CN119970018A - A blood sugar signal processing method for blood sugar detection - Google Patents
A blood sugar signal processing method for blood sugar detection Download PDFInfo
- Publication number
- CN119970018A CN119970018A CN202510131765.7A CN202510131765A CN119970018A CN 119970018 A CN119970018 A CN 119970018A CN 202510131765 A CN202510131765 A CN 202510131765A CN 119970018 A CN119970018 A CN 119970018A
- Authority
- CN
- China
- Prior art keywords
- blood glucose
- screening
- optical
- value
- electrochemical
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Landscapes
- Investigating Or Analysing Biological Materials (AREA)
- Measurement Of The Respiration, Hearing Ability, Form, And Blood Characteristics Of Living Organisms (AREA)
Abstract
本发明公开了一种用于血糖检测的血糖信号处理方法，涉及血糖信号处理技术领域，该方法包括：利用光学血糖传感器和电化学血糖传感器对目标用户在预测检测窗口的血糖进行同步连续检测，获得光学血糖信号检测值序列和电化学血糖信号检测值序列；获得光学血糖信号检测值子序列集合和电化学血糖信号检测值子序列集合；进行多模态交互融合分析，获得血糖融合特征；调用目标用户的历史血糖特征集合对血糖融合特征进行认证，获得认证结果，若认证结果为通过，则将血糖融合特征作为血糖信号处理结果。本发明解决了现有技术中血糖信号处理准确性低，无法有效反映用户血糖情况的技术问题。达到了提高血糖信号处理的可靠性的技术效果。
The present invention discloses a blood sugar signal processing method for blood sugar detection, which relates to the technical field of blood sugar signal processing. The method comprises: using an optical blood sugar sensor and an electrochemical blood sugar sensor to synchronously and continuously detect the blood sugar of a target user in a predicted detection window, and obtaining an optical blood sugar signal detection value sequence and an electrochemical blood sugar signal detection value sequence; obtaining an optical blood sugar signal detection value subsequence set and an electrochemical blood sugar signal detection value subsequence set; performing multimodal interactive fusion analysis to obtain a blood sugar fusion feature; calling the target user's historical blood sugar feature set to authenticate the blood sugar fusion feature, and obtaining an authentication result. If the authentication result is passed, the blood sugar fusion feature is used as the blood sugar signal processing result. The present invention solves the technical problem that the accuracy of blood sugar signal processing in the prior art is low and the blood sugar condition of the user cannot be effectively reflected. The technical effect of improving the reliability of blood sugar signal processing is achieved.
Description
Technical Field
The invention relates to the technical field of blood glucose signal processing, in particular to a blood glucose signal processing method for blood glucose detection.
Background
Diabetes is a common metabolic disease with an increasing incidence year by year, affecting a large population worldwide. Therefore, the management of diabetes requires the ability to accurately detect the blood glucose level of a patient in real time, thereby timely adjusting the treatment regimen according to blood glucose fluctuations. The traditional blood sugar detection method mainly relies on regular blood sampling and measurement by a blood sugar meter, and the intermittent mode is simple and feasible, but can not provide dynamic change information of blood sugar, so that accurate treatment of diabetes is limited.
The prior art has the technical problems that the accuracy of blood sugar signal processing is low, and the blood sugar condition of a user cannot be effectively reflected.
Disclosure of Invention
The application provides a blood sugar signal processing method for blood sugar detection, which is used for solving the technical problems that the blood sugar signal processing accuracy is low and the blood sugar condition of a user cannot be effectively reflected in the prior art.
In view of the above problems, the present application provides a blood glucose signal processing method for blood glucose detection, the method comprising:
the method comprises the steps that synchronous and continuous detection is carried out on blood sugar of a target user in a prediction detection window by using an optical blood sugar sensor and an electrochemical blood sugar sensor, and an optical blood sugar signal detection value sequence and an electrochemical blood sugar signal detection value sequence are obtained;
Dividing the optical blood glucose signal detection value sequence and the electrochemical blood glucose signal detection value sequence according to a preset dividing scale to obtain an optical blood glucose signal detection value subsequence set and an electrochemical blood glucose signal detection value subsequence set;
performing multi-mode interactive fusion analysis on the optical blood glucose signal detection value subsequence set and the electrochemical blood glucose signal detection value subsequence set to obtain blood glucose fusion characteristics;
And calling the historical blood glucose feature set of the target user to authenticate the blood glucose fusion feature, obtaining an authentication result, and taking the blood glucose fusion feature as a blood glucose signal processing result if the authentication result is passed.
One or more technical schemes provided by the application have at least the following technical effects or advantages:
According to the application, the optical blood glucose sensor and the electrochemical blood glucose sensor are utilized to synchronously and continuously detect the blood glucose of a target user in a prediction detection window to obtain an optical blood glucose signal detection value sequence and an electrochemical blood glucose signal detection value sequence, then the optical blood glucose signal detection value sequence and the electrochemical blood glucose signal detection value sequence are respectively divided according to a preset division scale to obtain an optical blood glucose signal detection value subsequence set and an electrochemical blood glucose signal detection value subsequence set, and then multimode interaction fusion analysis is carried out on the optical blood glucose signal detection value subsequence set and the electrochemical blood glucose signal detection value subsequence set to obtain blood glucose fusion characteristics, then a historical blood glucose characteristic set of the target user is called to authenticate the blood glucose fusion characteristics to obtain an authentication result, and if the authentication result is passed, the blood glucose fusion characteristics are used as blood glucose signal processing results. The technical effect of improving the blood sugar signal processing quality and reliability is achieved.
Drawings
Fig. 1 is a schematic flow chart of a blood glucose signal processing method for blood glucose detection according to an embodiment of the present invention.
Fig. 2 is a schematic flow chart of obtaining a blood glucose fusion characteristic in a blood glucose signal processing method for blood glucose detection according to an embodiment of the present invention.
