KR20240059583A - Method and system for providing weight management service based on artificial intelligence using biometric information - Google Patents

Abstract

A weight management service providing system according to an aspect according to the technical idea of the present disclosure includes a communication unit that receives biometric information of the user from the user's terminal or data collection device; And based on the received biometric information, obtain at least one of the weight change prediction models of a plurality of other users registered in the weight management service as a weight change prediction model for the user, and use the obtained weight change prediction model as the basis. and at least one processor that controls the communication unit to generate a blood sugar management guide corresponding to the user's target weight loss information and transmit the generated blood sugar management guide to the user's terminal, and the weight change prediction model is , It includes a machine learning-based model that is learned using a setting variable that includes at least blood sugar characteristic information and a response variable that includes weight change amount information.

Description

Method and system for providing artificial intelligence-based weight management service using biometric information {METHOD AND SYSTEM FOR PROVIDING WEIGHT MANAGEMENT SERVICE BASED ON ARTIFICIAL INTELLIGENCE USING BIOMETRIC INFORMATION}

The technical idea of this disclosure relates to a method and system for providing weight management services, and in particular, relates to a method and system for providing weight management services based on artificial intelligence using biometric information such as the user's blood sugar or ketone levels. .

Obesity refers to a condition in which excessive fat tissue is accumulated in the body, and its prevalence is increasing due to modern people's lack of exercise, stress, and increased intake of animal products. Obesity is one of the causes of various diseases such as diabetes, heart disease, and high blood pressure, and awareness that obesity itself is a fatal disease is gradually spreading. Accordingly, interest and efforts in weight management are increasing throughout society, and various weight management methods are emerging, such as management through diet or exercise, consumption of diet foods, and lipolysis or suction procedures.

However, conventional weight management methods may incur excessive costs or may not be sufficiently scientifically proven, which may lead to risks such as side effects. In particular, there is a problem in that large deviations may occur depending on the individual's biological characteristics. In other words, for more effective weight management for each individual, it may be important to determine what lifestyle habits or biological characteristics affect weight management for each individual.

Meanwhile, in recent years, continuous blood sugar measurement techniques have been developed and used to measure blood sugar periodically or continuously by attaching a device to a part of the body when measuring the user's blood sugar. However, in the past, continuous blood sugar measurement techniques were mostly used only for blood sugar management or diabetes. There were limitations to its use. Accordingly, research has recently been conducted to expand the scope of use, and in particular, various studies related to weight loss or obesity prevention through management of blood sugar levels are being conducted. Blood sugar, or glucose in the blood vessels, is our body's main energy source and is mainly stored in the liver and muscles. However, excess glucose that exceeds the storage capacity of the liver and muscles is accumulated as fat in the body, causing weight gain, and insulin, which is oversecreted when blood sugar levels rise rapidly, also promotes fat accumulation and increases weight. Therefore, appropriate management of blood sugar levels can help prevent weight loss or obesity by preventing an increase in body fat and also encouraging the use of fat accumulated in the body.

Additionally, weight management methods such as the ketogenic diet or intermittent fasting, which reduce body fat by putting the human body into a state of ketosis, have recently been attracting attention. However, the timing of entering the ketosis state may vary depending on each individual's physical characteristics and constitution, and if ketone levels increase excessively in the ketosis state, side effects such as ketoacidosis may occur.

1. Korean Patent Publication No. 10-1107062 (registered on January 11, 2012) 2. US published patent US 2018/0108272 (published on 2020.06.09) 3. Korea Patent Publication No. 10-2021-0062291 (published on May 31, 2021)

The problem that the present invention seeks to solve is to implement a weight management service that provides guidance from a blood sugar perspective for customized weight management for each user.

In addition, the problem that the present invention seeks to solve is to implement a method that can smoothly provide weight management services to new users who do not have a machine learning model for providing weight management services.

In order to achieve the above object, a weight management service providing system according to an aspect according to the technical idea of the present disclosure includes a communication unit that receives the user's biometric information from the user's terminal or data collection device; And based on the received biometric information, obtain at least one of the weight change prediction models of a plurality of other users registered in the weight management service as a weight change prediction model for the user, and use the obtained weight change prediction model as the basis. and at least one processor that controls the communication unit to generate a blood sugar management guide corresponding to the user's target weight loss information and transmit the generated blood sugar management guide to the user's terminal, and the weight change prediction model is , It includes a machine learning-based model that is learned using a setting variable that includes at least blood sugar characteristic information and a response variable that includes weight change amount information.

According to one embodiment, the user's biometric information includes the user's physical characteristic information and blood sugar data, and the physical characteristic information includes at least one of age, gender, height, weight, and whether or not a chronic disease occurs. It can be included.

According to one embodiment, the at least one processor extracts other users whose biometric information satisfies predefined similar judgment criteria with the user's biometric information from a database, and applies a classification model to the extracted other users. By applying the method to select a weight change prediction model for each of a predefined number of other users, a weight change prediction model for the user can be obtained.

According to one embodiment, the at least one processor obtains a plurality of weight prediction changes for a plurality of different blood sugar characteristic information using the obtained weight change prediction model, and each of the plurality of weight prediction changes is By comparing the target weight loss information, blood sugar characteristic information corresponding to the target weight loss information can be identified, and the blood sugar management guide can be generated based on the identified blood sugar characteristic information.

According to one embodiment, the target weight loss information includes a target weight loss and a target weight loss period, and the at least one processor converts each of the plurality of predicted weight change amounts to correspond to the target weight loss period to obtain the target weight loss. , and as a result of the comparison, blood sugar characteristic information corresponding to the predicted change in weight that has the smallest difference from the target weight loss can be identified as blood sugar characteristic information corresponding to the target weight loss information.

According to one embodiment, the blood sugar characteristic information may include at least one of a fasting blood sugar maintenance time, a blood sugar spike period, an excess fasting blood sugar period, and an under-fasting blood sugar period.

According to one embodiment, after providing the blood sugar management guide, the at least one processor collects the user's blood sugar data, monitors the user's blood sugar management status based on the collected blood sugar data, and reports the monitoring results. Based on this, the blood sugar management guide can be updated.

