Remote data management system and method for glycosylation monitor
By introducing a data tracking and suspended state recognition unit, combined with interpolation repair and constraint mapping modules, the problems of misidentification of intermediate state data and inconsistency of constraints in the sugar metabolism monitor were solved, achieving data consistency and efficient processing, and improving system performance and response speed.
Patent Information
- Application Number
- CN202511722510.4
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-11-21
- Publication Date
- 2026-02-27
- Estimated Expiration
- Not applicable · inactive patent
AI Technical Summary
In the existing remote data management system for glucose metabolism monitors, intermediate state data may be incorrectly identified as valid data, leading to inaccurate blood glucose analysis and reporting results. Furthermore, system upgrades or asynchronous node adjustments can cause inconsistent data constraints, affecting the accuracy and consistency of the data.
A data tracking unit, a suspended state data identification unit, and a dynamic adjustment mechanism are introduced. By identifying and marking suspended state data, combined with interpolation repair technology, data consistency is ensured. Furthermore, constraint alignment is performed through a constraint mapping module and a consensus module, and acquisition and transmission parameters are dynamically adjusted to reduce computing resource consumption.
It effectively prevents misidentification of intermediate state data, ensures data consistency and accuracy, improves system performance, reduces latency and computational burden, and guarantees real-time data processing results.
Smart Images
Figure CN121583430A_ABST
Abstract
Description
TECHNICAL FIELD
[0001] The present application relates to the field of data management, in particular to a data remote management system and method of a sugar monitoring instrument. BACKGROUND
[0002] With the development of remote medical technology, sugar monitoring equipment based on wireless sensors has become an important tool for blood glucose data collection. For example, in a typical continuous glucose monitoring system, the running process is as follows: after the sensor is attached, the glucose concentration or its changes in the position where the sensor is located (usually the intercellular fluid or interstitial fluid) are collected. The glucose-related signals (such as electrochemical reaction, electrode current, or optical / electromagnetic response) are converted into electrical signals by the sensor. The signals are transmitted to the receiving device through the wireless transmission module. The sensors used include electrochemical sensors, optical / near-infrared sensors, electromagnetic / microwave sensors, and other biological fluid sensors.
[0003] In the traditional remote data management system of the sugar monitoring instrument, a single blood glucose data usually goes through five stages from collection to cloud storage: collection, caching, transmission, storage, and cleaning / aggregation. In this process, network jitter, power failure, or system upgrade may cause some data packets to complete only part of the stage, forming "intermediate state" data. Although these intermediate data do not immediately cause system errors, when the system is restarted or aggregated queries (such as blood glucose curve drawing analysis) are performed, these incomplete data may be incorrectly identified as valid data, thereby affecting subsequent blood glucose analysis and reporting results.
[0004] In addition, the sugar monitoring system involves multiple layers of data processing nodes, and at each level, the data is subject to certain constraint conditions, including numerical accuracy (such as the number of decimal places), unit specification, and interval range. With the continuous upgrading of the system, the software and hardware environment and algorithm iteration, the constraint conditions of the data processing nodes are often not adjusted synchronously. For example, the sensor firmware may have improved the data accuracy to 0.01 mmol / L, but the cloud data processing is still rounded to 0.1 mmol / L. Such asynchronous constraint adjustment may cause the system constraint chain to be out of balance, causing systematic bias in the same blood glucose data after multiple levels of synchronization and writeback, thereby affecting the accuracy and consistency of the data. The existing technology usually eliminates the effects of asynchrony by introducing a dynamic adjustment mechanism, but in the sugar monitoring instrument as an edge device, it increases the complexity of the system, and each data processing level needs additional precision monitoring, synchronization, and adjustment, consuming a large amount of computing resources, affecting the performance and response time of the system, causing delays to increase, and even affecting the effect of real-time data processing. SUMMARY
[0005] In view of the deficiencies of the prior art, the present application provides a data remote management system and method of a sugar monitoring instrument, in order to achieve the above object, the present application is realized by the following technical scheme: a data remote management system of a sugar monitoring instrument, comprising: A data management module for monitoring the data state of blood glucose data during transmission, the data management module comprising a data acquisition unit, a data transmission unit, a data tracking unit and a suspended state data identification unit.
[0006] The data acquisition unit is used to collect blood glucose data from the sensor and dynamically adjust the collection frequency.
[0007] The data transmission unit is used to transmit the blood glucose data to each module for processing, and dynamically adjusts the transmission parameters based on the network state data.
[0008] The data tracking unit is used to record the data state of the blood glucose data at each time node.
[0009] The suspended state data identification unit is used to identify suspended state data generated during the transmission of blood glucose data.
[0010] A constraint mapping module for constraint mapping of blood glucose data during transmission, the constraint mapping module comprising a constraint space modeling unit, a constraint drift detection unit and a data repair adjustment unit.
[0011] The constraint space modeling unit is used to assign a high-dimensional constraint vector to each blood glucose data.
[0012] The constraint drift detection unit is used to detect changes in the constraint condition of the blood glucose data in the processing cycle, and calculate the constraint space vector distance to determine whether the blood glucose data has constraint drift.
[0013] The data repair adjustment unit is used to align the constraint of the blood glucose data with the constraint drift.
[0014] A constraint consensus module for consistency processing of blood glucose data after constraint mapping, the constraint consensus module comprising a state synchronization unit, a convergence control unit and a consensus adjustment unit.
[0015] The state synchronization unit is used to periodically synchronize the constraint state of the blood glucose data to the cloud.
[0016] The convergence control unit is used to control the convergence speed and convergence accuracy of the constraint consensus.
[0017] The consensus adjustment unit is used to dynamically adjust the constraint condition according to the constraint difference.
