User health state management method and system based on smart wearable device

Abstract

The application discloses a user health state management method and system based on an intelligent wearable device; first, a radio frequency transmitting module of the intelligent wearable device transmits electromagnetic wave signals of specific wavelengths to a preset body area of a user, and receives corresponding feedback signals; a pre-trained glucose feedback signal determination model is used to determine the received feedback signals; after confirming the existence of the glucose feedback signals, a target blood glucose state evaluation model is further called to process the signals, and autoregressive characteristics of the feedback signals are output; by comparing with pre-stored blood glucose health state characteristics, the blood glucose health state of the user is evaluated; thus, real-time monitoring and health state evaluation of biochemical indexes such as blood glucose of the user are realized, and the application efficiency of the intelligent wearable device in the health management field is improved.

Description

Technical Field

[0001] The present invention relates to the field of smart wearable devices, and in particular to a user health status management method and system based on smart wearable devices. Background Art

[0002] As people's focus on health continues to grow, the application of smart wearable devices in the field of health monitoring is becoming increasingly widespread. However, existing technologies have shortcomings in monitoring user health status, especially blood sugar health status. For example, existing monitoring methods are not accurate enough and lack real-time performance, making it difficult to accurately determine the user's blood sugar health status and unable to meet people's higher demands for health management. Summary of the Invention

[0003] The purpose of the present invention is to provide a user health status management method and system based on smart wearable devices.

[0004] In a first aspect, an embodiment of the present invention provides a user health status management method based on a smart wearable device, comprising:

[0005] In response to a health status assessment instruction triggered by a preset period, initializing the radio frequency transmission module and the feedback signal receiving module of the smart wearable device;

[0006] Transmitting an electromagnetic wave signal of a preset wavelength to a preset body area of ​​a target user through the radio frequency transmission module;

[0007] receiving, by the feedback signal receiving module, a feedback signal to be determined corresponding to the electromagnetic wave signal of the preset wavelength;

[0008] calling a pre-trained glucose feedback signal determination model to perform glucose feedback signal determination on the feedback signal to be determined, and obtaining a determination result indicating whether the feedback signal to be determined includes a glucose feedback signal;

[0009] If the determination result indicates that a glucose feedback signal exists, obtaining a glucose feedback signal to be processed;

[0010] calling a pre-trained target blood glucose state assessment model to process the glucose feedback signal and output an autoregressive feature of the feedback signal;

[0011] performing a comparison operation on the feedback signal autoregressive feature and the pre-stored blood glucose health status feature to obtain a comparison result;

[0012] When the comparison result indicates that the characteristic distance between the autoregressive characteristic of the feedback signal and the blood glucose health status characteristic is less than a preset characteristic distance threshold, the blood glucose health status corresponding to the glucose feedback signal is obtained.

[0013] In a second aspect, an embodiment of the present invention provides a server system, including a server, wherein the server is configured to execute the method described in the first aspect.

[0014] Compared with the existing technology, the beneficial effects provided by the present invention include: adopting a user health status management method and system based on a smart wearable device disclosed by the present invention, transmitting an electromagnetic wave signal of a specific wavelength to a preset body area of ​​the user through the radio frequency transmission module of the smart wearable device, and receiving a corresponding feedback signal. Using a pre-trained glucose feedback signal judgment model, the received feedback signal is subjected to glucose feedback signal judgment. After confirming the existence of a glucose feedback signal, the target blood glucose state evaluation model is further called to process the signal, and the autoregressive features of the feedback signal are output. The user's blood glucose health status is evaluated by comparing with the pre-stored blood glucose health status features. With this design, real-time monitoring of biochemical indicators such as blood glucose and health status evaluation of the user are realized, thereby improving the application efficiency of smart wearable devices in the field of health management. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] To more clearly illustrate the technical solutions of the embodiments of the present invention, the following briefly describes the drawings required for use in the embodiments. It should be understood that the following drawings illustrate only certain embodiments of the present invention and should not be construed as limiting the scope of the present invention. Those skilled in the art can, without inventive effort, derive other relevant drawings from these drawings.

[0016] Figure 1 A schematic diagram of the steps of a method for managing user health status based on a smart wearable device provided in an embodiment of the present invention;

[0017] Figure 2 A schematic block diagram of the structure of a computer device provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0018] To make the objectives, technical solutions, and advantages of the embodiments of the present invention more apparent, the technical solutions of the embodiments of the present invention will be described clearly and completely below in conjunction with the accompanying drawings of the embodiments of the present invention. It should be understood that the described embodiments are only a portion of the embodiments of the present invention, not all of them. Generally, the components of the embodiments of the present invention described and illustrated in the drawings herein may be arranged and designed in a variety of different configurations.

[0019] The specific embodiments of the present invention are described in detail below with reference to the accompanying drawings.

[0020] In order to solve the technical problems in the above background technology, Figure 1This is a flow chart of a method for managing the user health status based on a smart wearable device provided in an embodiment of the present disclosure. The method for managing the user health status based on a smart wearable device is introduced in detail below.

[0021] Step S201, in response to a health status assessment instruction triggered at a preset period, initializing a radio frequency transmitting module and a feedback signal receiving module of the smart wearable device;

[0022] Step S202, transmitting an electromagnetic wave signal of a preset wavelength to a preset body area of ​​the target user through the radio frequency transmission module;

[0023] Step S203, receiving, by the feedback signal receiving module, a feedback signal to be determined corresponding to the electromagnetic wave signal of the preset wavelength;

[0024] Step S204, calling a pre-trained glucose feedback signal determination model to perform glucose feedback signal determination on the feedback signal to be determined, and obtaining a determination result indicating whether the feedback signal to be determined includes a glucose feedback signal;

[0025] Step S205, if the determination result indicates that a glucose feedback signal exists, obtaining a glucose feedback signal to be processed;

[0026] Step S206, calling a pre-trained target blood glucose state assessment model to process the glucose feedback signal and output an autoregressive feature of the feedback signal;

[0027] Step S207, performing a comparison operation on the autoregressive feature of the feedback signal and the pre-stored blood glucose health status feature to obtain a comparison result;

[0028] Step S208: When the comparison result indicates that the characteristic distance between the autoregressive characteristic of the feedback signal and the blood glucose health status characteristic is less than a preset characteristic distance threshold, the blood glucose health status corresponding to the glucose feedback signal is obtained.

[0029] In an embodiment of the present invention, for example, the preset period is 8:00 a.m. every day. At 8:00 a.m., the server receives a health status assessment instruction. The server sends an initialization instruction to the smart wearable device. Upon receiving the instruction, the smart wearable device initializes its internal RF transmission module and feedback signal receiving module. For example, the smart wearable device can be a smart bracelet. When initializing the RF transmission module, the device resets parameters such as the transmission frequency and power to ensure that the transmitted electromagnetic wave signal is stable and meets preset standards. The feedback signal receiving module clears previously received buffered data and adjusts parameters such as the receiving sensitivity to the initial state in preparation for receiving new feedback signals. For example, the preset body area is the inner wrist, and the preset wavelength of the electromagnetic wave signal is a radio wave of a specific frequency. The server controls the RF transmission module of the smart wearable device to transmit an electromagnetic wave signal of this specific frequency to the inner wrist of the user. For example, the RF transmission module in the smart bracelet generates and transmits electromagnetic wave signals with a frequency of 1 GHz. These electromagnetic wave signals can penetrate the skin to a certain depth and interact with tissue and body fluids on the inner wrist. After the RF transmission module transmits electromagnetic wave signals, these signals are reflected, refracted, and scattered as they penetrate the tissue and body fluids on the inside of the user's wrist. The feedback signal receiving module receives these electromagnetic wave feedback signals after these interactions. For example, the signals received by the feedback signal receiving module may contain frequency and amplitude variations caused by factors such as tissue composition and blood flow. These variations are the feedback signals to be determined. The server pre-stores and loads a glucose feedback signal determination model trained with a large amount of data. After receiving the feedback signals to be determined, the server inputs these signals into the model for determination. For example, the model analyzes and calculates multiple features of the feedback signal, such as spectral characteristics, amplitude variations, and phase variations. If the model determines that these features match the characteristic patterns associated with glucose, the result is a glucose feedback signal; otherwise, it is determined not to be present. For example, after analyzing the feedback signal to be determined, the model finds that the intensity variations in certain frequency bands within its spectral characteristics match the expected pattern of interaction between glucose molecules and electromagnetic waves, thus determining that the feedback signal contains a glucose feedback signal. If the determination result indicates that the feedback signal to be determined contains a glucose feedback signal, the server further obtains these glucose feedback signals to be processed. For example, the server receives more detailed and accurate data about glucose feedback signals from smart wearable devices, including information such as the signal's duration and intensity profile. The server then processes the glucose feedback signals using a pre-trained target blood glucose state assessment model. For example, the model performs complex mathematical operations and feature extraction on the glucose feedback signals, ultimately outputting autoregressive features of the feedback signals. This autoregressive feature can reflect the changing trends and patterns of blood glucose levels.For example, an autoregressive feature is a vector containing multiple numerical values, each representing the changing trends and correlations of blood glucose levels at different time points. The server pre-stores feature data for various blood glucose health states. The newly obtained autoregressive feature of the feedback signal is compared with these pre-stored features. For example, the comparison operation can be performed by calculating the Euclidean distance or cosine similarity between the two feature vectors. For example, the pre-stored blood glucose health state features include characteristic patterns for normal blood glucose, hyperglycemia, and hypoglycemia. Through comparison, the calculated similarity between the autoregressive feature of the feedback signal and the normal blood glucose characteristic pattern is 0.8, the similarity with the hyperglycemia characteristic pattern is 0.2, and the similarity with the hypoglycemia characteristic pattern is 0.1. If the comparison results indicate that the characteristic distance between the autoregressive feature of the feedback signal and the blood glucose health state feature is less than a preset threshold, the server determines the user's current blood glucose health state. For example, the preset characteristic distance threshold is 0.5. Since the similarity between the autoregressive feature of the feedback signal and the normal blood glucose characteristic pattern is 0.8 (greater than 0.5), the server determines that the user's current blood glucose health state is normal.

[0030] In an embodiment of the present invention, the glucose feedback signal determination model is obtained in the following manner.

[0031] Calling the glucose feedback signal determination model to perform glucose feedback signal determination on the invalid instantaneous feedback signal instance and the valid instantaneous feedback signal instance, respectively, and obtaining an invalid instantaneous determination result and a valid instantaneous result accordingly;

[0032] The invalid instantaneous feedback signal instance is obtained by adding an instantaneous feedback signal to the invalid instance, and the valid instantaneous feedback signal instance is obtained by adding an instantaneous feedback signal to the valid instance corresponding to the invalid instance;

[0033] determining an error parameter of a first error function based on the instance target value of the invalid instance, the instance target value of the valid instance, the invalid instantaneous determination result, and the valid instantaneous result;

[0034] Calling the glucose feedback signal determination model to perform glucose feedback signal determination on the invalid continuous feedback signal instance and the valid continuous feedback signal instance, respectively, to obtain an invalid continuous determination result and a valid continuous determination result;

[0035] The invalid continuous feedback signal instance is obtained by adding a continuous feedback signal to the invalid instance, and the valid continuous feedback signal instance is obtained by adding a continuous feedback signal to the valid instance;

[0036] determining a first error parameter of a second error function based on the instance target value of the invalid continuous feedback signal instance, the instance target value of the valid continuous feedback signal instance, the invalid continuous determination result, and the valid continuous determination result;

[0037] Based on the error parameter of the first error function and the first error parameter of the second error function, network parameters of the glucose feedback signal determination model are optimized.