A method of providing a weight management service according to an aspect according to the technical idea of the present disclosure includes: acquiring, by a biometric information collection unit included in the at least one computing device, biometric information of a user; A prediction model acquisition unit included in the at least one computing device, based on the acquired biometric information, calculates at least one of weight change prediction models of other users registered in the weight management service as a weight change prediction model for the user. acquiring; generating, by a blood sugar management guide providing unit included in the at least one computing device, a blood sugar management guide corresponding to the user's target weight loss information based on an obtained weight change prediction model; and providing, by the blood sugar management guide providing unit, the generated blood sugar management guide to the user, wherein the weight change prediction model includes a setting variable including at least blood sugar characteristic information and a response variable including weight change amount information. It includes a machine learning-based model that is learned using .

The method and system for providing a weight management service according to the technical idea of the present disclosure provides a customized guide from the perspective of blood sugar for weight management for each user by learning the correlation between the user's blood sugar characteristics and weight change, compared to conventional weight management. Compared to other methods, side effects can be reduced and weight management effects can be maximized.

In addition, according to the technical idea of the present disclosure, even for new users who do not have a machine learning model for providing weight management services, the service is smoothly provided using models of users with similar biological conditions among already registered users. can do.

In addition, the weight management service provision method and system according to the technical idea of the present disclosure creates and provides a weight management guide (diet plan, etc.) that reflects the user's optimal fasting time, thereby maximizing the user's weight management effect and weight management. Side effects of poetry can be minimized.

The effects according to the technical idea of the present disclosure are not limited to the effects mentioned above, and other effects not mentioned can be clearly understood by those skilled in the art from the description below. There will be.

In order to more fully understand the drawings cited in this disclosure, a brief description of each drawing is provided.
1 is a conceptual diagram including a weight management service providing system and devices connected thereto according to an exemplary embodiment of the present disclosure.
FIG. 2 is a block diagram schematically showing the configuration of the weight management service providing system shown in FIG. 1.
Figure 3 is a flowchart for explaining a method of providing a weight management service according to an exemplary embodiment of the present disclosure.
Figures 4 and 5 are diagrams for explaining an embodiment of a process in which a weight change prediction model for each user is generated.
FIG. 6 is a diagram for explaining a specific example of the method for providing weight management services described above in FIG. 3.
FIG. 7 is a diagram illustrating a specific example of a classification model that selects a weight change prediction model for a user from weight change prediction models of other users.
Figure 8 is a flowchart for explaining a method of providing a weight management service according to an exemplary embodiment of the present disclosure.
Figure 9 is a diagram to explain an embodiment of the process of generating a ketone level prediction model for each user.
FIG. 10 is a diagram for explaining a specific example of the weight management service provision method described above in FIG. 8.
Figure 11 shows an example of a ketone level prediction graph over time obtained from a ketone level prediction model selected for a user.
FIG. 12 is a block diagram schematically showing the hardware configuration of a device included in the weight management service providing system shown in FIG. 1.

Illustrative embodiments according to the technical idea of the present disclosure are provided to more completely explain the technical idea of the present disclosure to those skilled in the art, and the examples below can be modified into various other forms. may be, and the scope of the technical idea of the present disclosure is not limited to the examples below. Rather, these embodiments are provided to make the present disclosure more faithful and complete and to completely convey the technical idea of the present invention to those skilled in the art.

Although the terms first, second, etc. are used in this disclosure to describe various members, regions, layers, portions, and/or components, these members, parts, regions, layers, portions, and/or components are referred to by these terms. It is obvious that it should not be limited by . These terms do not imply any particular order, superiority, inferiority, or superiority or inferiority, and are used only to distinguish one member, region, region, or component from another member, region, region, or component. Accordingly, the first member, region, portion, or component described in detail below may refer to the second member, region, portion, or component without departing from the teachings of the technical idea of the present disclosure. For example, a first component may be referred to as a second component without departing from the scope of the present disclosure, and similarly, the second component may also be referred to as a first component.

Unless otherwise defined, all terms used herein, including technical terms and scientific terms, have the same meaning as commonly understood by those skilled in the art in the technical field to which the concept of the present disclosure pertains. Additionally, commonly used terms, as defined in dictionaries, should be interpreted to have meanings consistent with what they mean in the context of the relevant technology, and should not be used in an overly formal sense unless explicitly defined herein. It should not be interpreted.

In cases where an embodiment can be implemented differently, a specific process sequence or process sequence may be performed differently from the described sequence. For example, two processes or processes described in succession may be performed substantially simultaneously, or may be performed in an order opposite to the order in which they are described.

In addition, terms such as “unit”, “unit”, “unit”, and “module” used in this specification refer to a unit that processes at least one function or operation, which refers to a processor, micro Processor (Micro Processer), Micro Controller, CPU (Central Processing Unit), GPU (Graphics Processing Unit), APU (Accelerate Processor Unit), DSP (Drive Signal Processor), ASIC (Application Specific Integrated Circuit), FPGA It may be implemented as hardware or software, such as a Field Programmable Gate Array, or a combination of hardware and software, and may also be implemented in a form combined with a memory that stores data necessary for processing at least one function or operation. .

Additionally, we would like to clarify that the division of components in this specification is merely a division according to the main function each component is responsible for. That is, two or more components, which will be described below, may be combined into one component, or one component may be divided into two or more components for more detailed functions. In addition to the main functions it is responsible for, each of the components described below may additionally perform some or all of the functions handled by other components, and some of the main functions handled by each component may be performed by other components. Of course, it can also be carried out exclusively by .

As used herein, the term 'and/or' includes each and every combination of one or more of the mentioned elements.

Hereinafter, embodiments based on the technical idea of the present disclosure will be described in detail with reference to the attached drawings.

1 is a conceptual diagram including a weight management service providing system and devices connected thereto according to an exemplary embodiment of the present disclosure.

The weight management service provision system 10 according to an embodiment of the present disclosure acquires an artificial intelligence-based prediction model, such as predicting a change in the user's weight or predicting a change in ketone levels, based on the user's blood sugar data and body characteristic information. And, using the obtained prediction model, services for weight management can be provided.

In FIG. 1, the weight management service providing system 10 is shown as consisting of one server, but the weight management service providing system 10 includes a plurality of computing devices (PCs, servers, data centers, etc.) connected to each other. It may be configured to include.

Meanwhile, when a new user wants to use this service, it may be difficult to obtain accurate prediction models for the user until sufficient data about the new user is secured. Therefore, in order to effectively attract service users, a method that can provide a smooth service may be required even before sufficient data on new users or users for whom a prediction model does not exist is secured. Accordingly, the weight management service providing system 10 can provide a smooth service even to users (such as new users) for whom a prediction model does not exist.

