Intelligent monitoring system for nutrition and health status

Abstract

The application provides a kind of nutrition and health state intelligent monitoring system, system includes meal image acquisition module, food recognition and quantity estimation module, nutrition intake calculation module, basic physiological data acquisition module, correlation correction module and state evaluation module;Through food recognition and quantity estimation to meal image, obtain initial nutrition intake result in combination with food composition database, and determine energy consumption result based on activity data and basal metabolic data, correlation correction is carried out to initial nutrition intake result, and nutrition state and health state monitoring result are output.The application carries out deviation determination to initial nutrition intake result obtained by food recognition and energy consumption result determined based on activity data and basal metabolic data, so as to be favorable to reduce the influence of food recognition error and quantity estimation deviation on nutrition intake result, improve the objectivity of nutrition intake result.

Description

Technical Field

[0001] This invention relates to the field of nutritional health monitoring technology, and in particular to an intelligent monitoring system for nutrition and health status. Background Technology

[0002] With the increasing demand for health management among residents, the monitoring of users' dietary intake and basic physiological status based on mobile terminals and wearable devices has been gradually applied to home health management and basic dietary management of chronic diseases.

[0003] In typical nutrition and health monitoring programs, one type of program mainly estimates the types of food, the amount of food consumed, and the intake of nutrients by having users manually enter dietary information or collect images of meals. The other type of program mainly uses wearable devices to collect basic physiological data such as heart rate, steps, activity duration, sleep duration, and basal metabolic rate to analyze the user's activity level and basic health status.

[0004] The above solution still has the following problems in practical applications: On the one hand, when estimating nutritional intake solely based on manual records or food image recognition results, it is easily affected by factors such as food obstruction, differences in shooting angles, differences in container shapes, and errors in portion estimation, leading to discrepancies between the nutritional intake results and the actual intake. On the other hand, nutritional analysis based solely on dietary intake results, or displaying dietary data and basic physiological data separately, lacks a processing mechanism to correlate and correct nutritional intake results based on basic physiological consumption results. This makes it difficult to objectively reflect the matching status between users' nutritional intake and actual energy consumption, affecting the accuracy and stability of nutritional status and health status monitoring results.

[0005] To address this, an intelligent monitoring system for nutrition and health status is proposed. Summary of the Invention

[0006] In view of this, the present invention provides an intelligent monitoring system for nutrition and health status to solve or alleviate the technical problems existing in the prior art, and at least provides a beneficial alternative.

[0007] The technical solution of the present invention is implemented as follows: a nutrition and health status intelligent monitoring system, including a dietary image acquisition module, a food identification and portion estimation module, a nutrient intake calculation module, a basic physiological data acquisition module, a correlation correction module, and a status assessment module; The dietary image acquisition module is used to acquire dietary images of the user within a preset monitoring period; The food identification and portion estimation module is connected to the dietary image acquisition module and is used to identify the food categories in the dietary image and obtain the corresponding intake portions. The nutrient intake calculation module is connected to the food identification and portion estimation module, and is used to calculate the initial nutrient intake result based on the food category and the intake portion according to the preset food composition database. The basic physiological data acquisition module is used to acquire the user's activity data and basal metabolic data within the preset monitoring period; The correlation correction module is connected to the nutrient intake calculation module and the basic physiological data acquisition module respectively. It is used to determine the energy consumption result corresponding to the preset monitoring period based on the activity data and the basic metabolic data, determine the deviation between the energy consumption result and the initial nutrient intake result, and when the deviation exceeds the preset threshold, adjust at least one of the intake portion and food component conversion parameters according to the preset correction rules to obtain the corrected nutrient intake result. The status assessment module is connected to the correlation correction module and is used to output the user's nutritional status and health status monitoring results based on the corrected nutritional intake results and the basic physiological data.

[0008] Furthermore, the preset monitoring period includes a single-meal monitoring period and a daily monitoring period, and the dietary image acquisition module is used to acquire at least one dietary image within the single-meal monitoring period. By monitoring by single meal or by day, a correspondence can be established between dietary intake data and physiological data of the same period, facilitating subsequent correlation analysis between nutrient intake and energy consumption.

