Intelligent blood glucose concentration detection method and device based on wearable device

Abstract

The application provides a blood glucose concentration intelligent detection method and device based on a wearable device, and relates to the technical field of medical instruments. The method can input the matching degree and size relationship corresponding to each remaining infrared receiver into a subcutaneous tissue feature determination model to determine the subcutaneous tissue features of the skin area where each remaining infrared receiver is located. The subcutaneous tissue features of the skin area where each infrared receiver is located and the first average blood glucose concentration of n initial blood glucose concentrations collected at n sampling time points are input into an actual blood glucose concentration determination model to determine the actual blood glucose concentrations corresponding to the m infrared receivers. The second average blood glucose concentration of the actual blood glucose concentrations corresponding to the m infrared receivers is determined as the blood glucose concentration collected by the wearable device, and the accuracy of the determined blood glucose concentration is high.

Description

Technical Field

[0001] This application relates to the field of medical device technology, and in particular to a method and device for intelligent detection of blood glucose concentration based on wearable devices. Background Technology

[0002] Diabetes is one of the three major chronic diseases worldwide. It is caused by insufficient insulin secretion or inability to utilize insulin, leading to abnormally high blood sugar levels. However, current medical treatments cannot cure diabetes, and it often presents with no obvious symptoms in its early stages, making it only detectable through blood glucose monitoring. Therefore, safe, rapid, accurate, and real-time blood glucose monitoring is crucial for controlling and treating diabetes.

[0003] Current non-invasive blood glucose testing methods use a combination of infrared spectroscopy and Raman laser spectroscopy to determine blood glucose concentration. However, the accuracy of the blood glucose concentration obtained by these methods remains low. Summary of the Invention

[0004] This application provides a smart blood glucose concentration detection method and device based on wearable devices to solve the problem of low accuracy of blood glucose concentration detection in the prior art.

[0005] In a first aspect, this application provides a smart blood glucose concentration detection method based on a wearable device. The wearable device includes a controller, an infrared light source, and m infrared receivers, where m is an integer greater than or equal to 10. The m infrared receivers are arranged at intervals around a circular ring. Each of the m infrared receivers includes a reference infrared receiver. A subcutaneous tissue detection component is arranged around the reference infrared receiver. The infrared light source is located at the center of the circular ring. The method provided in this application includes: The controller controls the infrared light source to emit infrared rays of the target wavelength for glucose detection at n sampling times within a preset time period, which are then irradiated into the human skin. The controller acquires m infrared spectra from m infrared receivers at each sampling time, where each infrared receiver acquires n infrared spectra within a preset time period. The controller determines the initial blood glucose concentration of each infrared receiver at each sampling time based on the infrared spectrum collected by each infrared receiver at each sampling time. The controller determines the difference in initial blood glucose concentration collected by each infrared receiver at every two adjacent sampling times within a preset duration; For any two adjacent sampling times, the controller determines the average and covariance of the difference in initial blood glucose concentrations corresponding to each infrared receiver; The controller determines the reliability of the blood glucose collection environment conditions corresponding to two adjacent sampling times based on the average value and covariance of the differences between two adjacent sampling times, the average value of the largest and smallest differences among the average values ​​of multiple differences, and the largest and smallest covariances among multiple covariances. The controller summarizes the reliability of the blood glucose collection environment conditions corresponding to two adjacent sampling times for each group; If the variance of the blood glucose collection reliability corresponding to two adjacent sampling times in each group is less than the set variance threshold, the controller determines that the blood glucose collection environment conditions of the wearable device within the preset time are in the standard blood glucose collection environment conditions. The controller receives subcutaneous tissue features of the skin region corresponding to the reference infrared receiver collected by the subcutaneous tissue detection component. The subcutaneous tissue features include at least the thickness of the stratum corneum, the thickness of the subcutaneous fat, and the melanin content of the skin region. The controller determines the difference between the initial blood glucose concentration collected by the reference infrared receiver and the remaining m-1 infrared receivers at any sampling time; The controller determines the matching degree between the subcutaneous tissue characteristics of the skin region where each of the remaining infrared receivers is located and the subcutaneous tissue characteristics of the skin region where the reference infrared receiver is located, based on the average and covariance of the differences corresponding to each of the remaining m-1 infrared receivers, the average of the largest and smallest differences among the averages of multiple differences, and the largest and smallest covariances among multiple covariances; and determines the relationship between the first average blood glucose concentration of the n initial blood glucose concentrations collected by each of the remaining infrared receivers at n sampling times and the first average blood glucose concentration of the n initial blood glucose concentrations collected by the reference infrared receiver at n sampling times. The controller inputs the matching degree and size relationship corresponding to each remaining infrared receiver into the pre-trained subcutaneous tissue feature determination model to determine the subcutaneous tissue features of the skin area where each remaining infrared receiver is located. The subcutaneous tissue feature determination model is trained by inputting multiple first training samples into the first neural network. Each first training sample includes the historical matching degree and historical size relationship corresponding to the historical infrared receiver, as well as the historical actual subcutaneous tissue features of the corresponding skin area. The controller inputs the subcutaneous tissue features of the skin region where each infrared receiver is located, and the first average blood glucose concentration of the n initial blood glucose concentrations collected at n sampling times, into a pre-trained actual blood glucose concentration determination model to determine the actual blood glucose concentration corresponding to the m infrared receivers. The actual blood glucose concentration determination model is trained by inputting multiple second training samples into a second neural network. Each second training sample includes the historical subcutaneous tissue features of the skin region where the historical infrared receivers are located, the historical first average blood glucose concentration of the n historical initial blood glucose concentrations collected at n historical sampling times, and the corresponding historical actual blood glucose concentration. The controller determines the second average blood glucose concentration corresponding to the actual blood glucose concentration of m infrared receivers as the blood glucose concentration collected by the wearable device.

[0006] In some implementations, the controller determines the reliability of the blood glucose collection environment conditions corresponding to two adjacent sampling times based on the average and covariance of the differences between each pair of adjacent sampling times, the average of the largest and smallest differences among the averages of multiple differences, and the largest and smallest covariances among multiple covariances. This includes: For each pair of adjacent sampling times, the controller uses the following formula: , Determine the reliability of the blood glucose collection environment conditions corresponding to two adjacent sampling times in this group, where, This is the average of the differences between two adjacent sampling times in each group. The covariance is the difference between two adjacent sampling times in each group. It is the average of the largest difference among multiple differences. It is the average of the smallest difference among multiple differences. The largest covariance among multiple covariances. The smallest covariance among multiple covariances. This represents the reliability of the blood glucose collection environment conditions corresponding to two adjacent sampling times in this group. This is a preset adjustment factor.

[0007] In some implementations, the controller determines the degree of matching between the subcutaneous tissue features of the skin region where each remaining infrared receiver is located and the subcutaneous tissue features of the skin region where a reference infrared receiver is located, based on the average and covariance of the differences corresponding to each of the remaining m-1 infrared receivers, the average of the largest and smallest differences among the averages of the multiple differences, and the largest and smallest covariances among the multiple covariances. This includes: For each remaining infrared receiver, the controller uses the following formula: The degree of matching between the subcutaneous tissue features of the skin region where the remaining infrared receiver is located and the subcutaneous tissue features of the skin region where the reference infrared receiver is located is determined. This represents the average of the differences corresponding to the remaining infrared receivers. Let be the covariance corresponding to the remaining infrared receiver. It is the average of the largest difference among multiple differences. It is the average of the smallest difference among multiple differences. The largest covariance among multiple covariances. The smallest covariance among multiple covariances. To assess the reliability of the blood glucose collection environment conditions corresponding to the remaining infrared receiver. This is a preset adjustment factor.