The specific configuration and operation of the weight management service providing system 10 will be described in detail later with reference to FIGS. 2 to 11.

The terminal 20 receives various data related to the use of the weight management service from the user or the data collection device 30 and transmits them to the weight management service providing system 10, or various data provided from the weight management service providing system 10. It may correspond to a device that outputs information. For example, an application for using services provided by the weight management service providing system 10 may be installed on the terminal 20 . The user can use the above-mentioned service by downloading and installing the application and executing the installed application. This terminal 20 may be implemented as an electronic device such as a smartphone, tablet PC, laptop, or PC.

According to one embodiment, the terminal 20 collects the user's personal information (ID, name, date of birth, etc.) and physical information (gender, age, height, weight, body mass index (BMI), skeletal muscle mass, body water, abdominal fat mass, abdominal Fat percentage, basal metabolic rate, waist circumference, occurrence of chronic diseases (diabetes, high blood pressure, dyslipidemia, etc.) can be obtained. For example, the personal information and physical information may be input directly from the user through the application, etc., or may be received from another server (e.g., health checkup data server). Additionally, in order to provide a weight management service to the user, the terminal 20 may receive target weight loss information including target weight loss and target weight loss period from the user.

Depending on the embodiment, the terminal 20 may obtain data (meal event data) related to the user's meal event. For example, the meal event data may include image data (or identification data such as a barcode, QR code, food name text, etc.) of food consumed or scheduled to be consumed by the user. As another example, the meal event data may include food intake information directly input by the user through the input means of the terminal 20. For example, the food intake information may include food name, intake start time, intake end time, amount consumed, energy (calories), and/or nutrient content (carbohydrates, fat, protein, sugar, sodium, cholesterol, etc.).

At least one data collection device 30 provides data for acquiring a user's weight change prediction model and/or a ketone level prediction model, or predicts a user's weight change through an obtained weight change prediction model or creates a ketone level prediction model. It may include a device that collects the data necessary to predict the user's ketone levels. For example, the data collection device 30 may include a biometric data measuring device 31, a wearable device 32, a scale 33, etc. Depending on the embodiment, some of the user's physical information or other information related to the user may be collected by the at least one data collection device 30.

The biometric data measuring device 31 may correspond to a device for measuring a user's biometric data. According to one embodiment, the biometric data measuring device 31 may measure the user's blood sugar level. In this case, the biometric data measuring device 31 may include a continuous glucose monitor that is attached to the user's body and acquires blood sugar data by periodically or continuously measuring blood sugar over a predetermined period of time. For example, continuous blood sugar data obtained through a continuous blood glucose meter may include blood sugar data for various states, such as fasting blood sugar data obtained when the user is in a fasting state and postprandial blood sugar data that changes depending on food intake. The biometric data measuring device 31 may transmit the acquired blood sugar data to the weight management service providing system 10 through the terminal 20 or a separate communication device (router, repeater, base station, etc.). Depending on the embodiment, the user's fasting blood sugar data among the blood sugar data may be directly input by the user through the terminal 20.

According to one embodiment, the biometric data measuring device 31 may measure the user's ketone level. In this case, the biometric data measuring device 31 may include a conventional ketone measuring device that measures ketone levels from the user's exhaled breath or blood. Depending on the embodiment, the biometric data measuring device 31 may be a device that is attached to the user's body and acquires ketone data by periodically or continuously measuring ketone levels in interstitial fluid, etc., over a predetermined period of time. Depending on the embodiment, ketone data may be directly input by the user through the terminal 20.

Depending on the embodiment, the biometric data measuring device 31 may be implemented as a single device capable of measuring blood sugar and ketone levels, or may be implemented as a device integrated with the wearable device 32.

The wearable device 32 is a type of device worn on a specific location of the user's body (wrist, etc.), and is at least one device for detecting various data such as the user's heart rate, number of steps, location information, body temperature, oxygen saturation, and sleep time. may include sensors. According to an embodiment of the present disclosure, the wearable device 32 detects whether an exercise event has occurred from data acquired through the at least one sensor, and transmits the exercise event data generated according to the detection result to the terminal 20 or a separate device. It can be transmitted to the weight management service providing system 10 through a communication device (router, repeater, base station, etc.). For example, the exercise event data may include exercise type, exercise type (aerobic/anaerobic, etc.), exercise frequency, exercise time (exercise start time, exercise end time), and exercise intensity. Depending on the embodiment, at least some of the exercise event data may be input by the user through an input means of the terminal 20 or the wearable device 32.

In another embodiment, the wearable device 32 transmits data acquired through the at least one sensor to the terminal 20, and the terminal 20 detects whether an exercise event has occurred through analysis of the received data and detects the occurrence of the exercise event. Exercise event data can also be generated. Alternatively, data acquired through the at least one sensor may be transmitted to the weight management service providing system 10, and the weight management service providing system 10 may analyze the data and generate the exercise event data.

As another example, the wearable device 32 may further include a sensor that measures the user's blood sugar or ketone levels.

The scale 33 may acquire the user's weight data and transmit the obtained weight data to the weight management service providing system 10 through the terminal 20 or a separate communication device. Depending on the embodiment, the weight data may further include data such as body fat percentage, muscle mass, and/or water content in addition to the user's weight, and specifically, body fat mass, skeletal muscle mass, body water, body cell, abdominal fat mass, abdominal fat ratio, and basal body fat mass. It may include data such as metabolic rate.

FIG. 2 is a block diagram schematically showing the configuration of the weight management service providing system shown in FIG. 1.

Referring to FIG. 2, the weight management service providing system 10 includes a biometric information collection unit 110, a prediction model acquisition unit 120, a blood sugar management guide providing unit 130, a weight management guide providing unit 140, and It may include a database 150. Depending on the embodiment, the weight management service providing system 10 may include more or fewer components than those shown in FIG. 2 .

The biometric information collection unit 110 may collect the user's biometric information from the user's terminal 20 and/or the data collection device 30. For example, the biometric information may include blood sugar data (continuous blood sugar data, etc.), ketone data, weight data, and/or physical characteristic information (age, gender, height, occurrence of chronic diseases (diabetes, high blood pressure, dyslipidemia, etc.), etc. ), etc. may be included.

Depending on the embodiment, the biometric information collection unit 110 may receive the user's personal information, meal event data, and/or exercise event data from the terminal 20 and/or the data collection device 30.