[0009] Furthermore, the dietary image acquisition module is located on a mobile terminal, and the dietary image includes at least one of the following: a main food area and a reference area for portion estimation and a tableware outline area. Introducing portion estimation reference information into the dietary image helps improve the stability of portion estimation.

[0010] Furthermore, the food identification and portion estimation module employs a target detection model to detect and classify food targets in the dietary image, and determines the intake portion based on at least one of the reference area, the tableware outline area, and image depth information. This enables food category identification and portion estimation under typical mobile terminal acquisition conditions, meeting the needs of ordinary users in home applications.

[0011] Furthermore, the initial nutrient intake result output by the nutrient intake calculation module includes at least total energy and one or more of the following: protein, fat, carbohydrates, dietary fiber, sodium, potassium, and phosphorus. By calling a preset food composition database for conversion, the dietary image recognition result can be transformed into a nutritionally meaningful intake result, providing a data foundation for subsequent status assessment.

[0012] Furthermore, the activity data and basal metabolic data acquired by the basic physiological data acquisition module include at least two or more of the following: dynamic heart rate, resting heart rate, steps, activity duration, sleep duration, weight, and basal metabolic rate. By introducing basic physiological data that can be stably acquired by conventional wearable devices, the user's basic physiological state can be continuously recorded without the need for complex detection equipment.

[0013] Furthermore, the correlation correction module determines the energy consumption result as the sum of basal metabolic rate and activity expenditure, wherein the basal metabolic rate is determined based on the basal metabolic data, and the activity expenditure is determined based on the activity data. This allows for a corresponding correlation between dietary intake data and physiological expenditure data, rather than relying solely on a static judgment based on a single dietary identification result.

[0014] Furthermore, the correlation correction module is used to generate an individualized correction coefficient based on historical deviation data when the deviation exceeds the preset threshold in multiple consecutive monitoring periods, and to update at least one of the intake portion and food composition conversion parameters using the individualized correction coefficient. By correcting the deviation within consecutive monitoring periods, the cumulative deviation caused by differences in single image acquisition angles, food occlusion, or portion estimation errors can be reduced, improving the system's adaptability to different individual usage scenarios.

[0015] Furthermore, the status assessment module is used to determine, based on the deviation of the corrected nutrient intake result from the target nutrient intake range and abnormal changes in the basic physiological data, whether the nutritional status is insufficient, balanced, or excessive, and whether the health status is normal, warning, or risky. This enables the system output to not only reflect nutrient intake but also to provide a corresponding health status judgment in conjunction with physiological changes.

[0016] Furthermore, the system also includes a suggestion generation module, which is connected to the status assessment module and is used to generate individualized dietary adjustment suggestions based on the nutritional status and health status. This allows for the output of specific adjustment directions beyond status monitoring, facilitating daily dietary management for users.

[0017] The embodiments of the present invention have the following advantages due to the adoption of the above technical solutions: I. This invention determines the deviation between the initial nutrient intake result obtained from dietary image recognition and the energy consumption result determined based on activity data and basal metabolic data, and adjusts at least one of the intake portion and food composition conversion parameters when the deviation exceeds a preset threshold. This helps to reduce the impact of food recognition error and portion estimation deviation on nutrient intake results and improve the objectivity of nutrient intake results.

[0018] Second, this invention combines the corrected nutritional intake results with basic physiological data to output the user's nutritional status and health status monitoring results, thereby improving the consistency and stability between nutritional monitoring and health status assessment, and enhancing the practicality of the system in home health monitoring and basic dietary management of chronic diseases.

[0019] The above overview is for illustrative purposes only and is not intended to be limiting in any way. In addition to the illustrative aspects, embodiments, and features described above, further aspects, embodiments, and features of the invention will become readily apparent from the accompanying drawings and the following detailed description. Attached Figure Description

[0020] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0021] Figure 1 This is a block diagram of the intelligent monitoring system for nutrition and health status of the present invention. Detailed Implementation

[0022] In the following description, only certain exemplary embodiments are briefly described. As those skilled in the art will recognize, the described embodiments can be modified in various ways without departing from the spirit or scope of the invention. Therefore, the drawings and description are considered to be exemplary in nature and not restrictive.