[0038] In an embodiment of the present invention, the server first prepares a large amount of sample data, including invalid and valid instances. These instances are obtained by screening and classifying a large amount of previously collected user data. To obtain an invalid instantaneous feedback signal instance, for example, the server selects a blood glucose monitoring data instance originally marked as invalid. This invalid instance may be due to data deviation caused by equipment failure, strong external interference, or other abnormal conditions. The server then adds a simulated instantaneous feedback signal to this invalid instance. This instantaneous feedback signal can be a short, high-intensity abnormal signal, such as a large voltage fluctuation occurring within a very short period of time. The server then invokes a glucose feedback signal determination model to perform a glucose feedback signal determination on this invalid instance with the instantaneous feedback signal added. After analysis and calculation, the model provides an invalid instantaneous determination result. For example, the model determines that this invalid instance with the instantaneous feedback signal still does not meet the characteristics of a glucose feedback signal and determines it as "invalid." For a valid instantaneous feedback signal instance, the server selects a blood glucose monitoring data instance originally marked as valid, corresponding to the invalid instance. This valid instance can be accurate data from a healthy individual or an individual with stable blood glucose under normal monitoring conditions. Then, a similar instantaneous feedback signal is added to this valid instance. The glucose feedback signal determination model is then called for determination. The model generates a valid instantaneous determination result, such as "valid," indicating that the model believes the valid instance with the instantaneous feedback signal meets the characteristics of a glucose feedback signal. Next, the server determines the error parameter of the first error function based on the instance target value of the invalid instance, the instance target value of the valid instance, and the invalid instantaneous determination result and valid instantaneous result just obtained. For example, the instance target value of the invalid instance is set to 0 (indicating invalid), while the instance target value of the valid instance is set to 1 (indicating valid). If the invalid instantaneous determination result is 0 and the valid instantaneous determination result is 1, then the instance target value is fully matched, and the error parameter of the first error function will be small, possibly close to 0. However, if the invalid instantaneous determination result is 1 or the valid instantaneous determination result is 0, a larger error will occur, and the error parameter of the first error function will increase accordingly. The server then continues to obtain invalid continuous feedback signal instances and valid continuous feedback signal instances. For example, for an invalid continuous feedback signal instance, the server reselects the previous invalid instance and adds a longer-duration feedback signal to it. This continuous feedback signal may simulate a continuous abnormality over a period of time, such as data changes caused by sustained low-intensity interference or chronic equipment failure. For valid continuous feedback signal instances, a similar continuous feedback signal is added to the corresponding valid instance.The server then invokes the glucose feedback signal determination model to determine the two continuous feedback signal instances, obtaining invalid continuous determination results and valid continuous determination results. For example, the model determines the invalid continuous feedback signal instance as "invalid" and the valid continuous feedback signal instance as "valid." Next, the server determines the first error parameter of the second error function based on the instance target value of the invalid continuous feedback signal instance (set to 0), the instance target value of the valid continuous feedback signal instance (set to 1), the invalid continuous determination results, and the valid continuous determination results. If the determination result is completely consistent with the instance target value, the first error parameter of the second error function will be small; conversely, if there is a deviation, the error parameter will increase. Finally, the server optimizes the network parameters of the glucose feedback signal determination model based on the error parameters of the first error function and the first error parameters of the second error function. For example, the server uses an optimization algorithm (such as a gradient descent algorithm) to calculate the direction and magnitude of network parameter adjustment based on these two error parameters. If the error parameter is large, indicating that the model's determination result deviates significantly from the actual target value, the server will make significant adjustments to the network parameters to improve the accuracy and reliability of the model. By repeating the above process continuously until the error parameter reaches an acceptably small value, the glucose feedback signal determination model can effectively determine various instantaneous and continuous feedback signal instances accurately.

[0039] In the embodiments of the present invention, the following implementation modes are also provided.

[0040] Calling the glucose feedback signal determination model to perform glucose feedback signal determination on the invalid instance and the valid instance corresponding to the invalid instance, and obtaining invalid instance determination results and valid instance determination results respectively;

[0041] determining a first error parameter of a third error function based on the instance target value of the invalid instance, the instance target value of the valid instance, the invalid instance determination result, and the valid instance determination result;

[0042] Optimizing the network parameters of the glucose feedback signal determination model based on the error parameter of the first error function and the first error parameter of the second error function includes:

[0043] Based on the error parameter of the first error function, the first error parameter of the second error function, and the first error parameter of the third error function, network parameters of the glucose feedback signal determination model are optimized.

[0044] In an embodiment of the present invention, for example, the server first obtains a series of invalid instances and corresponding valid instances. For example, the invalid instances may be data from individuals with serious illnesses resulting in abnormal blood glucose monitoring data, while the corresponding valid instances are normal blood glucose data from healthy individuals under the same monitoring conditions. The server then uses the glucose feedback signal judgment model to judge these invalid instances. For example, for a particular invalid instance, the model analyzes and processes its data and arrives at a judgment result of "invalid." Similarly, for the corresponding valid instance, the model also provides a judgment result, such as "valid." The server then determines the first error parameter of the third error function based on the instance target value of the invalid instance (e.g., 0 for invalid) and the instance target value of the valid instance (e.g., 1 for valid), as well as the invalid and valid instance judgment results obtained recently. For example, there are multiple pairs of invalid and valid instances. For the first invalid instance, its target value is 0, and the model judgment result is also 0, which is consistent, so the error in this part is small. For the second valid instance, its target value is 1, and the model judgment result is also 1, which is also no error. However, if for the third invalid instance, the target value is 0 but the model judgment result is 1, a large error occurs. By comprehensively calculating the target values ​​and judgment results for all these instances, the first error parameter of the third error function can be obtained. Next, when optimizing the network parameters of the glucose feedback signal judgment model, the server comprehensively considers the error parameter of the first error function, the first error parameter of the second error function, and the first error parameter of the third error function. For example, the server first normalizes these three error parameters so that they are compared and calculated at the same order of magnitude. For example, the error parameter of the first error function is 0.1, the first error parameter of the second error function is 0.2, and the first error parameter of the third error function is 0.15. The server then assigns weights to each of these three error functions. The weights can be determined based on experience, data characteristics, or through methods such as cross-validation. For example, the weight assigned to the first error function is 0.3, the weight assigned to the second error function is 0.4, and the weight assigned to the third error function is 0.3. Next, the server calculates the combined error: 0.1 × 0.3 + 0.2 × 0.4 + 0.15 × 0.3 = 0.165. Based on this combined error, the server uses an optimization algorithm, such as stochastic gradient descent, to adjust the network parameters of the glucose feedback signal determination model. If the combined error is large, indicating that the model's performance is not good enough, the server will make larger adjustments to the network parameters, such as increasing the learning rate to converge to better parameter values ​​more quickly. If the combined error is small, the server will appropriately reduce the learning rate and make more refined adjustments to avoid over-adjustment that may lead to model instability.In the process of adjusting the network parameters, the server will continuously repeat the above steps. For each round of new data, the server will recalculate the error parameters, update the comprehensive error, and adjust the network parameters accordingly. For example, after the first round of adjustment, the server will test the model again with a new batch of invalid instances and valid instances, recalculate the parameters of the three error functions, and obtain a new comprehensive error of 0.12. Compared with the previous 0.165, it has decreased, indicating that the performance of the model is gradually improving. However, the server will continue to adjust until the comprehensive error reaches a preset minimum value or after a certain number of iterations, the performance of the model reaches the expected level. By continuously optimizing the network parameters in this way, the glucose feedback signal judgment model can judge invalid instances and valid instances more and more accurately, improving its reliability and accuracy in practical applications.

[0045] In an embodiment of the present invention, the optimization of the network parameters of the glucose feedback signal determination model based on the error parameters of the first error function, the first error parameters of the second error function, and the first error parameters of the third error function can be implemented through the following examples.

[0046] respectively obtaining a weighting coefficient of the first error function, a weighting coefficient of the second error function, and a weighting coefficient of the third error function;

[0047] Obtaining an error parameter of a total error function based on the error parameter and weighting coefficient of the first error function, the first error parameter and weighting coefficient of the second error function, and the first error parameter and weighting coefficient of the third error function;

[0048] The network parameters of the glucose feedback signal determination model are optimized according to the error parameters of the total error function.

[0049] In an embodiment of the present invention, illustratively, the server has first obtained the error parameters of the first error function, the first error parameters of the second error function, and the first error parameters of the third error function through a series of operations. In order to more accurately optimize the network parameters of the glucose feedback signal determination model, the server needs to obtain the weighting coefficients of these three error functions respectively. For example, through analysis of historical data and multiple experimental verifications, the server determines that the weighting coefficient of the first error function is 0.2, which means that when the total error is comprehensively considered, the influence of the first error function accounts for 20%. Similarly, the server determines that the weighting coefficient of the second error function is 0.5 and the weighting coefficient of the third error function is 0.3. Next, the server obtains the error parameters of the total error function based on these error parameters and weighting coefficients. For example, the error parameter of the first error function is 0.1, the first error parameter of the second error function is 0.15, and the first error parameter of the third error function is 0.08. The error parameter of the total error function is calculated as follows: Total error = 0.1 × 0.2 + 0.15 × 0.5 + 0.08 × 0.3 = 0.02 + 0.075 + 0.024 = 0.119. After obtaining the error parameter of the total error function, 0.119, the server optimizes the network parameters of the glucose feedback signal determination model based on this value. The server uses an optimization algorithm, such as Stochastic Gradient Descent (SGD). In each iteration, the server calculates the gradient of the network parameters based on the error parameter of the total error function. For example, if the network parameters of the current model are a set of values ​​[w1, w2, w3, ..., wn], the server calculates the gradient of each parameter with respect to the total error by taking the derivative [dw1, dw2, dw3, ..., dwn]. The server then updates the network parameters based on the learning rate and gradient. The learning rate is an important factor that controls the step size of parameter updates. For example, the learning rate is 0.01. The new network parameters = original network parameters - learning rate × gradient. That is, the new network parameters are [w1 - 0.01 × dw1, w2 - 0.01 × dw2, w3 - 0.01 × dw3, ..., wn - 0.01 × dwn]. After updating the network parameters, the server tests again with new data, recalculates the error parameters of the first error function, the first error parameters of the second error function, and the first error parameters of the third error function, and calculates the new error parameters of the total error function using the above method. If the error parameters of the new total error function are smaller than the previous ones, it indicates that the model is moving towards a better optimization, and the server continues to update the network parameters in this way. However, if the error parameters of the total error function do not decrease or even increase, the server can adjust the learning rate or re-examine the data and model settings. For example, after the first round of optimization, the error parameter of the total error function decreased from 0.119 to 0.105.The server continues with the next round of optimization, repeating this process until the error parameter of the total error function reaches a stable minimum or meets the preset stopping condition. During this process, the server also performs some additional operations. For example, if it is found that a certain data subset has too large or too small an impact on the total error, the server will re-evaluate the corresponding error function weighting coefficient to ensure the rationality and effectiveness of the optimization process. For example, the server will also monitor the performance of the model on the validation set to prevent the model from overfitting. If the total error on the training set continues to decrease, but the performance on the validation set does not improve or even decreases, the server will implement regularization measures, such as L1 and L2 regularization, to limit the size of the network parameters and prevent the model from being too complex. Through this repeated calculation, update, and adjustment, the server continuously optimizes the network parameters of the glucose feedback signal judgment model, enabling the model to more accurately judge the glucose feedback signal.

[0050] In an embodiment of the present invention, before calling the glucose feedback signal determination model to perform glucose feedback signal determination on invalid instantaneous feedback signal instances and valid instantaneous feedback signal instances, the embodiment of the present invention further provides the following implementation manner.

[0051] Swapping the instance target value of the invalid instance and the instance target value of the valid instance to obtain the enhanced instance target value of the invalid instance and the enhanced instance target value of the valid instance;

[0052] determining a second error parameter of the third error function based on the enhanced instance target value of the invalid instance, the enhanced instance target value of the valid instance, the invalid instance determination result, and the valid instance determination result;

[0053] generating an instantaneous feedback signal based on a second error parameter of the third error function;

[0054] The instantaneous feedback signal is added to the invalid instance and the valid instance respectively, so as to obtain the invalid instantaneous feedback signal instance and the valid instantaneous feedback signal instance accordingly.

[0055] In an embodiment of the present invention, illustratively, the server first obtains invalid instances and valid instances. For example, an invalid instance is a set of data from a blood glucose monitoring device acquired under extreme interference, while a valid instance is accurate data acquired under ideal monitoring conditions. Before invoking the glucose feedback signal determination model to determine invalid and valid instantaneous feedback signal instances, the server swaps the instance target values ​​of the invalid and valid instances. Originally, the instance target value of the invalid instance may be marked as 0, while the instance target value of the valid instance is marked as 1. After the swap, the invalid instance receives an enhanced instance target value of 1, while the valid instance receives an enhanced instance target value of 0. Next, the server determines the second error parameter of the third error function based on the enhanced instance target value of the invalid instance, the enhanced instance target value of the valid instance, and the previous determination results for the invalid and valid instances. For example, the server first calculates the error for the invalid instance. For example, for an invalid instance, the previous determination result was "invalid," but due to the swapped instance target values, an error is now generated. The server calculates the error value for this portion using a specific error calculation method, comprehensively considering all relevant data and the enhanced instance target value for the invalid instance. Similarly, for valid instances, errors are calculated in the same manner. The server then combines these errors to obtain a second error parameter for the third error function. Based on this second error parameter, the server generates an instantaneous feedback signal. This instantaneous feedback signal can be an electrical signal with specific frequency and amplitude variations. The server then adds the generated instantaneous feedback signal to both invalid and valid instances. For example, for a data sequence [d1, d2, d3, ..., dn] for an invalid instance, the server adds the instantaneous feedback signal [f1, f2, f3, ..., fm] according to a specific rule to obtain a new data sequence [d1+f1, d2+f2, d3+f3, ..., dn+fm] for the invalid instantaneous feedback signal instance. Similar operations are performed for valid instances to obtain a valid instantaneous feedback signal instance. Through this processing, the server enriches the data used to train the glucose feedback signal determination model, enabling the model to better learn and identify glucose feedback signal characteristics under different circumstances, thereby improving the model's accuracy and robustness. In actual operation, the server processes a large number of invalid and valid instances, and fine-tunes and optimizes the generated transient feedback signals to ensure that the added invalid and valid transient feedback signal instances can effectively improve the model training effect. For example, in one round of processing, the server performed the above operations on 100 invalid instances and 100 valid instances. When calculating the error, the server considers multiple characteristic dimensions of the data, such as the peak value, mean, and variance of the signal. For the generated transient feedback signal, the server adjusts the frequency range and amplitude of the signal based on the performance of the previous model training and the characteristics of the data.If the model's training effectiveness isn't significantly improved after adding the instantaneous feedback signal, the server reanalyzes the error parameters and the generated instantaneous feedback signal to identify potential issues and make improvements. For another example, the server also visualizes the data for instances after adding the instantaneous feedback signal, visually observing the distribution and changes in the data to better understand and optimize the entire processing process. By continuously looping through these steps, the server gradually refines the model's training data, enabling the glucose feedback signal determination model to function more accurately and reliably.