According to the embodiment, the biometric information collection unit 110 divides the biometric information received for multiple users into databases 150 and provides them to the database 150, thereby allowing the database 150 to store and manage biometric information for each user. You can do it.

The prediction model acquisition unit 120 may obtain a weight change prediction model for the user based on the user's biometric information and/or personal information provided from the biometric information collection unit 110. According to one embodiment, the prediction model acquisition unit 120 may check whether a weight change prediction model for the user exists in the database 150 based on the user's personal information.

If a weight change prediction model for the user exists in the database 150, the prediction model acquisition unit 120 may load the corresponding weight change prediction model from the database 150. On the other hand, if the weight change prediction model for the user does not exist in the database 150, the prediction model acquisition unit 120 creates weight change prediction models for each of the other pre-registered users based on the user's biometric information. By selecting at least one of the above, a weight change prediction model for the user can be obtained.

Meanwhile, the weight change prediction model for each user can be generated through machine learning techniques, and the creation of the weight change prediction model will be explained later with reference to FIGS. 4 and 5.

Additionally, the prediction model acquisition unit 120 may obtain a ketone level prediction model for the user based on biometric information provided from the biometric information collection unit 110, similar to the weight change prediction model. The ketone level prediction model can be created through machine learning techniques and can predict changes in ketone levels according to the user's fasting time. The creation of the ketone level prediction model will be explained later with reference to FIG. 9.

The blood sugar management guide provider 130 may generate a blood sugar management guide corresponding to the user's target weight loss information based on a weight change prediction model obtained for the user and provide the blood sugar management guide to the user. The weight management guide providing unit 140 estimates the optimal fasting time for the user based on the ketone level prediction model obtained for the user, generates a weight management guide reflecting the estimated optimal fasting time, and provides it to the user. can do. Depending on the embodiment, the blood sugar management guide providing unit 130 and/or the weight management guide providing unit 140 may provide diagnostic information or health information (tips or common sense for health management, etc.) for weight management or health management of the user. It can also be provided together.

The database 150 stores various data/information received by the weight management service provision system 10 separately for each user, or various data, information, and algorithms (prediction models, etc.) generated by the weight management service provision system 10. can be saved. Depending on the embodiment, the database 150 may correspond to a third party's database server, data center, etc., and the weight management service providing system 10 may be connected to the database server or data center.

Through the following drawings, a method for providing weight management services according to various embodiments of the present disclosure will be described in detail.

Figure 3 is a flowchart for explaining a method of providing a weight management service using blood sugar data according to an exemplary embodiment of the present disclosure. Figures 4 and 5 are diagrams for explaining an embodiment of a process in which a weight change prediction model for each user is generated. FIG. 6 is a diagram for explaining a specific example of the method for providing weight management services described above in FIG. 3. FIG. 7 is a diagram illustrating a specific example of a classification model that selects a weight change prediction model for a user from weight change prediction models of other users.

Hereinafter, the main processes included in the weight management service provision method according to an embodiment of the present disclosure will be described assuming that the weight management service provision system 10 is the performer. However, depending on the embodiment, some of the above main processes may be performed by devices other than the weight management service providing system 10 (eg, the terminal 20 or a third party service server, etc.).

The embodiment of FIG. 3 may correspond to an embodiment applied to users for whom a weight change prediction model does not exist (eg, new users, etc.). Based on this, referring to FIG. 3, the weight management service providing system 10 can collect the user's biometric information (S300).

As described above in FIG. 2 , the weight management service providing system 10 may receive the user's biometric information from the terminal 20 and/or the data collection device 30. In the embodiment of FIG. 3, the biometric information may include the user's physical characteristic information (age, gender, height, weight, presence of chronic disease, etc.) and blood sugar data (continuous blood sugar data, fasting blood sugar information, etc.) .

The weight management service providing system 10 may extract a plurality of other users with similar biological conditions as the user based on the collected biometric information (S310).

The weight management service providing system 10 compares the biometric information of each other user stored in the database 150 with the collected biometric information, and based on the comparison result, other users who have similar biometric conditions (biometric information) to the user. can be extracted. For example, based on the user's biometric information, at least one other user whose biometric information stored in the database 150 satisfies a predefined similarity judgment standard in relation to the user's biometric information can be extracted. As a specific example, the weight management service provision system 10 may provide a plurality of other users who have the same information on gender and chronic disease onset as the user and whose differences in age, height, weight, and/or fasting blood sugar level are within a predetermined range. Users can be extracted as users with biological conditions similar to the user.

Depending on the embodiment, the weight management service providing system 10 may extract other users with similar blood sugar change patterns determined from continuous blood sugar data for a predetermined period (eg, two weeks, etc.) obtained from the user.

Depending on the embodiment, the weight management service providing system 10 may further consider the user's meal event data and/or exercise event data to determine other users with similar eating habits and/or exercise habits among other users with similar biological conditions. Users can also be extracted.

The weight management service providing system 10 may obtain at least one weight change prediction model among the extracted weight change prediction models for each of a plurality of other users as a weight change prediction model for the user (S320).

The database 150 of the weight management service providing system 10 may store extracted weight change prediction models for each of a plurality of other users. Before explaining step S320, contents related to the creation of the weight change prediction model will be explained with reference to FIGS. 4 and 5.

Referring to FIG. 4, the prediction model learning device 400 corresponds to a device that generates a weight change prediction model for the user according to machine learning techniques. The prediction model learning device 400 may be included in the weight management service providing system 10, but depending on the embodiment, it may be implemented as a separate device from the weight management service providing system 10.

The prediction model learning unit 410 is based on user physical characteristic information (age, gender, chronic disease onset, etc.), weight change information, and blood sugar data obtained from the user's terminal 20 and/or data collection device 30. Based on this, a weight change prediction model for the user can be learned.

User physical characteristic information can be used to set initial parameters of a weight change prediction model for the user. For example, the database 150 may store initial parameter information of a weight change prediction model by physical characteristics (age group, gender, presence of chronic disease, etc.). The initial parameter information may be obtained from weight change prediction models of multiple existing users, but this is not necessarily the case. The prediction model learning unit 410 may learn a weight change prediction model for the user by cumulatively inputting the user's weight change amount information and blood sugar data into a weight change prediction model for which initial parameters are set. In other words, the weight change prediction model may correspond to a model that learns the correlation between blood sugar characteristics and weight change and predicts the amount of weight change when blood sugar data is later input. The weight change prediction model can predict and output the user's weight change value based on the input blood sugar data, but depending on the embodiment, it predicts and outputs the user's weight change range, so that there is a difference between the predicted weight change and the actual weight change. Problems caused by differences can be resolved.