Detailed Description
The application will be further illustrated with reference to specific examples. It is to be understood that these examples are illustrative of the present application and are not intended to limit the scope of the present application. Furthermore, it should be understood that various changes and modifications can be made by one skilled in the art after reading the teachings of the present application, and such equivalents are intended to fall within the scope of the application as defined in the appended claims.
It should be noted that the terms "comprises" and "comprising" are intended to cover a non-exclusive inclusion, such that a process, method, system, article, or server that comprises a list of steps or elements is not necessarily limited to those steps or elements expressly listed or inherent to such process, method, article, or apparatus.
An embodiment, as shown in fig. 1, the present application provides a blood glucose signal processing method for blood glucose detection, where the method includes:
S1, synchronously and continuously detecting blood sugar of a target user in a prediction detection window by utilizing an optical blood sugar sensor and an electrochemical blood sugar sensor to obtain an optical blood sugar signal detection value sequence and an electrochemical blood sugar signal detection value sequence;
In one possible embodiment, the optical glucose sensor is a device that measures glucose levels by detecting light absorption, scattering, or reflection characteristics, typically with non-invasive or minimally invasive, continuous detection capabilities. The electrochemical blood glucose sensor is a device for measuring the blood glucose concentration by a current or voltage signal generated by glucose in an electrochemical reaction, and has the characteristic of high precision. The target user is an individual who receives blood glucose test, typically a diabetic or a population in need of blood glucose test.
The predictive detection window refers to a time range for blood glucose data acquisition and detection in a preset time period, and can provide continuous data support for dynamic blood glucose analysis, wherein the preset detection window can be set by a person skilled in the art. The optical blood glucose signal detection value sequence and the electrochemical blood glucose signal detection value sequence are blood glucose data recorded by an optical sensor and an electrochemical sensor respectively and are organized in a time sequence form.
The optical blood glucose sensor and the electrochemical blood glucose sensor are utilized to start the blood glucose level of the target user at the same time and continuously collect data in the same time period, so that errors caused by time offset are reduced, and consistency of the data is ensured. The optical sensor provides continuous blood glucose excursion information on a noninvasive premise, and the electrochemical sensor supplements the accuracy of the optical signal through high-precision data. The step ensures that the acquisition of blood glucose data can cover the dynamic change of a time axis, and simultaneously combines the noninvasive property and high precision, thereby providing high-quality data input for the subsequent processing steps (such as data partitioning and fusion analysis).
By utilizing the characteristic advantages of the optical blood glucose sensor and the electrical blood glucose sensor, the technical effect of making up possible defects of a single sensor is achieved, reliable detection of blood glucose is realized, and a data support target is provided for analysis of blood glucose fluctuation trend of a target user.
S2, dividing the optical blood glucose signal detection value sequence and the electrochemical blood glucose signal detection value sequence according to a preset dividing scale to obtain an optical blood glucose signal detection value subsequence set and an electrochemical blood glucose signal detection value subsequence set;
In one embodiment, the preset dividing scale is a time period of two adjacent dividing intervals preset by a person skilled in the art, such as dividing every 5 minutes or 10 minutes. And respectively dividing the optical blood glucose signal detection value sequence and the electrochemical blood glucose signal detection value sequence according to the preset dividing scale, so as to obtain the optical blood glucose signal detection value subsequence set and the electrochemical blood glucose signal detection value subsequence set.
And each subsequence in the optical blood glucose signal detection value subsequence set and the electrochemical blood glucose signal detection value subsequence set comprises continuous data in a time period with a preset division scale, so that subsequent processing and analysis are facilitated. By sub-sequence division, data support is provided for subsequent finer analysis, so that local variation trend of blood sugar, such as fluctuation amplitude, short-term fluctuation and other information, can be conveniently extracted, and the accuracy of blood sugar signal processing is improved.
S3, carrying out multi-mode interactive fusion analysis on the optical blood glucose signal detection value subsequence set and the electrochemical blood glucose signal detection value subsequence set to obtain blood glucose fusion characteristics;
In one possible embodiment, the optical glucose sensor and the electrochemical glucose sensor have different advantages and disadvantages, respectively, and the optical glucose sensor has good continuous detection performance, but is easily affected by the outside, and has low detection accuracy, while the electrochemical glucose sensor has high detection accuracy, but has poor continuous detection performance. Therefore, representative signal detection value screening is carried out on each subsequence in the optical blood glucose signal detection value subsequence set and the electrochemical blood glucose signal detection value subsequence set respectively to obtain an optical blood glucose signal screening value sequence and an electrochemical blood glucose signal screening value sequence, and the aims of reducing the dimension of data and improving the signal processing efficiency are fulfilled. And further, carrying out feature extraction based on the screened data, and determining the optical blood glucose screening features and the electrochemical blood glucose screening features in a preset detection window. The aim of extracting the characteristics of the blood glucose signals is fulfilled. And then carrying out multi-mode interactive fusion analysis on the optical blood glucose screening characteristics and the electrochemical blood glucose screening characteristics, and obtaining the blood glucose fusion characteristics through average value calculation by enhancing the optical blood glucose screening characteristics and the electrochemical blood glucose screening characteristics through interactive fusion analysis. The technical effects of providing more reliable data support for blood glucose detection and improving the processing precision of blood glucose signals are achieved.
And S4, calling the historical blood glucose feature set of the target user to authenticate the blood glucose fusion feature, obtaining an authentication result, and taking the blood glucose fusion feature as a blood glucose signal processing result if the authentication result is passed.
In one embodiment, the historical blood glucose feature set reflects the change in blood glucose of the target user over a historical time, and these features generally include fluctuations in blood glucose levels, trends, abrupt points, blood glucose concentrations, and the like, reflecting the historical pattern of blood glucose changes of the target user. After the blood glucose fusion feature is obtained, the blood glucose fusion feature is authenticated by utilizing the historical blood glucose feature set of the target user, so that whether the blood glucose fusion feature is reliable or not is determined, and if the authentication result is passed, the blood glucose fusion feature is used as the blood glucose signal processing result. Therefore, the technical effect of improving the accuracy of blood glucose signal processing is achieved.