[0008] In some implementations, the controller determines the initial blood glucose concentration corresponding to each infrared receiver at each sampling time based on the infrared spectrum collected by each infrared receiver at each sampling time, including: The controller inputs the infrared spectrum collected by each infrared receiver at each sampling time into a pre-trained initial blood glucose concentration determination model to obtain the initial blood glucose concentration corresponding to each infrared receiver at each sampling time. The initial blood glucose concentration determination model is trained by inputting multiple third training samples into a third neural network. Each third training sample includes historical infrared spectra collected by historical infrared receivers and corresponding historical blood glucose concentrations.

[0009] In some implementations, after the controller summarizes the reliability of the blood glucose collection environment conditions corresponding to two adjacent sampling times for each group, the method provided in this application further includes: If the variance of the blood glucose collection reliability corresponding to two adjacent sampling times in each group is greater than or equal to the set variance threshold, the controller will output a prompt message indicating that the reliability of the blood glucose collection environment conditions in which the wearable device is located within a preset time period is low.

[0010] Secondly, this application also provides a smart blood glucose concentration detection device based on a wearable device, configured in the wearable device. The wearable device includes a controller, an infrared light source, and m infrared receivers, where m is an integer greater than or equal to 10. The m infrared receivers are arranged at intervals around a ring, wherein the m infrared receivers include a reference infrared receiver. A subcutaneous tissue detection component is arranged around the reference infrared receiver. The infrared light source is located at the center of the ring. The device includes: The infrared control unit is used to control the infrared light source to emit infrared rays of the target wavelength for detecting glucose at n sampling times within a preset time period, which are then irradiated into the human skin. The spectrum acquisition unit is used to acquire m infrared spectra collected by m infrared receivers at each sampling time, wherein each infrared receiver collects n infrared spectra within a preset time period. The initial blood glucose concentration determination unit is used to determine the initial blood glucose concentration of each infrared receiver at each sampling time based on the infrared spectrum collected by each infrared receiver at each sampling time. The parameter determination unit is used to determine the difference in the initial blood glucose concentration collected by each infrared receiver at every two adjacent sampling times within a preset duration. The parameter determination unit is used to determine the average and covariance of the difference in initial blood glucose concentrations corresponding to each infrared receiver for any two adjacent sampling times. The reliability determination unit is used to determine the reliability of the blood glucose collection environment conditions corresponding to two adjacent sampling times based on the average value and covariance of the differences between two adjacent sampling times, the average value of the largest and smallest differences among the average values ​​of multiple differences, and the largest and smallest covariances among multiple covariances. The data aggregation unit is used to aggregate the reliability of the blood glucose collection environment conditions corresponding to two adjacent sampling times for each group; The environmental condition determination unit is used to determine that the blood glucose collection environment conditions of the wearable device within a preset time period are in the standard blood glucose collection environment conditions, provided that the variance of the blood glucose collection reliability corresponding to two adjacent sampling times in each group is less than the set variance threshold. The subcutaneous tissue feature acquisition unit is used to receive subcutaneous tissue features of the skin area corresponding to the reference infrared receiver acquired by the subcutaneous tissue detection component. The subcutaneous tissue features include at least the thickness of the stratum corneum, the thickness of the subcutaneous fat, and the melanin content of the skin area. The parameter determination unit is also used to determine the difference between the initial blood glucose concentration collected by the reference infrared receiver and the remaining m-1 infrared receivers at any sampling time. The matching degree determination unit is used to determine the matching degree between the subcutaneous tissue features of the skin region where each remaining infrared receiver is located and the subcutaneous tissue features of the skin region where the reference infrared receiver is located, based on the average value and covariance of the differences corresponding to each of the remaining m-1 infrared receivers, the average value of the largest and smallest differences among the average values ​​of multiple differences, and the largest and smallest covariances among the multiple covariances; and to determine the relationship between the first average blood glucose concentration of the n initial blood glucose concentrations collected by each remaining infrared receiver at n sampling times and the first average blood glucose concentration of the n initial blood glucose concentrations collected by the reference infrared receiver at n sampling times. The subcutaneous tissue feature determination unit is used to input the matching degree and size relationship corresponding to each remaining infrared receiver into the pre-trained subcutaneous tissue feature determination model to determine the subcutaneous tissue features of the skin area where each remaining infrared receiver is located. The subcutaneous tissue feature determination model is trained by inputting multiple first training samples into the first neural network. Each first training sample includes the historical matching degree and historical size relationship corresponding to the historical infrared receiver, as well as the historical actual subcutaneous tissue features of the corresponding skin area. The actual blood glucose concentration determination unit is used to input the subcutaneous tissue features of the skin region where each infrared receiver is located, and the first average blood glucose concentration of n initial blood glucose concentrations collected at n sampling times, into a pre-trained actual blood glucose concentration determination model to determine the actual blood glucose concentration corresponding to m infrared receivers. The actual blood glucose concentration determination model is trained by inputting multiple second training samples into a second neural network. Each second training sample includes historical subcutaneous tissue features of the skin region where the historical infrared receiver is located, the historical first average blood glucose concentration of n historical initial blood glucose concentrations collected at n historical sampling times, and the corresponding historical actual blood glucose concentration. The blood glucose concentration determination unit is used to determine the second average blood glucose concentration corresponding to the actual blood glucose concentration of m infrared receivers as the blood glucose concentration collected by the wearable device.

[0011] In some implementations, the reliability determination unit is specifically used to calculate the following formula for each pair of adjacent sampling times: , Determine the reliability of the blood glucose collection environment conditions corresponding to two adjacent sampling times in this group, where, This is the average of the differences between two adjacent sampling times in each group. The covariance is the difference between two adjacent sampling times in each group. It is the average of the largest difference among multiple differences. It is the average of the smallest difference among multiple differences. The largest covariance among multiple covariances. The smallest covariance among multiple covariances. This represents the reliability of the blood glucose collection environment conditions corresponding to two adjacent sampling times in this group. This is a preset adjustment factor.

[0012] In some implementations, the reliability determination unit, specifically for each remaining infrared receiver, uses an equation: The degree of matching between the subcutaneous tissue features of the skin region where the remaining infrared receiver is located and the subcutaneous tissue features of the skin region where the reference infrared receiver is located is determined. This is the average of the differences between two adjacent sampling times in each group. The covariance is the difference between two adjacent sampling times in each group. It is the average of the largest difference among multiple differences. It is the average of the smallest difference among multiple differences. The largest covariance among multiple covariances. The smallest covariance among multiple covariances. To assess the reliability of the blood glucose collection environment conditions corresponding to the remaining infrared receiver. This is a preset adjustment factor.

[0013] In some implementations, the initial blood glucose concentration is specifically used to input the infrared spectrum collected by each infrared receiver at each sampling time into a pre-trained initial blood glucose concentration determination model to obtain the initial blood glucose concentration corresponding to each infrared receiver at each sampling time. The initial blood glucose concentration determination model is obtained by training a third neural network by inputting multiple third training samples into it. Each third training sample includes historical infrared spectra collected by historical infrared receivers and corresponding historical blood glucose concentrations.