Depending on the embodiment, the prediction model learning unit 410 may learn a weight change prediction model by additionally utilizing the user's meal event data, exercise event data, etc. For example, the prediction model learning unit 410 provides information on the correlation between postprandial blood sugar characteristics and weight change according to dietary activity based on postprandial blood sugar data and weight change information extracted from blood sugar data based on meal event data. 1 You can learn a weight change prediction model. The postprandial blood sugar data may refer to continuous blood sugar data acquired over a predetermined period of time from the time of occurrence of a meal event. Alternatively, the prediction model learning unit 410 determines the second weight for the correlation between blood sugar characteristics and weight change due to exercise based on post-exercise blood sugar data and weight change information extracted from blood sugar data based on exercise event data. You can also learn a change prediction model. The post-exercise blood sugar data may refer to continuous blood sugar data acquired over a predetermined period of time from the time of occurrence of the exercise event. The first weight change prediction model and the second weight change prediction model may be implemented separately from each other, but depending on the embodiment, they may be implemented as one integrated weight change prediction model.

Meanwhile, the prediction model learning device 400 may acquire blood sugar characteristic information through analysis of blood sugar data, and use the obtained blood sugar characteristic information when learning the weight change prediction model. In this regard, referring to (a) of FIG. 5, the prediction model learning device 400 calculates fasting blood sugar 511, fasting blood sugar fluctuation range 512, 513, and fasting blood sugar from blood sugar data 510 for a predetermined period. Fasting blood sugar information including maintenance time 514 can be obtained. Fasting blood sugar 511 is a blood sugar level measured in a fasting state a predetermined time (e.g., 8 hours or more) has elapsed after a meal, and corresponds to a blood sugar level in a state where there is no or minimal effect from food. Meanwhile, the blood sugar level measured in a fasting state does not always maintain a constant level and may fluctuate within a certain range. Accordingly, the prediction model learning device 400 may determine fasting blood sugar fluctuation ranges 512 and 513, which represent the range of blood sugar levels measured in a fasting state, from the blood sugar data. In this case, the fasting blood sugar 511 may correspond to the average, mode, or median of blood sugar levels measured in a fasting state. The fasting blood sugar maintenance time 514 may mean a section in which the measured blood sugar level falls within the fasting blood sugar fluctuation ranges 512 and 513.

The prediction model learning device 400 may determine the fasting blood sugar 511 and the fasting blood sugar fluctuation range 512 and 513 based on blood sugar levels at a preset time point (for example, after bedtime and before a meal, etc.). According to the embodiment, the prediction model learning device 400 determines the user's fasting blood sugar 511 and fasting blood sugar fluctuation range 512 by analyzing blood sugar data based on the food intake time determined from the user's meal event data. You may. According to another embodiment, the prediction model learning device 400 calculates the user's continuous blood sugar dispersion from continuous blood sugar data, and calculates the calculated continuous blood sugar dispersion to biological characteristics similar to the user (age, gender, height, chronic disease incidence). It is also possible to determine the user's fasting blood sugar level (511) and fasting blood sugar fluctuation range (512, 513) by applying it to the fasting blood sugar dispersion range of the user group having the variable (whether or not, etc.). The method of determining the fasting blood sugar 511 and the fasting blood sugar fluctuation range 512 and 513 is not limited to the examples described above, and various methods that can be implemented by those skilled in the art may be applied. The prediction model learning device 400 may determine the fasting blood sugar maintenance time 514 from blood sugar data based on the fasting blood sugar fluctuation ranges 512 and 513.

Referring to (b) of FIG. 5, the prediction model learning device 400 determines the fasting blood sugar baseline 520 corresponding to the determined fasting blood sugar (or fasting blood sugar fluctuation range) and the blood sugar spike area corresponding to the blood sugar spike reference value. Based on the baseline 521, the fasting blood sugar exceedance section information 531, blood sugar spike section information 532, and under-fasting blood sugar section information 533 can be obtained from the blood sugar data of the predetermined section. The blood sugar spike reference value may vary depending on the user's fasting blood sugar level, but is not limited thereto.

The fasting blood sugar exceeding section information 531 may correspond to the area of a section having a blood sugar level exceeding the fasting blood sugar baseline 520 among the blood sugar data of a predetermined section. The blood sugar spike section information 532 may correspond to the area of a section (i.e., a blood sugar spike section) having a blood sugar level exceeding the blood sugar spike area baseline 521 among the blood sugar data of a predetermined section. The low fasting blood sugar section information 533 may correspond to the area of a section with a blood sugar level below the fasting blood sugar baseline 520 among the blood sugar data of a predetermined section.

According to the above-described embodiment, the blood sugar characteristic information used for learning the blood sugar change prediction model includes excess fasting blood sugar section information 531, blood sugar spike section information 532, under-fasting blood sugar section information 533, and It may include at least one of the fasting blood sugar maintenance times 514.

Referring again to FIG. 4, the weight change prediction model is a model that estimates the relationship between multiple explanatory variables and one response variable, including a multiple linear regression model and an elastic net regression model. model), regression tree, multi perceptron neural network model, etc. For example, the plurality of explanatory variables may include at least one of the blood sugar characteristic information, user physical characteristic information, food intake information, and exercise information, and the response variable may correspond to the amount of change in body weight in a predetermined period. In other words, when information corresponding to the explanatory variable is input, the generated weight change prediction model can predict and provide the amount of weight change (response variable) according to the input information.

Step S320 of FIG. 3 will be described again.

In order to provide an effective weight management service even in the absence of a weight change prediction model for the user, the weight management service providing system 10 uses weight change prediction models of other users with similar biological conditions to predict the weight change for the user. A weight change prediction model can be obtained.

Referring to FIG. 6 , the weight management service providing system 10 may apply a classification model to a plurality of extracted other users to select a weight change prediction model for each of a predefined number of other users. The classification model may include known classification models such as KNN (k-nearest neighborhood) or decision tree, but is not limited thereto.