Further, as shown in fig. 2, the multi-mode interactive fusion analysis is performed on the optical blood glucose signal detection value subsequence set and the electrochemical blood glucose signal detection value subsequence set to obtain a blood glucose fusion feature, and step S3 of the embodiment of the present application further includes:
The optical blood glucose signal detection value subsequence set and the electrochemical blood glucose signal detection value subsequence set are subjected to in-subsequence signal detection value centralized iterative screening respectively, and screening results are respectively ordered according to the sequence of detection time to obtain an optical blood glucose signal screening value sequence and an electrochemical blood glucose signal screening value sequence;
Extracting blood glucose signal characteristics from the optical blood glucose signal screening value sequence and the electrochemical blood glucose signal screening value sequence by utilizing a blood glucose signal extraction network layer to obtain optical blood glucose screening characteristics and electrochemical blood glucose screening characteristics;
And carrying out multi-mode interactive fusion analysis on the optical blood glucose screening characteristics and the electrochemical blood glucose screening characteristics to determine blood glucose fusion characteristics.
In one possible embodiment, each subsequence of the optical blood glucose signal detection value subsequence set and the electrochemical blood glucose signal detection value subsequence set is subjected to concentrated iterative screening to remove noise data or abnormal values, screening values which can most represent the conditions of the subsequence detection values in each subsequence are reserved, and further, the screened screening values are rearranged according to a time sequence to generate an optical signal screening value sequence and an electrochemical signal screening value sequence, so that the integrity of data and the consistency of a time axis are ensured.
In order to improve the blood glucose signal processing efficiency, the blood glucose signal extraction network layer is used for carrying out feature extraction on the optical blood glucose signal screening value sequence and the electrochemical blood glucose signal screening value sequence, and the network layer can capture the time dependence, the nonlinear relation and the deep dynamic characteristics of the signals, so that the optical blood glucose screening features and the electrochemical blood glucose screening features are generated. The optical blood sugar screening characteristics reflect blood sugar characteristics of a target user after the optical blood sugar sensor is used for detecting blood sugar of the target user, wherein the blood sugar characteristics of the target user comprise blood sugar concentration characteristics, blood sugar change rate characteristics, blood sugar fluctuation amplitude characteristics, blood sugar trend characteristics and the like in a preset detection window. The electrochemical blood sugar screening characteristics reflect blood sugar characteristics of a target user in a preset detection window after blood sugar detection of the target user by using an electrochemical blood sugar sensor, wherein the blood sugar characteristics comprise blood sugar concentration characteristics, blood sugar change rate characteristics, blood sugar fluctuation amplitude characteristics, blood sugar trend characteristics and the like
Furthermore, by combining the continuity of the optical signal and the accuracy of the electrochemical signal, that is, performing multi-mode interactive fusion analysis on the optical blood glucose screening feature and the electrochemical blood glucose screening feature, the blood glucose fusion feature which is more representative and can comprehensively reflect the blood glucose variation dynamics of the user is obtained. Therefore, the complete process from data screening, feature extraction and fusion analysis is completed, and the technical effects of improving the quality and reliability of blood glucose fusion features are achieved.
Further, the sub-sequence set of optical blood glucose signal detection values and the sub-sequence set of electrochemical blood glucose signal detection values are subjected to in-sequence signal detection value concentrated iterative screening respectively, and screening results are respectively ordered according to the sequence of detection time to obtain an optical blood glucose signal screening value sequence and an electrochemical blood glucose signal screening value sequence, and step S3 of the embodiment of the application further comprises:
Randomly extracting a first optical blood glucose signal detection value subsequence from the optical blood glucose signal detection value subsequence set;
Calculating the average value of the first optical blood glucose signal detection value subsequence to obtain a first optical blood glucose signal detection average value;
Taking the first optical blood glucose signal detection mean value as an initial screening value, and constructing an initial neighborhood in the first optical blood glucose signal detection value subsequence according to a preset centralized screening step;
calculating the average value of the first optical blood glucose signal detection values in the initial neighborhood to obtain an initial neighborhood average value;
Performing iterative screening authentication based on the difference value between the initial neighborhood mean value and the initial screening value, and randomly extracting a first optical blood glucose signal detection value from the neighborhood edge of the initial neighborhood as an iterative screening value if the iterative screening authentication result is passed;
iterating in the first optical blood glucose signal detection value subsequence based on the iteration screening value until the iteration screening authentication result is not passed, and taking a neighborhood mean value of an iteration neighborhood corresponding to the last iteration as a first optical signal screening value;
And by analogy, traversing the optical blood glucose signal detection value subsequence set to perform concentrated screening of signal detection values in the subsequence to obtain Q optical blood glucose signal screening values, and arranging the detection times of the optical blood glucose signal detection value subsequences corresponding to the Q optical blood glucose signal screening values in sequence from front to back to obtain the optical blood glucose signal screening value sequence, wherein Q is the number of the optical blood glucose signal detection value subsequences in the optical blood glucose signal detection value subsequence set;
and carrying out concentrated screening on the signal detection values in the subsequence of the electrochemical blood glucose signal detection value subsequence to obtain the electrochemical blood glucose signal screening value sequence.
Further, performing iterative screening authentication based on the difference value between the initial neighborhood mean value and the initial screening value, and if the iterative screening authentication result is passed, randomly extracting a first optical blood glucose signal detection value from the neighborhood edge of the initial neighborhood as an iterative screening value, wherein step S3 in the embodiment of the application further comprises:
Calculating the difference value between the initial neighborhood mean value and the initial screening value, judging whether the difference value is larger than or equal to a preset difference value, and if so, iteratively screening an authentication result to pass;
if not, the iterative filtering authentication result is not passed, and the initial neighborhood average value is used as the first optical signal filtering value.