[0056] In the embodiment of the present invention, the generation of the instantaneous feedback signal based on the second error parameter of the third error function can be implemented through the following examples.

[0057] determining, based on a second error parameter of the third error function, a parameter update direction of the third error function with respect to the network parameter, and obtaining an optimization guidance value of the glucose feedback signal determination model;

[0058] The optimized guidance amount of the glucose feedback signal determination model is determined as an instantaneous feedback signal.

[0059] In this embodiment of the present invention, the server has already obtained the second error parameter of the third error function. This error parameter reflects the degree of deviation in the current glucose feedback signal determination model when processing invalid and valid instances. First, the server determines the parameter update direction of the third error function for the network parameters based on this second error parameter. The server implements this process through a series of mathematical calculations and analyses. For example, the third error function is a complex mathematical expression related to the network parameters, such as E = f(w1, w2, w3, ..., wn), where w1 through wn are network parameters. By taking the partial derivative of this function with respect to each parameter, the server can determine the gradient direction of each parameter. If the second error parameter is large, indicating a large deviation in the model, the gradient value calculated by the server will also be relatively large, indicating that the network parameters require a larger adjustment. On the other hand, if the second error parameter is small, the gradient value will also be correspondingly small, and the adjustment range of the network parameters will be smaller. For example, for a network parameter w1, the server calculates a gradient of -0.5, which means that w1 needs to be updated in a decreasing direction to reduce the error. For another network parameter, w2, the gradient is 0.3, indicating that w2 needs to be updated in an increasing direction. The server comprehensively considers the gradient directions of all network parameters to determine the parameter update direction of the third error function for each network parameter. This update direction serves as the optimization guidance for the glucose feedback signal determination model. The server then determines this optimization guidance as an instantaneous feedback signal. For example, the server organizes the parameter update direction calculated above, including the adjustment amplitude and direction of each parameter, into a specific data format or signal form. This signal can be recognized and applied by subsequent processing. For example, this instantaneous feedback signal is represented as a set of values ​​[dw1, dw2, dw3, ..., dwn], where dw1 through dwn correspond to the adjustment values ​​for each network parameter. In actual operation, the server processes large amounts of data and complex network structures simultaneously. For each calculation and update, the server performs rigorous accuracy and plausibility checks to ensure that the generated instantaneous feedback signal effectively guides model optimization. For example, when processing a neural network with thousands of nodes and connections, the server utilizes parallel computing technology to accelerate gradient calculation and update direction determination. The server also compares and analyzes current calculation results with historical data to promptly identify potential anomalies, such as exploding or vanishing gradients. For example, the server dynamically adjusts the algorithms and parameters used to calculate gradients and determine update directions based on the data distribution and model complexity. If the data is noisy or uncertain, the server may employ more robust optimization algorithms or add data preprocessing steps to improve the accuracy and effectiveness of the instantaneous feedback signal.In this way, the server can generate an effective instantaneous feedback signal based on the second error parameter of the third error function, providing key guidance for the optimization of the glucose feedback signal determination model.

[0060] In the embodiment of the present invention, the adding of the instantaneous feedback signal to the invalid instance and the valid instance respectively to obtain the invalid instantaneous feedback signal instance and the valid instantaneous feedback signal instance correspondingly can be implemented through the following examples.

[0061] Determining a weighting coefficient of the instantaneous feedback signal, and performing a weighting operation on the instantaneous feedback signal based on the weighting coefficient to obtain a weighted instantaneous feedback signal;

[0062] The weighted instantaneous feedback signal is added to the invalid instance and the valid instance respectively to obtain the invalid instantaneous feedback signal instance and the valid instantaneous feedback signal instance.

[0063] In an embodiment of the present invention, illustratively, the server has generated an instantaneous feedback signal, which will then be added to the invalid instance and the valid instance. First, the server determines the weighting coefficient of the instantaneous feedback signal. The determination of this weighting coefficient can be based on a variety of factors, such as the previous model training history, the characteristic distribution of the data, or calculated by pre-set rules and algorithms. For example, the server determines that the weighting coefficient is 0.2 based on previous training experience and the characteristics of the current data. Then, the server performs a weighting operation on the instantaneous feedback signal based on this weighting coefficient. For example, the original value of the instantaneous feedback signal is [10, 20, 30, 40, 50]. By multiplying it by the weighting coefficient 0.2, the weighted instantaneous feedback signal is [2, 4, 6, 8, 10]. Next, the server adds this weighted instantaneous feedback signal to the invalid instance and the valid instance respectively. For example, an invalid instance is a set of blood glucose monitoring data, such as [120, 130, 125, 118, 122]. By adding the corresponding weighted instantaneous feedback signals, the invalid instantaneous feedback signal instance is obtained as [122, 134, 131, 126, 132]. Similarly, for a valid instance, for example, originally [90, 88, 92, 89, 91], after adding the weighted instantaneous feedback signal, the valid instantaneous feedback signal instance is obtained as [92, 92, 98, 97, 101]. In actual processing, the server can handle a large number of invalid and valid instances simultaneously. For example, the server may process 1000 invalid instances and 1000 valid instances at once. For each instance, the weighting and addition operations are precisely performed according to the above steps. When determining the weighting coefficients, the server considers the importance and sensitivity of different data types. For example, if the invalid instance data typically has large fluctuations, the server will choose a smaller weighting coefficient to avoid excessively interfering with the characteristics of the original data. For valid instances, if the data is relatively stable, the weighting coefficient will be slightly increased to enhance the impact of the feedback signal on model training. The server also monitors and evaluates instances after the weighted instantaneous feedback signal is added in real time. If the added instance causes deviation or instability in model training, the server will readjust the weighting coefficient or re-examine the generation method and characteristics of the instantaneous feedback signal. For example, after the server added the weighted instantaneous feedback signal in the first round, it found that the model was overly sensitive to certain types of blood sugar fluctuations in subsequent training. After analysis, the server determined that the weighting coefficient was too large. Therefore, the server adjusted the weighting coefficient to 0.1, regenerated and added the weighted instantaneous feedback signal, and repeated the model training to observe whether the results improved. For another example, the server dynamically adjusts the weighting coefficient and instantaneous feedback signal characteristics based on different time periods or user groups.For example, for morning blood glucose monitoring data, due to the unique physiological state of the human body, the server will adopt different weighting strategies and feedback signal generation methods to improve the model's accuracy and adaptability during this specific period. Through such meticulous operations and continuous optimization and adjustment, the server can accurately incorporate weighted instantaneous feedback signals into invalid and valid instances, providing richer and more effective data for training the glucose feedback signal judgment model, thereby continuously improving the model's performance and accuracy.

[0064] In an embodiment of the present invention, the glucose feedback signal determination is executed by calling a discrimination component in the glucose feedback signal determination model, and the glucose feedback signal determination model further includes a synthesis component;

[0065] Before calling the glucose feedback signal determination model to perform glucose feedback signal determination on the invalid continuous feedback signal instance and the valid continuous feedback signal instance, the embodiment of the present invention further provides the following implementation manner.

[0066] calling a synthesis component in a glucose feedback signal determination model to generate a continuous feedback signal based on an input preset continuous signal;

[0067] The continuous feedback signal is added to the invalid instance and the valid instance respectively, so as to obtain the invalid continuous feedback signal instance and the valid continuous feedback signal instance accordingly.

[0068] In an embodiment of the present invention, the server utilizes a glucose feedback signal determination model comprising a discrimination component and a synthesis component during glucose feedback signal determination. Before determining invalid and valid continuous feedback signal instances, the server first invokes the synthesis component within the model. For example, the preset continuous signal input to the synthesis component is a series of analog electrical signal values ​​sampled at specific time intervals, such as [5, 10, 15, 20, 25], with each value representing the signal strength at a specific moment. After receiving this preset continuous signal, the synthesis component begins generating the continuous feedback signal using its internal algorithms and logic. This generation process involves filtering, amplifying, and modulating the input signal. For example, the synthesis component may low-pass filter the input signal to remove high-frequency noise, then linearly amplify the filtered signal to increase its amplitude by a certain multiple. Subsequently, the processed signal is converted into a continuous feedback signal with specific characteristics through a specific modulation method, such as frequency modulation or phase modulation. For example, the continuous feedback signal generated after the above processing is [30, 40, 50, 60, 70]. Next, the server adds this continuous feedback signal to both invalid and valid instances. For example, an invalid instance is a set of blood glucose concentration measurements, such as [8.0, 9.0, 8.5, 7.5, 8.2]. By adding the corresponding continuous feedback signals, the resulting invalid continuous feedback signal instance is [38.0, 49.0, 58.5, 67.5, 78.2]. For a valid instance, for example, [5.0, 4.8, 5.2, 4.9, 5.1], after adding the continuous feedback signal, the resulting valid continuous feedback signal instance is [35.0, 44.8, 55.2, 64.9, 75.1]. In practice, the server can handle a large number of invalid and valid instances simultaneously. For example, there may be thousands or even more invalid and valid instances waiting to be processed. For each invalid and valid instance, the server strictly follows the above steps to generate and add the continuous feedback signal to ensure data accuracy and consistency. During the continuous feedback signal generation process, the server adjusts the parameters of the synthesis component based on different requirements and data characteristics. For example, if the input preset continuous signal has large fluctuations, the server will increase the filtering intensity to obtain a smoother continuous feedback signal. At the same time, the server will also perform quality checks on the instances after the continuous feedback signal is added. If it is found that some instances have abnormal values ​​or do not conform to the expected data pattern after being added, the server will re-examine the operation and addition process of the synthesis component, identify possible problems and correct them. For example, during the inspection process, the server found that the data in a certain invalid continuous feedback signal instance deviated significantly from the normal range. After analysis, it was found that the synthesis component had a parameter error during the modulation process. The server adjusted the parameters in a timely manner, regenerated and added the continuous feedback signal to ensure the rationality of the instance data.In addition, the server may also dynamically change the preset continuous signal input to the synthesis component according to different application scenarios and user needs. For example, for blood glucose monitoring of patients with specific diseases, a preset continuous signal with specific characteristics will be used to generate a continuous feedback signal that is more consistent with the characteristics of the disease, thereby improving the model's judgment accuracy for such special cases. Through such precise and accurate operations, the server can provide rich and accurate invalid continuous feedback signal instances and valid continuous feedback signal instances for subsequent glucose feedback signal judgments, thereby continuously optimizing and improving the performance and reliability of the glucose feedback signal judgment model.

[0069] In the embodiment of the present invention, the adding of the continuous feedback signal to the invalid instance and the valid instance respectively to obtain the invalid continuous feedback signal instance and the valid continuous feedback signal instance correspondingly can be implemented through the following examples.

[0070] Determining a weighting coefficient of the continuous feedback signal, and performing a weighting operation on the continuous feedback signal based on the weighting coefficient to obtain a weighted continuous feedback signal;

[0071] The weighted continuous feedback signal is added to the invalid instance and the valid instance respectively to obtain the invalid continuous feedback signal instance and the valid continuous feedback signal instance.