For example, when the classification model is KNN, the weight management service providing system 10 may quantify data included in the biometric information of each of the user and a plurality of extracted other users according to a predefined algorithm. In addition, the weight management service providing system 10 may obtain a weight change prediction model for the user by selecting a prediction model for each of a predefined number of other users whose numerical value is most similar to the numerical value for the user. The graph shown in FIG. 7 expresses numerical values for each user according to the above algorithm, and the greater the similarity of the extracted points 711 to 717 of the plurality of other users with the user's biometric information, the greater the user's points 700. ) can be displayed in a location close to . For example, if the predefined number is 1, another user's weight change prediction model corresponding to the first point 711 closest to the user's point 700 may be obtained as the weight change prediction model for the user. Alternatively, if the predefined number is 7, other users' weight change prediction models corresponding to the seven points (711 to 717) closest to the user's point (700) may be obtained as the weight change prediction model for the user. You can.

The weight management service providing system 10 generates a blood sugar management guide corresponding to the user's target weight loss information based on at least one weight change prediction model obtained (S330) and provides the generated blood sugar management guide to the user. You can do it (S340).

Referring to FIG. 6 together, the weight management service providing system 10 includes at least one weight change prediction model acquired for the user, and target loss information (target loss period and target loss weight) received from the terminal 20, etc. Based on this, a blood sugar management guide corresponding to the target weight loss information can be created and provided.

As described above, the weight change prediction model corresponds to a model that predicts and provides the amount of weight change (weight change value or weight change range) corresponding to the input information when blood sugar characteristic information (blood sugar data) is input. The weight management service providing system 10 may sequentially input the blood sugar characteristic information into each of the obtained at least one weight change prediction model while variously modifying the blood sugar characteristic information in order to generate a blood sugar management guide corresponding to the target weight loss information. Each of the at least one weight change prediction model may output a predicted weight change amount according to input blood sugar characteristic information and biometric information. The weight management service providing system 10 performs a calculation such as a weighted average for each of at least one predicted weight change amount output (e.g., assigning a high weight to the weight change prediction model of another user with high similarity in biological conditions). The predicted change in body weight can be obtained. The obtained predicted change in weight corresponds to the predicted change in weight over a predetermined period, and the weight management service providing system 10 converts the obtained predicted change in weight to correspond to the target loss period and compares it with the target loss weight. Information can be converted to correspond to the predetermined period and compared with the predicted change in body weight. As a result of the comparison, blood sugar characteristic information with the smallest difference between the predicted change in weight and the target weight loss can be identified, and the weight management service providing system 10 can generate a blood sugar management guide based on the obtained blood sugar characteristic information.

Depending on the embodiment, as described above in FIG. 4, the weight change prediction model includes a first weight change prediction model for the correlation between blood sugar characteristics after a meal and weight change, and a first weight change prediction model for the correlation between blood sugar characteristics and weight change after exercise. When the second weight change prediction model is included, the weight management service provision system 10 applies the above-described process to each of the first weight change prediction model and the second weight change prediction model, thereby managing postprandial blood sugar according to the target weight loss information. We can provide guidance and post-exercise blood sugar management guidance.

For example, the blood sugar management guide may include at least one blood sugar characteristic among the fasting blood sugar excess section, blood sugar spike section, low fasting blood sugar section, and fasting blood sugar maintenance time included in the blood sugar characteristic information, and depending on the embodiment, may be related to blood sugar. Additional information may be included. The blood sugar management guide may be provided in visual/auditory form through an application installed on the user's terminal 20. Additionally, the weight management service providing system 10 may also provide additional information (food intake/exercise/sleep guide, etc.) to encourage the user to live according to the blood sugar management guide.

Depending on the embodiment, the blood sugar management guide includes a management guide for any one of the above fasting blood sugar excess section information, blood sugar spike section information, under-fasting blood sugar section information, and fasting blood sugar maintenance time that has the greatest impact on the user's weight loss. Alternatively, it may include a management guide listed in the order that affects weight loss. However, the form of the blood sugar management guide is not limited to this, and the weight management service providing system 10 may provide the blood sugar characteristic information in various forms to achieve the user's weight loss goal.

Depending on the embodiment, the weight management service providing system 10 may monitor the user's blood sugar level to update a previously provided blood sugar management guide or provide information on the predicted change amount of the user's weight (S350).

When a user uses a biometric data meter (31; e.g., continuous blood sugar meter), the weight management service providing system 10 periodically/continuously receives the user's blood sugar data after providing a blood sugar management guide, and adds the received blood sugar data to the user's blood sugar data. Based on this, the user's blood sugar management status can be monitored.

The weight management service providing system 10 may update and provide a blood sugar management guide to achieve the target weight loss information based on the monitoring results. For example, if it is determined as a result of monitoring that the user fails to comply with the blood sugar management guide, the weight management service providing system 10 may update and provide the blood sugar management guide to achieve the target weight loss information for the remaining period. Alternatively, the weight management service providing system 10 may provide information about the predicted change in body weight expected according to the monitoring results. Depending on the embodiment, the weight management service providing system 10 may encourage the user to set a more realistic weight loss goal by requesting modification of the goal weight loss information according to the monitoring results.

According to the embodiments of FIGS. 3 to 7, the weight management service providing system 10 provides a weight management service using the user's blood sugar level, thereby providing a more accurate weight reflecting individual biometric characteristics and blood sugar effects on meals and exercise. Provision of management services may be possible. In addition, the system provides blood sugar management guidance to users (such as new users) for whom a blood sugar change prediction model does not exist, using the blood sugar change prediction model of other users with similar biological conditions or blood sugar change patterns, thereby facilitating smooth weight management. Services can be provided.

Figure 8 is a flow chart for explaining a method of providing a weight management service using ketone levels according to an exemplary embodiment of the present disclosure. Figure 9 is a diagram to explain an embodiment of the process of generating a ketone level prediction model for each user. FIG. 10 is a diagram for explaining a specific example of the weight management service provision method described above in FIG. 8. Figure 11 shows an example of a ketone level prediction graph over time obtained from a ketone level prediction model selected for a user.

Among various diet methods, recently popular diet methods include the ketogenic diet and intermittent fasting. Ketogenic diet and intermittent fasting are methods that use the principle of bringing the human body's metabolism to ketosis, a state in which fat accumulated in the body is used as the main energy source instead of carbohydrates. The ketosis state can be entered when the fasting state continues for more than a predetermined period of time. In the ketosis state, the ketone level in the body exceeds a predetermined value, and an increase in the ketone level means that the rate of body fat decomposition increases. You can.