[0018] A data remote management method of a sugar monitoring instrument, comprising: The data state of the blood glucose data in the transmission process is monitored, the blood glucose data is collected from the sensor, the collection frequency is dynamically adjusted, the blood glucose data is transmitted to each module for processing, the transmission parameters are dynamically adjusted based on the network state data, the data state of the blood glucose data at each time node is recorded, and the suspended state data generated in the blood glucose data transmission process is identified, wherein the suspended state data refers to the blood glucose data that has occupied a position in the system but does not have reliable effectiveness.
[0019] The blood glucose data in the transmission process is constrained and mapped, high-dimensional constraint vectors are allocated to each blood glucose data, the change of the constraint condition of the blood glucose data in the processing period is detected, the constraint space vector distance is calculated to determine whether the blood glucose data has constraint drift, and the blood glucose data with constraint drift is subjected to constraint alignment.
[0020] After the constraint mapping, the consistency of the blood glucose data is processed, the constraint state of the blood glucose data is periodically synchronized to the cloud, the convergence speed and convergence accuracy of the constraint consensus are controlled, and the constraint condition is dynamically adjusted according to the constraint difference.
[0021] Compared with the prior art, the embodiments of the present application have at least the following beneficial effects: The present application provides a data remote management system of a sugar monitoring instrument, which can monitor the state of blood glucose data in the transmission process and dynamically adjust the related parameters of collection, transmission and processing by introducing a data tracking unit, a suspended state data identification unit and a dynamic adjustment mechanism. In particular for intermediate state data, the system can track through an independent state chain and identify and mark the data as suspended state. If the data stays at a certain node for more than a predetermined threshold, the system can timely find and process these data to avoid being incorrectly identified as valid data, thereby effectively preventing false analysis and reporting results caused by intermediate state data. In this way, the system can greatly reduce the risk of blood glucose monitoring affected by misidentification of intermediate state data.
[0022] The present application introduces a constraint mapping module and a constraint consensus module to ensure that the constraint conditions of each blood glucose data at different nodes are consistently managed. The system allocates a high-dimensional constraint vector to each blood glucose data and monitors the change of the constraint condition of each node in real time during data processing, calculates the distance of the constraint space vector to determine whether there is constraint drift, and adjusts according to the constraint difference. The constraint consensus module can dynamically adjust the convergence speed and accuracy of the constraint consensus to ensure that different levels of the system can maintain high consistency when processing blood glucose data. In this way, the system not only maintains high data processing performance, but also effectively solves the inconsistency problem caused by system upgrade or node asynchronous adjustment, thereby ensuring the accuracy and consistency of the data.
[0023] The application greatly reduces the consumption of computing resources and the complexity of the system through the local lookup table and the adaptive acquisition frequency adjustment algorithm. In the data acquisition and transmission process, the system can automatically adjust dynamically according to the sensor state, network state and device performance without relying too much on the precision synchronization of each level. Through this optimization design, the system can significantly improve the performance of the system while maintaining the data precision and consistency, reduce the delay and computing burden caused by precision synchronization, and thus ensure the processing effect and response speed of real-time data.
[0024] The application proposes a targeted repair strategy for blood glucose data affected by drift (especially suspended state data). For suspended state data, not only regression methods are used for repair, but also interpolation repair techniques are combined to fill in data gaps by using interpolation of previous and subsequent data points, ensuring that the repair process is more accurate and reliable. At the same time, the repaired data is also aligned with the global constraint space to further ensure data consistency. If the repaired data still has significant differences with the global constraint space, the system can trigger further adjustment. Through this method, the system can accurately repair the data affected by drift and ensure the stability and consistency of the repaired data, thereby avoiding affecting the subsequent data analysis results.
[0025] Of course, implementing any product of the application does not necessarily need to achieve all the above advantages at the same time. BRIEF DESCRIPTION OF DRAWINGS
[0026] Figure 1 The figure is a schematic diagram of the system module of the application.
[0027] Figure 2 The figure is a schematic diagram of the method flow of the application.
[0028] Figure 3 The figure is a schematic diagram of the logic flow of the application.
[0029] Figure 4 The figure is a schematic diagram of the logic flow of the application. DETAILED DESCRIPTION
[0030] The technical solutions in the embodiments of the application will be described clearly and completely below with reference to the accompanying drawings in the embodiments of the application. Obviously, the described embodiments are only part of the embodiments of the application, not all the embodiments. Based on the embodiments in the application, all other embodiments obtained by those skilled in the art without creative labor are within the scope of protection of the application.
[0031] It should be noted that in the embodiments of the present application, the lookup table is a pre-stored mapping data structure for quickly mapping input parameters to corresponding output values, avoiding real-time calculation overhead, and the core function of the lookup table is to quickly obtain the corresponding control or correction value according to the input state parameter or deviation value. It can be regarded as a discrete function approximation for mapping continuous parameter space to executable system adjustment actions, realizing fast response, low delay and edge computing feasibility. In the system construction stage, the designer presets the lookup table by combining historical prior data.
[0032] Referring to Figure 1 The embodiments of the present application provide a data remote management system of a sugar monitoring instrument, which comprises: As Figure 3 and Figure 4 The logic flow chart in the embodiments of the present application is used to intuitively show the overall working process of the data remote management system of the sugar monitoring instrument and the logical relationship between the modules. The closed-loop processing logic of the data acquisition, transmission, suspension identification, constraint alignment and drift correction of the system is shown. It is described that from the blood glucose data acquisition, the sampling frequency is adjusted based on the reference model parameters and the dynamic state vector; then the network state is monitored in real time during the transmission of the blood glucose data, and the transmission parameters are dynamically adjusted according to the network condition; during the data transmission and processing, the suspended state data is identified, and the suspended or drifted data is processed through interpolation repair, constraint space mapping and alignment; finally, the processed data is synchronized to the cloud to realize data consistency and constraint consensus.