[0072] In an embodiment of the present invention, illustratively, in the process of processing the combination of the continuous feedback signal with the invalid instance and the valid instance, the server first needs to determine the weighting coefficient of the continuous feedback signal. For example, based on the previous model training effect and the characteristics of the data, the server determines that the weighting coefficient of the continuous feedback signal is 0.3 after a series of calculations and analyses. Next, the server performs a weighting operation on the continuous feedback signal based on this weighting coefficient. For example, the original value of the continuous feedback signal is [100, 200, 300, 400, 500]. By multiplying it by the weighting coefficient of 0.3, the weighted continuous feedback signal obtained is [30, 60, 90, 120, 150]. Then, the server adds this weighted continuous feedback signal to the invalid instance and the valid instance respectively. For example, an invalid instance is a set of blood glucose concentration data, such as [12.5, 13.0, 12.8, 12.2, 12.6]. By adding the corresponding weighted continuous feedback signals, the resulting invalid continuous feedback signal instance is [42.5, 73.0, 102.8, 132.2, 142.6]. For a valid instance, for example, the original data set is [5.5, 5.8, 6.0, 5.2, 5.6]. After adding the weighted continuous feedback signal, the resulting valid continuous feedback signal instance is [35.5, 65.8, 96.0, 115.2, 125.6]. In actual operation, the server processes a large number of invalid and valid instances simultaneously. For example, the server may have to process tens of thousands of invalid instances and the same number of valid instances at once. For each invalid and valid instance, the server precisely follows the above steps. When determining the weighting coefficient, the server considers multiple factors. For example, if the data for an invalid instance fluctuates significantly while the continuous feedback signal is strong, the server will select a smaller weighting coefficient to avoid excessively impacting the original data. Conversely, if the data for a valid instance is relatively stable and the continuous feedback signal is expected to have a more significant impact on model training, the weighting coefficient will be appropriately increased. The server also monitors and evaluates the instances after the weighted continuous feedback signal is added in real time. For example, the server checks whether the data distribution of newly generated invalid and valid continuous feedback signal instances is reasonable and conforms to the expected pattern. For example, if the server discovers during monitoring that the data of some invalid continuous feedback signal instances exhibits unusual clustering or dispersion after the weighted continuous feedback signal is added, the server will immediately reassess the rationality of the weighting coefficient and may adjust the weighting coefficient, regenerate, and add the weighted continuous feedback signal. Furthermore, the server dynamically adjusts the weighting coefficient and processing method based on different data sources and application scenarios. For example, for data from specific age groups or disease groups, the server will use different strategies to determine the weighting coefficient to ensure that the model can accurately adapt to different situations.Through such delicate and rigorous operations, the server can ensure that the generated invalid continuous feedback signal instances and valid continuous feedback signal instances are accurate and effective, providing reliable data support for the subsequent training and optimization of the glucose feedback signal determination model.

[0073] In an embodiment of the present invention, after optimizing the network parameters of the glucose feedback signal determination model based on the error parameters of the first error function and the first error parameters of the second error function, the embodiment of the present invention further provides the following implementation methods.

[0074] calling the optimized glucose feedback signal determination model to perform glucose feedback signal determination on the invalid continuous feedback signal instance and the valid continuous feedback signal instance, respectively, and obtaining an invalid continuous optimization determination result and a valid continuous optimization determination result respectively;

[0075] determining a second error parameter of the second error function based on the instance target value of the invalid instance, the instance target value of the valid instance, the invalid continuous optimization determination result, and the valid continuous optimization determination result;

[0076] Based on the error parameter of the first error function and the second error parameter of the second error function, network parameters of the optimized glucose feedback signal determination model are optimized.

[0077] In an embodiment of the present invention, illustratively, after the server completes preliminary optimization of the network parameters of the glucose feedback signal determination model based on the error parameters of the first error function and the first error parameters of the second error function, it proceeds to subsequent operations. First, the server invokes the preliminarily optimized glucose feedback signal determination model to perform glucose feedback signal determination on invalid and valid continuous feedback signal instances. For example, an invalid continuous feedback signal instance is a series of glucose-related data exhibiting specific characteristics over a long period of time, such as sustained low-intensity fluctuations that do not conform to normal glucose fluctuation patterns. A valid continuous feedback signal instance is data exhibiting normal and stable glucose fluctuations over the same period of time. After the optimized model determines these instances, the server obtains invalid continuous optimization determination results and valid continuous optimization determination results. For example, for an invalid continuous feedback signal instance, the model determines it as "invalid," while for a valid continuous feedback signal instance, the model determines it as "valid." Next, the server determines the second error parameter of the second error function based on the instance target value of the invalid instance, the instance target value of the valid instance, and the newly obtained invalid continuous optimization determination results and valid continuous optimization determination results. For example, the instance target value for an invalid instance is set to 0 (indicating invalid), and the instance target value for a valid instance is set to 1 (indicating valid). If the model's judgment results for a series of invalid continuous feedback signal instances are completely consistent with the instance target value, that is, all are judged as "invalid," then the error in this portion is relatively small. However, if some instances are mistakenly judged as "valid," a corresponding error will be generated. Similarly, for valid continuous feedback signal instances, if the model's judgment results are all consistent with the instance target value ("valid"), the error is relatively small; if some are mistakenly judged as "invalid," an error will be generated. The server comprehensively considers the differences between all these judgment results and the instance target value and calculates a second error parameter of the second error function using a specific algorithm. Then, based on the error parameter of the first error function and the second error parameter of the second error function, the server further optimizes the network parameters of the previously optimized glucose feedback signal judgment model. For example, the server performs a weighted sum of the error parameter of the first error function and the second error parameter of the second error function and adjusts the model's network parameters based on the summation result. For example, the first error function has an error parameter of 0.1 and a weight of 0.4; the second error function has a second error parameter of 0.05 and a weight of 0.6. The weighted sum is 0.4 × 0.1 + 0.6 × 0.05 = 0.07. Based on this result, the server uses an optimization algorithm, such as stochastic gradient descent (SGD), to adjust the model's network parameters.For example, if the calculated gradient of a network parameter is -0.02 and the learning rate is set to 0.01, the new network parameter value will be updated to the original parameter value + learning rate × gradient. In practice, the server processes a large number of invalid and valid continuous feedback signal instances simultaneously to obtain a more comprehensive and accurate error assessment. For example, the server processes 1,000 invalid and 1,000 valid continuous feedback signal instances. The server carefully records and analyzes the judgment results for each instance. If, during the optimization process, the model performance falls short of expectations, the server will further adjust the error function weights and learning rate or re-examine the data preprocessing steps to ensure that the model can more accurately judge glucose feedback signals. For example, the server will compare the current optimization results with previous optimization results to observe the trend of the error parameter. If the error parameter does not significantly decrease after multiple optimizations, the server will consider increasing the amount of training data or adopting a more complex model structure. Through this continuous iterative optimization process, the server gradually improves the accuracy and reliability of the glucose feedback signal judgment model.

[0078] In an embodiment of the present invention, before calling the glucose feedback signal determination model to perform glucose feedback signal determination on invalid instantaneous feedback signal instances and valid instantaneous feedback signal instances, the embodiment of the present invention further provides the following implementation manner.

[0079] Acquire basic invalid instances of multiple information sources, and acquire basic valid instances of multiple information sources corresponding to the basic invalid instances;

[0080] Performing an integration operation on the basic invalid instances of the multiple information sources to obtain the invalid instance;

[0081] An integration operation is performed on the basic valid instances of the multiple information sources to obtain the valid instance.

[0082] In an embodiment of the present invention, illustratively, before invoking the glucose feedback signal determination model to determine invalid and valid instantaneous feedback signal instances, the server first needs to obtain basic invalid and valid instances from multiple information sources and perform an integration operation on them. The server is connected to multiple different data sources, which may include different medical institutions, health monitoring equipment manufacturers, or research institutions. From the first information source, the server obtains a batch of basic invalid instances. For example, a series of blood glucose monitoring data is obtained from the database of a large hospital. This data is marked as inaccurate or invalid blood glucose measurement results due to equipment failure, special patient physiological conditions, or other abnormal conditions. For example, this data was collected at different times, in different departments, and for different patients, encompassing a variety of possible abnormal conditions. Next, from a second information source, such as a cloud platform of a professional health technology company, the server obtains another set of basic invalid instances. This data set may be collected using a new wearable health device with different measurement accuracy and data formats, but is also marked as invalid. The server obtains basic invalid instances from multiple information sources in the same manner. After obtaining basic invalid instances from multiple information sources, the server begins the integration operation. For example, the server first cleans these basic invalid instances from different information sources, removing any possible duplicate data, erroneous data, or data with inconsistent formats. The server then standardizes this data, converting data with different measurement units and measurement intervals to a standardized format. For example, if blood glucose values ​​from different sources are expressed in millimoles per liter (mmol / L) or milligrams per deciliter (mg / dL), the server will convert all data to a unified unit. For inconsistent measurement intervals, the server will perform interpolation or sampling to ensure that all data have consistent time intervals. After cleaning and standardization, the server merges and aggregates these basic invalid instances to obtain a comprehensive invalid instance that encompasses a wide range of invalid cases. Similarly, the server also obtains basic valid instances corresponding to the basic invalid instances from multiple information sources. These basic valid instances may come from blood glucose monitoring data from other medical institutions that has undergone rigorous quality control and verification, or from accurate data collected under controlled experimental conditions. The server performs similar data cleaning, standardization, and aggregation on these basic valid instances from multiple sources as it does on the basic invalid instances, ultimately obtaining a comprehensive valid instance that represents accurate and valid blood glucose monitoring results. In practice, servers can handle massive amounts of data and complex data formats. For example, a server may obtain data from dozens or even hundreds of different information sources, each of which may provide thousands of basic instances.During the data cleaning process, the server uses advanced data analysis algorithms and tools to quickly and accurately identify and remove abnormal data. During standardization, the server can reference internationally accepted medical data standards and specifications to ensure data consistency and comparability. After integration is complete, the server also conducts quality assessments and verifications on the invalid and valid instances obtained, for example through random sampling inspections and comparisons with known standard data sets, to ensure the accuracy and reliability of the integration results. Through such comprehensive and detailed data acquisition and integration operations, the server provides a high-quality, broadly representative data foundation for the subsequent training and optimization of the glucose feedback signal determination model.

[0083] In an embodiment of the present invention, the blood glucose status assessment model is obtained in the following manner.

[0084] performing a segmentation operation on the glucose feedback signal instance in the glucose feedback signal instance set to obtain a first glucose feedback signal segment and a second glucose feedback signal segment;

[0085] calling a first integrated unit in the blood glucose status assessment model, processing the first glucose feedback signal segment to obtain a first glucose feedback signal feature, and performing autoregressive processing and feature expansion on the first glucose feedback signal feature to obtain a first extended autoregressive feature;

[0086] calling a second integrated unit in the blood glucose status assessment model, processing the second glucose feedback signal segment to obtain a second glucose feedback signal feature, and performing autoregressive processing and feature expansion on the second glucose feedback signal feature to obtain a second extended autoregressive feature;

[0087] obtaining a first feature vector based on the first extended autoregressive feature and the second extended autoregressive feature of the same glucose feedback signal instance, and obtaining a second feature vector based on the first extended autoregressive feature of the first glucose feedback signal instance and the first extended autoregressive feature or the second extended autoregressive feature of the other glucose feedback signal instances in the set of glucose feedback signal instances;

[0088] Obtaining a first cost parameter according to the first eigenvector, the second eigenvector, and the number of glucose feedback signal instances in the glucose feedback signal instance set, and obtaining a second cost parameter according to the first glucose feedback signal feature and the second glucose feedback signal feature;

[0089] The first integrated unit is structurally optimized according to the first cost parameter and the second cost parameter to obtain a target blood glucose state assessment model for assessing the blood glucose health state of the target user's glucose feedback signal.

[0090] In an embodiment of the present invention, for example, the server first obtains a set of glucose feedback signal instances. This set includes a large number of glucose feedback signal instances from different users, at different times, and under different monitoring conditions. The server then performs a segmentation operation on each glucose feedback signal instance in the set. For example, the segmentation is based on time intervals, with the portion of each instance before a specific time point being labeled as the first glucose feedback signal segment, and the portion after that time point being labeled as the second glucose feedback signal segment. For example, for a glucose feedback signal instance lasting 10 minutes, the server uses the 5th minute as the segmentation point, with the first 5 minutes of data as the first glucose feedback signal segment, and the last 5 minutes of data as the second glucose feedback signal segment. Next, the server invokes the first integrated unit in the blood glucose status assessment model to process the first glucose feedback signal segment. The first integrated unit may include a series of feature extraction algorithms and modules. For example, the first integrated unit extracts a series of numerical features by performing frequency domain analysis, time domain analysis, and statistical calculations on the first glucose feedback signal segment. These features collectively constitute the first glucose feedback signal features. The first glucose feedback signal features are then subjected to autoregressive processing. This involves establishing a mathematical model to describe how these features change over time and predict their likely future values. Feature expansion builds on autoregressive processing by adding relevant auxiliary features, such as those related to physiological cycles and meal times, to produce the first extended autoregressive feature. Simultaneously, the server invokes the second integrated unit in the blood glucose status assessment model to process the second glucose feedback signal segment. This second integrated unit can employ a different, yet complementary, feature extraction method from the first.