However, the timing of entering the ketosis state may vary depending on each person's physical characteristics and constitution. In addition, if ketone levels increase excessively in ketosis, problems such as ketoacidosis may occur, so a method to guide individuals to maintain an optimal ketosis state for weight management may be required.

Referring to FIG. 8 as an embodiment of the present disclosure to solve the above-described problem, the weight management service providing system 10 collects the user's biometric information (S800) and collects a plurality of other biometric information having similar biometric conditions to the user. Users can be extracted (S810).

Since steps S800 and S810 are similar to steps S300 and S310 of FIG. 3, duplicate description thereof will be omitted.

The weight management service providing system 10 may obtain at least one of the extracted ketone level prediction models according to the fasting maintenance time of each of the plurality of other users as a ketone level prediction model for the user (S820).

The database 150 of the weight management service providing system 10 may store a ketone level prediction model according to the fasting maintenance time of each of a plurality of different extracted users. Prior to the description of step S820, contents related to the creation of the ketone numerical prediction model will be explained through FIGS. 9 to 10.

Referring to FIG. 9, the ketone level prediction model learning device 900 corresponds to a device that generates a ketone level prediction model for a user according to machine learning techniques. The ketone level prediction model learning device 900 may be included in the weight management service providing system 10, but depending on the embodiment, it may be implemented as a separate device from the weight management service providing system 10.

The ketone level prediction model learning unit 910 includes user physical characteristic information (age, gender, height, presence of chronic disease, etc.), ketone data, and /Or based on blood sugar data (blood sugar characteristic information, etc.), a ketone level prediction model according to the fasting time for the user can be learned. Depending on the embodiment, the ketone level prediction model learning unit 910 may additionally receive the user's meal event data, exercise event data, and/or body weight data to learn the ketone level prediction model.

According to the embodiment, as described above in FIG. 5, the ketone level prediction model learning device 900 acquires blood sugar characteristic information including fasting blood sugar maintenance time through analysis of the blood sugar data, and uses the obtained blood sugar characteristic information as shown in FIG. The ketone level prediction model can be learned using. The fasting blood sugar maintenance time may be related to the above-mentioned ketosis state and ketone level.

The ketone level prediction model may correspond to a model that estimates the relationship between multiple explanatory variables and one response variable, similar to the weight change prediction model described above. The plurality of explanatory variables may include at least some of blood sugar characteristic information (fasting blood sugar maintenance time), food intake information (eating time, etc.), and user's physical characteristic information, and the response variable may include ketone levels. .

Step S820 of FIG. 8 will be described again.

In order to provide an effective weight management service even in the absence of a ketone level prediction model for the user, the weight management service providing system 10 uses the ketone level prediction models of other users with similar biological conditions to provide information about the user. A ketone level prediction model can be obtained.

Referring to FIG. 10 together, the weight management service providing system 10 may apply a classification model to a plurality of other extracted users to select a ketone level prediction model for each of a predefined number of other users. The classification model may include known classification models such as KNN (k-nearest neighborhood) or decision tree, but is not limited thereto. An example related to the classification model is described in detail in FIG. 7.

The weight management service providing system 10 may generate a weight management guide for the user based on the obtained at least one ketone level prediction model (S830) and provide the generated weight management guide to the user (S840). ).

Referring to Figure 10, the weight management service providing system 10 provides prediction information on ketone levels according to the user's fasting maintenance time (fasting blood sugar maintenance time) based on at least one ketone level prediction model obtained for the user. It can be obtained. Specifically, the weight management service providing system 10 may input a predetermined fasting maintenance time value into the at least one ketone level prediction model. Each of the at least one ketone level prediction model may output a ketone level according to the input fasting maintenance time value. The weight management service providing system 10 calculates one weighted average for each of at least one output ketone level (e.g., assigning a high weight to the ketone level prediction model of another user with high similarity in biological conditions). Ketone levels can be obtained. The weight management service providing system 10 can obtain information on changes in ketone levels according to fasting maintenance time by inputting different fasting maintenance time values.

Referring to FIG. 11, the graph in FIG. 11 expresses information on changes in ketone levels according to the obtained fasting maintenance time. For example, if the fasting time is within the first time (T1), the ketone level is very low, which may mean that the user's metabolic state does not enter the ketosis state.

If the fasting time exceeds the first time (T1), the ketone level may gradually increase, which may mean that the user's metabolic state enters the ketosis state. Meanwhile, a section where the ketone level is less than a predetermined first value (K1) may have a weak degree of body fat decomposition, and a section where the ketone level exceeds the second value (K2) may correspond to a health risk section. Each of the first value K1 and the second value K2 may be set to the same value regardless of the user, but may be set differently depending on the user's biological characteristics, etc., depending on the embodiment. According to this embodiment, the optimal fasting maintenance time for the user is the second time (T2), which is the time when the ketone level is within a preset value range (between the first value (K1) and the second value (K2)) It may correspond to the third time (T3).

The weight management service providing system 10 may generate a weight management guide (diet plan, etc.) that reflects the optimal fasting time. For example, when creating a weight management guide including a diet and meal schedule, the weight management service providing system 10 may generate the interval between meals to correspond to the optimal fasting time. The user can maximize the effect of weight management by maintaining the optimal fasting time for the user according to the weight management guide.

FIG. 12 is a block diagram schematically showing the hardware configuration of a device included in the weight management service providing system shown in FIG. 1.

The hardware configuration of the device 1200 shown in FIG. 12 may correspond to the hardware configuration of each of at least one device constituting the weight management service providing system 10 shown in FIG. 1. Additionally, the hardware configuration of the device 1200 may also correspond to the hardware configuration of the terminal 20 and/or the data collection device 30 shown in FIG. 1 . That is, the weight management service providing system 10, the terminal 20, and/or the data collection device 30 may include at least some of the hardware configurations of the device 1200 shown in FIG. 12, and in the embodiment Additional configurations may be included accordingly.

Referring to FIG. 12 , the device 1200 may include a communication unit 1210, an input unit 1220, an output unit 1230, a control unit 1240, and a memory 1250. The control configuration shown in FIG. 12 is an example for convenience of explanation, and the device 1200 may include more or less configurations than the configuration shown in FIG. 12 .

The communication unit 1210 may include one or more communication modules that enable communication with other terminals or servers by connecting the device 1200 to a network. For example, the communication module may include a mobile communication module such as LTE, 5G, etc., a wireless communication module such as Wi-Fi, and/or various other wired or wireless communication modules.