[0014] In some embodiments, the apparatus provided in this application further includes: The prompt information output unit is used to output a prompt message indicating that the reliability of the blood glucose collection environment conditions in which the wearable device is located within a preset time period is low when the variance of the blood glucose collection reliability corresponding to two adjacent sampling times in each group is greater than or equal to a set variance threshold.

[0015] This application provides a method and apparatus for intelligent blood glucose concentration detection based on a wearable device. It can determine the difference in initial blood glucose concentration between two adjacent sampling times within a preset duration (e.g., 5 minutes) for each infrared receiver. For example, for infrared receiver 1, the difference in initial blood glucose concentration between the first and second sampling times within the preset duration is determined; for infrared receiver 2, the difference in initial blood glucose concentration between the first and second sampling times within the preset duration is determined, and so on, for infrared receiver m, the difference in initial blood glucose concentration between the first and second sampling times within the preset duration is determined.

[0016] For any two adjacent sampling times in a given group, determine the average and covariance of the differences in initial blood glucose concentrations corresponding to each infrared receiver. For example, for two adjacent sampling times in the first group, determine the average and covariance of the differences in initial blood glucose concentrations corresponding to each infrared receiver; for two adjacent sampling times in the second group, determine the average and covariance of the differences in initial blood glucose concentrations corresponding to each infrared receiver, ..., for two adjacent sampling times in the (n-1)th group, determine the average and covariance of the differences in initial blood glucose concentrations corresponding to each infrared receiver.

[0017] The reliability of the blood glucose collection environment conditions corresponding to two adjacent sampling times is determined by the average value and covariance of the differences between two adjacent sampling times, the average value of the largest and smallest differences among the average values ​​of multiple differences, and the largest and smallest covariances among multiple covariances.

[0018] Summarize the reliability of the blood glucose sampling environment conditions for each group of two adjacent sampling times. For example, summarize the reliability of the blood glucose sampling environment conditions for two adjacent sampling times in group 1, group 2, ..., and group (n-1). Since the blood glucose concentration collected by the user wearing the wearable device on their wrist is easily affected by the blood glucose sampling environment conditions (such as changes in ambient light, relative displacement between the skin and the infrared receiver due to user movement, leading to changes in the optical path length of the infrared light source, etc.), it is necessary to summarize the reliability of the blood glucose sampling environment conditions for each group of two adjacent sampling times.

[0019] If the variance of the blood glucose collection reliability corresponding to two adjacent sampling times in each group is less than the set variance threshold, then the stability and reliability of the current blood glucose collection environment are high, and it is determined that the blood glucose collection environment conditions in which the wearable device is located within the preset time period are in the standard blood glucose collection environment conditions.

[0020] The device receives subcutaneous tissue features of the skin region corresponding to the reference infrared receiver, which are acquired by the subcutaneous tissue detection component. The subcutaneous tissue features include at least the thickness of the stratum corneum, the thickness of the subcutaneous fat, and the melanin content of the skin region.

[0021] The difference between the initial blood glucose concentration collected by the reference infrared receiver and the remaining m-1 infrared receivers at any given sampling time is determined. For example, at each sampling time, the difference between the initial blood glucose concentration collected by the reference infrared receiver and the first remaining infrared receiver is determined, the difference between the reference infrared receiver and the second remaining infrared receiver is determined, and so on, until the initial blood glucose concentration collected by the reference infrared receiver and the (m-1)th remaining infrared receiver is determined. Understandably, the accuracy of blood glucose concentration measurement is also affected by the characteristics of the subcutaneous tissue. Therefore, the matching degree between the subcutaneous tissue characteristics of the skin region where each remaining infrared receiver is located and the subcutaneous tissue characteristics of the skin region where the reference infrared receiver is located can be determined based on the average value and covariance of the differences corresponding to each of the remaining m-1 infrared receivers, the average value of the largest and smallest differences among the average values ​​of multiple differences, and the largest and smallest covariances among the multiple covariances.

[0022] And determine the relationship between the first average blood glucose concentration of the n initial blood glucose concentrations collected by each remaining infrared receiver at n sampling times and the first average blood glucose concentration of the n initial blood glucose concentrations collected by the reference infrared receiver at n sampling times.

[0023] The matching degree and size relationship corresponding to each remaining infrared receiver are input into a pre-trained subcutaneous tissue feature determination model to determine the subcutaneous tissue features of the skin region where each remaining infrared receiver is located. The subcutaneous tissue feature determination model is trained by inputting multiple first training samples into a first neural network. Each first training sample includes the historical matching degree and historical size relationship corresponding to historical infrared receivers, as well as the historical actual subcutaneous tissue features of the corresponding skin region. Since wearable devices are typically small, only one subcutaneous tissue detection component is needed to collect the subcutaneous tissue features of the skin region where each infrared receiver is located, while maintaining the small size of the wearable device, saving costs, and achieving high accuracy.

[0024] The subcutaneous tissue features of the skin region where each infrared receiver is located, and the first average blood glucose concentration of n initial blood glucose concentrations collected at n sampling times, are input into a pre-trained actual blood glucose concentration determination model to determine the actual blood glucose concentration corresponding to m infrared receivers. Since the actual blood glucose concentration determination model is trained by inputting multiple second training samples into a second neural network, each second training sample includes historical subcutaneous tissue features of the skin region where the historical infrared receivers are located, the historical first average blood glucose concentration of n historical initial blood glucose concentrations collected at n historical sampling times, and the corresponding historical actual blood glucose concentration. Thus, the reliability of the determined actual blood glucose concentrations corresponding to the m infrared receivers is high. Furthermore, the accuracy of determining the second average blood glucose concentration of the actual blood glucose concentrations corresponding to the m infrared receivers as the blood glucose concentration collected by the wearable device is also high. Attached Figure Description

[0025] 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 some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0026] Figure 1 This is a schematic diagram of the structure of a wearable device provided in an embodiment of this application; Figure 2 This is a schematic diagram of the intelligent blood glucose concentration detection method based on wearable devices provided in the embodiments of this application. Detailed Implementation

[0027] To make the objectives, technical solutions, and advantages of the embodiments of this application clearer, the technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, not all embodiments. Based on the embodiments of this application, all other embodiments made by those skilled in the art under the guidance of these embodiments are within the scope of protection of this application.

[0028] The terms “first,” “second,” “third,” “fourth,” etc. (if present) in the specification, claims, and accompanying drawings of this application are used to distinguish similar objects and are not necessarily used to describe a particular order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of this application described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms “comprising” and “having,” and any variations thereof, are intended to cover a non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0029] This application provides a method for intelligent blood glucose concentration detection based on wearable devices, applied to wearable devices. For example... Figure 1 As shown, the wearable device includes a controller (not shown in the attached figure), an infrared light source 101, and m infrared receivers 102, where m is an integer greater than or equal to 10. The m infrared receivers 102 are arranged at intervals around a circular ring 103. The m infrared receivers 102 include a reference infrared receiver 104. A subcutaneous tissue detection component 105 is arranged around the reference infrared receiver 104. The infrared light source 101 is located at the center of the circular ring 103. The diameter of the circular ring 103 can be, but is not limited to, 40mm, 50mm, or 60mm, etc., and is not limited here. Figure 2 As shown, the method provided in this application embodiment includes: S201: The controller controls the infrared light source 101 to emit infrared rays of the target wavelength for detecting glucose at n sampling times within a preset time period, which are then irradiated into the human skin.

