating the first variable is judged as an error and deleted.

[0164] Next, a process of determining an error in the above-mentioned biological component measurement data by the error determination unit 170 will be described.

[0165] After the first variable is generated in the step (S310), it can be determined whether the subject's age in months is less than or equal to a first standard (S511), and the first standard can be set to 60 months.

[0166] If the subject's age is less than or equal to the first standard (S511: Yes), it can be determined whether the difference between the first variable and a value in the biological component data is greater than or equal to 20% (S513). A value in the biological component data can be set as weight data 1101-3.

[0167] If the difference between the first variable and the weight data 1101 - 3 is greater than or equal to 20% ( S513 : Yes), the biological component data is deleted by the data deletion unit 180 ( S711 ), and the refined second data is generated by the second data generation unit 190 ( S712 ).

[0168] In addition, if the difference between the first variable and the weight data (1101-3) is not greater than 20% (S513: No), the second data generating unit 190 may generate refined second data (S712).

[0169] In addition, if the subject's age in months exceeds the first standard (S511: No), it can be determined whether the subject's age in months exceeds the first standard and falls within a range below the second standard (S521), and the second standard can be set to 75 months.

[0170] If the age of the subject falls between the first standard and the second standard (S511: Yes), it can be determined whether the difference between the first variable and a value in the biological component data is greater than or equal to 15% (S523).

[0171] Since the age of the subject that is misjudged in step (S523) is higher than the age of the subject that is misjudged in step (S513), the range of change in body composition is relatively small.

[0172] Therefore, the standard for judging the error of the biological component data entered in the step (S523) (Is the difference between the first variable and the weight data greater than or equal to 15%?) can be set higher than the standard for judging the error of the biological component data entered in the step (513) (Is the difference between the first variable and the weight data greater than or equal to 20%?).

[0173] Setting the standard higher may mean that, in this embodiment, the range for judging the difference between the first variable and the weight data as an error is narrower. That is, in the preprocessing method of biological component measurement data of this embodiment, as the age of the subject increases, the standard for judging an error can be set higher.

[0174] If the difference between the first variable and the weight data 1101 - 3 is greater than or equal to 15% ( S523 : Yes), the biological component data may be deleted by the data deleting unit 180 ( S721 ), and the refined second data may be generated by the second data generating unit 190 ( S722 ).

[0175] In addition, if the difference between the first variable and the weight data 1101 - 3 is not greater than or equal to 15% ( S523 : No), the second data generating unit 190 may generate refined second data ( S723 ).

[0176] In addition, if the subject's age is not between the first standard and the second standard (S521: No), it can be determined whether the subject's age exceeds the second standard and is within the range below the third standard (S531), and the third standard can be set to 100 months.

[0177] If the age of the subject falls between the second standard and the third standard (S531: Yes), it can be determined whether the difference between the first variable and a value in the biological component data is greater than or equal to 10%.

[0178] If the difference between the first variable and the weight data 1101 - 3 is greater than or equal to 10% ( S533 : Yes), the biological component data may be deleted by the data deletion unit 180 ( S731 ), and the refined second data may be generated by the second data generation unit 190 ( S732 ).

[0179] In addition, if the difference between the first variable and the weight data 1101 - 3 is not greater than or equal to 10% ( S533 : No), the second data generating unit 190 may generate refined second data ( S733 ).

[0180] In addition, if the age of the subject does not fall between the second standard and the third standard (S531: No), it can be determined whether the age of the subject exceeds the third standard (S541).

[0181] If the age of the subject exceeds the third criterion (S541: Yes), it can be determined whether the difference between the first variable and a value in the biological component data is greater than or equal to 3% (S543).

[0182] If the difference between the first variable and the weight data 1101 - 3 is greater than or equal to 3% ( S543 : Yes), the biological component data may be deleted by the data deleting unit 180 ( S741 ), and refined second data may be generated by the second data generating unit 190 ( S742 ).

[0183] In addition, if the difference between the first variable and the weight data 1101 - 3 is not greater than or equal to 3% ( S543 : No), the second data generating unit 190 may generate refined second data ( S743 ).

[0184] In addition, if the subject's age in months does not exceed the third standard (S541: No), there is an error in the subject's age in months data 1201-3, or an error occurs in the process of determining whether the subject's age in months data 1201-3 belongs to the range defined by the first standard to the third standard, and therefore, the process can be terminated. Alternatively, the process of determining whether the subject's age in months is below the first standard (S511) can be repeated.

[0185] The second data generated as described above is refined data that has been subjected to a process of judging errors based on the biological component data 1101 and the identification data 1201 of the input subject, and therefore can be used as input data for predicting growth or generating a solution by predicting growth. At this time, the second data can be used as input to Figure 2 The data of the input unit 10.