[0023] The embodiments of the present invention will now be described in detail with reference to the accompanying drawings.

[0024] like Figure 1 As shown in the figure, this embodiment of the invention provides an intelligent monitoring system for nutrition and health status, which mainly includes a dietary image acquisition module, a food recognition and portion estimation module, a nutrient intake calculation module, a basic physiological data acquisition module, a correlation correction module, a status assessment module, and a suggestion generation module. Each module can be deployed in the same terminal device, or separately in a mobile terminal, a wearable terminal, and a server, and data interaction can be achieved through wired or wireless communication.

[0025] In this invention, a dietary image acquisition module is used to acquire dietary images of the user within a preset monitoring period; a food identification and portion estimation module is used to identify the food categories in the dietary images and obtain the corresponding intake portions; a nutrient intake calculation module is used to calculate the initial nutrient intake result based on a preset food composition database, according to the food category and intake portion; a basic physiological data acquisition module is used to acquire the user's activity data and basal metabolic data within the preset monitoring period; a correlation correction module is used to determine the energy consumption result for the corresponding preset monitoring period based on the activity data and basal metabolic data, and to determine the deviation between the energy consumption result and the initial nutrient intake result. When the deviation exceeds a preset threshold, at least one of the intake portion and food composition conversion parameters is adjusted according to a preset correction rule to obtain a corrected nutrient intake result; a status assessment module is used to output the user's nutritional status and health status monitoring results based on the corrected nutrient intake result and basic physiological data; and a suggestion generation module is used to generate individualized dietary adjustment suggestions based on the nutritional status and health status.

[0026] The preset monitoring period in this invention can be set to a single-meal monitoring period or a daily monitoring period. Under a single-meal monitoring period, the system processes dietary intake data corresponding to a single meal and related physiological data before and after that meal. Under a daily monitoring period, the system processes cumulative nutrient intake data from multiple meals throughout the day and corresponding basic physiological data. By setting these two monitoring periods, it can be adapted to application scenarios such as home health monitoring for the general population, basic dietary management for patients with chronic diseases, and continuous weight management.

[0027] Example 1 This embodiment provides an intelligent monitoring system for nutrition and health status in ordinary home health monitoring scenarios.

[0028] Users take photos of their meals before breakfast, lunch, and dinner using a mobile device, capturing at least one frontal view image each time. If there is occlusion in the food, overlapping dishes, or significant gravy, an additional image at an angle is allowed to improve the stability of food category recognition and portion estimation. The meal images retain the outline of at least one of the following: plate, bowl, soup bowl, or spoon, for reference in subsequent portion estimation.

[0029] After receiving the dietary image, the food recognition and portion estimation module uses a target detection model to detect and classify food categories such as rice, pasta, meat, eggs, vegetables, fruits, dairy products, and beverages in the image. It also estimates the area ratio and volume relationship of each food target by combining the outline information of tableware, thereby obtaining the intake portion corresponding to each food category.

[0030] In this embodiment, the food identification and portion estimation module can use existing mature target detection models to realize food target detection and classification, without limiting the specific model name.

[0031] The nutrient intake calculation module matches the identified food categories and their intake portions with a preset food composition database to calculate the initial nutrient intake result for that single meal. The initial nutrient intake result includes at least total energy, protein, fat, and carbohydrates, and may further include one or more of dietary fiber, sodium, potassium, and phosphorus. The initial nutrient intake results for breakfast, lunch, and dinner are summed to obtain the user's initial nutrient intake result for the day.