The input unit 1220 is a component for acquiring information such as user input, video, and audio, and may include various input means such as various mechanical/electronic input means, cameras, and microphones. The output unit 1230 is used to provide information to users by generating output related to vision, hearing, or tactile senses, and may include a display, speaker, vibration module, etc.

The control unit 1240 may control the overall operation of the device 1200. The control unit 1240 may process signals, data, and information input or output through the above-described components, or may provide certain information or functions according to various applications or algorithms stored in the memory 1250.

For example, the control unit 1240 performs operations related to providing a blood sugar management guide for weight control according to an embodiment of the present disclosure (obtaining a weight change prediction model, generating a blood sugar management guide corresponding to target weight loss information, and providing a blood sugar management guide. provision, etc.) can control the process for at least some of them. In addition, the control unit 1240 performs operations related to providing a weight management guide using ketone levels according to an embodiment of the present disclosure (obtaining a ketone level prediction model, generating and providing a weight management guide reflecting the optimal fasting maintenance time, etc.). You can control at least some of the processes.

The control unit 1240 may include at least one processor, and the processor may be implemented with hardware such as a CPU, an application processor (AP), an MCU, an integrated circuit, an ASIC, or an FPGA.

The memory 1250 may store programs and data necessary for the operation of the device 1200. Additionally, the memory 1250 may store data generated or acquired through the control unit 1240. The memory 1250 may be composed of a storage medium such as ROM, RAM, flash memory, SSD, or HDD, or a combination of storage media.

The description of the above-described embodiments is merely an example with reference to the drawings for a more thorough understanding of the present disclosure, and should not be construed as limiting the technical idea of the present disclosure.

In addition, it will be clear to those skilled in the art to which this disclosure pertains that various changes and modifications can be made without departing from the basic principles of the present disclosure.

Claims (11)

In a weight management service providing system including at least one computing device,
a communication unit that receives the user's biometric information from the user's terminal or data collection device; and
Based on the received biometric information, obtain at least one of weight change prediction models of a plurality of other users registered in the weight management service as a weight change prediction model for the user,
Based on the obtained weight change prediction model, generate a blood sugar management guide corresponding to the user's goal weight loss information,
At least one processor that controls the communication unit to transmit the generated blood sugar management guide to the user's terminal,
The weight change prediction model is,
A machine learning-based model that is learned using a setting variable containing at least blood sugar characteristic information and a response variable containing weight change information,
Weight management service delivery system.
According to paragraph 1,
The user's biometric information is,
Contains physical characteristic information and blood sugar data of the user,
The physical characteristic information includes at least one of age, gender, height, weight, and whether or not a chronic disease occurs.
Weight management service delivery system.
According to paragraph 1,
The at least one processor,
From the database, extract other users whose biometric information satisfies predefined similar judgment criteria with the user's biometric information,
Obtaining a weight change prediction model for the user by applying a classification model to the extracted other users to select a weight change prediction model for each of a predefined number of other users,
Weight management service delivery system.
According to paragraph 1,
The at least one processor,
Using the obtained weight change prediction model, obtain a plurality of weight prediction changes for a plurality of different blood sugar characteristic information,
Compare each of the plurality of predicted weight change amounts with the target loss information to identify blood sugar characteristic information corresponding to the target loss information,
Generating the blood sugar management guide based on the identified blood sugar characteristic information,
Weight management service delivery system.
According to clause 4,
The target loss information includes target loss weight and target loss period,
The at least one processor,
Converting each of the plurality of predicted weight change amounts to correspond to the target weight loss period and comparing them with the target weight loss weight,
As a result of the comparison, the blood sugar characteristic information corresponding to the predicted change in weight with the smallest difference from the target weight loss is identified as the blood sugar characteristic information corresponding to the target weight loss information,
Weight management service delivery system.
According to clause 5,
The blood sugar characteristic information includes at least one of a fasting blood sugar maintenance time, a blood sugar spike section, an over-fasting blood sugar section, and an under-fasting blood sugar section.
Weight management service delivery system.
According to paragraph 1,
The at least one processor,
After providing the blood sugar management guide, collect blood sugar data of the user,
Monitoring the user's blood sugar management status based on the collected blood sugar data,
Updating the blood sugar management guide based on monitoring results,
Weight management service delivery system.
In a method of providing weight management service through at least one computing device,
Obtaining, by a biometric information collection unit included in the at least one computing device, biometric information of a user;
A prediction model acquisition unit included in the at least one computing device, based on the acquired biometric information, calculates at least one of weight change prediction models of other users registered in the weight management service as a weight change prediction model for the user. acquiring;
generating, by a blood sugar management guide providing unit included in the at least one computing device, a blood sugar management guide corresponding to the user's target weight loss information based on an obtained weight change prediction model; and
The blood sugar management guide providing unit includes providing the generated blood sugar management guide to the user,
The weight change prediction model is,
A machine learning-based model that is learned using a setting variable containing at least blood sugar characteristic information and a response variable containing weight change information,
How to provide weight management services.
According to clause 8,
The step of obtaining at least one of the weight change prediction models of other users as a weight change prediction model for the user,
extracting, from a database, other users whose biometric information satisfies predefined similar criteria to the user's biometric information; and
Comprising the step of obtaining a weight change prediction model for the user by applying a classification model to the extracted other users to select a weight change prediction model for each of a predefined number of other users,
How to provide weight management services.
According to clause 8,
The step of creating the blood sugar management guide is,
Obtaining a plurality of predicted weight change amounts for a plurality of different blood sugar characteristic information using the obtained weight change prediction model;
Comparing each of the plurality of predicted weight change amounts with the target loss information and identifying blood sugar characteristic information corresponding to the target loss information; and
Comprising the step of generating the blood sugar management guide based on the identified blood sugar characteristic information,
How to provide weight management services.
According to clause 10,
The target weight loss information includes target weight loss and target weight loss period,
The step of identifying blood sugar characteristic information corresponding to the target weight loss information,
Converting each of the plurality of predicted weight change amounts to correspond to the target weight loss period and comparing them with the target weight loss; and
As a result of the comparison, identifying blood sugar characteristic information corresponding to the predicted change in weight with the smallest difference from the target loss weight as blood sugar characteristic information corresponding to the target weight loss information,
How to provide weight management services.

Images