In one possible embodiment, one optical blood glucose signal detection value sub-sequence is randomly extracted from the set of optical blood glucose signal detection value sub-sequences as the first optical blood glucose signal detection value sub-sequence. And (3) carrying out concentrated iterative screening on the signal detection values in the first optical signal detection value subsequence, removing noise and abnormal points, and screening out a representative first optical blood glucose signal detection value subsequence. And traversing and calculating the average value of the detection values in the first optical blood glucose signal detection value subsequence to obtain a first optical blood glucose signal detection average value.
Furthermore, the first optical blood glucose signal detection mean value is used as an initial screening value for constructing an initial neighborhood. Preferably, the first optical blood glucose signal detection values in the preset centralized screening step, which are obtained by adding the difference value between the first optical blood glucose signal detection value subsequence and the first optical blood glucose signal detection mean value, into an initial empty set, are obtained. Wherein, the preset screening step is the preset amplitude of the concentrated iterative screening which is performed once by a person skilled in the art.
And further, calculating the average value of the first optical blood glucose signal detection values in the initial neighborhood to obtain an initial neighborhood average value. Wherein the initial neighborhood average reflects an average level of optical glucose signal detection values within the initial neighborhood.
Preferably, the difference value between the initial neighborhood mean value and the initial screening value is calculated, the difference value reflects the degree of dispersion of the data in the initial neighborhood, and the larger the difference value is, the more dispersion of the data in the initial neighborhood is indicated, and the data capable of reflecting the common condition of the first optical blood glucose signal detection value subsequence is more likely not to be in the initial neighborhood. Further judging whether the difference value is larger than or equal to a preset difference value (the minimum difference value for carrying out centralized iterative screening is preset by a person skilled in the art), if so, judging that the iterative screening authentication result is passed, indicating that the centralized iterative screening is needed to be carried out at the moment, and if not, judging that the iterative screening authentication result is not passed, indicating that the centralized iterative screening is needed to be stopped at the moment, wherein the data distribution in an initial neighborhood is centralized and has higher representativeness, and taking the initial neighborhood mean value as the first optical signal screening value.
Preferably, if the iterative filtering authentication result is passing, randomly extracting a first optical blood glucose signal detection value from the neighborhood edge of the initial neighborhood as an iterative filtering value. And further, using the iteration screening value as a new neighborhood center to construct an iteration screening neighborhood. And further, based on the same principle of iterative screening authentication according to the difference value between the initial neighborhood mean value and the initial screening value, calculating whether the difference value between the iterative screening neighborhood mean value and the iterative screening value is larger than or equal to a preset difference value, if so, randomly extracting a first optical blood glucose signal detection value from the edge of the iterative screening neighborhood to update the iterative screening value, and then continuing iteration in the first optical blood glucose signal detection value subsequence according to the updated iterative screening value until the iterative screening authentication result is not passed, stopping iteration, and taking the neighborhood mean value of the iterative neighborhood corresponding to the last iteration as the first optical signal screening value.
And similarly, based on the same principle, carrying out in-subsequence signal detection value centralized screening on each optical blood glucose signal detection value subsequence set in the optical blood glucose signal detection value subsequence set to obtain Q optical blood glucose signal screening values, and arranging detection times of the optical blood glucose signal detection value subsequences corresponding to the Q optical blood glucose signal screening values in sequence from front to back to obtain the optical blood glucose signal screening value sequence, wherein Q is the number of the optical blood glucose signal detection value subsequences in the optical blood glucose signal detection value subsequence set. And further, based on the same principle as the optical blood glucose signal screening value sequence, performing in-subsequence signal detection value centralized screening on the electrochemical blood glucose signal detection value subsequence to obtain the electrochemical blood glucose signal screening value sequence.
Through centralized iterative screening, noise or abnormal points in optical and electrochemical signals can be effectively removed, quality and representativeness of screening values are ensured, the refinement degree of the screening process is controlled through a neighborhood mean value and a screening step, the screened signal sequence is smoother, and errors caused by data fluctuation are reduced. Furthermore, the rationality of the signal detection value is verified through gradual iteration in the screening and authentication process, the representative value with statistical significance can be stably extracted, and the reliability of the signal analysis result is improved. The technical effect of laying a foundation for the accuracy of blood glucose signal processing is achieved.
Further, step S3 of the embodiment of the present application further includes:
Acquiring a plurality of sample optical blood glucose signal screening value sequences and a plurality of corresponding sample optical blood glucose screening characteristics as optical training data, performing supervision training on a framework constructed based on a convolutional neural network by utilizing the optical training data, and learning a one-to-one mapping relation between the optical blood glucose signal screening value sequences and the sample optical blood glucose screening characteristics in the training until the training converges, so as to acquire an optical blood glucose signal extraction branch after the training is completed;
Acquiring a plurality of sample electrochemical blood glucose signal screening value sequences and a plurality of corresponding sample electrochemical blood glucose screening characteristics as electrochemical training data, performing supervision training on a framework constructed based on a convolutional neural network by utilizing the optical training data, and learning a one-to-one mapping relation between the sample electrochemical blood glucose signal screening value sequences and the sample electrochemical blood glucose screening characteristics in the training until the training converges, so as to obtain an electrochemical blood glucose signal extraction branch after the training is completed;
And connecting the optical blood glucose signal extraction branch and the electrochemical blood glucose signal extraction branch in parallel to obtain the blood glucose signal extraction network layer.
In one possible embodiment, the plurality of sample optical glucose signal screening value sequences are sequences extracted from a plurality of sample optical glucose signal detection value subsequences after a screening process, and are used for training and testing a neural network model. The plurality of sample optical blood glucose screening features refers to feature data corresponding to a sample optical blood glucose signal screening value sequence. And taking the plurality of sample optical blood glucose signal screening value sequences and the corresponding plurality of sample optical blood glucose screening characteristics as optical training data.