[0032] Meanwhile, the basic physiological data acquisition module collects the user's dynamic heart rate, resting heart rate, steps, activity duration, sleep duration, and basal metabolic rate throughout the day via a wearable terminal. The correlation correction module determines basal metabolic rate consumption based on basal metabolic rate and activity consumption based on steps, dynamic heart rate, and activity duration, thus obtaining the daily energy consumption result. Subsequently, the correlation correction module determines the deviation between the daily energy consumption result and the initial nutritional intake result; when the deviation does not exceed a preset threshold, the initial nutritional intake result is maintained as the final result; when the deviation exceeds the preset threshold, at least one of the intake portion and food composition conversion parameters is adjusted according to preset correction rules to obtain the corrected nutritional intake result.

[0033] The status assessment module outputs the user's nutritional status and health status for the day based on the corrected nutritional intake results and basic physiological data.

[0034] For example, if the corrected total energy intake is consistently below the lower limit of the target intake range, and the dynamic heart rate fluctuation is small with normal activity levels, the nutritional status can be determined as insufficient intake. Conversely, if the corrected total energy intake is above the upper limit of the target intake range, and weight shows an upward trend over multiple consecutive monitoring periods, the nutritional status can be determined as excessive intake, and the health status will be classified as a warning state. The suggestion generation module will then output corresponding dietary adjustment suggestions, such as reducing the intake of high-energy snacks, increasing the proportion of high-quality protein, controlling the intake of staple foods at dinner, or increasing daytime activity levels.

[0035] In this embodiment, instead of directly providing a nutritional assessment based solely on the user's photographic results, the dietary intake results are correlated with the basic physiological consumption results and necessary corrections are made. This reduces the impact of food identification bias and portion estimation bias on the results, making it suitable for daily home nutrition and health status monitoring of the general population.

[0036] Example 2 This embodiment provides an intelligent monitoring system for nutrition and health status in the context of basic dietary management for patients with chronic kidney disease.

[0037] The target users are chronic kidney disease patients in a stable management phase. In addition to monitoring total energy intake, these users also need to pay close attention to the intake of nutrients such as protein, sodium, potassium, and phosphorus. Therefore, in this embodiment, the nutrient intake calculation module, besides calculating total energy, protein, fat, and carbohydrates, also focuses on outputting the initial intake results of sodium, potassium, and phosphorus.

[0038] In practical use, users capture images of their three meals and snacks via mobile devices. These images include staple foods, vegetables, meat and poultry, soy products, soups, and fruits. After identifying each food category, the food recognition and portion estimation module calculates the protein, sodium, potassium, and phosphorus content based on a pre-set food composition database, generating the initial nutritional intake result for each meal.

[0039] For soups, stews, mixed dishes, or dishes with a lot of sauce, users can add simple dish tags on their mobile devices to improve the accuracy of food category matching. Simple dish tags can include basic category information such as "stew," "soup," "braised," "pickled," and "high-salt processed food."

[0040] The basic physiological data acquisition module obtains basic physiological data such as dynamic heart rate, resting heart rate, steps, activity duration, sleep duration, and weight through a wearable terminal. The correlation correction module calculates the daily energy expenditure based on the basal metabolic data and activity data, and determines the deviation between the daily energy expenditure result and the initial nutrient intake result. When the system determines that the deviation between the total energy intake result and the energy expenditure result is too large, it corrects the relevant food intake portions according to preset correction rules. For users with excessive deviations for multiple consecutive monitoring periods, the correlation correction module further generates individualized correction coefficients based on historical deviation data to update at least one of the subsequent intake portions and food composition conversion parameters, so that the subsequent monitoring results are more consistent with the user's actual dietary habits and image acquisition characteristics.

[0041] In this embodiment, the status assessment module not only outputs the nutritional status corresponding to the total energy, but also focuses on analyzing whether the target nutrients deviate from the target nutrient intake range.

[0042] For example, when the corrected sodium intake consistently exceeds the upper limit of the target intake range, the nutritional status is determined to be insufficient salt restriction; when the corrected protein intake consistently falls below the lower limit of the target intake range, the nutritional status is determined to be insufficient protein intake; when the corrected potassium or phosphorus intake repeatedly approaches or exceeds the upper limit of the target intake range, the health status is determined to be in a warning state. The suggestion generation module outputs corresponding dietary adjustment suggestions based on the above results, including reducing the intake of high-sodium seasonings, controlling the intake of high-phosphorus processed foods, adjusting the frequency of high-potassium fruit and vegetable intake, and increasing the intake of high-quality, low-burden protein sources.