[0186] In addition, the second data can be used as learning data for predicting growth or a model for predicting growth. In this case, the second data can be used as input to Figure 7 The data of the prediction model in .

[0187] The preferred embodiments of the present invention are described above. All embodiments and condition examples disclosed in this specification are intended to help those skilled in the art understand the principles and concepts of the present invention. Therefore, it should be understood that those skilled in the art can implement variations within the scope of the essential features of the present invention.

[0188] Therefore, the disclosed embodiments should be considered from an illustrative point of view rather than a restrictive point of view. The scope of the present invention should be determined by the scope of the claims rather than the above description, and all differences within the scope of equivalence thereof should be included in the present invention.

[0189] In addition, the methods according to various embodiments of the present invention can be implemented by a program and then provided to a server or device. Therefore, each device can connect to a server or device storing a program and download the program.

[0190] In addition, the above-mentioned methods according to various embodiments of the present invention can be implemented by programs and stored in various non-transitory computer readable media (non-transitory computer readable medium). Non-transitory computer readable media refers to media that semi-permanently stores data and can be read by a device, rather than media that store data for a short time such as registers, caches, and memories. Specifically, the above-mentioned various applications or programs can be stored and provided on non-transitory computer readable media, such as CDs, DVDs, hard disks, Blu-ray discs, USBs, memory cards, ROMs, etc.

[0191] Although preferred embodiments of the present invention are illustrated above, the present invention is not limited to the above-mentioned specific embodiments, and various modifications may be made by a person skilled in the art without departing from the spirit of the present invention as claimed in the claims, and these modifications should not be understood separately from the technical ideas or prospects of the present invention.

Claims

1. A method for preprocessing biological component measurement data for predicting growth, which is a method for preprocessing biological component measurement data for predicting growth performed by a computing device, characterized in that: include: The step of inputting the physical data of the subject; The step of generating a first variable based on the physical data of the subject; as well as The step of comparing the first variable with a preset value to determine errors in the body data.

2. The method for preprocessing biological component measurement data for predicting growth according to claim 1, characterized in that: The step of inputting the physical data of the subject comprises: A step of inputting biological component data of the subject; and The step of obtaining identification data of the subject is input.

3. The method for preprocessing biological component measurement data for predicting growth according to claim 2, characterized in that: The method further includes the step of connecting the biological component data of the subject and the identification data of the subject to generate first data after the physical data of the subject is input.

4. The method for preprocessing biological component measurement data for predicting growth according to claim 3, characterized in that: The method further includes a step of deleting the biological component data if a preset value is measured in the biological component data after the first data is generated.

5. The method for preprocessing biological component measurement data for predicting growth according to claim 3, characterized in that: The first variable is generated based on biological composition data of the subject.

6. The method for preprocessing biological component measurement data for predicting growth according to claim 5, characterized in that: The first variable is generated as the sum of body fat mass, muscle mass, and bone mineral mass in the biological composition data of the subject.

7. The method for preprocessing biological component measurement data for predicting growth according to claim 3, characterized in that: The step of determining an error includes the step of comparing the first variable with a value in the biological component data.

8. The method for preprocessing biological component measurement data for predicting growth according to claim 7, characterized in that: The error determination step also includes a step of determining whether the age of the subject in the identification data meets a preset standard before comparing the first variable with a value in the biological component data.

9. The method for preprocessing biological component measurement data for predicting growth according to claim 7, characterized in that: The method further includes the step of deleting the biological component data when a result value of comparing the first variable with a value in the biological component data falls within a preset range.

10. The method for preprocessing biological component measurement data for predicting growth according to claim 7, characterized in that: The method further includes generating second data when a result value of comparing the first variable with a value in the biological component data does not fall within a preset range.

11. A program stored in a computer-readable recording medium including a program code for executing the method for preprocessing biological component measurement data for predicting growth according to any one of claims 1 to 10.

12. A computer-readable recording medium having recorded therein a program for executing the method for preprocessing biological component measurement data for predicting growth according to any one of claims 1 to 10.

13. A preprocessing device for biological component measurement data for predicting growth, which is a preprocessing device for biological component measurement data for predicting growth, characterized in that: include: An input unit for inputting the physical data of the subject; A variable generating unit, which generates a first variable based on the physical data of the subject; as well as An error judging unit is configured to judge an error of the physical data by comparing the first variable with a preset value.

14. The preprocessing device for biological component measurement data for predicting growth according to claim 13, characterized in that: The input unit comprises: A first input unit for inputting biological component data of the subject; and The second input unit receives the identification data of the subject.

15. The preprocessing device for biological component measurement data for predicting growth according to claim 14, characterized in that: The device further includes a connection unit for connecting data inputted through the first input unit and the second input unit.

Images