Further, model construction was performed using a Convolutional Neural Network (CNN) framework. CNNs automatically learn and extract spatial and temporal features in data through multi-layer structures such as convolutional layers, pooling layers, fully connected layers, and the like. The input to the network is a sequence of optical glucose signal screening values and the output is a glucose screening feature corresponding to each sequence. Preferably, the plurality of sample optical glucose screening features are labeled. In the training process, the network performs training by calculating the difference (i.e. error) between the network output and the real label according to the input optical blood glucose signal screening value sequence (training data) and the corresponding optical blood glucose screening characteristics (label data). The error is back-propagated to various levels in the network by a back-propagation algorithm, and network parameters (e.g., convolution kernel weights, etc.) are adjusted to minimize the error. In each iteration, the convolutional neural network updates the model parameters through an optimization algorithm (such as gradient descent) to gradually reduce the prediction error. After each training round, the network can gradually adjust the weight and the bias of the network, so that the mapping relation between the input optical blood glucose signal screening value sequence and the output optical blood glucose screening characteristics is more and more accurate. After multiple rounds of training, the error of the model tends to be stable, reaches a minimum value or a convergence state, and indicates that the network has learned the optimal mapping relation from the optical blood glucose signal screening value sequence to the blood glucose screening characteristics. After training is completed and convergence is achieved, the obtained model is the optical blood glucose signal extraction branch after training is completed. The optical blood sugar signal extraction branch can automatically extract key features in the optical blood sugar signal sequence and map the key features with blood sugar screening features in a one-to-one correspondence manner.
Based on the same construction principle as the optical blood glucose signal extraction branch, taking a plurality of sample electrochemical blood glucose signal screening value sequences and a plurality of corresponding sample electrochemical blood glucose screening characteristics as electrochemical training data, performing supervised training on a framework constructed based on a convolutional neural network by utilizing the optical training data, and learning a one-to-one mapping relation between the electrochemical blood glucose signal screening value sequences and the sample electrochemical blood glucose screening characteristics in the training until the training converges, so as to obtain the trained electrochemical blood glucose signal extraction branch. The electrochemical blood sugar signal extraction branch can automatically extract key features in an electrochemical blood sugar signal sequence and map the key features with blood sugar screening features in a one-to-one correspondence manner.
Furthermore, the optical blood glucose signal extraction branch after training is combined with the electrochemical blood glucose signal extraction branch in parallel to form a multi-input network layer (blood glucose signal extraction network layer), so that two different types of signals can be processed simultaneously, and the blood glucose related characteristics of the signals can be extracted.
Further, performing multi-mode interactive fusion analysis on the optical blood glucose screening feature and the electrochemical blood glucose screening feature to determine blood glucose fusion features, wherein step S3 of the embodiment of the application further comprises:
calculating the similarity of the optical blood glucose screening characteristics and the electrochemical blood glucose screening characteristics to obtain a screening characteristic similarity set;
Normalizing the screening feature similarity set, and adding a processing result into an initial empty matrix to obtain a multi-mode interaction fusion matrix;
Performing convolution calculation on the optical blood glucose screening characteristics and the multi-mode interaction fusion matrix to obtain optical blood glucose fusion characteristics;
performing convolution calculation on the electrochemical blood glucose screening characteristics and the multi-mode interaction fusion matrix to obtain electrochemical blood glucose fusion characteristics;
and obtaining the average value of the optical blood glucose fusion characteristic and the electrochemical blood glucose fusion characteristic to obtain the blood glucose fusion characteristic.
Further, step S3 of the embodiment of the present application further includes:
Obtaining a normalization function, wherein the normalization function is as follows:
Wherein Mix [ lim (x i,yi) ] is a normalization value corresponding to the ith screening feature similarity in the screening feature similarity set, e is the bottom of natural logarithm, n is the total number of screening feature similarities in the screening feature similarity set, and lim (x i,yi) is the ith screening feature similarity in the screening feature similarity set;
And carrying out normalization processing on the screening feature similarity set by using the normalization function to obtain the processing result.
In one possible embodiment, the optical glucose screening feature and the electrochemical glucose screening feature are mutually enhanced for a multimodal interactive fusion analysis of the optical glucose screening feature and the electrochemical glucose screening feature. Firstly, calculating the similarity of the corresponding features in the optical blood glucose screening feature and the electrochemical blood glucose screening feature by using a cosine calculation formula to obtain a screening feature similarity set. Wherein the set of screening feature similarities reflects a degree of similarity between the optical blood glucose screening feature and the electrochemical blood glucose screening feature.
And further, carrying out normalization processing on the screening feature similarity set by using a normalization function, eliminating dimension differences, and adding a processing result into an initial empty matrix to obtain the multi-mode interaction fusion matrix. The multimodal interaction fusion matrix reflects an interaction relationship between the optical blood glucose screening feature and the multimodal interaction fusion matrix.
And performing convolution operation on the optical blood glucose screening characteristics and the multi-mode interaction fusion matrix, extracting high-order interaction characteristics between the optical blood glucose screening characteristics and the multi-mode interaction fusion matrix, and generating the optical blood glucose fusion characteristics. Preferably, a sample optical blood glucose screening feature set, a sample multi-mode interaction fusion matrix and a sample optical blood glucose fusion feature are obtained to serve as training data, and a framework constructed based on a convolutional neural network is subjected to supervised training by utilizing the training data until training converges, so that an optical convolutional network layer with the training completed is obtained. And carrying out convolution analysis on the optical blood glucose screening characteristics and the multi-mode interaction fusion matrix by using the optical convolution network layer to obtain the optical blood glucose fusion characteristics.