[0033] The data management module is used to monitor the data state of the blood glucose data in the transmission process, and the data management module comprises a data acquisition unit, a data transmission unit, a data tracking unit and a suspended state data identification unit.
[0034] The data acquisition unit is used to acquire blood glucose data from the sensor and dynamically adjust the acquisition frequency.
[0035] The blood glucose data is acquired from the sensor, and the acquisition frequency is dynamically adjusted, and the specific process is as follows: When the blood glucose data is first acquired from the sensor, the default sampling frequency is used, and the reference model construction parameters are recorded at the same time, including the initial baseline blood glucose value, the body surface temperature and the sensor signal-to-noise ratio.
[0036] The initial baseline blood glucose value reflects the starting point of the blood glucose level of the blood glucose data. If the baseline blood glucose value is high, the sampling frequency needs to be increased to capture the rapidly changing data points.
[0037] The body surface temperature will affect the optical response of the sensor. When the body surface temperature deviates from the optimal working interval of the sensor, the measurement may be unstable or the drift increases, and it is necessary to increase the sampling frequency to obtain more data.
[0038] Based on the reference model construction parameters, a collection reference model is constructed, and the collection reference model outputs a reference sampling frequency.
[0039] The construction of the collection reference model is to create a preliminary data collection model according to the above recorded parameters (baseline blood glucose value, body surface temperature, sensor signal-to-noise ratio). The collection reference model is used to predict the frequency of collecting blood glucose data under different conditions. For example, if the sensor signal-to-noise ratio is high, a higher sampling frequency can be used, and if the signal is weak, the sampling frequency may need to be reduced.
[0040] The reference model construction parameters, including the initial baseline blood glucose value, body surface temperature and sensor signal-to-noise ratio, are converted into dimensionless indicators, specifically baseline blood glucose deviation, temperature deviation and signal-to-noise ratio indicators, and each dimensionless indicator is normalized to [0, 1]. The specific conversion process is: ; ; ; Wherein, is the baseline blood glucose deviation, is the temperature deviation, and S is the signal-to-noise ratio indicator, is the collected initial baseline blood glucose value, is the default baseline blood glucose value, is the blood glucose normalization scale, which is used to describe the upper limit of typical fluctuations and is preset by the model. In the embodiments of the present application, is 5mmol / L. In actual application, it needs to be adjusted based on design requirements. is the collected body surface temperature, is the default body surface temperature, is the temperature normalization scale, and the acquisition method is the same as . is the collected sensor signal-to-noise ratio, is the best signal-to-noise ratio that the device can achieve, and the acquisition method is the factory calibration of the device.
[0041] ; Wherein, is the comprehensive construction indicator. In the embodiments of the present application, since the normalization has been performed in the previous step, the weighting coefficient is no longer set. In actual application, the corresponding weighting coefficient needs to be set based on design requirements.
[0042] ; Wherein, is the reference sampling frequency, is the minimum collection frequency, is the default sampling frequency, is the maximum acquisition frequency, is the adjustment coefficient in the clipping limit function, used to limit the parameter range, is the comprehensive construction index, output by the acquisition reference model, is the clipping limit function.
[0043] For example, = 1 / 1 min (i.e., once per minute), = 1 / 10 min (i.e., once every 10 minutes), = 1 / 10 s (i.e., once every 10 seconds), = 3. If = 0.4, then is 2.2 / 1 min (i.e., 2.2 times per minute), and then quantize f (because the device can only support discrete gears) to the closest frequency.
[0044] After the second acquisition, the blood glucose data is statistically obtained, including the blood glucose change rate, the outlier index, and the sensor optical reflection stability, and the blood glucose data is quantized as a dynamic acquisition state vector.
[0045] The blood glucose change rate refers to the rate of change of blood glucose, i.e., the rate of increase or decrease of blood glucose value. When the change rate is large, more frequent data acquisition may be needed to capture rapid fluctuations.
[0046] The outlier index measures the degree of deviation of the current blood glucose value from the normal range. If the blood glucose value is too high or too low, the system can consider adjusting the sampling frequency to take more samples to better track abnormal fluctuations.
[0047] The sensor optical reflection stability is used to measure the stability of the sensor reading optical signals. Instability of the sensor can lead to reduced data quality, so when the optical reflection stability is poor, the acquisition frequency may need to be increased.
[0048] Based on the dynamic acquisition state vector, the corresponding acquisition frequency adjustment value is found in the local lookup table, and the reference sampling frequency is adjusted.
[0049] The data transmission unit is used to transmit blood glucose data to each module for processing, and dynamically adjusts the transmission parameters based on network state data.
[0050] The blood glucose data is transmitted to each module for processing, and the transmission parameters are dynamically adjusted based on network state data, the specific process being: When the blood glucose data is collected from the sensor, it enters the network transmission phase, and the network status data between the real-time monitoring device and the cloud is monitored in real time, including RTT (round trip time), Ping fluctuation, network packet loss rate, and network bandwidth change rate.
[0051] RTT (Round Trip Time) refers to the time it takes for a data packet to travel from the sender to the receiver and back. A higher RTT indicates a larger network delay, which can cause delays in data transmission.
[0052] Ping is a test of network delay by sending an ICMP echo request. Ping fluctuation refers to the degree of instability of network delay. Large fluctuations can affect the stability of data.
[0053] Network packet loss rate refers to the proportion of data packets lost during data transmission. A high packet loss rate indicates poor network quality, and there is a risk of data loss during transmission.
[0054] Network bandwidth change rate refers to the rate of change of network bandwidth. Large or unstable changes in bandwidth can affect the transmission rate and reliability of data.
[0055] The network status data is compared with the corresponding network status deterioration indicator data, which includes RTT deterioration indicator, Ping fluctuation deterioration indicator, network packet loss rate deterioration indicator, and network bandwidth change rate deterioration indicator. If any network status data exceeds the corresponding network status deterioration indicator data, the degradation transmission is started. The network status data that exceeds the corresponding network status deterioration indicator data is recorded as network deterioration data. The deviation between the network deterioration data and the corresponding network status deterioration indicator data is obtained and recorded as network deterioration deviation. The network deterioration deviation is normalized and input into the pre-stored lookup table to obtain the corresponding transmission parameter adjustment data set.
[0056] The RTT deterioration indicator is used to detect whether the network delay exceeds the normal range. RTT (Round-Trip Time) reflects the total time it takes for data to be transmitted from the source to the target and returned. If the RTT value abnormally increases, it indicates that the network delay increases, which can cause data transmission to slow down.
[0057] The Ping fluctuation deterioration indicator is used to detect whether the communication stability exceeds the normal range. Ping fluctuation refers to the fluctuation of network delay. Large fluctuations usually indicate that the network is unstable, which can affect the stability of real-time applications and data streams.
[0058] The network packet loss rate deterioration index is used to identify whether the data loss in transmission exceeds the normal range. Higher packet loss rate means that there is more data loss in the network transmission process, which may cause information loss, transmission error or retransmission, and increase the transmission delay.
[0059] The network bandwidth change rate deterioration index is used to identify whether the large change of bandwidth exceeds the normal range, especially when the bandwidth suddenly decreases. Such fluctuations may cause bottlenecks in data flow, affecting the transmission efficiency of large data.
[0060] It should be noted that the RTT deterioration index for judging network delay anomaly, the Ping fluctuation deterioration index for detecting communication stability, the network packet loss rate deterioration index for identifying transmission reliability decline, and the network bandwidth change rate deterioration index for identifying sudden bandwidth drop are set by designers based on device operation history or experience rules.
[0061] It should be noted that by normalizing the deterioration deviation, it is ensured that network state data of different types can be compared on the same scale. This process will make the network deterioration deviation value between 0 and 1, which is convenient for further processing. The normalized network deterioration deviation value will be input into the local lookup table, and in the embodiment of the present application, the lookup table has pre-stored corresponding transmission parameter adjustment data set. The transmission parameter adjustment data set includes sampling frequency adjustment value, data compression rate adjustment value, etc.
[0062] By dynamically adjusting the network transmission parameters, the system can automatically optimize the data transmission strategy in the case of network quality deterioration.
[0063] The data tracking unit is used to record the data state of blood glucose data at each time node.
[0064] The hanging state data identification unit is used to identify the hanging state data generated in the blood glucose data transmission process. The hanging state data refers to the blood glucose data that has occupied a position in the system but does not have reliable effectiveness.
[0065] The hanging state data generated in the blood glucose data transmission process is identified, which specifically includes: The hanging identification parameters of each piece of blood glucose data are obtained, including the transmission queue state, ACK receipt time, data version number and residence time of each node in the processing period. It should be noted that each piece of blood glucose data has an independent state chain, and in the embodiment of the present application, the independent state chain is collection-->send-->confirm-->warehouse.
[0066] Based on the transmission queue state of the blood glucose data and the residence time of each node in the processing period, if the residence time of a piece of blood glucose data in a certain node exceeds the pre-set residence time threshold, the piece of blood glucose data will be put into the candidate hanging pool and recorded as candidate hanging blood glucose data.
[0067] In the candidate suspension pool, each candidate suspension blood glucose data is further detected by a suspension state detection condition, the suspension state detection condition includes sequence interruption, time expiration, CRC inconsistency and existence of repeated version number, and the suspension identification parameter of each candidate suspension blood glucose data is compared with the suspension state detection condition. If the suspension identification parameter of a certain candidate suspension blood glucose data meets any suspension state detection condition, the candidate suspension blood glucose data is determined as suspension state data.
[0068] It should be noted that the sequence number in data transmission should be continuous. If the sequence number is interrupted during data transmission, it may mean that some data is lost or transmission error occurs. If the sequence number of a certain data is not continuous with the sequence number of the previous or next data, the data is determined as sequence interruption.
[0069] It should also be noted that in blood glucose data transmission, each data has a corresponding time stamp for recording the time of data collection and transmission. If the time stamp of a certain blood glucose data exceeds the predetermined maximum allowed time, it means that the data has been stranded in the network for too long and cannot be confirmed or processed in time.
[0070] CRC (Cyclic Redundancy Check) is a method for detecting whether an error occurs in the data transmission process. In the data transmission process, each data will have a CRC check value, and the receiving end will recalculate and verify this check value. If the CRC check value of the received data is inconsistent with the calculated value, it means that the data has been damaged in the transmission process. If it is found that the CRC check of a certain blood glucose data fails, it means that the data has an error in the transmission process.
[0071] Blood glucose data usually has a unique version number in the system. If the system finds that the version number of a certain data is repeated, it means that the same data is sent or processed multiple times, which is usually caused by network delay or error.
[0072] If the suspension identification parameter of a certain candidate suspension blood glucose data does not meet all suspension state detection conditions, the candidate suspension blood glucose data is released from the candidate suspension pool.
[0073] The constraint mapping module is used for constraint mapping of blood glucose data in the transmission process, and the constraint mapping module includes a constraint space modeling unit, a constraint drift detection unit and a data repair adjustment unit.
[0074] The constraint space modeling unit is used for assigning a high-dimensional constraint vector to each blood glucose data.
[0075] Each blood glucose data is assigned a high-dimensional constraint vector, specifically including: The high-dimensional constraint vector is used to describe the consistency of each blood glucose data in the system, including format consistency, version consistency and processing state consistency in the embodiment of the application, and the consistency range can be specified based on the specific processing requirements of the data in actual application.
[0076] Before collection, a static constraint dimension is established for the blood glucose data based on the sensor type, including precision constraint dimension, unit constraint dimension, interval range constraint dimension and time constraint dimension, and an initial constraint vector is constructed based on the static constraint dimension.
[0077] The maximum / minimum measurement range, nominal precision, supported minimum resolution and unit calibrated at the factory of the sensor are obtained as the basis for constructing the static constraint dimension. The constraint dimension is a set of quantifiable boundary rules for limiting the blood glucose data collection format, precision, value range and time logic, and the combination thereof forms a constraint vector, which is used to drive the data validity judgment and dynamic control in the collection process.
[0078] After generating the corresponding initial constraint vector, the initial constraint vector is used to identify the data baseline. The initial constraint vector is written into the constraint management chain in the local and cloud, and serves as a reference framework for all subsequent collected data. Thereafter, each blood glucose data carries or is associated with the constraint vector when transmitted to the system, so as to verify whether the data conforms to the collection environment and measurement specification recognized by the system.
[0079] After collection, the initial constraint vector is updated based on the real-time information of the sensor, and a high-dimensional constraint vector of each blood glucose data is output, wherein the real-time information of the sensor includes sensor device state parameters, network state parameters and algorithm state parameters.
[0080] After the blood glucose data collection is completed, the initial constraint vector is dynamically modified in combination with the current running environment and performance feedback of the sensor. Specifically, the real-time information of the sensor is collected, normalized and compared with the performance degradation threshold corresponding to the real-time information of the sensor, and then the influence degree of the real-time information of the sensor on the current data reliability is calculated based on an influence factor weight model. The performance degradation threshold is provided through the sensor specification, which usually lists the performance specifications and use limitations of the sensor, and the performance degradation threshold is calibrated based on the use limitations.
[0081] The influence factor weight model is a mathematical model for quantifying the influence degree of each real-time parameter on the blood glucose data constraint vector, which can map the real-time information of the sensor to a standardized influence value of [0, 1]. In the system construction stage, the influence factor weight model can be pre-set by the designer, or a large amount of sensor information data can be input to fit the influence of the real-time information of the sensor on the data error using a machine learning method.
[0082] According to the influence degree, each constraint dimension in the initial constraint vector is modified, including multiplying the normalized influence value with the initial constraint vector, thereby obtaining the high-dimensional constraint vector of each blood glucose data.
[0083] For example, assuming that the initial constraint of a blood glucose data is: Precision: 0.1 mmol / L; Measurement range: 1.0-30.0 mmol / L; Sampling frequency: once per minute; Unit: mmol / L; After the update of the device temperature rise and the battery power drop, the constraint vector becomes: Precision: 1.0 mmol / L; Measurement range: 2.0-28.0 mmol / L; Sampling frequency: once every 5 minutes; Unit: mmol / L; The high-dimensional constraint vector of each blood glucose data is written into the state chain.
[0084] The constraint drift detection unit is used to detect the change of the constraint condition of the blood glucose data in the processing period, calculate the constraint space vector distance, and determine whether the blood glucose data has constraint drift.
[0085] The constraint condition of the blood glucose data in the processing period is changed, the constraint space vector distance is calculated, and it is determined whether the blood glucose data has constraint drift, including: At each node in the processing period, a state snapshot is generated for each blood glucose data, and after the dimension alignment, a comparable vector of the blood glucose data at each node is formed, the constraint space vector distance is calculated,
[0086] wherein, is the constraint space vector distance of the blood glucose data between the i-th node and the j-th node, is the comparable vector of the blood glucose data at the i-th node, is the comparable vector of the blood glucose data at the j-th node, i and j are the node numbers in the processing period, i=1, 2, 3,..., N, j=1, 2, 3,..., N, i≠j, and N is the total number of nodes in the processing period.
[0087] The constraint space vector distance is compared with the dynamic threshold value, if the constraint space vector distance is greater than the dynamic threshold value, it is determined that there is constraint drift, if the constraint space vector distance is less than or equal to the dynamic threshold value, it is determined that there is no constraint drift.
[0088] The dynamic threshold is a variable threshold value for determining whether a constraint drift occurs, and is dynamically adjusted based on real-time information of the sensor.
[0089] The data repair adjustment unit is used for constraint alignment of the blood glucose data with constraint drift.
[0090] The constraint alignment of the blood glucose data with constraint drift includes: The blood glucose data with constraint drift is recorded as drift-affected data, a constraint space vector distance of the drift-affected data is extracted, a drift distance of the drift-affected data is obtained by subtracting the dynamic threshold from the constraint space vector distance, and a corresponding drift-affected coefficient is obtained by inputting the drift distance into a locally pre-stored lookup table.
[0091] The drift-affected data refers to blood glucose data with data deviation caused by changes in system constraints (such as precision, data format, range, etc.) during data processing. Constraint drift usually causes systematic errors or distortion of data. If the drift-affected data is pending data, data repair and constraint alignment are performed simultaneously. The constraint space vector is a high-dimensional vector used to describe the position of the blood glucose data in the constraint space, indicating the characteristics of the data in terms of precision, format, unit, range, etc. The constraint space vector distance refers to the difference between the blood glucose data and the target constraint (such as ideal constraint or standard constraint) in the constraint space. This distance is used to quantify the degree of drift of the data. The drift-affected coefficient is used to quantify the drift impact. The drift-affected coefficient adjusts the value of the blood glucose data according to the size and type of the drift, thereby correcting the errors caused by the drift. The pre-stored lookup table records the correction coefficients (drift-affected coefficients) corresponding to different drift distances. When the drift-affected data is identified, the system looks up the correction coefficient corresponding to the drift distance and uses it to repair the data.
[0092] If the drift-affected data is not pending data, the drift-affected data is aligned based on the drift-affected coefficient of the drift-affected data using a regression method.
[0093] The regression method is a fitting method used in the embodiments of the present application to predict the constraint vector adjustment amount that needs to be corrected according to the drift influence coefficient. The regression model is selected, and in the embodiments of the present application, a linear regression is selected because there is a linear relationship between the drift influence coefficient and the constraint vector adjustment amount. In actual applications, the corresponding regression model such as a weighted regression model and a nonlinear regression model needs to be selected according to the calculation requirements. It should be noted that the linear relationship between the drift influence coefficient and the constraint vector adjustment amount is that the larger the drift influence coefficient, the more serious the drift of the blood glucose data, and the greater the constraint change required. In the system construction stage, the linear regression model is fitted by inputting a large amount of sample data. The drift influence coefficient is input, and the regression model outputs the constraint vector correction amount, which represents the amplitude of adjustment of each constraint dimension. The constraint vector correction amount is superimposed with the original constraint vector to complete the constraint alignment operation.
[0094] Meanwhile, data repair and constraint alignment are performed, and the specific process is as follows: The interpolation repair is performed on the suspended state data, the suspended data gap is filled by interpolation using the data points before and after the suspended state data, the repaired suspended state data is mapped to the constraint space, the vector distance between the repaired suspended state data and the global constraint space is calculated, and if the vector distance between the repaired suspended state data and the global constraint space exceeds the dynamic threshold, further adjustment of data repair and constraint space alignment is triggered until the vector distance between the repaired suspended state data and the global constraint space does not exceed the dynamic threshold.
[0095] If the vector distance between the repaired suspended state data and the global constraint space does not exceed the dynamic threshold, the repaired suspended state data and the adjusted constraint space state are synchronized to the cloud.
[0096] The constraint consensus module is used for consistency processing of the blood glucose data after constraint mapping. The constraint consensus module includes a state synchronization unit, a convergence control unit, and a consensus adjustment unit.
[0097] The state synchronization unit is used for periodically synchronizing the constraint state of the blood glucose data to the cloud.
[0098] The convergence control unit is used for controlling the convergence speed and convergence accuracy of the constraint consensus.
[0099] The convergence speed and convergence accuracy of the constraint consensus are controlled, and the specific processing conditions are as follows: The convergence speed of the constraint consensus is controlled, and the specific process is as follows: In the processing cycle, whether the acceleration convergence is required is judged based on the change rate of the blood glucose data of each node, if the change rate of the blood glucose data of a node exceeds the preset change rate threshold, it is judged that the node requires acceleration convergence, and the convergence speed adjustment value is found in the local lookup table based on the change rate difference between the change rate of the blood glucose data of the node and the change rate threshold. The change rate of the blood glucose data of each node refers to the rate of change of the blood glucose data in the processing cycle.
[0100] If the change rate of the blood glucose data of a node is less than or equal to the preset change rate threshold, it is judged that the node does not require acceleration convergence.
[0101] The convergence precision of the constraint consensus is controlled, and the specific process is: The change rate of the blood glucose data of each node is input into the local lookup table to find the error tolerance precision adjustment value, and the error tolerance precision is adjusted.
[0102] In the process of constraint consensus, when the difference between the constraint state of all nodes and the global constraint state in the processing cycle is less than the error tolerance precision, it is judged that the constraint consensus is completed, and when the difference between the constraint state of a node and the global constraint state is greater than or equal to the error tolerance precision, the convergence precision adjustment value is found in the local lookup table based on the difference between the state difference and the error tolerance precision.
[0103] The consensus adjustment unit is used to dynamically adjust the constraint condition according to the constraint difference.
[0104] In this embodiment, the present application provides a data remote management method of a sugar monitoring instrument, comprising: The data state of the blood glucose data in the transmission process is monitored, the blood glucose data is collected from the sensor, and the collection frequency is dynamically adjusted, the blood glucose data is transmitted to each module for processing, and the transmission parameter is dynamically adjusted based on the network state data, the data state of the blood glucose data at each time node is recorded, and the suspended state data generated in the blood glucose data transmission process is identified. Suspended state data refers to blood glucose data that has occupied a position in the system but does not have reliable effectiveness.
[0105] The blood glucose data is constrained and mapped, each blood glucose data is allocated a high-dimensional constraint vector, the change of the constraint condition of the blood glucose data in the processing cycle is detected, the constraint space vector distance is calculated to judge whether the blood glucose data has constraint drift, and the blood glucose data with constraint drift is subjected to constraint alignment.
[0106] The blood glucose data is subjected to consistency processing, the constraint state of the blood glucose data is periodically synchronized to the cloud, the convergence speed and convergence precision of the constraint consensus are controlled, and the constraint condition is dynamically adjusted according to the constraint difference.
[0107] It is to be understood that the terminology used herein is for the purpose of describing particular embodiments only and is not intended to be limiting; it is not intended to exclude myriad other embodiments of the present application that other present or future devices, platforms, modules, components and systems can utilize, nor is it intended to exclude prior art contrary to this application. It must be noted that as used herein, the singular form "a", "an" and "the" include plural references unless the context clearly dictates otherwise. As such, the terms "comprise" (and grammatical variations thereof, such as "comprising" and "comprises"), "have" (and grammatical variations thereof, such as "having" and "has"), "include" (and grammatical variations thereof, such as "including" and "includes") or the like, are used herein not to limit the component, element, or method or process or steps as defined or implicit by such terms, but rather only to discern that such components or elements or methods or processes or steps are either included in, or encompassed by, the described embodiments. As used herein, "and / or" means and. Unless otherwise defined, all terms (including technical and scientific terms) used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application belongs. It will be further understood that terms, such as those defined in commonly used dictionaries, should be interpreted as having a meaning that is consistent with their meaning in the context of the relevant art and the present disclosure, and will not be interpreted in an overly literal or overly formal sense unless expressly so defined herein.
[0108] The preferred embodiments of the application disclosed above are only to help explain the principles of the present application. The preferred embodiments do not describe all the details of the present application, nor limit the present application to the specific embodiments. It is apparent that many modifications and variations can be made to the present application according to the contents of the present specification. The present specification selects and specifically describes these embodiments in order to better explain the principles and practical applications of the present application, so that those skilled in the art can well understand and utilize the present application. Any modifications and variations that do not deviate from the structure of the present application or exceed the scope defined by the present application shall be within the scope of protection of the present application.
Claims
1. A remote data management system for a glucose metabolism monitor, characterized in that, include: The data management module is used to monitor the data status of blood glucose data during transmission. The data management module includes a data acquisition unit, a data transmission unit, a data tracking unit, and a suspended state data identification unit. The data acquisition unit is used to acquire blood glucose data from the sensor and dynamically adjust the acquisition frequency; The data transmission unit is used to transmit blood glucose data to each module for processing and to dynamically adjust transmission parameters based on network status data. The data tracking unit is used to record the data status of blood glucose data at each time point; The suspended state data identification unit is used to identify suspended state data generated during blood glucose data transmission; The constraint mapping module is used to perform constraint mapping on blood glucose data during transmission. The constraint mapping module includes a constraint space modeling unit, a constraint drift detection unit, and a data repair and adjustment unit. The constraint space modeling unit is used to assign high-dimensional constraint vectors to each blood glucose data. The constraint drift detection unit is used to detect changes in the constraint conditions of blood glucose data during the processing cycle and to calculate the constraint space vector distance to determine whether there is constraint drift in the blood glucose data. The data repair and adjustment unit is used to perform constraint alignment on the constrained drift blood glucose data; The constraint consensus module is used to perform consistency processing on blood glucose data after constraint mapping. The constraint consensus module includes a state synchronization unit, a convergence control unit, and a consensus adjustment unit. The status synchronization unit is used to periodically synchronize the constraint status of blood glucose data to the cloud. The convergence control unit is used to control the convergence speed and convergence accuracy of the constraint consensus. The consensus adjustment unit is used to dynamically adjust the constraint conditions based on the constraint differences.
2. The remote data management system for a glucose metabolism monitor according to claim 1, characterized in that: The specific process of acquiring blood glucose data from the sensor and dynamically adjusting the acquisition frequency is as follows: When collecting blood glucose data from the sensor for the first time, the default sampling frequency is used, and reference model construction parameters, including the initial baseline blood glucose value, body surface temperature and sensor signal-to-noise ratio, are recorded at the time of collection. Based on the parameters of the reference model, a data acquisition reference model is constructed, and the data acquisition reference model outputs a reference sampling frequency. After secondary acquisition, blood glucose data were obtained, including blood glucose change rate, outlier index and sensor optical reflection stability. The blood glucose data was then quantified into a dynamic acquisition state vector. Based on the dynamic acquisition state vector, the corresponding acquisition frequency adjustment value is retrieved from the local lookup table, and the reference sampling frequency is adjusted accordingly.
3. The remote data management system for a glucose metabolism monitor according to claim 1, characterized in that: The process of transmitting blood glucose data to various modules for processing and dynamically adjusting transmission parameters based on network status data is as follows: Once the blood glucose data is collected from the sensor, it enters the network transmission stage, and the network status data between the device and the cloud is monitored in real time. The network status data includes RTT, Ping fluctuation, network packet loss rate and network bandwidth change rate. The network status data is compared with the corresponding network status deterioration index data, which includes RTT deterioration index, Ping fluctuation deterioration index, network packet loss rate deterioration index, and network bandwidth change rate deterioration index. If any network status data exceeds the corresponding network status deterioration index data, degraded transmission is initiated. The network status data that exceeds the corresponding network status deterioration index data is recorded as network deterioration data. The deviation between the network deterioration data and the corresponding network status deterioration index data is obtained and recorded as network deterioration deviation. After normalizing the network deterioration deviation, it is input into a locally pre-stored lookup table to obtain the corresponding transmission parameter adjustment dataset.
4. The remote data management system for a glucose metabolism monitor according to claim 1, characterized in that: The identification of suspended data generated during blood glucose data transmission specifically includes: Obtain the suspension identification parameters for each blood glucose data, including transmission queue status, ACK receipt time, data version number, and dwell time at each node in the processing cycle; Based on the transmission queue status of blood glucose data and the dwell time of each node in the processing cycle, if the dwell time of a blood glucose data in a certain node exceeds the preset dwell time threshold, the blood glucose data will be put into the candidate suspension pool and recorded as candidate suspended blood glucose data. In the candidate suspension pool, each candidate suspended blood glucose data is detected by suspension state detection conditions. The suspension state detection conditions include sequence interruption, time expiration, CRC inconsistency, and the existence of duplicate version numbers. The suspension identification parameters of each candidate suspended blood glucose data are compared with the suspension state detection conditions. If the suspension identification parameters of a candidate suspended blood glucose data meet any suspension state detection condition, then the candidate suspended blood glucose data is determined as suspended state data. If the suspension identification parameters of a candidate suspended blood glucose data do not meet all suspension state detection conditions, the candidate suspended blood glucose data will be released from the candidate suspension pool.