[0030] For example, infrared rays of the target wavelength for glucose detection are emitted and irradiated into the skin of a human body at 1-second intervals within a 5-minute period. For instance, for infrared receiver 1021, the difference between the initial blood glucose concentration collected at the first and second sampling times within a preset time period is determined; for infrared receiver 1022, the difference between the initial blood glucose concentration collected at the first and second sampling times within a preset time period is determined, ..., for infrared receiver 102m, the difference between the initial blood glucose concentration collected at the first and second sampling times within a preset time period is determined.

[0031] S202: The controller acquires m infrared spectra collected by m infrared receivers 102 at each sampling time, wherein each infrared receiver 102 collects n infrared spectra within a preset time period.

[0032] S203: The controller determines the initial blood glucose concentration of each infrared receiver 102 at each sampling time based on the infrared spectrum collected by each infrared receiver 102 at each sampling time.

[0033] Specifically, the controller inputs the infrared spectrum collected by each infrared receiver 102 at each sampling time into a pre-trained initial blood glucose concentration determination model to obtain the initial blood glucose concentration corresponding to each infrared receiver 102 at each sampling time. The initial blood glucose concentration determination model is trained by inputting multiple third training samples into a third neural network. Each third training sample includes historical infrared spectra collected by the historical infrared receiver 102 and the corresponding historical blood glucose concentration.

[0034] S204: The controller determines the difference in initial blood glucose concentration collected by each infrared receiver 102 at every two adjacent sampling times within a preset duration.

[0035] For example, determine the difference in initial blood glucose concentration collected by the first infrared receiver 102 at two adjacent sampling times in the first group of preset durations; determine the difference in initial blood glucose concentration collected by the first infrared receiver 102 at two adjacent sampling times in the second group of preset durations, ..., determine the difference in initial blood glucose concentration collected by the first infrared receiver 102 at two adjacent sampling times in the (n-1)th group of preset durations; For example, determine the difference in initial blood glucose concentration collected by the second infrared receiver 102 at two adjacent sampling times in the first group of preset durations; determine the difference in initial blood glucose concentration collected by the second infrared receiver 102 at two adjacent sampling times in the second group of preset durations, ..., determine the difference in initial blood glucose concentration collected by the second infrared receiver 102 at two adjacent sampling times in the (n-1)th group of preset durations; For example, the difference in initial blood glucose concentration collected by the m-th infrared receiver 102 at two adjacent sampling times in the first group of preset durations is determined; the difference in initial blood glucose concentration collected by the m-th infrared receiver 102 at two adjacent sampling times in the second group of preset durations is determined; ..., the difference in initial blood glucose concentration collected by the m-th infrared receiver 102 at two adjacent sampling times in the (n-1)-th group of preset durations is determined.

[0036] S205: For any two adjacent sampling times, the controller determines the average value and covariance of the difference in initial blood glucose concentration corresponding to each infrared receiver 102.

[0037] For example, for two adjacent sampling times in the first group, the average value and covariance of the difference in initial blood glucose concentration corresponding to each infrared receiver 102 are determined; for two adjacent sampling times in the second group, the average value and covariance of the difference in initial blood glucose concentration corresponding to each infrared receiver 102 are determined, ..., for two adjacent sampling times in the (n-1)th group, the average value and covariance of the difference in initial blood glucose concentration corresponding to each infrared receiver 102 are determined.

[0038] S206: The controller determines the reliability of the blood glucose collection environment conditions corresponding to two adjacent sampling times based on the average value and covariance of the differences between two adjacent sampling times in each group, the average value of the largest and smallest differences among the average values ​​of multiple differences, and the largest and smallest covariances among multiple covariances.

[0039] For example, for each pair of adjacent sampling times, the controller uses the following formula: , Determine the reliability of the blood glucose collection environment conditions corresponding to two adjacent sampling times in this group, where, This is the average of the differences between two adjacent sampling times in each group. The covariance is the difference between two adjacent sampling times in each group. It is the average of the largest difference among multiple differences. It is the average of the smallest difference among multiple differences. The largest covariance among multiple covariances. The smallest covariance among multiple covariances. This represents the reliability of the blood glucose collection environment conditions corresponding to two adjacent sampling times in this group. This is a preset adjustment factor.

[0040] S207: The controller summarizes the reliability of the blood glucose collection environment conditions corresponding to two adjacent sampling times for each group.

[0041] For example, the reliability of the blood glucose collection environment conditions corresponding to two adjacent sampling times in the first group, the reliability of the blood glucose collection environment conditions corresponding to two adjacent sampling times in the second group, ..., and the reliability of the blood glucose collection environment conditions corresponding to two adjacent sampling times in the (n-1)th group are summarized. Since the blood glucose concentration collected by the user wearing the wearable device on their wrist is easily affected by the blood glucose collection environment conditions (such as changes in ambient light, relative displacement between the skin and the infrared receiver 102 due to user activity, leading to changes in the optical path length of the infrared light source 101, etc.), it is necessary to summarize the reliability of the blood glucose collection environment conditions corresponding to two adjacent sampling times in each group.

[0042] S208: If the variance of the blood glucose collection reliability corresponding to two adjacent sampling times in each group is less than the set variance threshold, the controller determines that the blood glucose collection environment conditions of the wearable device within the preset time period are in the standard blood glucose collection environment conditions.

[0043] In addition, the method provided in this application embodiment further includes: when the variance of the blood glucose collection reliability corresponding to two adjacent sampling times in each group is greater than or equal to a set variance threshold, the controller outputs a prompt message indicating that the reliability of the blood glucose collection environment conditions in which the wearable device is located within a preset time period is low.

[0044] S209: The controller receives subcutaneous tissue features of the skin region corresponding to the reference infrared receiver 104 collected by the subcutaneous tissue detection component 105, wherein the subcutaneous tissue features include at least the thickness of the stratum corneum, the thickness of the subcutaneous fat, and the melanin content of the skin region.

[0045] S210: The controller determines the difference between the initial blood glucose concentration collected by the reference infrared receiver 104 and the remaining m-1 infrared receivers 102 at any sampling time.

[0046] For example, at any sampling time, the difference between the initial blood glucose concentration collected by the reference infrared receiver 104 and the remaining first infrared receiver 102 is determined; the difference between the initial blood glucose concentration collected by the reference infrared receiver 104 and the remaining second infrared receiver 102 is determined; ..., the difference between the initial blood glucose concentration collected by the reference infrared receiver 104 and the remaining (m-1)th infrared receiver 102 is determined. Understandably, the accuracy of blood glucose concentration collection is also affected by the characteristics of the subcutaneous tissue.

[0047] S211: The controller determines the matching degree between the subcutaneous tissue characteristics of the skin region where each of the remaining infrared receivers 102 is located and the subcutaneous tissue characteristics of the skin region where the reference infrared receiver 104 is located, based on the average value and covariance of the differences corresponding to each of the remaining m-1 infrared receivers 102, the average value of the largest and smallest differences among the average values ​​of multiple differences, and the largest and smallest covariances among the multiple covariances; and determines the relationship between the first average blood glucose concentration of the n initial blood glucose concentrations collected by each of the remaining infrared receivers 102 at n sampling times and the first average blood glucose concentration of the n initial blood glucose concentrations collected by the reference infrared receiver 104 at n sampling times.

[0048] In some implementations, the controller determines the degree of matching between the subcutaneous tissue features of the skin region where each of the remaining m-1 infrared receivers 102 is located and the subcutaneous tissue features of the skin region where the reference infrared receiver 104 is located, based on the average and covariance of the differences corresponding to each of the remaining m-1 infrared receivers 102, the average of the largest and smallest differences among the averages of multiple differences, and the largest and smallest covariances among the multiple covariances. For each remaining infrared receiver 102, the controller uses the following formula: The degree of matching between the subcutaneous tissue characteristics of the skin region where the remaining infrared receiver 102 is located and the subcutaneous tissue characteristics of the skin region where the reference infrared receiver 104 is located is determined. This is the average of the differences corresponding to the remaining infrared receivers 102. Let the covariance be the covariance corresponding to the remaining infrared receiver 102. It is the average of the largest difference among multiple differences. It is the average of the smallest difference among multiple differences. The largest covariance among multiple covariances. The smallest covariance among multiple covariances. To assess the reliability of the blood glucose collection environment conditions corresponding to the remaining infrared receiver 102, This is a preset adjustment factor.

[0049] S212: The controller inputs the matching degree and size relationship corresponding to each remaining infrared receiver 102 into the pre-trained subcutaneous tissue feature determination model to determine the subcutaneous tissue features of the skin area where each remaining infrared receiver 102 is located. The subcutaneous tissue feature determination model is obtained by inputting multiple first training samples into the first neural network for training. Each first training sample includes the historical matching degree and historical size relationship corresponding to the historical infrared receiver 102, as well as the historical actual subcutaneous tissue features of the corresponding skin area.

[0050] Since wearable devices are typically small in size, only one subcutaneous tissue detection component 105 is needed to collect the subcutaneous tissue features of the skin area where each infrared receiver 102 is located, thus ensuring the small size of the wearable device, saving costs and achieving high accuracy.

[0051] It should be noted that the first neural network can be, but is not limited to, a convolutional neural network. A convolutional neural network includes an input layer, a first convolutional layer, a first pooling layer, a second convolutional layer, a second pooling layer, a fully connected layer, and an output layer. The first convolutional layer includes 64 convolutional kernels, each with a size of 3×3, and uses ReLU as the activation function. The first pooling layer is a 1D convolutional layer with a pooling window size of 2. The second convolutional layer includes 128 convolutional kernels, each with a size of 3×3, and uses ReLU as the activation function. The second pooling layer is also a 1D convolutional layer with a pooling window size of 2. The fully connected layer is a Dense layer containing 50 neurons and uses ReLU as the activation function. The output layer uses a linear activation function.

[0052] S213: The controller inputs the subcutaneous tissue features of the skin region where each infrared receiver 102 is located, and the first average blood glucose concentration of the n initial blood glucose concentrations collected at n sampling times, into the pre-trained actual blood glucose concentration determination model to determine the actual blood glucose concentrations corresponding to the m infrared receivers 102; wherein, the actual blood glucose concentration determination model is obtained by inputting multiple second training samples into the second neural network for training, and each second training sample includes the historical subcutaneous tissue features of the skin region where the historical infrared receiver 102 is located, the historical first average blood glucose concentration of the n historical initial blood glucose concentrations collected at n historical sampling times, and the corresponding historical actual blood glucose concentration.

[0053] It should be noted that the second neural network can also be a convolutional neural network, which will not be elaborated here.

[0054] S214: The controller determines the second average blood glucose concentration corresponding to the actual blood glucose concentration of m infrared receivers 102 as the blood glucose concentration collected by the wearable device.

[0055] In summary, the intelligent blood glucose concentration detection method based on a wearable device provided in this application embodiment can determine the difference in initial blood glucose concentration between two adjacent sampling times for each infrared receiver 102 within a preset duration (e.g., 5 minutes). For example, for infrared receiver 1021, the difference in initial blood glucose concentration between the first and second sampling times within the preset duration is determined; for infrared receiver 1022, the difference in initial blood glucose concentration between the first and second sampling times within the preset duration is determined, ..., for infrared receiver 102m, the difference in initial blood glucose concentration between the first and second sampling times within the preset duration is determined.

[0056] For any two adjacent sampling times in a given group, determine the average and covariance of the differences in initial blood glucose concentrations corresponding to each infrared receiver 102. For example, for two adjacent sampling times in the first group, determine the average and covariance of the differences in initial blood glucose concentrations corresponding to each infrared receiver 102; for two adjacent sampling times in the second group, determine the average and covariance of the differences in initial blood glucose concentrations corresponding to each infrared receiver 102, ..., for two adjacent sampling times in the (n-1)th group, determine the average and covariance of the differences in initial blood glucose concentrations corresponding to each infrared receiver 102.

[0057] The reliability of the blood glucose collection environment conditions corresponding to two adjacent sampling times is determined by the average value and covariance of the differences between two adjacent sampling times, the average value of the largest and smallest differences among the average values ​​of multiple differences, and the largest and smallest covariances among multiple covariances.

[0058] Summarize the reliability of the blood glucose sampling environment conditions for each group of two adjacent sampling times. For example, summarize the reliability of the blood glucose sampling environment conditions for two adjacent sampling times in the first group, the second group, ..., and the (n-1)th group. Since the blood glucose concentration collected by the user wearing the wearable device on their wrist is easily affected by the blood glucose sampling environment conditions (such as changes in ambient light, relative displacement between the skin and the infrared receiver 102 due to user activity, leading to changes in the optical path length of the infrared light source 101, etc.), it is necessary to summarize the reliability of the blood glucose sampling environment conditions for each group of two adjacent sampling times.

[0059] If the variance of the blood glucose collection reliability corresponding to two adjacent sampling times in each group is less than the set variance threshold, then the stability and reliability of the current blood glucose collection environment are high, and it is determined that the blood glucose collection environment conditions in which the wearable device is located within the preset time period are in the standard blood glucose collection environment conditions.

[0060] The device receives subcutaneous tissue features of the skin region corresponding to the reference infrared receiver 104 collected by the subcutaneous tissue detection component 105. The subcutaneous tissue features include at least the thickness of the stratum corneum, the thickness of the subcutaneous fat, and the melanin content of the skin region.

[0061] The difference between the initial blood glucose concentration collected by the reference infrared receiver 104 and the remaining m-1 infrared receivers 102 at any given sampling time is determined. For example, at any given sampling time, the difference between the initial blood glucose concentration collected by the reference infrared receiver 104 and the remaining first infrared receiver 102 is determined, the difference between the reference infrared receiver 104 and the remaining second infrared receiver 102 is determined, and so on, until the difference between the reference infrared receiver 104 and the remaining m-1 infrared receivers 102 is determined. Understandably, the accuracy of blood glucose concentration measurement is also affected by the characteristics of the subcutaneous tissue. Therefore, the matching degree between the subcutaneous tissue characteristics of the skin region where each of the remaining m-1 infrared receivers 102 is located and the subcutaneous tissue characteristics of the skin region where the reference infrared receiver 104 is located can be determined based on the average value and covariance of the differences corresponding to each of the remaining m-1 infrared receivers 102, the average value of the largest and smallest differences among the average values ​​of multiple differences, and the largest and smallest covariances among the multiple covariances.

[0062] And determine the relationship between the first average blood glucose concentration of the n initial blood glucose concentrations collected by each remaining infrared receiver 102 at n sampling times and the first average blood glucose concentration of the n initial blood glucose concentrations collected by the reference infrared receiver 104 at n sampling times.

[0063] The matching degree and size relationship corresponding to each remaining infrared receiver 102 are input into a pre-trained subcutaneous tissue feature determination model to determine the subcutaneous tissue features of the skin region where each remaining infrared receiver 102 is located. The subcutaneous tissue feature determination model is trained by inputting multiple first training samples into a first neural network. Each first training sample includes the historical matching degree and historical size relationship corresponding to the historical infrared receiver 102, as well as the historical actual subcutaneous tissue features of the corresponding skin region. Since wearable devices are typically small, only one subcutaneous tissue detection component 105 is needed to collect the subcutaneous tissue features of the skin region where each infrared receiver 102 is located, while maintaining the small size of the wearable device, saving costs, and achieving high accuracy.

[0064] The subcutaneous tissue features of the skin region where each infrared receiver 102 is located, and the first average blood glucose concentration of n initial blood glucose concentrations collected at n sampling times, are input into a pre-trained actual blood glucose concentration determination model to determine the actual blood glucose concentration corresponding to m infrared receivers 102. Since the actual blood glucose concentration determination model is trained by inputting multiple second training samples into a second neural network, each second training sample includes historical subcutaneous tissue features of the skin region where the historical infrared receivers 102 are located, the historical first average blood glucose concentration of n historical initial blood glucose concentrations collected at n historical sampling times, and the corresponding historical actual blood glucose concentration. Thus, the reliability of the determined actual blood glucose concentrations corresponding to the m infrared receivers 102 is high. Furthermore, the accuracy of determining the second average blood glucose concentration of the actual blood glucose concentrations corresponding to the m infrared receivers 102 as the blood glucose concentration collected by the wearable device is also high.

[0065] In addition, this application embodiment also provides a smart blood glucose concentration detection device based on a wearable device, configured on the wearable device. It should be noted that the basic principle and technical effect of the smart blood glucose concentration detection device based on a wearable device provided in this application embodiment are the same as those in the above embodiments. For the sake of brevity, any parts not mentioned in this application embodiment can be referred to the corresponding content in the above embodiments. The wearable device includes a controller, an infrared light source 101, and m infrared receivers 102, where m is an integer greater than or equal to 10. The m infrared receivers 102 are arranged at intervals around a ring 103. The m infrared receivers 102 include a reference infrared receiver 104. A subcutaneous tissue detection component 105 is arranged around the reference infrared receiver 104. The infrared light source 101 is located at the center of the ring 103. The device provided in this application embodiment includes: The infrared control unit is used to control the infrared light source 101 to emit infrared rays of the target wavelength for detecting glucose at n sampling times within a preset time period, which are then irradiated into the human skin. The target wavelength can be in the range of 780nm-2500nm, for example, it can be 1000nm, 1500nm or 2000nm, etc., and is not limited here.

[0066] The spectrum acquisition unit is used to acquire m infrared spectra collected by m infrared receivers 102 at each sampling time, wherein each infrared receiver 102 collects n infrared spectra within a preset time period. The initial blood glucose concentration determination unit is used to determine the initial blood glucose concentration of each infrared receiver 102 at each sampling time based on the infrared spectrum collected by each infrared receiver 102 at each sampling time. The parameter determination unit is used to determine the difference in the initial blood glucose concentration collected by each infrared receiver 102 at every two adjacent sampling times within a preset duration. The parameter determination unit is used to determine the average value and covariance of the difference in initial blood glucose concentration corresponding to each infrared receiver 102 for any two adjacent sampling times. The reliability determination unit is used to determine the reliability of the blood glucose collection environment conditions corresponding to two adjacent sampling times based on the average value and covariance of the differences between two adjacent sampling times, the average value of the largest and smallest differences among the average values ​​of multiple differences, and the largest and smallest covariances among multiple covariances. The data aggregation unit is used to aggregate the reliability of the blood glucose collection environment conditions corresponding to two adjacent sampling times for each group; The environmental condition determination unit is used to determine that the blood glucose collection environment conditions of the wearable device within a preset time period are in the standard blood glucose collection environment conditions, provided that the variance of the blood glucose collection reliability corresponding to two adjacent sampling times in each group is less than the set variance threshold. The subcutaneous tissue feature acquisition unit is used to receive subcutaneous tissue features of the skin area corresponding to the reference infrared receiver 104 acquired by the subcutaneous tissue detection component 105. The subcutaneous tissue features include at least the thickness of the stratum corneum, the thickness of the subcutaneous fat, and the melanin content of the skin area. The parameter determination unit is also used to determine the difference between the initial blood glucose concentration collected by the reference infrared receiver 104 and the remaining m-1 infrared receivers 102 at any sampling time. The matching degree determination unit is used to determine the matching degree between the subcutaneous tissue features of the skin region where each remaining infrared receiver 102 is located and the subcutaneous tissue features of the skin region where the reference infrared receiver 104 is located, based on the average value and covariance of the differences corresponding to each of the remaining m-1 infrared receivers 102, the average value of the largest and smallest differences among the average values ​​of multiple differences, and the largest and smallest covariances among the multiple covariances; and to determine the relationship between the first average blood glucose concentration of the n initial blood glucose concentrations collected by each remaining infrared receiver 102 at n sampling times and the first average blood glucose concentration of the n initial blood glucose concentrations collected by the reference infrared receiver 104 at n sampling times. The subcutaneous tissue feature determination unit is used to input the matching degree and size relationship corresponding to each remaining infrared receiver 102 into the pre-trained subcutaneous tissue feature determination model to determine the subcutaneous tissue features of the skin area where each remaining infrared receiver 102 is located. The subcutaneous tissue feature determination model is trained by inputting multiple first training samples into the first neural network. Each first training sample includes the historical matching degree and historical size relationship corresponding to the historical infrared receiver 102, as well as the historical actual subcutaneous tissue features of the corresponding skin area. The actual blood glucose concentration determination unit is used to input the subcutaneous tissue features of the skin region where each infrared receiver 102 is located, and the first average blood glucose concentration of n initial blood glucose concentrations collected at n sampling times, into a pre-trained actual blood glucose concentration determination model to determine the actual blood glucose concentration corresponding to m infrared receivers 102; wherein, the actual blood glucose concentration determination model is trained by inputting multiple second training samples into a second neural network, and each second training sample includes historical subcutaneous tissue features of the skin region where the historical infrared receiver 102 is located, historical first average blood glucose concentration of n historical initial blood glucose concentrations collected at n historical sampling times, and corresponding historical actual blood glucose concentration; The blood glucose concentration determination unit is used to determine the second average blood glucose concentration corresponding to the actual blood glucose concentration of m infrared receivers 102 as the blood glucose concentration collected by the wearable device.

[0067] In some implementations, the reliability determination unit is specifically used to calculate the following formula for each pair of adjacent sampling times: , Determine the reliability of the blood glucose collection environment conditions corresponding to two adjacent sampling times in this group, where, This is the average of the differences between two adjacent sampling times in each group. The covariance is the difference between two adjacent sampling times in each group. It is the average of the largest difference among multiple differences. It is the average of the smallest difference among multiple differences. The largest covariance among multiple covariances. The smallest covariance among multiple covariances. This represents the reliability of the blood glucose collection environment conditions corresponding to two adjacent sampling times in this group. This is a preset adjustment factor.

[0068] In some implementations, the reliability determination unit, specifically for each remaining infrared receiver 102, uses the following formula: The degree of matching between the subcutaneous tissue characteristics of the skin region where the remaining infrared receiver 102 is located and the subcutaneous tissue characteristics of the skin region where the reference infrared receiver 104 is located is determined. This is the average of the differences corresponding to the remaining infrared receivers 102. Let the covariance be the covariance corresponding to the remaining infrared receiver 102. It is the average of the largest difference among multiple differences. It is the average of the smallest difference among multiple differences. The largest covariance among multiple covariances. The smallest covariance among multiple covariances. To assess the reliability of the blood glucose collection environment conditions corresponding to the remaining infrared receiver 102, This is a preset adjustment factor.

[0069] In some implementations, the initial blood glucose concentration is specifically used to input the infrared spectrum collected by each infrared receiver 102 at each sampling time into a pre-trained initial blood glucose concentration determination model to obtain the initial blood glucose concentration corresponding to each infrared receiver 102 at each sampling time. The initial blood glucose concentration determination model is obtained by inputting multiple third training samples into a third neural network for training. Each third training sample includes historical infrared spectra collected by the historical infrared receiver 102 and the corresponding historical blood glucose concentration.

[0070] In some embodiments, the apparatus provided in this application further includes: The prompt information output unit is used to output a prompt message indicating that the reliability of the blood glucose collection environment conditions in which the wearable device is located within a preset time period is low when the variance of the blood glucose collection reliability corresponding to two adjacent sampling times in each group is greater than or equal to a set variance threshold.

[0071] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of this application, and are not intended to limit them. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some or all of the technical features therein. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of this application.

Claims

1. A smart method for detecting blood glucose concentration based on wearable devices, characterized in that, An application is made in a wearable device, the wearable device comprising a controller, an infrared light source, and m infrared receivers, wherein m is an integer greater than or equal to 10, the m infrared receivers being arranged at intervals around a circular ring, wherein the m infrared receivers include a reference infrared receiver, a subcutaneous tissue detection component being arranged around the reference infrared receiver, and the infrared light source being located at the center of the circular ring, the method comprising: The controller controls the infrared light source to emit infrared rays of the target wavelength for detecting glucose at n sampling times within a preset time period, which are then irradiated into the human skin. The controller acquires m infrared spectra from m infrared receivers at each sampling time, wherein each infrared receiver acquires n infrared spectra within the preset time period; The controller determines the initial blood glucose concentration corresponding to each infrared receiver at each sampling time based on the infrared spectrum collected by each infrared receiver at each sampling time; The controller determines the difference in initial blood glucose concentration collected by each of the infrared receivers at every two adjacent sampling times within the preset duration; For any two adjacent sampling times, the controller determines the average value and covariance of the difference in initial blood glucose concentration corresponding to each infrared receiver; The controller determines the reliability of the blood glucose collection environment conditions corresponding to the two adjacent sampling times based on the average value and covariance of the differences between the two adjacent sampling times, the average value of the largest and smallest differences among the average values ​​of multiple differences, and the largest and smallest covariances among the multiple covariances. The controller summarizes the reliability of the blood glucose collection environment conditions corresponding to two adjacent sampling times for each group; If the variance of the reliability of blood glucose collection at two adjacent sampling times in each group is less than the set variance threshold, the controller determines that the blood glucose collection environment conditions of the wearable device within the preset time period are in the standard blood glucose collection environment conditions. The controller receives subcutaneous tissue features of the skin region corresponding to the reference infrared receiver from the subcutaneous tissue detection component, wherein the subcutaneous tissue features include at least the stratum corneum thickness, subcutaneous fat thickness, and skin melanin content of the skin region; The controller determines the difference between the initial blood glucose concentration collected by the reference infrared receiver and the remaining m-1 infrared receivers at any sampling time. The controller determines the matching degree between the subcutaneous tissue features of the skin region where each remaining infrared receiver is located and the subcutaneous tissue features of the skin region where the reference infrared receiver is located, based on the average value and covariance of the differences corresponding to each of the remaining m-1 infrared receivers, the average value of the largest and smallest differences among the average values ​​of multiple differences, and the largest and smallest covariances among the multiple covariances; and determines the relationship between the first average blood glucose concentration of the n initial blood glucose concentrations collected by each remaining infrared receiver at the n sampling times and the first average blood glucose concentration of the n initial blood glucose concentrations collected by the reference infrared receiver at the n sampling times. The controller inputs the matching degree and size relationship corresponding to each remaining infrared receiver into a pre-trained subcutaneous tissue feature determination model to determine the subcutaneous tissue features of the skin region where each remaining infrared receiver is located. The subcutaneous tissue feature determination model is obtained by inputting multiple first training samples into a first neural network for training. Each first training sample includes the historical matching degree and historical size relationship corresponding to the historical infrared receiver, as well as the historical actual subcutaneous tissue features of the corresponding skin region. The controller inputs the subcutaneous tissue features of the skin region where each infrared receiver is located, and the first average blood glucose concentration of the n initial blood glucose concentrations collected at the n sampling times, into a pre-trained actual blood glucose concentration determination model to determine the actual blood glucose concentrations corresponding to the m infrared receivers; wherein, the actual blood glucose concentration determination model is obtained by training a second neural network by inputting multiple second training samples into it, and each second training sample includes the historical subcutaneous tissue features of the skin region where the historical infrared receiver is located, the historical first average blood glucose concentration of the n historical initial blood glucose concentrations collected at the n historical sampling times, and the corresponding historical actual blood glucose concentration; The controller determines the second average blood glucose concentration corresponding to the actual blood glucose concentration of the m infrared receivers as the blood glucose concentration collected by the wearable device.

2. The method according to claim 1, characterized in that, The controller determines the reliability of the blood glucose collection environment conditions corresponding to the two adjacent sampling times based on the average and covariance of the differences between each pair of adjacent sampling times, the average of the largest and smallest differences among the averages of multiple differences, and the largest and smallest covariances among multiple covariances. This includes: For each pair of adjacent sampling times, the controller uses the following formula: ， Determine the reliability of the blood glucose collection environment conditions corresponding to two adjacent sampling times in this group, where, This is the average of the differences between two adjacent sampling times in each group. The covariance is the difference between two adjacent sampling times in each group. It is the average of the largest difference among multiple differences. It is the average of the smallest difference among multiple differences. The largest covariance among multiple covariances. The smallest covariance among multiple covariances. This represents the reliability of the blood glucose collection environment conditions corresponding to two adjacent sampling times in this group. This is a preset adjustment factor.

3. The method according to claim 1, characterized in that, The controller determines the matching degree between the subcutaneous tissue features of the skin region where each remaining infrared receiver is located and the subcutaneous tissue features of the skin region where the reference infrared receiver is located, based on the average and covariance of the differences corresponding to each of the remaining m-1 infrared receivers, the average of the largest and smallest differences among the averages of multiple differences, and the largest and smallest covariances among the multiple covariances. This includes: For each remaining infrared receiver, the controller uses the following formula: The degree of matching between the subcutaneous tissue features of the skin region where the remaining infrared receiver is located and the subcutaneous tissue features of the skin region where the reference infrared receiver is located is determined, wherein, This represents the average of the differences corresponding to the remaining infrared receivers. Let be the covariance corresponding to the remaining infrared receiver. It is the average of the largest difference among multiple differences. It is the average of the smallest difference among multiple differences. The largest covariance among multiple covariances. The smallest covariance among multiple covariances. To assess the reliability of the blood glucose collection environment conditions corresponding to the remaining infrared receiver. This is a preset adjustment factor.