[0043] In this embodiment, by combining the results of basic dietary image recognition with basic physiological data obtained from conventional wearable terminals, joint monitoring of the basic nutrition and health status of patients with chronic kidney disease in a home setting is achieved. The entire monitoring process does not rely on complex non-invasive metabolite detection equipment, has a low barrier to entry, and is suitable for long-term continuous application. At the same time, with the help of the correlation correction module and individualized correction coefficients, dietary estimation biases caused by differences in soups, mixed dishes, and home-cooked meals in chronic disease dietary management can be reduced.

[0044] Example 3 This embodiment provides an intelligent monitoring system for nutrition and health status in scenarios requiring continuous monitoring and individualized correction.

[0045] The user uses the system of this invention for monitoring over seven consecutive days. The dietary image acquisition module acquires dietary images of breakfast, lunch, and dinner each day; the food identification and portion estimation module identifies and estimates the food categories and intake portions for the three meals each day; and the nutrient intake calculation module calculates the initial daily nutrient intake results. The basic physiological data acquisition module simultaneously acquires the user's dynamic heart rate, resting heart rate, steps, activity duration, sleep duration, weight, and basal metabolic rate over the seven days.

[0046] In the monitoring results of the first two days, the correlation correction module found that the user's initial nutrient intake was consistently higher than their energy expenditure, and the deviation exceeded a preset threshold. Combined with dietary image retrospective analysis, it was found that the user habitually used deeper bowls and plates to hold staple foods and soups, resulting in significant food edge occlusion in the images and an overestimation of initial portion sizes. Therefore, in the continuous monitoring from the third to the seventh day, the correlation correction module used historical deviation data to generate individualized correction coefficients and updated the subsequent intake portion conversion parameters. After the update, the estimated portions of staple foods and soups under the same tableware conditions were more closely closer to the user's actual eating habits.

[0047] In this embodiment, the status assessment module does not only provide static results based on single-day data, but also performs trend assessment by combining corrected nutrient intake results and basic physiological data over multiple consecutive monitoring periods. For example, if the system finds that the corrected total energy intake result is higher than the upper limit of the target nutrient intake range for seven consecutive days, and the weight shows a continuous increasing trend, the nutritional status is judged as continuously excessive, and the health status is judged as a risk state; conversely, if the system finds that the corrected total energy intake result gradually approaches the target nutrient intake range, and the weight change tends to stabilize, the health status is adjusted from a risk state to a warning state or a normal state.

[0048] The suggestion generation module outputs individualized dietary adjustment recommendations based on continuous trend assessment results. In this embodiment, the suggestion generation module does not use a fixed template to directly output uniform recommendations, but instead generates targeted recommendations based on the changing trends of corrected nutrient intake results. For example, for users with high dinner intake, the module outputs the suggestion to "reduce the proportion of staple food and high-fat side dishes at dinner"; for users with low protein content at breakfast, the module outputs the suggestion to "increase the proportion of eggs or dairy products at breakfast"; and for users with consistently low nighttime activity levels but high total energy intake, the module outputs the suggestion to "reduce nighttime snacks and appropriately increase light daytime physical activity".

[0049] In this embodiment, an individualized correction coefficient is generated using historical deviation data from multiple consecutive monitoring periods. This correction coefficient is then used to update at least one of the conversion parameters for intake portion and food composition. This allows the same user to gradually obtain monitoring results that better match their own collection habits and dietary structure over a long period of use. Therefore, it not only enables basic nutrition and health status monitoring but also improves the consistency and individual adaptability of results in continuous use scenarios.