[0091] For example, the second integrated unit may focus more on analyzing the signal's morphology and trends, extracting a second glucose feedback signal feature that is different from but complementary to the first integrated unit. Similarly, the second glucose feedback signal feature is subjected to autoregressive processing and feature expansion to obtain a second extended autoregressive feature. The server then obtains a first feature vector based on the first and second extended autoregressive features of the same glucose feedback signal instance. For example, the first extended autoregressive feature is a vector containing 10 values, and the second extended autoregressive feature is a vector containing 8 values. The server concatenates them to form a first feature vector containing 18 values. For the first glucose feedback signal instance in the set, the server obtains a second feature vector based on its first extended autoregressive feature and the first or second extended autoregressive features of other glucose feedback signal instances. For example, the server compares and calculates the first extended autoregressive feature of the first glucose feedback signal instance with the corresponding features of the other 100 instances, and obtains a second feature vector containing multiple values ​​using mathematical methods (such as calculating the mean, variance, covariance, etc.). Next, the server obtains a first cost parameter based on the first eigenvector, the second eigenvector, and the number of glucose feedback signal instances in the glucose feedback signal instance set. For example, each value in the first and second eigenvectors is transformed and weighted using a function, and then summed, averaged, or calculated based on the number of instances to obtain a first cost parameter that reflects the performance of the model in the current training phase. Simultaneously, the server obtains a second cost parameter based on the first and second glucose feedback signal features. For example, by calculating the difference, similarity, or distance metric between the two features, the server obtains the second cost parameter, which reflects the accuracy of the model during the feature extraction phase. Finally, the server optimizes the structure of the first integrated unit based on the first and second cost parameters. This involves adjusting parameters such as the neuron connection weights, number of layers, and activation function in the first integrated unit. For example, if the first cost parameter is large, indicating that the overall model performance is suboptimal, the server will significantly adjust the structure of the first integrated unit, such as increasing the number of layers or adjusting the neuron connection structure. If the second cost parameter is large, indicating that feature extraction is inaccurate, the server will optimize the parameters of the feature extraction algorithm or replace certain feature extraction modules. In actual operation, the server will perform multiple iterations of this process. For example, after the first round of optimization, the server recalculates the cost parameters and finds that the model performance has improved, but it still has not reached the ideal state. Therefore, it continues with the next round of optimization, constantly adjusting the model structure and parameters until the first cost parameter and the second cost parameter both reach an acceptable small value. At this time, the blood glucose status assessment model obtained can accurately assess the blood glucose health status of the target user's glucose feedback signal.In addition, the server may also consider other factors during the optimization process, such as computing resource limitations, model operation efficiency, adaptability to new data, etc., to ensure that the final target blood glucose status assessment model has good performance and practicality in actual applications.

[0092] In an embodiment of the present invention, the glucose feedback signal instance in the glucose feedback signal instance set includes a first glucose feedback signal instance;

[0093] The performing of the segmentation operation on the glucose feedback signal instance in the glucose feedback signal instance set to obtain the first glucose feedback signal segment and the second glucose feedback signal segment may be implemented through the following example.

[0094] performing a cutting operation on the first glucose feedback signal instance according to a first cutting interval to obtain a first glucose feedback signal segment;

[0095] A cutting operation is performed on the first glucose feedback signal instance according to a second cutting interval to obtain a second glucose feedback signal segment; wherein the first cutting interval is smaller than the second cutting interval.

[0096] In an embodiment of the present invention, the server illustratively obtains a set of multiple glucose feedback signal instances, including a first glucose feedback signal instance. For example, the first glucose feedback signal instance is a blood glucose data curve obtained over 30 minutes of continuous monitoring. The server sets the first segmentation interval to the first 10 minutes and the second segmentation interval to the first 20 minutes. Based on the first segmentation interval, the server performs a segmentation operation on the first glucose feedback signal instance. This means that the server selects the data segment for the first 10 minutes from the start of monitoring and marks it as the first glucose feedback signal segment. For example, the data within these 10 minutes may show relatively stable blood glucose level fluctuations or some slight fluctuations. Then, based on the second segmentation interval, the server again segments the first glucose feedback signal instance, this time selecting the data segment for the first 20 minutes and marking it as the second glucose feedback signal segment. Compared to the first glucose feedback signal segment, the second glucose feedback signal segment contains more information, including some initial trend changes or short-term blood glucose fluctuation patterns. In actual operation, the server may simultaneously process a large number of similar glucose feedback signal instances. For example, another first glucose feedback signal instance may contain 60 minutes of monitoring data. The server also performs segmentation according to the set first segmentation interval (e.g., the first 15 minutes) and the second segmentation interval (e.g., the first 30 minutes). When processing these segmentation operations, the server will accurately record the timestamp and corresponding blood glucose value of each data point to ensure the accuracy of the segmentation. The server will also perform further analysis and processing on the segmented fragments. For example, for the first glucose feedback signal segment, the server will calculate the basic statistical features such as the average value, maximum value, minimum value, and slope of change of the blood glucose value during this period. For the second glucose feedback signal segment, in addition to the above basic statistical features, more complex features such as the periodicity of blood glucose changes during this period and the correlation with specific events (such as diet, exercise) may also be analyzed. By performing such segmentation and analysis on different first glucose feedback signal instances, the server can accumulate rich data features and patterns, providing strong support for the training and optimization of the subsequent blood glucose status assessment model. In addition, the server will dynamically adjust the length of the first segmentation interval and the second segmentation interval according to different user groups, monitoring device types, or other relevant factors. For example, for users with a specific type of diabetes, the server will shorten the first cutting interval to 5 minutes and extend the second cutting interval to 25 minutes to focus more on capturing long-term blood sugar trends. For another example, when a new type of high-precision blood sugar monitoring device is used that can provide more intensive and accurate data, the server will fine-tune the cutting interval accordingly to make full use of the newly added data information. Throughout the process, the server will continuously monitor and evaluate the effectiveness of the cutting operation.If it's discovered that certain cut-off interval settings are causing poor model training or inaccurate evaluation results, the server will promptly adjust and optimize. Through this sophisticated and flexible processing approach, the server can effectively extract valuable information from the collection of glucose feedback signal instances, continuously improving the accuracy and reliability of blood sugar status assessments.

[0097] In an embodiment of the present invention, the first integrated unit in the blood glucose status assessment model is called, the first glucose feedback signal segment is processed to obtain a first glucose feedback signal feature, and the first glucose feedback signal feature is autoregressively processed and feature expanded to obtain a first extended autoregressive feature, which can be implemented through the following examples.

[0098] calling a feature extraction component of a first integrated unit in a blood glucose status assessment model to perform feature extraction on the first glucose feedback signal segment to obtain a first glucose feedback signal feature;

[0099] calling the autoregressive component of the first integrated unit to perform autoregressive processing on the first glucose feedback signal feature to obtain a first autoregressive feature;

[0100] A feature conversion component in the first integrated unit is called to perform a feature conversion operation on the first autoregressive feature to obtain the first extended autoregressive feature.

[0101] In an embodiment of the present invention, illustratively, after receiving a first glucose feedback signal segment, the server begins invoking the first integrated unit in the blood glucose status assessment model for processing. First, the server invokes the feature extraction component of the first integrated unit to perform feature extraction on the first glucose feedback signal segment. For example, the first glucose feedback signal segment may be a collection of continuous blood glucose measurement data over a period of time. The feature extraction component analyzes the statistical characteristics of this data, such as the mean, variance, and median; analyzes the frequency characteristics of the data to identify the dominant frequency components; and extracts trend characteristics of the data, such as an upward, downward, or stable trend. Through these analyses, the feature extraction component obtains a series of numerical values ​​that characterize the first glucose feedback signal segment. These numerical values ​​constitute the first glucose feedback signal signature. For example, for a first glucose feedback signal segment containing 100 blood glucose measurement values, the feature extraction component calculates a mean of 8.5 mmol / L, a variance of 1.2, and a median of 8.3 mmol / L, and determines that the data exhibits a slightly upward trend. These numerical values ​​together constitute the first glucose feedback signal signature. Next, the server invokes the autoregressive component of the first integrated unit to perform autoregressive processing on the first glucose feedback signal signature. The autoregressive component builds a mathematical model to describe how these features change over time. For example, the autoregressive component uses a second-order autoregressive model to analyze historical data of the first glucose feedback signal feature and calculate autoregressive coefficients, thereby generating an autoregressive model capable of predicting future feature values. After processing this model, a first autoregressive feature is obtained. For example, based on the previously extracted feature values, the coefficients calculated by the autoregressive model result in a slightly upward trend in future feature values. The server then invokes the feature conversion component in the first integrated unit to perform feature conversion on the first autoregressive feature. The feature conversion component performs linear or nonlinear transformations on the first autoregressive feature to increase its diversity and expressiveness. It also combines the first autoregressive feature with other relevant physiological or environmental features to obtain a richer feature representation. These operations yield a first extended autoregressive feature. For example, the feature conversion component performs a logarithmic transformation on the first autoregressive feature and combines it with features such as the current user's age and weight to obtain a more comprehensive first extended autoregressive feature. In actual processing, the server simultaneously processes a large number of first glucose feedback signal segments. Each segment is processed strictly according to the above steps to ensure accurate and consistent feature extraction and transformation. For example, if the server receives 1,000 first glucose feedback signal segments within an hour, each segment is processed sequentially by the feature extraction component, the autoregressive component, and the feature transformation component. During processing, the server performs quality checks and verification on the output of each component to ensure there are no outliers or incorrect calculation results.If it is found during the feature conversion process that some of the first autoregressive features do not meet expectations, the server will go back to the previous steps to check whether there are any problems with the feature extraction and autoregressive processing, and make corresponding adjustments and recalculations. At the same time, the server will dynamically adjust the parameters and algorithms of each component based on the characteristics and data characteristics of different users. For example, different feature conversion methods will be used for young users and elderly users; the order and parameters of the autoregressive model will also be different for users with large blood sugar fluctuations and relatively stable users. Through such meticulous and comprehensive processing, the server can extract the first extended autoregressive features with rich information and high expressive power from the first glucose feedback signal segment, providing strong support for subsequent blood sugar status assessment.

[0102] In an embodiment of the present invention, the second integrated unit in the blood glucose status assessment model is called, the second glucose feedback signal segment is processed to obtain a second glucose feedback signal feature, and the second glucose feedback signal feature is autoregressively processed and feature expanded to obtain a second extended autoregressive feature. This can be implemented through the following example.

[0103] calling a feature extraction component of a second integrated unit in the blood glucose status assessment model to perform feature extraction on the second glucose feedback signal segment to obtain a second glucose feedback signal feature;

[0104] calling the autoregressive component of the second integrated unit to perform autoregressive processing on the second glucose feedback signal feature to obtain a second autoregressive feature;

[0105] A feature conversion component in the second integrated unit is called to perform a feature conversion operation on the second autoregressive feature to obtain the second extended autoregressive feature.

[0106] In an embodiment of the present invention, illustratively, after receiving the second glucose feedback signal segment, the server initiates processing by invoking the second integrated unit in the blood glucose status assessment model. First, the server invokes the feature extraction component of the second integrated unit to extract features from the second glucose feedback signal segment. For example, the second glucose feedback signal segment may represent blood glucose monitoring data over a longer period of time, such as continuous measurements over an hour. The feature extraction component employs various methods to extract features, such as analyzing the frequency and temporal characteristics of the data through wavelet transforms, or automatically learning hidden features within the data using convolutional neural networks (CNNs) within deep learning. Through these methods, the feature extraction component can obtain a set of numerical values ​​describing the unique properties of the second glucose feedback signal segment, which are referred to as the second glucose feedback signal features. For example, the feature extraction component, through wavelet transforms, may discover that the second glucose feedback signal segment exhibits high-frequency fluctuations within certain time periods, while exhibiting relatively stable low-frequency trends within other time periods. This information is then converted into specific feature values. Next, the server invokes the autoregressive component of the second integrated unit to perform autoregressive processing on the second glucose feedback signal features. The autoregressive component constructs a mathematical model appropriate for these features, taking into account the temporal correlations and dependencies between the features. For example, the autoregressive component uses a high-order autoregressive model that can capture the complex dynamic patterns in the second glucose feedback signal feature. By learning from historical data and estimating parameters, an autoregressive model is derived that can predict future feature changes, thereby generating a second autoregressive feature. For example, the autoregressive model discovers periodic variations in the second glucose feedback signal feature and, based on this, predicts the likely trend of future feature values. The server then invokes the feature conversion component in the second integrated unit to perform feature conversion on the second autoregressive feature. The feature conversion component uses methods such as principal component analysis (PCA) to reduce the dimensionality of the second autoregressive feature to remove redundant information and highlight key features. Alternatively, the feature can be expanded and enhanced by fusing and correlating the second autoregressive feature with other relevant physiological indicators (such as insulin levels and exercise intensity). These processes result in a more representative and comprehensive second expanded autoregressive feature. For example, the feature conversion component uses principal component analysis to reduce the high-dimensional second autoregressive feature into several key components. Furthermore, the feature is expanded by integrating the user's insulin injection history, making it more comprehensive in reflecting blood glucose status. In practice, the server can simultaneously process a large number of second glucose feedback signal segments, each from a different user or during a different monitoring period. For each segment, the server strictly follows the aforementioned process to ensure the accuracy and reliability of feature extraction, autoregressive processing, and feature conversion. For example, the server may receive thousands of second glucose feedback signal segments in a single day. During processing, the server performs quality assessment and verification on the output of each step.If the feature extraction is found to be inaccurate, the parameters of the feature extraction component will be adjusted or the feature extraction method will be changed. If the effect of the autoregressive processing is not ideal, the structure and parameters of the autoregressive model will be re-optimized. At the same time, the server will dynamically adjust the various components of the second integrated unit according to different application scenarios and user needs. For example, for clinical research scenarios that require more accurate assessment of blood sugar status, a more complex and sophisticated feature conversion method will be adopted; while for daily health monitoring scenarios, more emphasis will be placed on computational efficiency and simplicity, and a relatively simple but effective processing method will be selected. Through such a comprehensive and sophisticated processing flow, the server can obtain rich and valuable second extended autoregressive features from the second glucose feedback signal segment, providing strong support for accurate assessment of blood sugar health status.