And carrying out convolution operation on the electrochemical blood glucose screening characteristics and the multi-mode interaction fusion matrix, extracting the interaction characteristics of the electrochemical blood glucose screening characteristics and the multi-mode interaction fusion matrix, and generating the electrochemical blood glucose fusion characteristics. Preferably, a sample electrochemical blood glucose screening feature set, a sample multi-mode interaction fusion matrix and sample electrochemical blood glucose fusion features are obtained to serve as training data, and a framework constructed based on a convolutional neural network is subjected to supervised training by utilizing the training data until training converges, so that an electrochemical convolutional network layer with complete training is obtained. And carrying out convolution analysis on the electrochemical blood glucose screening characteristics and the multi-mode interaction fusion matrix by using the electrochemical convolution network layer to obtain the electrochemical blood glucose fusion characteristics. Through convolution calculation, deep interactive relations are extracted from input features and interactive matrixes, and more accurate and comprehensive feature expression is provided for blood glucose detection and prediction.
Further, the step S4 of the embodiment of the present application further includes invoking the historical blood glucose feature set of the target user to authenticate the blood glucose fusion feature, obtaining an authentication result, and if the authentication result is passed, taking the blood glucose fusion feature as a blood glucose signal processing result:
respectively carrying out similarity calculation on the blood glucose fusion characteristics and the historical blood glucose characteristic set to obtain a historical similarity set;
And authenticating the historical similarity set by using a preset historical similarity threshold, and enabling the authentication result to pass when the number of the historical similarity sets smaller than the preset historical similarity threshold is smaller than the ratio of the total number of the historical similarities in the historical similarity set to be smaller than a preset ratio.
In a possible embodiment, the historical blood glucose feature set is accumulated for blood glucose detection data of past target users, and includes blood glucose features at a plurality of historical time points. Each historical blood glucose feature consists of multiple dimensions of data representing the user's blood glucose state at a certain time. The blood glucose fusion feature is a processing result of optical and electrochemical blood glucose signal fusion and represents the current blood glucose state of the target user. The feature set is also typically multi-dimensional data.
And respectively calculating the distance between the blood glucose fusion feature and each historical blood glucose feature in the historical blood glucose feature set by using Euclidean distance, wherein the smaller the distance is, the higher the similarity is. And taking the reciprocal of the calculation result as a historical similarity set. The historical similarity set reflects the similarity degree of the blood glucose fusion characteristic and the historical blood glucose condition of the target user.
Further, a predetermined historical similarity threshold (e.g., 0.6) is set by those skilled in the art, which is used to determine whether the similarity of the current blood glucose fusion profile to the historical blood glucose profile is sufficiently high. A similarity above the threshold value indicates that the current blood glucose state is very similar to the historical blood glucose state, and a lower threshold value indicates that a large change in blood glucose state has occurred, which may be a blood glucose signal processing error.
And comparing the number of the history similarity sets smaller than the preset history similarity threshold value with the ratio of the total number of the history similarity sets smaller than the preset ratio, and enabling the authentication result to pass. Preferably, the ratio calculation formula may be that the ratio=the number of similarities in the historical similarity set smaller than a preset historical similarity threshold value/the total number of the historical similarity sets. The ratio represents the degree of matching of the current glycemic fusion feature to the historical glycemic features.
If the ratio is smaller than the preset ratio (for example, 0.6), the authentication result is that the current blood glucose fusion characteristic and the historical blood glucose characteristic are matched with each other to a high enough degree, and the processing requirements are met. If the ratio is greater than the preset ratio, the authentication is not passed, which means that the current blood glucose fusion characteristic and the historical blood glucose characteristic have large differences and need further processing.
In summary, the embodiment of the application has at least the following technical effects:
1. The application utilizes the dual detection of the optical blood glucose sensor and the electrochemical blood glucose sensor, can complement each other, reduces the error possibly caused by a single sensor, and can better extract the useful information in the multi-source data by carrying out multi-mode interactive fusion analysis, thereby achieving the technical effect of further improving the accuracy of blood glucose signal processing.
2. According to the application, the historical blood glucose feature set of the target user is invoked to authenticate the blood glucose fusion feature, so that the consistency of the current blood glucose signal and the historical data is ensured. The process can effectively avoid error analysis caused by sudden events or abnormal fluctuation, ensure accurate reflection of blood glucose signals and ensure more reliable blood glucose processing results.
It should be noted that the sequence of the embodiments of the present application is only for description, and does not represent the advantages and disadvantages of the embodiments. And the foregoing description has been directed to specific embodiments of this specification. In addition, the processes depicted in the accompanying figures do not necessarily require the particular order shown, or sequential order, to achieve desirable results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.
The foregoing description of the preferred embodiments of the application is not intended to limit the application to the precise form disclosed, and any such modifications, equivalents, and alternatives falling within the spirit and scope of the application are intended to be included within the scope of the application.
The specification and figures are merely exemplary illustrations of the present application and are considered to cover any and all modifications, variations, combinations, or equivalents that fall within the scope of the application. It will be apparent to those skilled in the art that various modifications and variations can be made to the present application without departing from the scope of the application. Thus, the present application is intended to include such modifications and alterations insofar as they come within the scope of the application or the equivalents thereof.
Claims (8)
1. A method of processing a blood glucose signal for blood glucose testing, the method comprising:
the method comprises the steps that synchronous and continuous detection is carried out on blood sugar of a target user in a prediction detection window by using an optical blood sugar sensor and an electrochemical blood sugar sensor, and an optical blood sugar signal detection value sequence and an electrochemical blood sugar signal detection value sequence are obtained;
Dividing the optical blood glucose signal detection value sequence and the electrochemical blood glucose signal detection value sequence according to a preset dividing scale to obtain an optical blood glucose signal detection value subsequence set and an electrochemical blood glucose signal detection value subsequence set;
performing multi-mode interactive fusion analysis on the optical blood glucose signal detection value subsequence set and the electrochemical blood glucose signal detection value subsequence set to obtain blood glucose fusion characteristics;
And calling the historical blood glucose feature set of the target user to authenticate the blood glucose fusion feature, obtaining an authentication result, and taking the blood glucose fusion feature as a blood glucose signal processing result if the authentication result is passed.