[0050] In this embodiment, not only can the continuous monitoring of the user's nutritional intake and health status be completed, but the previous estimation deviation can also be gradually corrected as the monitoring period increases, so that the subsequent monitoring results are more consistent with the user's actual intake and basic physiological changes, making it more suitable for home health monitoring and basic dietary management scenarios that require long-term follow-up management.

[0051] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any person skilled in the art can easily conceive of various variations or substitutions within the technical scope disclosed in the present invention, and these should all be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. A smart monitoring system for nutrition and health status, characterized in that, include: The dietary image acquisition module is used to acquire dietary images of the user within a preset monitoring period; The food identification and portion estimation module is connected to the dietary image acquisition module and is used to identify the food categories in the dietary image and obtain the corresponding intake portions. The nutrient intake calculation module is connected to the food identification and portion estimation module and is used to calculate the initial nutrient intake result based on the food category and the intake portion according to the preset food composition database. The basic physiological data acquisition module is used to acquire the user's activity data and basal metabolic data within the preset monitoring period; The correlation correction module is connected to the nutrient intake calculation module and the basic physiological data acquisition module, respectively. It is used to determine the energy consumption result corresponding to the preset monitoring period based on the activity data and the basic metabolic data, determine the deviation between the energy consumption result and the initial nutrient intake result, and when the deviation exceeds the preset threshold, adjust at least one of the intake portion and food component conversion parameters according to the preset correction rules to obtain the corrected nutrient intake result. The status assessment module, connected to the correlation correction module, is used to output the user's nutritional status and health status monitoring results based on the corrected nutritional intake results and the basic physiological data.

2. The intelligent monitoring system for nutrition and health status according to claim 1, characterized in that: The preset monitoring period includes a single-meal monitoring period and a daily monitoring period. The dietary image acquisition module is used to acquire at least one dietary image within the single-meal monitoring period.

3. The intelligent monitoring system for nutrition and health status according to claim 1, characterized in that: The dietary image acquisition module is installed on a mobile terminal, and the dietary image includes at least one of the following: the main area of ​​the meal, the reference area for portion estimation, and the outline area of ​​the tableware.

4. The intelligent monitoring system for nutrition and health status according to claim 1, characterized in that: The food recognition and portion estimation module uses a target detection model to detect and classify food targets in the dietary image, and determines the intake portion based on at least one of the reference area, tableware outline area, and image depth information.

5. The intelligent monitoring system for nutrition and health status according to claim 1, characterized in that: The initial nutrient intake result output by the nutrient intake calculation module includes at least total energy and one or more of the following: protein, fat, carbohydrates, dietary fiber, sodium, potassium, and phosphorus.

6. The intelligent monitoring system for nutrition and health status according to claim 1, characterized in that: The activity data and basal metabolic data acquired by the basic physiological data acquisition module include at least two or more of the following: dynamic heart rate, resting heart rate, steps, activity duration, sleep duration, weight, and basal metabolic rate.

7. The intelligent monitoring system for nutrition and health status according to claim 1, characterized in that: The correlation correction module determines the energy consumption result as the sum of basal metabolic consumption and activity consumption, wherein the basal metabolic consumption is determined based on the basal metabolic data, and the activity consumption is determined based on the activity data.

8. The intelligent monitoring system for nutrition and health status according to claim 7, characterized in that: The correlation correction module is used to generate an individualized correction coefficient based on historical deviation data when the deviation exceeds the preset threshold in multiple consecutive monitoring periods, and to update at least one of the intake portion and food composition conversion parameters using the individualized correction coefficient.

9. The intelligent monitoring system for nutrition and health status according to claim 1, characterized in that: The status assessment module is used to determine the nutritional status as one of insufficient intake, balanced intake, or excessive intake based on the deviation of the corrected nutritional intake result from the target nutritional intake range and the abnormal changes in the basic physiological data, and to determine the health status as one of normal state, warning state, or risk state.

10. The intelligent monitoring system for nutrition and health status according to claim 9, characterized in that: It also includes a suggestion generation module, which is connected to the status assessment module and is used to generate individualized dietary adjustment suggestions based on the nutritional status and the health status.

Images