4. The method according to claim 1, characterized in that, The controller determines the initial blood glucose concentration corresponding to each infrared receiver at each sampling time based on the infrared spectrum collected by each infrared receiver at each sampling time, including: The controller inputs the infrared spectrum collected by each infrared receiver at each sampling time into a pre-trained initial blood glucose concentration determination model to obtain the initial blood glucose concentration corresponding to each infrared receiver at each sampling time. The initial blood glucose concentration determination model is obtained by training multiple third training samples into a third neural network. Each third training sample includes historical infrared spectra collected by historical infrared receivers and corresponding historical blood glucose concentrations.

5. The method according to claim 1, characterized in that, After the controller summarizes the reliability of the blood glucose collection environment conditions corresponding to two adjacent sampling times for each group, the method further includes: If the variance of the reliability of blood glucose collection at two adjacent sampling times in each group is greater than or equal to a set variance threshold, the controller outputs a prompt message indicating that the reliability of the blood glucose collection environment conditions in which the wearable device is located within the preset time period is low.

6. A smart blood glucose concentration detection device based on a wearable device, characterized in that, The device is configured for use in a wearable device, which includes a controller, an infrared light source, and m infrared receivers, where m is an integer greater than or equal to 10. The m infrared receivers are arranged at intervals around a circular ring, and each of the m infrared receivers includes a reference infrared receiver. A subcutaneous tissue detection component is arranged around the reference infrared receiver. The infrared light source is located at the center of the circular ring. The device includes: An infrared control unit is used to control the infrared light source to emit infrared rays of the target wavelength for detecting glucose at n sampling times within a preset time period, which are then irradiated into the skin of the human body. The spectral acquisition unit is used to acquire m infrared spectra collected by m infrared receivers at each sampling time, wherein each infrared receiver collects n infrared spectra within the preset time period; An initial blood glucose concentration determination unit is used to determine the initial blood glucose concentration corresponding to each infrared receiver at each sampling time based on the infrared spectrum collected by each infrared receiver at each sampling time. The parameter determination unit is used to determine the difference in initial blood glucose concentration collected by each of the infrared receivers at every two adjacent sampling times within the preset duration. The parameter determination unit is used to determine the average value and covariance of the difference in initial blood glucose concentration corresponding to each of the infrared receivers for any two adjacent sampling times. The reliability determination unit is used to determine the reliability of the blood glucose collection environment conditions corresponding to the two adjacent sampling times based on the average value and covariance of the differences corresponding to the two adjacent sampling times in each group, the average value of the largest difference and the average value of the smallest difference among the average values ​​of multiple differences, and the largest covariance and the smallest covariance among multiple covariances. The data aggregation unit is used to aggregate the reliability of the blood glucose collection environment conditions corresponding to two adjacent sampling times for each group; The environmental condition determination unit is used to determine that the blood glucose collection environmental conditions of the wearable device within the preset time period are standard blood glucose collection environmental conditions when the variance of the blood glucose collection reliability corresponding to two adjacent sampling times in each group is less than a set variance threshold. The subcutaneous tissue feature acquisition unit is used to receive subcutaneous tissue features of the skin area corresponding to the reference infrared receiver acquired by the subcutaneous tissue detection component, wherein the subcutaneous tissue features include at least the stratum corneum thickness, subcutaneous fat thickness and skin melanin content of the skin area; The parameter determination unit is also used to determine the difference between the initial blood glucose concentration collected by the reference infrared receiver and the remaining m-1 infrared receivers at any of the sampling times. The matching degree determination unit is used to determine the matching degree between the subcutaneous tissue features of the skin region where each remaining infrared receiver is located and the subcutaneous tissue features of the skin region where the reference infrared receiver is located, based on the average value and covariance of the differences corresponding to each of the remaining m-1 infrared receivers, the average value of the largest and smallest differences among the average values ​​of multiple differences, and the largest and smallest covariances among the multiple covariances; and to determine the relationship between the first average blood glucose concentration of the n initial blood glucose concentrations collected by each remaining infrared receiver at the n sampling times and the first average blood glucose concentration of the n initial blood glucose concentrations collected by the reference infrared receiver at the n sampling times. The subcutaneous tissue feature determination unit is used to input the matching degree and size relationship corresponding to each remaining infrared receiver into a pre-trained subcutaneous tissue feature determination model to determine the subcutaneous tissue features of the skin area where each remaining infrared receiver is located. The subcutaneous tissue feature determination model is obtained by inputting multiple first training samples into a first neural network for training. Each first training sample includes the historical matching degree and historical size relationship corresponding to the historical infrared receiver, as well as the historical actual subcutaneous tissue features of the corresponding skin area. The actual blood glucose concentration determination unit is used to input the subcutaneous tissue features of the skin region where each infrared receiver is located, and the first average blood glucose concentration of the n initial blood glucose concentrations collected at the n sampling times, into a pre-trained actual blood glucose concentration determination model to determine the actual blood glucose concentration corresponding to m infrared receivers; wherein, the actual blood glucose concentration determination model is obtained by inputting multiple second training samples into a second neural network for training, and each second training sample includes historical subcutaneous tissue features of the skin region where the historical infrared receiver is located, the historical first average blood glucose concentration of the n historical initial blood glucose concentrations collected at the n historical sampling times, and the corresponding historical actual blood glucose concentration; The blood glucose concentration determination unit is used to determine the second average blood glucose concentration corresponding to the actual blood glucose concentration of the m infrared receivers as the blood glucose concentration collected by the wearable device.

7. The apparatus according to claim 6, characterized in that, The reliability determination unit is specifically used to apply the following formula to each pair of adjacent sampling times: ， Determine the reliability of the blood glucose collection environment conditions corresponding to two adjacent sampling times in this group, where, This is the average of the differences between two adjacent sampling times in each group. The covariance is the difference between two adjacent sampling times in each group. It is the average of the largest difference among multiple differences. It is the average of the smallest difference among multiple differences. The largest covariance among multiple covariances. The smallest covariance among multiple covariances. This represents the reliability of the blood glucose collection environment conditions corresponding to two adjacent sampling times in this group. This is a preset adjustment factor.

8. The apparatus according to claim 6, characterized in that, The reliability determination unit is specifically used for the controller to employ the following formula for each remaining infrared receiver: The degree of matching between the subcutaneous tissue features of the skin region where the remaining infrared receiver is located and the subcutaneous tissue features of the skin region where the reference infrared receiver is located is determined, wherein, This is the average of the differences between two adjacent sampling times in each group. The covariance is the difference between two adjacent sampling times in each group. It is the average of the largest difference among multiple differences. It is the average of the smallest difference among multiple differences. The largest covariance among multiple covariances. The smallest covariance among multiple covariances. To assess the reliability of the blood glucose collection environment conditions corresponding to the remaining infrared receiver. This is a preset adjustment factor.

9. The apparatus according to claim 6, characterized in that, The initial blood glucose concentration is specifically used to input the infrared spectrum collected by each of the infrared receivers at each sampling time into a pre-trained initial blood glucose concentration determination model to obtain the initial blood glucose concentration corresponding to each of the infrared receivers at each sampling time. The initial blood glucose concentration determination model is obtained by inputting multiple third training samples into a third neural network for training. Each third training sample includes historical infrared spectra collected by historical infrared receivers and corresponding historical blood glucose concentrations.

10. The apparatus according to claim 6, characterized in that, The device further includes: The prompt information output unit is used to output a prompt information indicating that the reliability of the blood glucose collection environment conditions of the wearable device within the preset time period is low when the variance of the blood glucose collection reliability corresponding to two adjacent sampling times in each group is greater than or equal to a set variance threshold.

Images