2. The method for processing blood glucose signals for blood glucose testing according to claim 1, wherein performing a multi-modal interactive fusion analysis on the optical blood glucose signal detection value subsequence set and the electrochemical blood glucose signal detection value subsequence set to obtain blood glucose fusion features comprises:
The optical blood glucose signal detection value subsequence set and the electrochemical blood glucose signal detection value subsequence set are subjected to in-subsequence signal detection value centralized iterative screening respectively, and screening results are respectively ordered according to the sequence of detection time to obtain an optical blood glucose signal screening value sequence and an electrochemical blood glucose signal screening value sequence;
Extracting blood glucose signal characteristics from the optical blood glucose signal screening value sequence and the electrochemical blood glucose signal screening value sequence by utilizing a blood glucose signal extraction network layer to obtain optical blood glucose screening characteristics and electrochemical blood glucose screening characteristics;
And carrying out multi-mode interactive fusion analysis on the optical blood glucose screening characteristics and the electrochemical blood glucose screening characteristics to determine blood glucose fusion characteristics.
3. The method for processing blood glucose signals for blood glucose testing according to claim 2, wherein the steps of performing in-subsequence iterative screening of signal detection values in the subsequence set and the electrochemical blood glucose signal detection value subsequence set, respectively, and sorting the screening results according to the time sequence of the testing, respectively, to obtain an optical blood glucose signal screening value sequence and an electrochemical blood glucose signal screening value sequence, comprise:
Randomly extracting a first optical blood glucose signal detection value subsequence from the optical blood glucose signal detection value subsequence set;
Calculating the average value of the first optical blood glucose signal detection value subsequence to obtain a first optical blood glucose signal detection average value;
Taking the first optical blood glucose signal detection mean value as an initial screening value, and constructing an initial neighborhood in the first optical blood glucose signal detection value subsequence according to a preset centralized screening step;
calculating the average value of the first optical blood glucose signal detection values in the initial neighborhood to obtain an initial neighborhood average value;
Performing iterative screening authentication based on the difference value between the initial neighborhood mean value and the initial screening value, and randomly extracting a first optical blood glucose signal detection value from the neighborhood edge of the initial neighborhood as an iterative screening value if the iterative screening authentication result is passed;
iterating in the first optical blood glucose signal detection value subsequence based on the iteration screening value until the iteration screening authentication result is not passed, and taking a neighborhood mean value of an iteration neighborhood corresponding to the last iteration as a first optical signal screening value;
And by analogy, traversing the optical blood glucose signal detection value subsequence set to perform concentrated screening of signal detection values in the subsequence to obtain Q optical blood glucose signal screening values, and arranging the detection times of the optical blood glucose signal detection value subsequences corresponding to the Q optical blood glucose signal screening values in sequence from front to back to obtain the optical blood glucose signal screening value sequence, wherein Q is the number of the optical blood glucose signal detection value subsequences in the optical blood glucose signal detection value subsequence set;
and carrying out concentrated screening on the signal detection values in the subsequence of the electrochemical blood glucose signal detection value subsequence to obtain the electrochemical blood glucose signal screening value sequence.
4. The method for processing blood glucose signals for blood glucose detection of claim 3, wherein performing iterative screening authentication based on a difference between the initial neighborhood mean value and the initial screening value, and randomly extracting a first optical blood glucose signal detection value from a neighborhood edge of the initial neighborhood as an iterative screening value if the iterative screening authentication result is passed, comprises:
Calculating the difference value between the initial neighborhood mean value and the initial screening value, judging whether the difference value is larger than or equal to a preset difference value, and if so, iteratively screening an authentication result to pass;
if not, the iterative filtering authentication result is not passed, and the initial neighborhood average value is used as the first optical signal filtering value.
5. A blood glucose signal processing method for blood glucose testing as recited in claim 2, comprising:
Acquiring a plurality of sample optical blood glucose signal screening value sequences and a plurality of corresponding sample optical blood glucose screening characteristics as optical training data, performing supervision training on a framework constructed based on a convolutional neural network by utilizing the optical training data, and learning a one-to-one mapping relation between the optical blood glucose signal screening value sequences and the sample optical blood glucose screening characteristics in the training until the training converges, so as to acquire an optical blood glucose signal extraction branch after the training is completed;
Acquiring a plurality of sample electrochemical blood glucose signal screening value sequences and a plurality of corresponding sample electrochemical blood glucose screening characteristics as electrochemical training data, performing supervision training on a framework constructed based on a convolutional neural network by utilizing the optical training data, and learning a one-to-one mapping relation between the sample electrochemical blood glucose signal screening value sequences and the sample electrochemical blood glucose screening characteristics in the training until the training converges, so as to obtain an electrochemical blood glucose signal extraction branch after the training is completed;
And connecting the optical blood glucose signal extraction branch and the electrochemical blood glucose signal extraction branch in parallel to obtain the blood glucose signal extraction network layer.
6. The method of claim 2, wherein performing a multimodal interactive fusion analysis of the optical glucose screening feature and the electrochemical glucose screening feature to determine a glucose fusion feature comprises:
calculating the similarity of the optical blood glucose screening characteristics and the electrochemical blood glucose screening characteristics to obtain a screening characteristic similarity set;
Normalizing the screening feature similarity set, and adding a processing result into an initial empty matrix to obtain a multi-mode interaction fusion matrix;
Performing convolution calculation on the optical blood glucose screening characteristics and the multi-mode interaction fusion matrix to obtain optical blood glucose fusion characteristics;
performing convolution calculation on the electrochemical blood glucose screening characteristics and the multi-mode interaction fusion matrix to obtain electrochemical blood glucose fusion characteristics;
and obtaining the average value of the optical blood glucose fusion characteristic and the electrochemical blood glucose fusion characteristic to obtain the blood glucose fusion characteristic.