5. The remote data management system for a glucose metabolism monitor according to claim 1, characterized in that: The process of assigning high-dimensional constraint vectors to each blood glucose data point specifically includes: Before data collection, static constraint dimensions are established for blood glucose data based on sensor type, including accuracy constraint dimension, unit system constraint dimension, interval range constraint dimension and time constraint dimension, and an initial constraint vector is constructed based on the static constraint dimensions. After data collection, the initial constraint vector is updated based on real-time sensor information, and the high-dimensional constraint vector of each blood glucose data is output. The real-time sensor information includes sensor device status parameters, network status parameters, and algorithm status parameters. Write the high-dimensional constraint vectors of each blood glucose data into the state chain.
6. The remote data management system for a glucose metabolism monitor according to claim 1, characterized in that: The changes in constraints of the detected blood glucose data during the processing cycle are analyzed, and the constraint space vector distance is calculated to determine whether constraint drift exists in the blood glucose data. Specifically, this includes: At each node in the processing cycle, a state snapshot is generated for each blood glucose data. After performing dimension alignment, a comparable vector of blood glucose data is formed at each node, and the constraint space vector distance of each node is calculated. The constraint space vector distance is compared with the dynamic threshold. If the constraint space vector distance of a node is greater than the dynamic threshold, the node is determined to have constraint drift. If the constraint space vector distance is less than or equal to the dynamic threshold, the node is determined not to have constraint drift.
7. The remote data management system for a glucose metabolism monitor according to claim 1, characterized in that: The constraint alignment of the blood glucose data to address constraint drift specifically includes: Blood glucose data with constraint drift is recorded as drift impact data. The constraint space vector distance of the drift impact data is extracted and subtracted from the dynamic threshold to obtain the drift distance of the drift impact data. The drift distance is then entered into a locally stored lookup table to find the corresponding drift impact coefficient. If the data affected by the drift is suspended data, then data repair and constraint alignment are performed simultaneously. If the drift impact data is not suspended data, then the drift impact data is constrained and aligned using a regression method based on the drift impact coefficient of the drift impact data.
8. The remote data management system for a glucose metabolism monitor according to claim 1, characterized in that: The convergence speed and accuracy of the control constraint consensus are specifically processed under the following conditions: The convergence speed of the constraint consensus is controlled, and the specific process is as follows: During the processing cycle, based on the rate of change of blood glucose data at each node, it is determined whether accelerated convergence is required. If the rate of change of blood glucose data at a certain node exceeds the preset rate of change threshold, it is determined that the node needs accelerated convergence. Based on the difference between the rate of change of blood glucose data at that node and the rate of change threshold, the convergence speed adjustment value is obtained by looking up the value in the local lookup table. If the rate of change of blood glucose data at a certain node is less than or equal to the preset rate of change threshold, then it is determined that the node does not need to accelerate convergence. The convergence accuracy of the constraint consensus is controlled through the following process: During the processing cycle, the rate of change of blood glucose data at each node is entered into a local lookup table to find the error tolerance precision adjustment value, and the error tolerance precision is adjusted accordingly. During the constraint consensus process, when the difference between the constraint state of all nodes within the processing cycle and the global constraint state is less than the error tolerance precision, the constraint consensus is considered complete. When the difference between the constraint state of a node and the global constraint state is greater than or equal to the error tolerance precision, the convergence precision adjustment value is obtained by looking up the value in the local lookup table based on the difference between the state difference and the error tolerance precision.
9. The remote data management system for a glucose metabolism monitor according to claim 7, characterized in that: The process of simultaneously performing data repair and constraint alignment is as follows: Interpolation repair is performed on the suspended data. Data points before and after the suspended data are interpolated to fill the gaps in the suspended data. The repaired suspended data is mapped to the constraint space. The vector distance between the repaired suspended data and the global constraint space is calculated. If the vector distance between the repaired suspended data and the global constraint space exceeds the dynamic threshold, further adjustments to the data repair and constraint space alignment are triggered. If the vector distance between the repaired suspended state data and the global constraint space does not exceed the dynamic threshold, the repaired suspended state data and the adjusted constraint space state will be synchronized to the cloud.
10. A method for a remote data management system applied to a glucose metabolism monitor according to any one of claims 1-9, characterized in that: The system monitors the data status of blood glucose data during transmission, collects blood glucose data from sensors, dynamically adjusts the collection frequency, transmits blood glucose data to various modules for processing, dynamically adjusts transmission parameters based on network status data, records the data status of blood glucose data at each time point, and identifies suspended data generated during blood glucose data transmission. Suspended data refers to blood glucose data that has occupied a position in the system but does not have reliable validity. Constraint mapping is performed on blood glucose data during transmission, high-dimensional constraint vectors are assigned to each blood glucose data, changes in constraint conditions of blood glucose data are detected during the processing cycle, constraint space vector distance is calculated to determine whether there is constraint drift in blood glucose data, and constraint alignment is performed on blood glucose data with constraint drift. After constraint mapping, the blood glucose data is processed for consistency. The constraint status of the blood glucose data is periodically synchronized to the cloud to control the convergence speed and accuracy of constraint consensus. The constraint conditions are dynamically adjusted according to the constraint differences.