[0107] In an embodiment of the present invention, the glucose feedback signal instances in the glucose feedback signal instance set include a first glucose feedback signal instance and other glucose feedback signal instances;

[0108] The first feature vector is obtained based on the first extended autoregressive feature and the second extended autoregressive feature of the same glucose feedback signal instance, and the second feature vector is obtained based on the first extended autoregressive feature of the first glucose feedback signal instance and the first extended autoregressive feature or the second extended autoregressive feature of the other glucose feedback signal instances in the glucose feedback signal instance set. This can be implemented through the following examples.

[0109] obtaining a first feature vector according to a first extended autoregressive feature of the first glucose feedback signal instance, a second extended autoregressive feature of the first glucose feedback signal instance, and a preset adjustment coefficient;

[0110] A second feature vector is obtained according to the first extended autoregressive feature of the first glucose feedback signal instance, a preset adjustment coefficient, and the first extended autoregressive feature or the second extended autoregressive feature of the other glucose feedback signal instance.

[0111] In an embodiment of the present invention, the server illustratively processes a set of glucose feedback signal instances, including a first glucose feedback signal instance and other glucose feedback signal instances. First, for the first glucose feedback signal instance, the server obtains its first extended autoregressive feature and second extended autoregressive feature. For example, the first extended autoregressive feature is a set of values ​​describing the long-term trend and periodicity of blood glucose changes, such as [1.2, 0.8, 1.5, 0.9, 1.1]. The second extended autoregressive feature is a set of values ​​reflecting short-term fluctuations and specific patterns of blood glucose changes, such as [0.3, 0.2, 0.4, 0.1, 0.5]. The server also sets a preset adjustment coefficient, such as [0.6, 0.4]. The server performs a weighted calculation on the first and second extended autoregressive features according to the preset adjustment coefficient. For example, the first value of the first extended autoregressive feature, 1.2, is multiplied by the first value of the preset adjustment coefficient, 0.6. The first value of the second extended autoregressive feature, 0.3, is multiplied by the second value of the preset adjustment coefficient, 0.4. These two products are then added together to obtain the first value of the first feature vector. The other values ​​are calculated using the same method to ultimately obtain the first eigenvector, such as [0.84, 0.48, 1.02, 0.46, 0.86]. Next, the server obtains the first extended autoregressive feature of the first glucose feedback signal instance, such as [1.2, 0.8, 1.5, 0.9, 1.1], and the first extended autoregressive feature or the second extended autoregressive feature of the other glucose feedback signal instances. For example, the corresponding features of the other glucose feedback signal instances are [1.0, 0.7, 1.3, 0.8, 1.2] and [0.2, 0.3, 0.5, 0.1, 0.4]. Similarly, the server processes the first extended autoregressive feature of the first glucose feedback signal instance and the features of the other glucose feedback signal instances according to the preset adjustment coefficient. For example, for the first value, the value of the first extended autoregressive feature of the first glucose feedback signal instance, 1.2, is multiplied by a preset adjustment coefficient (e.g., 0.5). The first value of each other glucose feedback signal instance, 1.0, is multiplied by another preset adjustment coefficient (e.g., 0.5). These two products are then added together to obtain the first value of the second eigenvector. The other values ​​are calculated in the same manner, ultimately obtaining the second eigenvector, such as [1.1, 0.75, 1.4, 0.85, 1.15]. In actual processing, the server may process a large number of glucose feedback signal instances simultaneously. For example, in a large-scale health monitoring project, the server receives glucose feedback signal instances from thousands of users. For each user's first glucose feedback signal instance, the server accurately calculates the first eigenvector and the second eigenvector according to the above steps. The server strictly monitors and verifies the calculation process to ensure the accuracy of the numerical calculations.If data anomalies are discovered during the calculation process, such as when the value of an extended autoregressive feature exceeds a reasonable range, the server will clean the data and recalculate. At the same time, the server will dynamically adjust the value of the preset adjustment coefficient based on the characteristics and monitoring needs of different user groups. For example, for user groups with specific diseases, long-term trend characteristics will be given a higher weight, and the preset adjustment coefficient will be adjusted accordingly. Through such precise and accurate calculations, the server can obtain the first and second eigenvectors that accurately reflect the characteristics of blood sugar status, providing strong data support for subsequent blood sugar status assessment and model optimization.

[0112] In the embodiment of the present invention, obtaining the first cost parameter according to the first eigenvector, the second eigenvector, and the number of glucose feedback signal instances in the glucose feedback signal instance set may be implemented through the following example.

[0113] Obtaining an intermediate parameter according to the first eigenvector and the second eigenvector;

[0114] The first cost parameter is obtained based on the intermediate parameter, the number of glucose feedback signal instances in the glucose feedback signal instance set, and a reference value.

[0115] In this embodiment of the present invention, for example, the server has obtained the first and second eigenvectors, as well as the number of glucose feedback signal instances in the set. First, the server obtains an intermediate parameter based on the first and second eigenvectors. For example, the first eigenvector is [0.2, 0.3, 0.5, 0.4, 0.6], and the second eigenvector is [0.1, 0.2, 0.4, 0.3, 0.5]. The server may obtain the intermediate parameter by calculating the inner product, Euclidean distance, or other mathematical operations between the two vectors. For example, by calculating the inner product of the two vectors, the server multiplies the values ​​at corresponding positions in the first and second eigenvectors, then adds these products to obtain an intermediate parameter of 0.9. Next, the server obtains a first cost parameter based on this intermediate parameter, the number of glucose feedback signal instances in the set, and a baseline value. For example, if there are 1000 glucose feedback signal instances in the set, and the baseline value is set to 0.5, the server may first perform some mathematical transformation on the intermediate parameter, such as multiplying it by a specific coefficient or performing a logarithmic operation. Then, further calculations are performed based on the number of instances and the baseline value. For example, the server multiplies the intermediate parameter by 100 to obtain 90. Then, (90 - 0.5 × 1000)² / 1000 is calculated to obtain the first cost parameter. In actual processing, the server can simultaneously process multiple different sets of glucose feedback signal instances. For example, in a large-scale medical data analysis project, the server may be faced with multiple sets of people from different regions, age groups, and health conditions. For each set, the server accurately calculates the intermediate parameter and the first cost parameter according to the above steps. During the calculation process, the server conducts strict quality control and verification of the data. If an abnormal value is found in a feature vector, the server will check the data source and processing process to ensure the accuracy and reliability of the data. Furthermore, the server dynamically adjusts the baseline value and the specific algorithm for calculating the first cost parameter based on different application scenarios and analysis objectives. For example, a relatively simple calculation method can be used during initial model training, while a more complex and precise algorithm will be adopted for fine-tuning the model. Through such rigorous and meticulous calculations, the server can obtain accurate first cost parameters, providing an important basis for the subsequent optimization and improvement of the blood sugar status assessment model.

[0116] In the embodiments of the present invention, the following implementation modes are also provided.

[0117] framing the glucose feedback signal instance in the glucose feedback signal instance set to obtain a plurality of glucose feedback signals of the glucose feedback signal instance;

[0118] performing interference processing on the plurality of glucose feedback signals to obtain glucose feedback signal instances containing interference data;

[0119] The performing a segmentation operation on the glucose feedback signal instance in the glucose feedback signal instance set to obtain a first glucose feedback signal segment and a second glucose feedback signal segment includes:

[0120] performing a segmentation operation on the glucose feedback signal instance containing the interference data to obtain a first glucose feedback signal segment;

[0121] A segmentation operation is performed on the glucose feedback signal instance that does not contain interference data to obtain a second glucose feedback signal segment.

[0122] In an embodiment of the present invention, the server first obtains a set of glucose feedback signal instances, which contains a large number of glucose feedback signal instances. The server performs a framing operation on each glucose feedback signal instance in the set. For example, each glucose feedback signal instance is a continuous time series data. The server segments each instance into multiple segments according to a fixed time length, such as one frame per second, thereby obtaining multiple glucose feedback signals for that glucose feedback signal instance. For example, a glucose feedback signal instance lasting 10 seconds is framed by the server into 10 glucose feedback signals each lasting 1 second. Next, the server performs interference processing on the multiple glucose feedback signals obtained by these frame processing. Interference processing can include operations such as adding random noise, simulating signal attenuation, and introducing spike pulses, thereby obtaining glucose feedback signal instances containing interference data. For example, for a glucose feedback signal after frame processing, the server adds a certain degree of random noise to its value, causing some fluctuation and uncertainty in the originally stable signal. The server then performs a segmentation operation on the glucose feedback signal instance containing interference data to obtain a first glucose feedback signal segment. For example, segmentation is based on a time point or signal strength threshold. From a glucose feedback signal instance containing interference data, the server selects a specific time period or a portion that meets specific conditions as the first glucose feedback signal segment. For example, for a 5-second glucose feedback signal instance with interference, the server uses the second second as the starting point and selects the following 2 seconds of data as the first glucose feedback signal segment. Simultaneously, the server performs segmentation on the glucose feedback signal instance without interference data, obtaining a second glucose feedback signal segment. This segmentation can be based on the same rules as for instances with interference data, but because the data itself is free of interference, the segmented second glucose feedback signal segment has clearer and more stable features. For example, for an undisturbed 6-second glucose feedback signal instance, the server uses the third second as the starting point and selects the following 3 seconds of data as the second glucose feedback signal segment. In practice, the server may process thousands of glucose feedback signal instances simultaneously. For example, in a large medical research project, the server needs to process continuous monitoring data from hundreds of patients. For each patient's multiple glucose feedback signal instances, the server strictly follows the steps of framing, interference processing, and segmentation. During the framing process, the server precisely controls the time interval to ensure that each frame is of consistent length and accurately reflects signal changes. During interference processing, the server randomly generates interference data based on preset interference patterns and intensities and adds it to the original signal to simulate various interference scenarios that may occur in a real-world environment.During the cutting operation, the server will flexibly adjust the starting point, end point, and length of the cutting according to the specific needs and objectives of the research to ensure that the first and second glucose feedback signal segments obtained are representative and of research value. At the same time, the server will record and mark each processed segment in detail, including the source of the original instance, the type and intensity of the interference processing, the parameters of the cutting, and other information for subsequent analysis and backtracking. If certain data are found to be abnormal or not in line with expectations during the processing process, the server will perform error troubleshooting and data cleaning, and re-perform the corresponding processing steps to ensure the quality and availability of the final signal segment. Through such a comprehensive and sophisticated processing flow, the server can provide a rich, diverse, and practical data set for subsequent blood sugar status assessment and model training.

[0123] In the embodiment of the present invention, the calling of the first integrated unit in the blood glucose status assessment model and the processing of the first glucose feedback signal segment to obtain the first glucose feedback signal feature can be implemented through the following examples.

[0124] calling a first integrated unit in the blood glucose status assessment model to process the interference data to obtain a first glucose feedback signal feature;

[0125] The calling of the second integrated unit in the blood glucose status assessment model to process the second glucose feedback signal segment to obtain a second glucose feedback signal feature includes:

[0126] obtaining data at the same position in the second glucose feedback signal segment based on interference data in the first glucose feedback signal segment;

[0127] The second integrated unit in the blood glucose status assessment model is called to process the data to obtain a second glucose feedback signal feature.

[0128] In an embodiment of the present invention, illustratively, when processing the first and second glucose feedback signal segments, the server invokes an integrated unit in the blood glucose status assessment model according to a specific process to perform feature extraction. For the first glucose feedback signal segment, the server invokes the first integrated unit in the blood glucose status assessment model to process the interference data therein. For example, the first glucose feedback signal segment may contain interference data generated by factors such as device noise and environmental interference, such as abnormal spikes or large fluctuations at certain time points. The first integrated unit may include algorithms and modules specifically designed to process interference data. For example, it may use filtering algorithms to remove high-frequency noise or outlier detection and correction methods to smooth abnormal spikes. Through these processes, the server extracts key features that reflect blood glucose status from the interference data and defines them as first glucose feedback signal features. For example, after processing, the server extracts features such as the average trend of blood glucose value changes, the main fluctuation period, and differences from common interference patterns from the interference data as first glucose feedback signal features. For the second glucose feedback signal segment, the server first obtains data at the same location in the second glucose feedback signal segment based on the interference data in the first glucose feedback signal segment. For example, if the interference data in the first glucose feedback signal segment exhibits strong fluctuations at a specific location on the timeline, the server will search for similar data patterns at the corresponding time location in the second glucose feedback signal segment. The server then invokes the second integrated unit in the blood glucose status assessment model to process this data. The second integrated unit can employ different but complementary processing methods and feature extraction strategies than the first. For example, the second integrated unit may focus more on analyzing local details of the data, similarities with historical data, or correlations with other physiological indicators. By processing this data, the server obtains the second glucose feedback signal features. In actual processing scenarios, the server may be faced with a large number of first and second glucose feedback signal segments simultaneously. For example, in a large-scale health monitoring system, the server needs to process a large amount of blood glucose data from thousands of users every day. For each data segment of each user, the server strictly follows the above steps. When processing interference data, the server continuously optimizes and adjusts the parameters of the processing algorithm to adapt to the different types and intensities of interference. For example, if the interference in a batch of data is primarily low-frequency noise, the server will adjust the cutoff frequency of the filtering algorithm accordingly. Simultaneously, the server monitors and evaluates the quality of the processing results of the first and second integrated units in real time. If the extracted features are found to be inconsistent with expectations or inconsistent with other relevant data, the server will backtrack to check the data processing steps and algorithm application, and make necessary adjustments and recalculations. In addition, the server also takes into account individual differences and changes in health status of users.For users with specific medical conditions or at unique physiological stages, the server employs customized processing strategies and feature extraction methods to enhance the accuracy and specificity of blood glucose assessment. Through this sophisticated and comprehensive processing, the server fully leverages the information from the first and second glucose feedback signal segments, extracting valuable features that provide strong support for accurate blood glucose assessment.

[0129] In the embodiment of the present invention, obtaining the second cost parameter according to the first glucose feedback signal characteristic and the second glucose feedback signal characteristic can be implemented through the following example.

[0130] obtaining an absolute value of a difference between the first glucose feedback signal characteristic and the second glucose feedback signal characteristic;

[0131] The second cost parameter is obtained based on the absolute value of the difference and a reference value.

[0132] In this embodiment of the present invention, the server has already obtained a first glucose feedback signal feature and a second glucose feedback signal feature. First, the server calculates the absolute value of the difference between the first glucose feedback signal feature and the second glucose feedback signal feature. For example, the first glucose feedback signal feature is a set of values, such as [1.5, 2.0, 3.5, 4.0, 5.0], and the second glucose feedback signal feature is [1.2, 2.2, 3.0, 4.5, 5.5]. The server subtracts the values ​​at corresponding positions in the two features and takes the absolute value of the difference. For example, the absolute value of the difference at the first position is |1.5 - 1.2| = 0.3, the absolute value of the difference at the second position is |2.0 - 2.2| = 0.2, and so on, resulting in a set of absolute difference values ​​of [0.3, 0.2, 0.5, 0.5, 0.5]. Next, the server derives a second cost parameter based on the absolute values ​​of these differences and a baseline value. For example, the baseline value is set to 0.4. The server first performs an operation such as summing or averaging the absolute values ​​of the differences. For example, the server first calculates the average of the absolute values ​​of the aforementioned difference values, obtaining (0.3 + 0.2 + 0.5 + 0.5 + 0.5) / 5 = 0.4. The server then compares this average value with the baseline value and performs the calculation. For example, the square of (average value - baseline value) can be calculated, i.e., (0.4 - 0.4)² = 0, and this can be used as the second cost parameter. In practice, the server can simultaneously process a large number of first and second glucose feedback signal feature pairs. For example, in a large-scale health monitoring project, the server receives multiple sets of feature data from thousands of users daily. For each set of feature data, the server strictly follows the above steps. When calculating the absolute values ​​of the differences, the server uses high-precision numerical calculation methods to ensure the accuracy of the results. If the absolute values ​​of certain differences are found to be too large or too small, the server conducts further analysis. It checks for anomalies in the data collection process or whether the feature extraction algorithm needs adjustment. Furthermore, the server dynamically adjusts the baseline value based on different application scenarios and needs. For example, for patients with more severe conditions, the baseline value may be lowered to increase sensitivity to blood glucose changes. The server also uses the calculated second cost parameter for subsequent model optimization and evaluation. If the second cost parameter remains consistently high, the server will deem the model to have problems processing these two features and will require model improvement and adjustment. Through such precise and accurate calculation and analysis, the server can effectively leverage the difference between the first and second glucose feedback signal features to derive a reliable second cost parameter, providing strong support for the improvement and optimization of the blood glucose status assessment model.

[0133] In the embodiment of the present invention, the embodiment of the present invention also provides the following implementation methods.

[0134] obtaining a third feature vector according to the first extended autoregressive feature, the second extended autoregressive feature, and the number of glucose feedback signal instances in the glucose feedback signal instance set of the same glucose feedback signal instance;

[0135] obtaining a correlation matrix based on the first extended autoregressive feature and the second extended autoregressive feature of the same glucose feedback signal instance, and obtaining a fourth eigenvector based on the correlation matrix;

[0136] Obtaining a third cost parameter according to the third eigenvector and the fourth eigenvector;

[0137] The structural optimization of the first integrated unit according to the first cost parameter and the second cost parameter includes:

[0138] The first integrated unit is structurally optimized according to the first cost parameter, the second cost parameter, and the third cost parameter.

[0139] In an embodiment of the present invention, the server first processes each instance in the set of glucose feedback signal instances. For a particular glucose feedback signal instance, the server has already obtained its first extended autoregressive feature and second extended autoregressive feature. For example, the first extended autoregressive feature of this glucose feedback signal instance is a set of values ​​[1.2, 0.8, 1.5, 0.9, 1.1], and the second extended autoregressive feature is [0.7, 0.6, 0.8, 0.5, 0.9]. The server obtains a third feature vector based on the first and second extended autoregressive features of the same glucose feedback signal instance and the number of glucose feedback signal instances in the set of glucose feedback signal instances. For example, the server first calculates statistics of the two extended autoregressive features, such as the mean and variance. The first extended autoregressive feature has a mean of 1.0 and a variance of 0.1; the second extended autoregressive feature has a mean of 0.7 and a variance of 0.1. For example, the set of glucose feedback signal instances contains a total of 500 instances. The server then combines and calculates these statistics and the number of instances, for example, multiplying the mean by the number of instances to obtain [500, 350]. It then performs some normalization or transformation processing to obtain a third eigenvector, for example, [0.5, 0.35]. Next, the server obtains a correlation matrix based on the first and second extended autoregressive features of the same glucose feedback signal instance. The server calculates the correlation between these two features to obtain a correlation matrix. For example, this correlation matrix is ​​[[1, 0.8], [0.8, 1]]. The server then performs further processing based on this correlation matrix, such as calculating the matrix's eigenvalues ​​or performing a formal decomposition, to obtain a fourth eigenvector. For example, the fourth eigenvector obtained after processing is [0.6, 0.4]. The server obtains a third cost parameter based on the third and fourth eigenvectors. For example, the server can calculate a distance or difference metric between the two eigenvectors, such as the Euclidean distance. For example, the third eigenvector is [0.5, 0.35], and the fourth eigenvector is [0.6, 0.4]. Their Euclidean distance is √((0.5-0.6)²+(0.35-0.4)²)=0.141. The server can directly use this distance as the third cost parameter, or perform some transformations and adjustments on it. When optimizing the structure of the first integrated unit, the server comprehensively considers the first cost parameter, the second cost parameter, and the third cost parameter. For example, the first cost parameter reflects the error of the model in the overall prediction, with a value of 0.2; the second cost parameter reflects the accuracy of feature extraction, with a value of 0.15; and the third cost parameter reflects the degree of matching between the features, with a value of 0.1. The server will assign different weights to these cost parameters, such as a weight of 0.5 for the first cost parameter, 0.3 for the second cost parameter, and 0.2 for the third cost parameter.The server then calculates the weighted sum: 0.2 × 0.5 + 0.15 × 0.3 + 0.1 × 0.2 = 0.145. Based on this comprehensive cost evaluation result, the server performs structural optimization on the first integrated unit. This may include adjusting neuron connection weights, adding or removing neurons, and changing the number of network layers. In actual operation, the server may process a large number of glucose feedback signal instances simultaneously, performing the above feature calculations and cost evaluations on each instance. For example, on a dataset containing 1,000 instances, the server processes each instance in turn, accumulates statistical information on the cost parameters, and determines the direction and extent of structural optimization based on the overall evaluation results. If the cost parameters do not decrease significantly during the optimization process, the server will try different weight allocation strategies or conduct in-depth analysis of data characteristics to identify potential problems. The server also considers computing resources and time costs, selecting appropriate optimization algorithms and parameters while ensuring optimization results. Through this comprehensive consideration and optimization process, the server can continuously improve the performance and accuracy of the blood glucose status assessment model.

[0140] In the embodiment of the present invention, the embodiment of the present invention also provides the following implementation methods.

[0141] performing type identification on a first autoregressive feature obtained by performing autoregressive processing on the first glucose feedback signal feature to obtain a first confidence level;

[0142] performing type recognition on a second autoregressive feature obtained by performing autoregressive processing on the second glucose feedback signal feature to obtain a second confidence level;

[0143] Obtaining a fourth cost parameter based on the first confidence level and the second confidence level;

[0144] The structural optimization of the first integrated unit according to the first cost parameter, the second cost parameter, and the third cost parameter includes:

[0145] The first integrated unit is structurally optimized according to the first cost parameter, the second cost parameter, the third cost parameter, and the fourth cost parameter.

[0146] In an embodiment of the present invention, the server, while processing the glucose feedback signal features, exemplarily performs a series of complex operations to optimize the model structure. First, the server performs autoregressive processing on the first glucose feedback signal feature to obtain a first autoregressive feature. Then, the server applies a specific type recognition algorithm to analyze the first autoregressive feature. For example, if the first autoregressive feature exhibits a relatively stable and predictable pattern, the type recognition algorithm determines that it belongs to a specific type, such as "normal blood sugar fluctuation type," by matching and comparing the pattern. The algorithm also generates a numerical value representing the degree of certainty in this determination, namely a first confidence level. For example, a first confidence level of 0.85 indicates that the server has a high degree of certainty in this type determination. Similarly, the server performs autoregressive processing on the second glucose feedback signal feature to obtain a second autoregressive feature, and then performs type recognition on this feature to obtain a second confidence level. For example, if the second autoregressive feature exhibits a relatively complex and less common pattern, it may be classified as "potential dysglycemia type" after type recognition, corresponding to a second confidence level of 0.7. Based on the first and second confidence levels, the server determines a fourth cost parameter. For example, the server can set a rule to convert and weight the confidence level to obtain the fourth cost parameter. For example, the server multiplies the confidence level by 10 and takes the absolute value of the difference, namely |0.85×10-0.7×10|=1.5, and uses this value as the fourth cost parameter. When optimizing the structure of the first integrated unit, the server no longer relies solely on the first, second, and third cost parameters, but instead comprehensively considers the fourth cost parameter. For example, the first cost parameter reflects the deviation of the overall model prediction and has a value of 0.2; the second cost parameter reflects the accuracy of feature extraction and has a value of 0.15; the third cost parameter reflects the degree of matching between features and has a value of 0.1; and the fourth cost parameter, calculated as above, is 1.5. The server assigns a corresponding weight to each cost parameter. For example, the first cost parameter has a weight of 0.4, the second cost parameter has a weight of 0.3, the third cost parameter has a weight of 0.2, and the fourth cost parameter has a weight of 0.1. The server then calculates their weighted sum: 0.2×0.4+0.15×0.3+0.1×0.2+1.5×0.1=0.345. Based on this comprehensive cost evaluation result, the server performs structural optimization on the first integrated unit. This involves adjusting the connection weights between neurons. For example, if it is found that certain connection weights have a significant negative impact on the cost evaluation result, the server will reduce the value of these weights. Alternatively, the server may add or delete neurons to change the processing power and complexity of the first integrated unit. For example, the first integrated unit currently has 100 neurons. Based on the cost evaluation results, the server decides to delete 10 neurons that contribute less to the result to simplify the model structure and improve computational efficiency. In addition, the server may also change the number of layers in the network.If the server finds that the current number of layers is insufficient to effectively process the data, it may increase or decrease the number of layers to optimize model performance. In practice, the server may simultaneously process a large volume of diverse glucose feedback signal data. For example, a large medical database contains blood glucose monitoring data from tens of thousands of patients. For each data set, the server performs the aforementioned feature processing, confidence calculation, and cost evaluation, and continuously optimizes the first integrated unit based on the combined results. During the optimization process, if the server finds that the cost parameter has not been significantly reduced despite multiple adjustments to the first integrated unit, it will re-examine the feature processing and type recognition algorithms to identify any errors or areas for improvement. The server also considers the characteristics of different patient groups and differences in data distribution. For specific patient groups, such as children with diabetes or elderly patients with diabetes, the server performs individual data analysis and optimization based on their specific characteristics to ensure that the model provides accurate blood glucose status assessment for each group. Through this comprehensive, detailed, and dynamic optimization process, the server can continuously improve the performance of the first integrated unit, thereby enhancing the accuracy and reliability of the entire blood glucose status assessment model.

[0147] The embodiment of the present invention provides a computer device 100, which includes a processor and a non-volatile memory storing computer instructions. When the computer instructions are executed by the processor, the computer device 100 executes the aforementioned user health status management method based on the smart wearable device. Figure 2 As shown, Figure 2 This is a block diagram of the structure of a computer device 100 provided in an embodiment of the present invention. Computer device 100 includes a memory 111, a processor 112, and a communication unit 113. To enable data transmission or exchange, memory 111, processor 112, and communication unit 113 are electrically connected to each other, directly or indirectly. For example, these components can be electrically connected via one or more communication buses or signal lines.

[0148] For illustrative purposes, the foregoing description has been made with reference to specific embodiments. However, the above illustrative discussion is not intended to be exhaustive or to limit the present disclosure to the precise forms disclosed. Numerous modifications and variations are possible in light of the above teachings. These embodiments have been selected and described in order to best illustrate the principles of the present disclosure and its practical application, thereby enabling those skilled in the art to best utilize the present disclosure and to utilize various embodiments with various modifications as appropriate for the specific application contemplated.

Claims

1. A server system, characterized in that: The method includes a server configured to execute a user health status management method based on a smart wearable device, the method comprising: In response to a health status assessment instruction triggered by a preset period, initializing the radio frequency transmission module and the feedback signal receiving module of the smart wearable device; Transmitting an electromagnetic wave signal of a preset wavelength to a preset body area of ​​a target user through the radio frequency transmission module; receiving, by the feedback signal receiving module, a feedback signal to be determined corresponding to the electromagnetic wave signal of the preset wavelength; calling a pre-trained glucose feedback signal determination model to perform glucose feedback signal determination on the feedback signal to be determined, and obtaining a determination result indicating whether the feedback signal to be determined includes a glucose feedback signal; If the determination result indicates that a glucose feedback signal exists, obtaining a glucose feedback signal to be processed; calling a pre-trained target blood glucose state assessment model to process the glucose feedback signal and output an autoregressive feature of the feedback signal; performing a comparison operation on the feedback signal autoregressive feature and the pre-stored blood glucose health status feature to obtain a comparison result; When the comparison result indicates that the characteristic distance between the autoregressive characteristic of the feedback signal and the blood glucose health status characteristic is less than a preset characteristic distance threshold, obtaining the blood glucose health status corresponding to the glucose feedback signal; The glucose feedback signal determination model is obtained by the following methods, including: Calling the glucose feedback signal determination model to perform glucose feedback signal determination on the invalid instantaneous feedback signal instance and the valid instantaneous feedback signal instance, respectively, and obtaining an invalid instantaneous determination result and a valid instantaneous result accordingly; The invalid instantaneous feedback signal instance is obtained by adding an instantaneous feedback signal to the invalid instance, and the valid instantaneous feedback signal instance is obtained by adding an instantaneous feedback signal to the valid instance corresponding to the invalid instance; determining an error parameter of a first error function based on the instance target value of the invalid instance, the instance target value of the valid instance, the invalid instantaneous determination result, and the valid instantaneous result; Calling the glucose feedback signal determination model to perform glucose feedback signal determination on the invalid continuous feedback signal instance and the valid continuous feedback signal instance, respectively, to obtain an invalid continuous determination result and a valid continuous determination result; The invalid continuous feedback signal instance is obtained by adding a continuous feedback signal to the invalid instance, and the valid continuous feedback signal instance is obtained by adding a continuous feedback signal to the valid instance; determining a first error parameter of a second error function based on the instance target value of the invalid continuous feedback signal instance, the instance target value of the valid continuous feedback signal instance, the invalid continuous determination result, and the valid continuous determination result; Optimizing network parameters of the glucose feedback signal determination model based on the error parameter of the first error function and the first error parameter of the second error function; calling the optimized glucose feedback signal determination model to perform glucose feedback signal determination on the invalid continuous feedback signal instance and the valid continuous feedback signal instance, respectively, and obtaining an invalid continuous optimization determination result and a valid continuous optimization determination result respectively; determining a second error parameter of the second error function based on the instance target value of the invalid instance, the instance target value of the valid instance, the invalid continuous optimization determination result, and the valid continuous optimization determination result; Optimizing network parameters of the optimized glucose feedback signal determination model based on the error parameter of the first error function and the second error parameter of the second error function to obtain a trained glucose feedback signal determination model; The target blood glucose state assessment model is obtained by the following methods, including: performing a segmentation operation on a glucose feedback signal instance in a glucose feedback signal instance set to obtain a first glucose feedback signal segment and a second glucose feedback signal segment; the glucose feedback signal instances in the glucose feedback signal instance set include the first glucose feedback signal instance and the other glucose feedback signal instances; calling a first integrated unit in the blood glucose status assessment model, processing the first glucose feedback signal segment to obtain a first glucose feedback signal feature, and performing autoregressive processing and feature expansion on the first glucose feedback signal feature to obtain a first extended autoregressive feature; calling a feature extraction component of a second integrated unit in the blood glucose status assessment model to perform feature extraction on the second glucose feedback signal segment to obtain a second glucose feedback signal feature; calling the autoregressive component of the second integrated unit to perform autoregressive processing on the second glucose feedback signal feature to obtain a second autoregressive feature; Calling the feature conversion component in the second integrated unit to perform a feature conversion operation on the second autoregressive feature to obtain a second extended autoregressive feature; obtaining a first feature vector according to a first extended autoregressive feature of the first glucose feedback signal instance, a second extended autoregressive feature of the first glucose feedback signal instance, and a preset adjustment coefficient; Obtaining a second feature vector according to the first extended autoregressive feature of the first glucose feedback signal instance, a preset adjustment coefficient, and the first extended autoregressive feature or the second extended autoregressive feature of the other glucose feedback signal instance; Obtaining an intermediate parameter according to the first eigenvector and the second eigenvector; obtaining a first cost parameter based on the intermediate parameter, the number of glucose feedback signal instances in the glucose feedback signal instance set, and a reference value, and obtaining a second cost parameter based on the first glucose feedback signal feature and the second glucose feedback signal feature; The first integrated unit is structurally optimized according to the first cost parameter and the second cost parameter to obtain a target blood glucose state assessment model for assessing the blood glucose health state of the target user's glucose feedback signal.

2. The server system according to claim 1, wherein: The method further comprises: Calling the glucose feedback signal determination model to perform glucose feedback signal determination on the invalid instance and the valid instance corresponding to the invalid instance, and obtaining invalid instance determination results and valid instance determination results respectively; determining a first error parameter of a third error function based on the instance target value of the invalid instance, the instance target value of the valid instance, the invalid instance determination result, and the valid instance determination result; Optimizing the network parameters of the glucose feedback signal determination model based on the error parameter of the first error function and the first error parameter of the second error function includes: respectively obtaining a weighting coefficient of the first error function, a weighting coefficient of the second error function, and a weighting coefficient of the third error function; Obtaining an error parameter of a total error function based on the error parameter and weighting coefficient of the first error function, the first error parameter and weighting coefficient of the second error function, and the first error parameter and weighting coefficient of the third error function; The network parameters of the glucose feedback signal determination model are optimized according to the error parameters of the total error function.

3. The server system according to claim 2, wherein: Before calling the glucose feedback signal determination model to perform glucose feedback signal determination on the invalid instantaneous feedback signal instance and the valid instantaneous feedback signal instance, the method further includes: Swapping the instance target value of the invalid instance and the instance target value of the valid instance to obtain the enhanced instance target value of the invalid instance and the enhanced instance target value of the valid instance; determining a second error parameter of the third error function based on the enhanced instance target value of the invalid instance, the enhanced instance target value of the valid instance, the invalid instance determination result, and the valid instance determination result; determining, based on a second error parameter of the third error function, a parameter update direction of the third error function with respect to the network parameter, and obtaining an optimization guidance value of the glucose feedback signal determination model; determining the optimized guidance amount of the glucose feedback signal determination model as an instantaneous feedback signal; Determining a weighting coefficient of the instantaneous feedback signal, and performing a weighting operation on the instantaneous feedback signal based on the weighting coefficient to obtain a weighted instantaneous feedback signal; adding the weighted instantaneous feedback signal to the invalid instance and the valid instance respectively to obtain the invalid instantaneous feedback signal instance and the valid instantaneous feedback signal instance; Before calling the glucose feedback signal determination model to perform glucose feedback signal determination on the invalid instantaneous feedback signal instance and the valid instantaneous feedback signal instance, the method further includes: Acquire basic invalid instances of multiple information sources, and acquire basic valid instances of multiple information sources corresponding to the basic invalid instances; Performing an integration operation on the basic invalid instances of the multiple information sources to obtain the invalid instance; An integration operation is performed on the basic valid instances of the multiple information sources to obtain the valid instance.

4. The server system according to claim 1, wherein: The glucose feedback signal determination is executed by calling a discrimination component in the glucose feedback signal determination model, and the glucose feedback signal determination model further includes a synthesis component; Before calling the glucose feedback signal determination model to perform glucose feedback signal determination on the invalid continuous feedback signal instance and the valid continuous feedback signal instance, the method further includes: calling a synthesis component in a glucose feedback signal determination model to generate a continuous feedback signal based on an input preset continuous signal; Determining a weighting coefficient of the continuous feedback signal, and performing a weighting operation on the continuous feedback signal based on the weighting coefficient to obtain a weighted continuous feedback signal; The weighted continuous feedback signal is added to the invalid instance and the valid instance respectively to obtain the invalid continuous feedback signal instance and the valid continuous feedback signal instance.

5. The server system according to claim 1, wherein: The calling of the first integrated unit in the blood glucose state assessment model, processing the first glucose feedback signal segment to obtain a first glucose feedback signal feature, and performing autoregressive processing and feature expansion on the first glucose feedback signal feature to obtain a first extended autoregressive feature, includes: calling a feature extraction component of a first integrated unit in a blood glucose status assessment model to perform feature extraction on the first glucose feedback signal segment to obtain a first glucose feedback signal feature; calling the autoregressive component of the first integrated unit to perform autoregressive processing on the first glucose feedback signal feature to obtain a first autoregressive feature; A feature conversion component in the first integrated unit is called to perform a feature conversion operation on the first autoregressive feature to obtain the first extended autoregressive feature.

6. The server system according to claim 1, wherein: The performing a segmentation operation on the glucose feedback signal instance in the glucose feedback signal instance set to obtain a first glucose feedback signal segment and a second glucose feedback signal segment includes: performing a cutting operation on the first glucose feedback signal instance according to a first cutting interval to obtain a first glucose feedback signal segment; A cutting operation is performed on the first glucose feedback signal instance according to a second cutting interval to obtain a second glucose feedback signal segment; wherein the first cutting interval is smaller than the second cutting interval.

7. The server system according to claim 1, wherein: The method further comprises: framing the glucose feedback signal instance in the glucose feedback signal instance set to obtain a plurality of glucose feedback signals of the glucose feedback signal instance; performing interference processing on the plurality of glucose feedback signals to obtain glucose feedback signal instances containing interference data; The performing a segmentation operation on the glucose feedback signal instance in the glucose feedback signal instance set to obtain a first glucose feedback signal segment and a second glucose feedback signal segment includes: performing a segmentation operation on the glucose feedback signal instance containing the interference data to obtain a first glucose feedback signal segment; performing a segmentation operation on the glucose feedback signal instance that does not include interference data to obtain a second glucose feedback signal segment; The calling of the first integrated unit in the blood glucose status assessment model to process the first glucose feedback signal segment to obtain a first glucose feedback signal feature includes: calling a first integrated unit in the blood glucose status assessment model to process the interference data to obtain a first glucose feedback signal feature; The calling of the second integrated unit in the blood glucose status assessment model to process the second glucose feedback signal segment to obtain a second glucose feedback signal feature includes: obtaining data at the same position in the second glucose feedback signal segment based on interference data in the first glucose feedback signal segment; calling a second integrated unit in the blood glucose status assessment model to process the data to obtain a second glucose feedback signal feature; The obtaining of a second cost parameter according to the first glucose feedback signal characteristic and the second glucose feedback signal characteristic includes: obtaining an absolute value of a difference between the first glucose feedback signal characteristic and the second glucose feedback signal characteristic; Obtaining the second cost parameter based on the absolute value of the difference and a reference value; The method further comprises: obtaining a third feature vector according to the first extended autoregressive feature, the second extended autoregressive feature, and the number of glucose feedback signal instances in the glucose feedback signal instance set of the same glucose feedback signal instance; obtaining a correlation matrix based on the first extended autoregressive feature and the second extended autoregressive feature of the same glucose feedback signal instance, and obtaining a fourth eigenvector based on the correlation matrix; Obtaining a third cost parameter according to the third eigenvector and the fourth eigenvector; The structural optimization of the first integrated unit according to the first cost parameter and the second cost parameter includes: performing structural optimization on the first integrated unit according to the first cost parameter, the second cost parameter, and the third cost parameter; The method further comprises: performing type identification on a first autoregressive feature obtained by performing autoregressive processing on the first glucose feedback signal feature to obtain a first confidence level; performing type recognition on a second autoregressive feature obtained by performing autoregressive processing on the second glucose feedback signal feature to obtain a second confidence level; Obtaining a fourth cost parameter based on the first confidence level and the second confidence level; The structural optimization of the first integrated unit according to the first cost parameter, the second cost parameter, and the third cost parameter includes: The first integrated unit is structurally optimized according to the first cost parameter, the second cost parameter, the third cost parameter, and the fourth cost parameter.

Images