7. A blood glucose signal processing method for blood glucose testing as recited in claim 6, comprising:
Obtaining a normalization function, wherein the normalization function is as follows:
Wherein Mix [ lim (x i,yi) ] is a normalization value corresponding to the ith screening feature similarity in the screening feature similarity set, e is the bottom of natural logarithm, n is the total number of screening feature similarities in the screening feature similarity set, and lim (x i,yi) is the ith screening feature similarity in the screening feature similarity set;
And carrying out normalization processing on the screening feature similarity set by using the normalization function to obtain the processing result.
8. The method for processing a blood glucose signal for blood glucose detection of claim 1, wherein invoking the historical blood glucose feature set of the target user to authenticate the blood glucose fusion feature to obtain an authentication result, and if the authentication result is passed, taking the blood glucose fusion feature as the blood glucose signal processing result, comprising:
respectively carrying out similarity calculation on the blood glucose fusion characteristics and the historical blood glucose characteristic set to obtain a historical similarity set;
And authenticating the historical similarity set by using a preset historical similarity threshold, and enabling the authentication result to pass when the number of the historical similarity sets smaller than the preset historical similarity threshold is smaller than the ratio of the total number of the historical similarities in the historical similarity set to be smaller than a preset ratio.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202510131765.7A CN119970018B (en) | 2025-02-06 | 2025-02-06 | Blood sugar signal processing method for blood sugar detection |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202510131765.7A CN119970018B (en) | 2025-02-06 | 2025-02-06 | Blood sugar signal processing method for blood sugar detection |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN119970018A true CN119970018A (en) | 2025-05-13 |
| CN119970018B CN119970018B (en) | 2025-08-15 |
Family
ID=95642679
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202510131765.7A Active CN119970018B (en) | 2025-02-06 | 2025-02-06 | Blood sugar signal processing method for blood sugar detection |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN119970018B (en) |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20160029966A1 (en) * | 2014-07-31 | 2016-02-04 | Sano Intelligence, Inc. | Method and system for processing and analyzing analyte sensor signals |
| CN111631730A (en) * | 2020-05-15 | 2020-09-08 | 中国科学院深圳先进技术研究院 | Hypoglycemia warning method and system based on fusion of sensor data and physiological information |
| CN118844956A (en) * | 2024-06-26 | 2024-10-29 | 清华大学 | Human multi-parameter integrated detection device and wearable device containing the same |
-
2025
- 2025-02-06 CN CN202510131765.7A patent/CN119970018B/en active Active
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20160029966A1 (en) * | 2014-07-31 | 2016-02-04 | Sano Intelligence, Inc. | Method and system for processing and analyzing analyte sensor signals |
| CN111631730A (en) * | 2020-05-15 | 2020-09-08 | 中国科学院深圳先进技术研究院 | Hypoglycemia warning method and system based on fusion of sensor data and physiological information |
| CN118844956A (en) * | 2024-06-26 | 2024-10-29 | 清华大学 | Human multi-parameter integrated detection device and wearable device containing the same |
Also Published As
| Publication number | Publication date |
|---|---|
| CN119970018B (en) | 2025-08-15 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US20250104813A1 (en) | Genome-wide prediction method based on deep learning by using genome-wide data and bioinformatics features | |
| CN113030001B (en) | Fruit sugar degree detection method and system | |
| CN107137093B (en) | A kind of continuous blood sugar monitoring device comprising abnormal plasma glucose probability alarm | |
| CN111128327A (en) | Hypoglycemia early warning method and device | |
| FR3109455A1 (en) | Determination of health states of systems equipped with sensors | |
| CN119541840A (en) | Intelligent remote follow-up method and system for tumor patients | |
| CN115482932A (en) | Multivariate blood glucose prediction algorithm based on transfer learning and glycosylated hemoglobin | |
| CN119361171B (en) | Blood glucose concentration prediction method combined with neural network | |
| CN119073917A (en) | Consciousness disorder course monitoring system | |
| CN120067980A (en) | Multi-source big data intelligent analysis fusion processing method based on foundation detection equipment | |
| CN119400431B (en) | Epileptic seizure prediction method based on lightweight depth statistical model | |
| CN120280154A (en) | Diabetes mellitus and nephropathy risk assessment prediction system based on platelet index | |
| CN118913321B (en) | A fiber optic gyroscope fault diagnosis method based on MiniRocket network model | |
| Lyu et al. | Classification of coronary artery disease using radial artery pulse wave analysis via machine learning | |
| CN118709004B (en) | A user data efficient storage method and system for security assessment system | |
| CN118557187B (en) | A blood glucose concentration prediction method based on ELM-BiLSTM-attention | |
| CN119517428A (en) | A method for establishing a pSS prediction model based on multiple clinical test indicators and a pSS prediction system | |
| CN119970018A (en) | A blood sugar signal processing method for blood sugar detection | |
| KR102472848B1 (en) | METHOD AND APPARATUS FOR MEASURING NON-INVASIVE HbA1c USING LEARNING ALGORITHM | |
| CN114548259B (en) | A PISA fault identification method based on semi-supervised Semi-KNN model | |
| CN119155335B (en) | Intelligent well lid data analysis method and device based on Internet of things and storage medium | |
| CN119961849B (en) | Time series anomaly detection method, device, equipment and medium based on large model | |
| CN120087482B (en) | Method and system for analyzing and processing chronic disease quality control data based on big data | |
| CN118015430B (en) | Intelligent government affair service training method and system based on deep learning | |
| Li et al. | Diabetes prediction and analysis using machine learning models |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |