Machine learning based medical monitoring sensor data correlation storage method and system

Abstract

The application provides a medical monitoring sensor data correlation storage method and system based on machine learning, relating to the field of data processing. By utilizing machine learning technology, medical monitoring data from different sensors are automatically identified and correlatedly stored. By constructing a data correlation model, the connection between data can be accurately found, providing a comprehensive data analysis perspective for medical workers. The application improves the utilization efficiency and value of medical monitoring data, and helps to discover potential medical problems and risks.

Description

Technical Field

[0001] This application relates to the field of data processing, and more specifically, to a method and system for associating and storing medical monitoring sensor data based on machine learning. Background Technology

[0002] With the widespread adoption of wearable devices, collecting and analyzing user sensor data has become an important means of improving the quality of medical services and providing personalized treatment. User sensor data typically includes various types of sensor signals, such as heart rate, blood pressure, steps, blood oxygen, and HRV. This data is crucial for understanding a patient's health status, predicting disease risk, and developing personalized treatment plans.

[0003] In practical applications, effectively extracting key data related to specific health events or symptoms from massive and complex sensor data remains a significant challenge. Traditional data analysis methods often treat all data in a one-size-fits-all manner, ignoring the differences between various types of sensor signals and their varying contributions to health events. This can not only lead to the omission of crucial information but also obscure potential patterns and trends within the data due to homogenized processing. Summary of the Invention

[0004] The purpose of this application is to provide a method and system for associating and storing medical monitoring sensor data based on machine learning.

[0005] This application is implemented as follows:

[0006] In a first aspect, this application provides a method for associating and storing medical monitoring sensor data based on machine learning, the method comprising:

[0007] Acquire target sensor signals related to the sensor signals to be processed, wherein the target sensor signals cover different categories of sensor signals;

[0008] Based on the influence factors of the different categories of sensor signals, the correlation between the different categories of sensor signals covered by the target sensor signal and the sensor signal to be processed is obtained;

[0009] Based on the correlation between the different categories of sensor signals and the sensor signal to be processed, the selected category of sensor signal is obtained from the different categories of sensor signals;

[0010] The signal segments from the selected category of sensor signals that are related to the sensor signal to be processed are associated and stored.

[0011] The step of associating and storing signal segments from the selected category of sensor signals that are related to the sensor signal to be processed includes:

[0012] Multiple signal segments related to the sensor signal to be processed are retrieved from the selected category of sensor signals;

[0013] The continuity evaluation parameters between the multiple signal segments are determined by the signal coordinates of the multiple signal segments in the selected category of sensor signals;

[0014] When the continuity evaluation parameters among the multiple signal segments reach a preset standard, the signal segments in the selected category of sensor signals that are related to the sensor signal to be processed are associated and stored.

[0015] Optionally, obtaining the correlation between the different categories of sensor signals covered by the target sensor signal and the sensor signal to be processed, based on the influence factors of the different categories of sensor signals, includes:

[0016] The sensor signal to be processed is segmented into signal segments to obtain the signal segments to be processed;

[0017] Based on the signal segment to be processed, obtain the basic correlation between the different types of sensor signals and the sensor signal to be processed;

[0018] Based on the basic correlation between the different types of sensor signals and the sensor signal to be processed, and the influence factors of the different types of sensor signals, the correlation between the different types of sensor signals and the sensor signal to be processed is obtained.

[0019] Optionally, obtaining the basic correlation between the different categories of sensor signals and the sensor signal to be processed based on the signal segment to be processed includes:

[0020] The target signal segment related to the signal segment to be processed is retrieved from the different categories of sensor signals;

[0021] Based on the priority of the signal segment to be processed and the priority of the target signal segments related to the signal segment to be processed among the different categories of sensor signals, the basic correlation between the different categories of sensor signals and the sensor signal to be processed is determined.

[0022] Optionally, the target sensor signals can be multiple, and the method further includes:

[0023] Based on the correlation mechanism between the sensor signal to be processed and each target sensor signal, the influence factor of each target sensor signal is determined;

[0024] Based on the influence factor of each target sensor signal and the correlation between the different categories of sensor signals covered by each target sensor signal and the sensor signal to be processed, the correlation between each target sensor signal and the sensor signal to be processed is obtained;

[0025] Based on the correlation between each target sensor signal and the sensor signal to be processed, the multiple target sensor signals are sorted.

[0026] The acquisition of the target sensor signal related to the sensor signal to be processed includes:

[0027] The sensor signal to be processed is segmented into signal segments to obtain a set of signal segments to be processed;

[0028] The signal segments in the set of signal segments to be processed are aggregated to obtain multiple signal segment clusters;

[0029] The target sensor signals associated with the multiple signal segment clusters are acquired respectively, and each signal segment cluster corresponds to a related mechanism.

[0030] Optionally, the step of filtering sensor signals of a selected category from the different categories of sensor signals based on the correlation between the different categories of sensor signals and the sensor signal to be processed includes:

[0031] Sensor signals of the target category are selected from the different categories of sensor signals, as well as sensor signals of unselected categories other than the target category;

[0032] Based on the correlation between the different categories of sensor signals and the sensor signals to be processed, sensor signals of unselected categories are filtered out from the sensor signals of unselected categories;

[0033] The sensor signal of the undetermined category and the sensor signal of the target category are used as the sensor signal of the selected category;

[0034] The method of associating and storing signal segments from the selected category of sensor signals that are related to the sensor signal to be processed also includes:

[0035] The target category signal segment is obtained by filtering the set of signal segments to be processed, and the set of signal segments to be processed includes signal segments obtained by segmenting the sensor signal to be processed.

[0036] The signal segments of the selected category of sensor signals that are related to the signal segment of the target category are associated and stored.

[0037] Optionally, the step of filtering sensor signals of a pending category from the unselected sensor signals based on the correlation between the different categories of sensor signals and the sensor signal to be processed includes:

[0038] Based on the correlation between the different categories of sensor signals and the sensor signals to be processed, sensor signals of the original undetermined category with a correlation greater than or equal to a preset threshold are selected from the sensor signals of the unselected category.

[0039] Among the sensor signals of the original undetermined category with a correlation greater than or equal to a preset threshold, the top N sensor signals with the highest correlation are taken as the sensor signals of the undetermined category, where N is a positive integer greater than or equal to 1.

[0040] Optionally, the step of associating and storing signal segments from the selected category of sensor signals that are related to the sensor signal to be processed further includes:

[0041] Identify unselected category signal segments in the set of signal segments to be processed, other than the target category signal segments;

[0042] The signal segments in the selected category of sensor signals that are related to the signal segments of the unselected category are selected as the candidate signal segment set;

[0043] M signal segments are selected from the candidate signal segment set and stored in association, where M is a positive integer greater than or equal to 1.

[0044] Optionally, retrieving multiple signal segments related to the sensor signal to be processed from the selected category of sensor signals includes:

[0045] In the sensor signal to be processed, obtain the first signal segment and the second signal segment whose signal coordinates are adjacent;

[0046] A third signal segment related to the first signal segment and a fourth signal segment related to the second signal segment are retrieved from the sensor signals of the selected category;

[0047] The step of determining the continuity evaluation parameters between the multiple signal segments by using the signal coordinates of the multiple signal segments in the selected category of sensor signals includes:

[0048] Based on the signal coordinates of the third signal segment in the selected category of sensor signals and the signal coordinates of the fourth signal segment in the selected category of sensor signals, a continuity evaluation parameter between the third signal segment and the fourth signal segment is determined.

[0049] When the continuity evaluation parameters among the multiple signal segments reach a preset standard, the signal segments in the selected category of sensor signals related to the sensor signal to be processed are associated and stored, including:

[0050] When the continuity evaluation parameter between the third signal segment and the fourth signal segment reaches the preset standard, the signal segment related to the sensor signal to be processed is selected from the selected category of sensor signals, and the selected signal segment is associated and stored.

[0051] The step of determining the continuity evaluation parameters between the third signal segment and the fourth signal segment based on the signal coordinates of the third signal segment and the fourth signal segment in the selected category of sensor signals includes:

[0052] Based on the signal coordinates of the third signal segment in the selected category of sensor signals and the signal coordinates of the fourth signal segment in the selected category of sensor signals, the signal spacing between the third signal segment and the fourth signal segment is determined.

[0053] The continuity evaluation parameters between the third signal segment and the fourth signal segment are determined based on the signal coordinate reference of the third signal segment and the fourth signal segment, as well as the spacing between the third signal segment and the fourth signal segment.

[0054] In a second aspect, this application provides a computer system comprising: one or more processors; a memory; and one or more computer programs; wherein the one or more computer programs are stored in the memory and configured to be executed by the one or more processors, and when the one or more computer programs are executed by the processors, they implement the method described above.

[0055] The beneficial effects of this application are as follows: This application determines the correlation between the target sensor signal and the sensor signal to be processed by combining the influence factors of different types of sensor signals. The obtained correlation can characterize the contribution of different types of sensor signals to the sensor signal to be processed, thus improving the accuracy of the obtained correlation. By using the correlation between different types of sensor signals and the sensor signal to be processed, the selected types of sensor signals are selected from the different types of sensor signals covered by the target sensor signal. The signal segments of the selected types of sensor signals that are related to the sensor signal to be processed are associated and stored. This can accurately find the relationship between data, provide medical workers with a comprehensive data analysis perspective, improve the utilization efficiency and value of medical monitoring data, and help to discover potential medical problems and risks. Attached Figure Description

[0056] Figure 1 This is a flowchart of a machine learning-based medical monitoring sensor data association and storage method provided in an embodiment of this application.

[0057] Figure 2 This is a schematic diagram of the composition of a computer system provided in an embodiment of this application. Detailed Implementation

[0058] The embodiments of this application are described below with reference to the accompanying drawings. The terminology used in the implementation section of this application is for explaining specific embodiments only and is not intended to limit the scope of this application.

[0059] In this embodiment of the medical monitoring sensor data association and storage method based on machine learning, the executing entity is a computer system, including but not limited to servers, personal computers, laptops, tablets, and smartphones. The computer system can run independently to implement this application, or it can be connected to a network and implement this application through interactive operations with other computer systems on the network. The network where the computer system is located includes, but is not limited to, the Internet, wide area networks, metropolitan area networks, local area networks, and VPN networks.

[0060] like Figure 1 As shown, the method includes:

[0061] Step S110: Obtain the target sensor signal related to the sensor signal to be processed, which includes different categories of sensor signals.

[0062] In step S110, the computer system acquires the target sensor signal related to the sensor signal to be processed. First, the computer system clarifies the specific content of the sensor signal to be processed. The sensor signal to be processed may be one or more specific physiological indicators, behavioral activities, or medical events, such as a patient's heart rate changes, daily activity levels, medication usage, etc. Taking heart rate changes as an example, suppose we want to analyze the reason for a patient's abnormally elevated heart rate over a specific time period; then the heart rate data is the sensor signal to be processed.

[0063] Next, the computer system searches its database for target sensor signals related to the sensor signal to be processed. This data typically originates from patient health monitoring devices, electronic medical record systems, or other medical information systems. The target sensor signals not only contain information directly related to the sensor signal to be processed (such as the patient's electrocardiogram record), but may also encompass other factors that may influence the sensor signal, such as the patient's age, gender, weight, medical history, and lifestyle habits. This information is organized into different categories of sensor signals to facilitate subsequent analysis and processing.

[0064] To illustrate more specifically, let's assume the sensor signal to be processed is a sudden increase in a patient's heart rate at a certain time. The computer system searches its database for the patient's electrocardiogram (ECG) records as directly relevant target sensor signals. Simultaneously, it also retrieves indirectly relevant sensor signals such as the patient's age (a physiological factor that may affect heart rate changes), gender (a physiological difference that may affect heart rate changes), weight (which may affect cardiac workload and heart rate), medical history (such as heart disease, hypertension, and other diseases that may affect heart rate), and lifestyle habits (such as regular exercise, smoking, and alcohol consumption, which may affect daily sensory factors related to heart rate).

[0065] In acquiring target sensor signals, computer systems can utilize machine learning models to optimize the search and filtering process. For example, it can train a classification model that automatically identifies and retrieves the most relevant target sensor signal category based on the input sensor signal to be processed (such as heart rate changes). This model can be trained using historical data, where the input features include the type and value of the sensor signal to be processed, and the output is a probability distribution representing the degree of association between each sensor signal category and the sensor signal to be processed.

[0066] When training a classification model, computer systems can use supervised learning algorithms such as Support Vector Machines (SVM), Random Forests, or Gradient Boosting Machines. These algorithms learn a decision boundary or function from labeled training data to distinguish the degree of association between different categories of sensor signals and the sensor signal to be processed. During training, the computer system continuously adjusts the model parameters to minimize prediction error and improve classification accuracy. After the model is trained, it is applied to a real-world data retrieval task. For a new sensor signal to be processed, the computer system inputs it into the model, and the model outputs a probability distribution indicating the degree of association between each sensor signal category and the sensor signal to be processed. Based on this probability distribution, the computer system can select the sensor signal categories with the highest degree of association as target sensor signals for subsequent analysis.

[0067] In addition to classification models, computer systems can utilize clustering algorithms to further organize target sensor signals. Clustering algorithms group similar sensor signals into one category, making it easier for analysts to identify patterns and trends in the data. For example, the K-means clustering algorithm can divide sensor signals into K clusters based on their numerical characteristics, with each cluster containing numerically similar sensor signals. In this way, the computer system can organize the target sensor signals into different clusters, each representing a specific sensor pattern or patient group.

[0068] Step S120: Based on the influence factors of different categories of sensor signals, obtain the correlation between the different categories of sensor signals covered by the target sensor signal and the sensor signal to be processed.

[0069] In step S120, the computer system calculates the correlation between different categories of sensor signals in the target sensor signal and the sensor signal to be processed. This process is based on the influence factors (i.e., weights) of different categories of sensor signals, aiming to assess the contribution of each sensor signal to the sensor signal to be processed, thereby selecting the most relevant sensor signals for analysis. First, the computer system determines the influence factors of different categories of sensor signals. These influence factors, also known as weights, reflect the importance of each sensor signal in a specific analysis scenario. The determination of influence factors can be based on various factors, including but not limited to statistical analysis of historical data, expert opinions, and prediction results from machine learning models.

[0070] Taking heart rate and blood pressure as examples, suppose we are analyzing a patient's abnormal heart rate. In this case, both heart rate and blood pressure signals are important sensor signal categories. However, their influence on the abnormal heart rate may differ. Statistical analysis of historical data reveals a higher correlation between heart rate signals and abnormal heart rate, thus allowing for a higher weighting. Conversely, while blood pressure signals are also related to heart health, they may not be a primary factor in the current analytical context, thus warranting a lower weighting.

[0071] Besides statistical analysis, influencing factors can also be determined using machine learning models. For example, a regression model can be trained that uses each sensor signal as input features and the sensor signal to be processed as the target variable, optimizing the model parameters by minimizing the prediction error. During training, the model automatically learns the degree of influence of each input feature (i.e., sensor signal) on the target variable (i.e., the sensor signal to be processed) and represents these influences as weights. After training, this model can be used to calculate the correlation between each sensor signal in a new dataset and the sensor signal to be processed.

[0072] Once the influencing factors of each sensor signal are determined, the computer system can proceed to the correlation calculation stage. This process typically involves segmenting the target sensor signal and calculating the correlation between each segment and the signal to be processed. Correlation calculation can be based on various methods, such as cosine similarity, Euclidean distance, and Pearson correlation coefficient. Each method has its advantages and disadvantages, and the choice of method depends on the specific application scenario and data characteristics.

[0073] Taking cosine similarity as an example, suppose we have two vectors A and B, representing the sensor signal to be processed and a certain sensor signal segment, respectively. The formula for calculating cosine similarity is:

[0074] ;

[0075] in, This represents the dot product of vectors A and B. Let A and B represent the magnitudes of vectors A and B, respectively. The cosine similarity value ranges from -1 to 1, with values ​​closer to 1 indicating greater similarity between the two vectors.

[0076] In practical applications, the sensor signal to be processed and the sensor signal segment can be represented as feature vectors, where each feature corresponds to a value of a sensor signal. Then, the cosine similarity between them is calculated using the formula above, serving as a measure of correlation. It should be noted that since an influence factor (weight) has been determined for each sensor signal category, the corresponding feature values ​​need to be weighted when calculating the dot product.

[0077] Specifically, assume there are three sensor signal categories: heart rate, blood pressure, and activity level, with influence factors of 0.6, 0.3, and 0.1, respectively. Also, assume the sensor signal to be processed and the sensor signal segment can be represented by the following feature vectors: .

[0078] When calculating cosine similarity, the feature values ​​of the sensor signal segments need to be weighted first: ; Then, cosine similarity is calculated using the weighted sensor signal segment and the sensor signal to be processed:

[0079] ;

[0080] By calculating the cosine similarity between each sensor signal segment and the sensor signal to be processed, a correlation matrix can be obtained, where each element represents the degree of correlation between the corresponding sensor signal segment and the sensor signal to be processed. Finally, the computer system can use this correlation matrix to select the most relevant sensor signal segments for subsequent analysis.

[0081] Step S130: Based on the correlation between different types of sensor signals and the sensor signal to be processed, select the sensor signal of the selected category from the different types of sensor signals.

[0082] In step S130, the computer system selects the sensor signal category that is most relevant and has the highest correlation with the sensor signal to be processed from among many sensor signal categories, based on the correlation between the different categories of sensor signals and the sensor signal to be processed calculated in step S120, and uses this as the selected category for subsequent analysis.

[0083] First, the computer system sorts the correlations calculated in step S120. Assume a correlation matrix is ​​already obtained, where each row represents a sensor signal category, each column represents a sensor signal segment to be processed, and the elements in the matrix are the correlations between the corresponding sensor signal category and the sensor signal segment to be processed. The computer system can sort this matrix by row (or column), placing the sensor signal categories with the highest correlations first. Taking heart rate, blood pressure, and activity level as examples, assume the correlations between these three sensor signal categories and a certain sensor signal segment to be processed (such as abnormal heart rate over a certain period) have already been calculated.

[0084] Next, the computer system sets a filtering threshold or selects a certain number of sensor signal categories as selected categories based on the sorted relevance matrix. The filtering threshold can be set according to the actual analysis needs and data characteristics. For example, it can select all sensor signal categories with a relevance higher than a certain value, or select the top N sensor signal categories with the highest relevance (N is a preset integer).

[0085] Continuing with the above example, assuming a filtering threshold of 0.7, only sensor signal categories with a correlation higher than 0.7 will be selected as the chosen category. In this case, only the heart rate signal meets the condition, therefore, the heart rate signal will be selected as the chosen category for subsequent analysis.

[0086] Furthermore, if the top N sensor signal categories with the highest relevance are selected as the chosen categories, the computer system will select them sequentially from highest to lowest relevance until the number of selected sensor signal categories reaches N. For example, if the top two sensor signal categories with the highest relevance are selected, then heart rate and blood pressure signals will be chosen as the selected categories.

[0087] When setting screening thresholds or selecting the number of sensor signal categories, computer systems can consider the impact of factors such as data sparsity and noise interference on correlation. If the data contains a large number of missing or outlier values, these values ​​can reduce the accuracy of the correlation. To address this, computer systems can employ data preprocessing techniques (such as interpolation, smoothing, and denoising) to improve data integrity and accuracy, thereby ensuring the reliability of correlation calculations.

[0088] Furthermore, in certain complex scenarios, computer systems can utilize machine learning models to assist in selecting sensor signals of a chosen category. For example, a classification model can be trained to predict which sensor signal categories are most relevant to the sensor signal to be processed. This model can be trained based on historical data, where input features include the correlation between each sensor signal category and the sensor signal to be processed, statistical characteristics of the sensor signal categories, etc., and the output is a binary label (indicating whether the sensor signal category is selected). After the model is trained, the computer system can input new data into the model for prediction and select sensor signals of the chosen category based on the prediction results.

[0089] Step S140: Associate and store the signal segments in the selected category of sensor signals that are related to the sensor signal to be processed.

[0090] In step S140, the computer system associates and stores the signal segments from the selected category of sensor signals that are related to the sensor signal to be processed. This step aims to help users (such as medical professionals) quickly identify the sensor signal segments that are most relevant and valuable to the target analysis, thereby obtaining key information more efficiently and making accurate judgments.

[0091] First, the computer system determines which signal segments are relevant to the sensor signal being processed. This typically involves further subdividing the selected category of sensor signals into smaller segments and evaluating the correlation between each segment and the sensor signal being processed. This correlation can be measured using various metrics, such as temporal synchronicity, numerical similarity, and consistency of trends. For example, in heart rate monitoring, if the heart rate variability pattern of a particular signal segment matches the sensor signal being processed (e.g., a patient suddenly experiencing palpitations), then that signal segment can be considered relevant.

[0092] In determining relevant signal segments, computer systems can utilize advanced signal processing techniques, such as filtering, smoothing, and feature extraction, to improve signal quality and reduce noise interference. These techniques help to more accurately identify key features and trends in the signal.

[0093] Once the relevant signal segments are identified, the computer system stores them in association.

[0094] In summary, by comprehensively utilizing signal processing, machine learning, and other technologies, step S140 enables the computer system to accurately identify the relationships between data, providing medical professionals with a comprehensive data analysis perspective.

[0095] In one implementation, step S140 involves associating and storing signal segments from the selected category of sensor signals that are related to the sensor signal to be processed, including:

[0096] Step S141: Retrieve multiple signal segments related to the sensor signal to be processed from the selected category of sensor signals;

[0097] Step S142: Determine the continuity evaluation parameters between multiple signal segments by using the signal coordinates of multiple signal segments in the selected category of sensor signals;

[0098] Step S143: When the continuity evaluation parameters between multiple signal segments reach the preset standard, the signal segments related to the sensor signal to be processed in the selected category of sensor signals are associated and stored.

[0099] In step S141, the computer system identifies signal segments from the selected category of sensor signals that are highly correlated with the sensor signal to be processed. Correlation is typically assessed based on the similarity between the signal segment and the sensor signal to be processed in terms of value, trend, pattern, etc. To achieve this, the computer system can employ various methods, including but not limited to signal matching algorithms and pattern recognition techniques. For example, suppose the electrocardiogram data of a heart patient is being analyzed, where the selected category of sensor signal is the heart rate signal. The sensor signal to be processed is the heart rate change during a heart attack in this patient. The computer system retrieves signal segments from the heart rate signal that are similar in pattern to the heart rate change during this heart attack as target signal segments.

[0100] First, the computer system divides the heart rate signal into multiple shorter signal segments, each containing continuous heart rate data points. For each segment, the system extracts a series of features, such as mean, standard deviation, kurtosis, and skewness, which summarize the segment's main statistical characteristics. Then, the system calculates the similarity between each segment and the sensor signal to be processed. This can be achieved by calculating the distance (e.g., Euclidean distance) or correlation (e.g., Pearson correlation coefficient) between the segment's features and the sensor signal's features. A similarity threshold is set; only segments with similarities exceeding this threshold are selected as target segments. Through these steps, the computer system can identify multiple signal segments associated with the sensor signal to be processed.

[0101] In step S142, after the signal segments are identified, the computer system assesses the continuity and tightness between these segments. The continuity assessment parameter is a quantitative indicator used to measure whether the signal segments are closely connected without significant discontinuities or jumps. This is crucial for understanding the continuity and stability of patient behavior.

[0102] Continuity assessment parameters can be calculated through the following steps: For each signal segment, the computer system records its start and end coordinates (e.g., timestamps) within a selected category of sensor signals. For any two adjacent signal segments, the time interval (or spatial distance, if the signal is multidimensional) between them is calculated. One or more reference coordinates are selected, which can represent the importance of the signal segment or points of interest in the analysis. For example, in a heart rate signal, points of rapid heart rate changes can be selected as reference coordinates. The continuity assessment parameters are then calculated, for example, by first calculating a weighted sum of the intervals between all adjacent signal segments, with weights set according to the importance of the signal segment or the analytical requirements. For example, a signal segment containing critical information (such as a sudden change in heart rate) can be given a higher weight. The influence of the reference coordinates on the continuity of the signal segment is considered. If a signal segment contains reference coordinates, or is very close to reference coordinates, the value of the continuity assessment parameter can be increased. The weighted sum of the signal segment intervals is combined with the influence of the reference coordinates to obtain the final continuity assessment parameters. For example, the following formula can be used:

[0103] ;

[0104] in, It is the weight of the spacing between the i-th signal segments. It is the spacing between the i-th and i+1-th signal segments, and m is the number of reference coordinates. It is the effect of the j-th reference coordinate on continuity.

[0105] For example, continuing with the heart rate signal, suppose three signal segments have been identified: A, B, and C, with their start and end coordinates as follows:

[0106] Signal segment A: Signal segment B: Signal segment C: .

[0107] in, Suppose that signal segment B contains a significant heart rate variability point, and therefore it is given higher weight.

[0108] The steps for calculating the continuity assessment parameters are as follows:

[0109] 1. Calculate the signal segment spacing:

[0110] 2. Determine the influence of reference coordinates: Since signal segment B contains a point of sudden heart rate change, additional reference coordinate influence values ​​are given.

[0111] 3. Calculate continuity assessment parameters: ;

[0112] in, These are the spacing weights between signal segments A and B, and between signal segments B and C, respectively. It is the effect of the sudden heart rate change point contained in signal segment B on continuity.

[0113] In step S143, after determining the continuity evaluation parameters, the computer system compares them with preset standards. If the continuity evaluation parameters meet or exceed the preset standards, it indicates that these signal segments have high continuity and tightness, and therefore it is worthwhile to perform associated storage.

[0114] For example, continuing with the heart rate signal, assume the preset continuity assessment parameter standard is 10. If the continuity assessment parameter calculated through the above steps is greater than or equal to 10, the computer system will associate and store signal segments A, B, and C.

[0115] The purpose of correlated storage is not only to store signal segments related to the sensor signals being processed, but more importantly, to help users understand the overall patterns and trends of patient behavior by storing the continuity and coherence between these signal segments. This is especially important for healthcare professionals, as they need to develop personalized treatment plans and interventions based on patient behavior patterns.

[0116] In one implementation, step S120 involves obtaining the correlation between the target sensor signal and the sensor signal to be processed, based on the influence factors of different categories of sensor signals, including:

[0117] Step S121: Decompose the sensor signal to be processed into signal segments to obtain the signal segments to be processed;

[0118] Step S122: Obtain the basic correlation between different types of sensor signals and the sensor signals to be processed based on the signal segments to be processed;

[0119] Step S123: Based on the basic correlation between different types of sensor signals and the sensor signal to be processed, and the influence factors of different types of sensor signals, obtain the correlation between different types of sensor signals and the sensor signal to be processed.

[0120] In step S121, the computer system breaks down the sensor signal to be processed into smaller signal segments to facilitate subsequent basic correlation calculations. Signal segmentation can be based on time windows, the number of data points, or other custom rules. Each segment represents a subsequence of the sensor signal to be processed, used for correlation analysis with different categories of sensor signals.

[0121] For example, suppose the sensor signal to be processed is a continuous heart rate data segment spanning 24 hours. To segment the signal, the computer system can set a fixed-length time window (e.g., 1 hour) and divide the heart rate data according to this time window, resulting in multiple 1-hour heart rate signal segments. Each signal segment contains heart rate data points within that time period.

[0122] In step S122, after the signal segmentation is completed, the computer system calculates the basic correlation between each signal segment to be processed and the sensor signals of different categories. The basic correlation measures the similarity or association between signal segments and can be calculated by various methods, such as cosine similarity, Euclidean distance, and dynamic time warping (DTW).

[0123] For example, continuing with heart rate data, suppose there are two categories of sensor signals: blood pressure signals and step count signals. For each segment of heart rate signal to be processed, the computer system can calculate its basic correlation with the blood pressure signal segments and step count signal segments within the corresponding time period.

[0124] Calculate cosine similarity: Assume that the heart rate signal segment, blood pressure signal segment, and step count signal segment can all be represented as vectors, such as the heart rate signal segment vector. Blood pressure signal segment vector Step number signal segment vector Cosine similarity can be calculated using the following formula:

[0125] ;

[0126] ;

[0127] Where · represents the vector dot product, and |·| represents the vector magnitude.

[0128] In some cases, certain parts of a signal segment may be more important than others. For example, in a heart rate signal, outliers (such as heart rate jumps) may be more valuable for analysis than normal values. Therefore, when calculating baseline correlation, a computer system can assign higher weights to these important parts. This can be achieved by adjusting the element values ​​of the signal segment vector or by introducing additional weight vectors.

[0129] In step S123, after obtaining the basic correlation, the computer system combines the influence factors of different categories of sensor signals to calculate the final correlation. The influence factors reflect the importance of each category of sensor signal in a specific analysis scenario and are usually determined based on expert knowledge, historical data, or machine learning models.

[0130] Influence factors can be determined in several ways. First, domain experts can score different categories of sensor signals based on their experience, with higher scores indicating greater importance. Second, by analyzing historical data, a computer system can identify the sensor signal categories most relevant to the signal being processed and assign them higher influence factors. Third, a classification or regression model can be trained, using the sensor signal category and the signal being processed as input, and actual analysis results or expert evaluations as output, to learn the influence factors for each category of sensor signal.

[0131] Once the impact factors are determined, the computer system can calculate the final relevance using the following formula: . in, Indicates the type of the i-th sensor signal. This indicates the basic correlation between this type of signal and the sensor signal to be processed. This represents the influence factor of this type of signal.

[0132] For example, suppose the baseline correlation between heart rate and blood pressure signals is 0.8, and the baseline correlation with step count signals is 0.6. Also suppose that in the heart rate signal analysis scenario, the influence factor for blood pressure signals is 1.2 (because blood pressure and heart rate have a direct physiological relationship), and the influence factor for step count signals is 0.8 (because while step count is related to health, its direct correlation with heart rate is weak). Then the final correlation calculations are as follows: Blood pressure signal correlation: 0.8 × 1.2 = 0.96. Step count signal correlation: 0.6 × 0.8 = 0.48.

[0133] By comparing these final correlation scores, the computer system can determine which categories of sensor signals are most relevant to the sensor signal being processed, thus providing support for subsequent analysis and decision-making.

[0134] In one implementation, step S122, obtaining the basic correlation between different categories of sensor signals and the sensor signals to be processed based on the signal segment to be processed, includes:

[0135] Step S1221: Retrieve the target signal segment related to the signal segment to be processed from different categories of sensor signals;

[0136] Step S1222: Based on the priority of the signal segment to be processed and the priority of the target signal segments related to the signal segment to be processed in different categories of sensor signals, determine the basic correlation between the sensor signals of different categories and the sensor signals to be processed.

[0137] In step S1221, the computer system's task is to find target signal segments that are related to the signal segment to be processed in terms of time, content, or features. These target signal segments will be used for subsequent basic correlation calculations. To achieve this goal, the computer system takes a series of measures to ensure the correlation of the signal segments.

[0138] Criteria for determining correlation can be, for example, time synchronization. For time-series data such as heart rate and blood pressure, a computer system can check the degree of temporal overlap between the target signal segment and the signal segment to be processed. If two signal segments have significant temporal overlap and similar trends, they are, for instance, correlated. Alternatively, content similarity can be used. For non-time-series data such as text and images, a computer system can determine their correlation by calculating the similarity between signal segments (e.g., cosine similarity, Jaccard similarity, etc.). The higher the similarity, the more likely the signal segments are to be correlated. Another criterion can be feature consistency. For example, in some cases, a computer system can extract features of a signal segment (e.g., mean, variance, peak value, etc.) and determine the correlation of the signal segments by comparing the consistency of these features.

[0139] For example, suppose the signal segment to be processed is a continuous heart rate data segment spanning from 9:00 AM to 10:00 AM. The computer system finds the signal segment related to this heart rate data from blood pressure and step count signals.

[0140] For blood pressure signals, the computer system can examine the blood pressure data segment between 9 a.m. and 10 a.m. If this blood pressure data segment significantly overlaps with the heart rate data in time, and the trend of blood pressure change is similar to that of heart rate change (e.g., blood pressure also increases when heart rate increases), then this blood pressure data segment can be considered a relevant target signal segment.

[0141] For step count signals, computer systems can calculate the similarity between step count signals and heart rate signals. For example, the Dynamic Time Warping (DTW) algorithm can be used to compare the shape similarity between a segment of step count signal and a segment of heart rate signal. If the similarity is high, this segment of step count data can also be considered a relevant target signal segment.

[0142] In step S1222, after finding the relevant target signal segment, the next step of the computer system is to determine the basic relevance based on the priority of the signal segment. The priority reflects the importance or criticality of the signal segment in the analysis and can be determined based on the content, characteristics, or analysis results of the signal segment.

[0143] Prioritization can be based on several methods. For example, it can be based on signal segment content. Some signal segments may contain crucial information such as outliers, peaks, or troughs, which are essential for the analysis. Therefore, signal segments containing this crucial information can be assigned higher priority. Alternatively, it can be based on signal segment characteristics. Certain characteristics of a signal segment (such as mean, variance, and kurtosis) may be more important for the analysis. For instance, in heart rate analysis, an abnormally high mean heart rate may be more valuable for analysis than a normal mean heart rate. Another method is based on the analysis results. In some cases, the computer system may have already performed a preliminary analysis of the signal segments and drawn certain conclusions or findings. These conclusions or findings can guide the priority determination. For example, if a signal segment is analyzed to be highly correlated with a certain disease state, it should be assigned higher priority.

[0144] Once the priority of the signal segments is determined, the computer system can use the following formula to calculate the basic correlation: ;in, Let n represent the i-th sensor signal category, and n represent the number of relevant target signal segments found in that category. This represents the priority weight of the j-th relevant target signal segment. This represents the j-th relevant target signal segment. Indicates the signal segment to be processed. This represents a function that calculates the similarity between two signal segments.

[0145] For example, continuing with heart rate data, suppose we have already identified target signal segments in the blood pressure and step count signals that are related to the heart rate signal segment to be processed. Now, we need to calculate the baseline correlation based on the priority of these signal segments.

[0146] Assume that blood pressure signal segment 1 has a priority weight of 0.8 (because it contains a significant blood pressure peak), and blood pressure signal segment 2 has a priority weight of 0.6 (because it only represents ordinary blood pressure fluctuations). Assume that step count signal segment 1 has a priority weight of 0.5 (because its similarity to the heart rate signal segment is not particularly high), and step count signal segment 2 has a priority weight of 0.7 (because it is highly synchronized with the heart rate signal segment in time). Assume that the similarity between each relevant target signal segment and the heart rate signal segment to be processed has already been calculated.

[0147] The above formula can now be used to calculate the baseline correlation between blood pressure and step count signals and the heart rate signal to be processed: .

[0148] By comparing these two basic correlations, the computer system can determine which category of sensor signal is more relevant to the sensor signal to be processed, thus providing support for subsequent analysis and decision-making.

[0149] In summary, steps S1221 and S1222 together constitute the process of calculating the basic correlation between different categories of sensor signals and the sensor signal to be processed. Through correlation judgment and priority determination, the computer system can accurately assess the similarity and importance between signal segments, providing strong data support for subsequent analysis.

[0150] In one implementation scheme, if there are multiple target sensor signals, then the method also includes:

[0151] Step S150: Determine the influence factor of each target sensor signal based on the correlation mechanism between the sensor signal to be processed and each target sensor signal;

[0152] Step S160: Based on the influence factor of each target sensor signal and the correlation between the different categories of sensor signals covered by each target sensor signal and the sensor signal to be processed, obtain the correlation between each target sensor signal and the sensor signal to be processed;

[0153] Step S170: Sort the multiple target sensor signals according to the correlation between each target sensor signal and the sensor signal to be processed.

[0154] In step S150, the computer system determines the influence factor of each target sensor signal. The influence factor is a quantitative indicator used to measure the importance or weight of the target sensor signal in the overall analysis. To accurately determine the influence factor, the computer system considers the correlation mechanism between the sensor signal to be processed and each target sensor signal.

[0155] Correlation mechanisms refer to the ways or rules by which a connection is established between a sensor signal to be processed and a target sensor signal. These mechanisms can be based on various factors, including but not limited to time synchronization, content similarity, and feature consistency. For example, regarding time synchronization, if two signals significantly overlap in time and have similar trends of change, they are, for example, correlated. Regarding content similarity, if two signals contain the same or similar information (such as keywords, patterns, etc.), they are also, for example, correlated.

[0156] Computer systems can assign different weights to each target sensor signal based on the strength and complexity of the correlation mechanism. For example, if the correlation mechanism between a target sensor signal and the sensor signal to be processed is very clear and strong (such as direct time synchronization and content correlation), then that signal can be assigned a higher weight.

[0157] In some cases, computer systems can use machine learning models to predict the impact factor of each target sensor signal. These models can be trained on historical data and learn how to map relevant mechanisms to the impact factor. For example, regression or classification models can be used to predict the impact factor, where the input features may include multiple aspects of relevant mechanisms (such as time synchronization, content similarity, feature consistency, etc.), and the output is the impact factor of the target sensor signal.

[0158] In step S160, the computer system combines the correlation between the different types of sensor signals and the sensor signal to be processed calculated in step S120, as well as the influence factor determined in step S150, to calculate the overall correlation between each target sensor signal and the sensor signal to be processed.

[0159] The overall relevance can be calculated using the following formula:

[0160] ; in, Let represent the signal of the i-th target sensor, and n represent the number of different categories of sensor signals included in this signal. This represents the influence factor of the j-th category sensor signal in the i-th target sensor signal. This represents the correlation between the sensor signal of the j-th category and the sensor signal to be processed (calculated in step S120).

[0161] In step S170, the computer system sorts the multiple target sensor signals according to the overall correlation calculated in step S160 and lists them in order. This helps medical workers quickly identify the target sensor signal most relevant to the sensor signal to be processed, thereby enabling more efficient subsequent analysis.

[0162] The computer system can sort the overall relevance from high to low, placing the target sensor signals with the highest relevance first. This allows medical professionals to focus first on signals most relevant to the sensor signal being processed.

[0163] In some cases, the computer system can also set a correlation threshold. Only target sensor signals with an overall correlation exceeding this threshold will be listed. This helps reduce noise interference and improve analysis efficiency.

[0164] For example, suppose there are three target sensor signals (patient A, patient B, and patient C), each signal encompassing two categories of sensor signals (e.g., heart rate and blood pressure). In step S120, the correlation between each category of sensor signal and the sensor signal to be processed has been calculated. In step S150, an influence factor is assigned to each category of sensor signal in the target sensor signal according to the correlation mechanism. Now, the overall correlation between each target sensor signal and the sensor signal to be processed is calculated and ranked:

[0165] For patient A:

[0166] Heart rate signal correlation: 0.8; Blood pressure signal correlation: 0.6; Heart rate signal influence factor: 0.7 (based on correlation mechanism allocation); Blood pressure signal influence factor: 0.5 (based on correlation mechanism allocation); Overall correlation (patient A): 0.7×0.8 +0.5×0.6=0.86.

[0167] For patient B:

[0168] Correlation of heart rate signal: 0.7; Correlation of blood pressure signal: 0.5; Influence factor of heart rate signal: 0.8 (based on correlation mechanism allocation); Influence factor of blood pressure signal: 0.4 (based on correlation mechanism allocation); Overall correlation (patient B): 0.8×0.7 +0.4×0.5=0.76.

[0169] For patient C:

[0170] Correlation of heart rate signal: 0.6; Correlation of blood pressure signal: 0.4; Influence factor of heart rate signal: 0.6 (based on correlation mechanism allocation); Influence factor of blood pressure signal: 0.3 (based on correlation mechanism allocation); Overall correlation (patient C): 0.6×0.6 +0.3×0.4=0.48.

[0171] The target sensor signals are listed below in descending order of overall relevance:

[0172] 1. Patient A (overall correlation: 0.86); 2. Patient B (overall correlation: 0.76); 3. Patient C (overall correlation: 0.48).

[0173] Through the above steps, the signal most relevant to the sensor signal to be processed can be accurately identified from multiple target sensor signals, providing orderly data support for medical workers.

[0174] In one implementation, step S110, acquiring the target sensor signal related to the sensor signal to be processed, includes:

[0175] Step S111: Decompose the sensor signal to be processed into signal segments to obtain a set of signal segments to be processed;

[0176] Step S112: Divide the signal segments in the set of signal segments to be processed into clusters to obtain multiple signal segment clusters;

[0177] Step S113: Acquire the target sensor signals associated with multiple signal segment clusters respectively, with one signal segment cluster corresponding to one related mechanism.

[0178] In step S111, the computer system splits the sensor signal to be processed into multiple smaller signal segments to form a set of signal segments to be processed. Signal segment splitting is one of the key steps in data preprocessing, which helps the computer system to analyze the characteristics and patterns of the sensor signal to be processed in greater detail.

[0179] Signal segmentation methods include time-window-based segmentation. The computer system can set a fixed-length time window (such as 1 minute, 5 minutes, etc.) and then segment the sensor signal to be processed according to this time window. For example, if the sensor signal to be processed is a continuous heart rate data over a period of 24 hours, the computer system can segment it into 288 10-minute signal segments (assuming there are 6 10-minute signal segments per hour).

[0180] Alternatively, based on event-triggered segmentation, in some cases, the sensor signal being processed may contain specific event markers (such as outliers, peaks, etc.). The computer system can then segment the signal into different segments based on these event markers. For example, in heart rate data, the computer system can segment the signal into a new segment whenever the heart rate exceeds a certain threshold.

[0181] Alternatively, in addition to time windows, computer systems can also divide signals based on the number of data points. For example, each signal segment can be set to contain 100 data points, and the sensor signal to be processed can be divided according to this number.

[0182] For example, suppose the sensor signal to be processed is a continuous heart rate data segment spanning one hour. The computer system sets a 5-minute time window and divides the heart rate data into 12 signal segments. Each signal segment contains heart rate data points within that time period, forming a set of signal segments to be processed.

[0183] In step S112, after obtaining the set of signal segments to be processed, the next step of the computer system is to cluster these signal segments to form multiple signal segment clusters. The purpose of clustering is to group signal segments with similar characteristics or patterns into one category, so that the target sensor signal related to the sensor signal to be processed can be identified more accurately in the subsequent process.

[0184] Clustering can be achieved through features-based clustering. The computer system can extract features from each signal segment (such as mean, variance, kurtosis, skewness, etc.) and then use clustering algorithms (such as K-means, DBSCAN, etc.) to group these feature vectors into clusters. Signal segments with similar features will be grouped together.

[0185] Alternatively, pattern-based clustering can be used. In addition to features, computer systems can also consider pattern similarity between signal segments. For example, the Dynamic Time Warping (DTW) algorithm can be used to compare the shape similarity between signal segments and group signal segments with similar shapes into a single cluster.

[0186] Alternatively, clustering can be based on expert knowledge. In some cases, domain experts may be able to divide signal segments into different clusters based on experience. The computer system can accept expert input and perform clustering accordingly. For example, continuing with heart rate data, the computer system extracts the mean and variance of each signal segment as features, and then uses the K-means clustering algorithm to cluster these feature vectors. Suppose that three clusters are ultimately obtained, representing low heart rate, normal heart rate, and high heart rate signal segments, respectively.

[0187] In step S113, after obtaining multiple signal segment clusters, the final step of the computer system is to acquire the target sensor signal associated with each signal segment cluster. The correlation mechanism refers to the method or rule for establishing a connection between the signal segment cluster and the target sensor signal. Different signal segment clusters may correspond to different correlation mechanisms.

[0188] For example, time synchronization is used; that is, for time series data such as heart rate and blood pressure, the time synchronization between signal segment clusters and target sensor signals is an important correlation mechanism. If two signals have significant temporal overlap and similar trends, they are, for example, correlated.

[0189] Alternatively, content similarity can be used. In some cases, the content similarity between signal segment clusters and target sensor signals may be more important. For example, in text analysis, if two signal segment clusters contain the same keywords or phrases, they may be related to the same target sensor signal.

[0190] Alternatively, feature consistency can be employed; the consistency of features between signal segment clusters and target sensor signals is another possible correlation mechanism. For example, in image analysis, if two signal segment clusters have similar texture, color, or other features, they may be correlated with the same target image signal. Continuing with the heart rate data example, for signal segment clusters representing low heart rates, the computer system can search for target sensor signals that also contain low heart rate signal segments as correlated signals. For signal segment clusters representing normal heart rates, it can search for target sensor signals that contain normal heart rate signal segments. And for signal segment clusters representing high heart rates, it can search for target sensor signals that contain abnormal heart rate signal segments (such as sudden increases or decreases in heart rate). These correlation mechanisms may be determined based on factors such as time synchronization, content similarity, or feature consistency.

[0191] Through the above steps, the computer system can accurately identify the signal segment most relevant to the sensor signal to be processed from a large number of target sensor signals, and provide strong support for subsequent data analysis and pattern recognition.

[0192] In one implementation, step S130 involves filtering sensor signals of a selected category from the different categories of sensor signals based on the correlation between the different categories of sensor signals and the sensor signal to be processed, including:

[0193] Step S131: Select sensor signals of the target category and sensor signals of unselected categories from sensor signals of different categories;

[0194] Step S132: Based on the correlation between different categories of sensor signals and the sensor signals to be processed, filter out sensor signals of the unselected categories to obtain sensor signals of the undetermined categories;

[0195] Step S133: Use the sensor signal of the undetermined category and the sensor signal of the target category as the sensor signal of the selected category.

[0196] In the implementation of step S130, the goal of the computer system is to filter out the signal category most relevant to the sensor signal to be processed from different categories of sensor signals, as the selected category for subsequent analysis.

[0197] In step S131, the computer system performs a preliminary screening of all different categories of sensor signals to determine which categories are target categories and which are unselected categories. Target categories are usually predetermined based on domain knowledge, expert opinions, or historical data, and they have a high correlation or importance with the sensor signals to be processed.

[0198] Target category determination can be based on domain knowledge, for example. In some cases, domain experts may be able to directly determine which sensor signal categories are most relevant to the sensor signal being processed, based on their expertise. For instance, in heart rate monitoring, heart rate and blood pressure signals are often the target categories when analyzing heart rate abnormalities.

[0199] Alternatively, based on historical data, computer systems can analyze historical data to determine which sensor signal categories have a high correlation with the sensor signal to be processed. For example, association rule mining algorithms (such as the Apriori algorithm) can be used to discover frequent itemsets between sensor signal categories and the sensor signal to be processed, thereby determining the target category.

[0200] Alternatively, based on machine learning models, in certain complex scenarios, computer systems can use machine learning models to predict which sensor signal categories are most likely related to the sensor signal to be processed. For example, a classification model can be trained, using sensor signal categories as input features and the presence or absence of the sensor signal to be processed as the output label, to determine the target category through model prediction. For instance, suppose we are analyzing the electrocardiogram data of a heart patient, and the sensor signal to be processed is the patient's heart rate changes during a heart attack. Based on domain knowledge and historical data, heart rate and blood pressure signals can be identified as the target categories highly correlated with heart attacks. Other sensor signals (such as step count, sleep quality, etc.) are considered as unselected categories outside the target categories.

[0201] In step S132, after determining the target category and the unselected categories, the next step of the computer system is to filter out signal categories with high correlation to the sensor signals to be processed from the unselected categories based on the correlation calculated in step S120, and designate them as undetermined categories. Undetermined categories refer to sensor signal categories that, although not initially identified as target categories, exhibit high correlation in actual data analysis.

[0202] One method for filtering pending categories is to set a relevance threshold. The computer system can set a relevance threshold, and only unselected categories with a relevance exceeding this threshold will be filtered as pending categories. The threshold can be adjusted according to actual needs and data characteristics.

[0203] The computer system can sort the relevance of all unselected categories and select the top N categories with the highest relevance as candidate categories (N is a preset integer). This method ensures that the selected candidate categories have high relevance and importance.

[0204] Alternatively, machine learning models can be used for filtering. In some cases, computer systems can employ machine learning models to assist in filtering candidate categories. For example, a regression model can be trained to predict the correlation between each unselected category and the sensor signal to be processed, and the candidate categories can be filtered based on the predictions. For instance, let's continue with the example of electrocardiogram (ECG) data from a heart patient. Assume the correlation between heart rate signals, blood pressure signals, and other unselected categories (such as steps, sleep quality, etc.) and the heart rate change signal to be processed has already been calculated. By setting a correlation threshold or a ranking selection method, the step count signal can be filtered out from the unselected categories as a candidate category, because the step count signal may be correlated with heart rate changes during a heart attack (e.g., reduced activity before a heart attack may lead to a decrease in steps).

[0205] In step S133, the computer system merges the sensor signals of the undetermined category with the sensor signals of the target category to form the final selected category set. The selected category set contains all sensor signal categories that are highly correlated with the sensor signals to be processed, and these will be the focus of subsequent data analysis.

[0206] Category merging can be achieved by directly combining the candidate category and the target category into a new set, which is then designated as the selected category. Alternatively, in some cases, the computer system can assign different weights to different category signals based on their importance or relevance, and consider these weights during the merging process. For example, a weighted average method can be used to calculate the overall relevance or importance score of the selected category.

[0207] For example, let's continue using electrocardiogram (ECG) data from a heart disease patient. In step S132, the step count signal has already been selected as a candidate category. Now, the step count signal is merged with the heart rate and blood pressure signals (target categories) to form the final set of selected categories. This set will be the focus of subsequent data analysis to further explore the relationship between heart attacks and heart rate, blood pressure, and step count.

[0208] Through the above steps, the computer system can accurately filter out the signal category most relevant to the sensor signal to be processed from different categories of sensor signals, providing strong support for subsequent data analysis and pattern recognition.

[0209] In one implementation, step S132, based on the correlation between different categories of sensor signals and the sensor signal to be processed, filters sensor signals of the unselected category to obtain the sensor signals of the undetermined category, including:

[0210] Step S1321: Based on the correlation between different categories of sensor signals and the sensor signals to be processed, select sensor signals of the original unselected categories whose correlation is higher than the correlation threshold from the unselected sensor signals;

[0211] Step S1322: Among the sensor signals of the original undetermined category with a correlation greater than or equal to a preset threshold, the top N sensor signals with the highest correlation are taken as the sensor signals of the undetermined category, where N is a positive integer greater than or equal to 1.

[0212] In step S1321, the computer system determines a correlation threshold and, based on this threshold, filters out signal categories with higher correlation from the sensor signals that have not been selected as categories. These signal categories will be regarded as sensor signals of the original undetermined categories, and they may be further confirmed as undetermined categories in subsequent steps.

[0213] One method for determining the correlation threshold is based on statistical distribution. The computer system can calculate the statistical distribution (such as mean, standard deviation, etc.) of the correlation between all unselected sensor signals and the sensor signal to be processed, and determine a reasonable correlation threshold based on this distribution. For example, the mean plus one or two times the standard deviation can be chosen as the threshold to ensure that the selected signal categories have a certain degree of statistical significance.

[0214] Alternatively, a domain-knowledge-based approach may be used. In some cases, domain experts may directly provide a reasonable relevance threshold based on their expertise. This threshold may be based on past research experience, experimental results, or industry standards.

[0215] Alternatively, based on machine learning models, the computer system can use these models to predict a reasonable correlation threshold. This model can be trained on historical data to learn the relationship between correlation and signal category importance, and output a predicted threshold. For example, suppose we are analyzing the blood glucose data of a diabetic patient, and the sensor signal to be processed is an event where the patient's blood glucose level rose abnormally. There is a series of sensor signals without selected categories, including diet records, exercise records, sleep records, etc. First, the computer system calculates the correlation between these signals and the sensor signal to be processed. Then, based on statistical distribution methods, it selects the mean plus one standard deviation as the correlation threshold. In this way, all signal categories with correlations higher than this threshold (such as diet records) are filtered out as the original undetermined categories of sensor signals.

[0216] In step S1322, after determining the initial undetermined sensor signals, the next step of the computer system is to sort these signal categories according to their correlation and select the top N signal categories with the highest correlation as the final undetermined categories. This process helps ensure that the selected undetermined categories have the highest correlation with the sensor signals to be processed.

[0217] The computer system sorts all the sensor signals initially categorized into undetermined classes according to their relevance, from highest to lowest. This places the signal categories with the highest relevance at the top. After sorting, the system selects the top N most relevant signal categories as the undetermined classes. The value of N can be determined based on actual needs and data characteristics. For example, if the dataset contains many signal categories but only a few of the most important categories are desired for analysis, N can be set relatively small; conversely, if a more comprehensive understanding of the dataset is desired, N can be set relatively large.

[0218] For example, continuing with the example of blood glucose data from diabetic patients. In step S1321, dietary records have already been selected as sensor signals for the initial pending category. Assume that other sensor signals not yet selected (such as exercise records, sleep records, etc.) also reach the correlation threshold and are selected as sensor signals for the initial pending category. Now, the computer system sorts these signal categories according to their correlation from high to low. Assume the sorting result is: dietary records (correlation 0.9), exercise records (correlation 0.7), and sleep records (correlation 0.5). If N=2 is selected, then dietary records and exercise records will be selected as the final pending categories because they have the highest correlation with the sensor signals to be processed.

[0219] Through the steps described above, the computer system can filter out signal categories with high correlation to the sensor signal to be processed from the unselected sensor signals, and use these as candidate categories. This process not only considers the correlation between the signal category and the sensor signal to be processed, but also ensures that the selected candidate categories have the highest importance through sorting and selection.

[0220] In one implementation, step S140, which involves associating and storing signal segments from the selected category of sensor signals related to the sensor signal to be processed, further includes:

[0221] Step S144: Select target category signal segments from the set of signal segments to be processed. The set of signal segments to be processed includes signal segments obtained by segmenting the sensor signals to be processed.

[0222] Step S145: Associate and store the signal segments in the selected category of sensor signals that are related to the target category signal segment.

[0223] In step S144, the computer system filters target category signal segments from the set of signal segments to be processed. The set of signal segments to be processed is obtained in step S111 by segmenting the sensor signals to be processed into signal segments. It contains multiple signal segments, each representing a subsequence of the sensor signals to be processed. Target category signal segments refer to those signal segments that are closely related to a specific analysis target or point of interest.

[0224] Computer systems can extract features of each signal segment (such as mean, variance, kurtosis, skewness, etc.) and filter signal segments into target categories based on these features. For example, in heart rate monitoring, if the analysis goal is to identify abnormal heart rate events, the computer system can filter out signal segments with heart rate values ​​outside the normal range (such as less than 60 beats per minute or more than 100 beats per minute) as target category signal segments.

[0225] In addition to feature-based filtering methods, computer systems can also use pattern correlation algorithms to identify target category signal segments. For example, the Dynamic Time Warping (DTW) algorithm can be used to compare the shape similarity between signal segments and filter out signal segments with high correlation to a preset pattern or template as target category signal segments.

[0226] In some cases, domain experts can directly determine which signal segments belong to the target category based on their expertise. The computer system can accept expert input and perform the filtering accordingly.

[0227] For example, suppose we are analyzing the electrocardiogram (ECG) data of a heart patient, and the sensor signals to be processed are the heart rate changes during a heart attack. In step S111, the heart rate data has been divided into multiple signal segments, forming a set of signal segments to be processed. Now, the goal of the analysis is to identify abnormal heart rate events during a heart attack. Based on this goal, a feature-based screening method can be used to filter out signal segments with heart rate values ​​outside the normal range (e.g., below 60 beats per minute or above 120 beats per minute) as the target category signal segments.

[0228] In step S145, after filtering out the target category signal segments, the computer system finds signal segments related to these signal segments in the selected category of sensor signals and stores them in association. Correlation is typically assessed based on the similarity or association between signal segments, which can be achieved through various methods. The computer system can calculate the similarity (e.g., cosine similarity, Euclidean distance, etc.) between signal segments in the selected category of sensor signals and signal segments in the target category, and assess correlation based on this similarity. Higher similarity indicates a stronger correlation between the two signal segments. For time-series data, such as heart rate and blood pressure, the computer system can examine the temporal overlap between signal segments in the selected category of sensor signals and signal segments in the target category. If two signal segments have significant temporal overlap and similar trends, they are, for example, correlated. In addition to similarity and temporal synchronization, the computer system can also extract features of the signal segments (e.g., mean, variance, etc.) and compare the consistency of these features to assess correlation. Higher feature consistency indicates a greater likelihood of correlation between the two signal segments.

[0229] Computer systems can associate and store related signal segments by changing their visual attributes, such as color, thickness, and flashing frequency. For example, different colors can be used to distinguish different categories of signal segments, or bold lines can be used to highlight related signal segments.

[0230] In addition to visual emphasis, computer systems can also provide interactive display functions, allowing users to view details of relevant signal segments through zooming, panning, and other operations. For example, interactive chart libraries (such as D3.js, ECharts, etc.) can be used to create dynamic, interactive signal display interfaces.

[0231] To further help users understand the meaning and importance of relevant signal segments, the computer system can also generate corresponding text descriptions and explanations. These descriptions and explanations can be written based on the characteristics, patterns, or analysis results of the signal segments and presented on the user interface.

[0232] For example, let's continue using electrocardiogram (ECG) data from a heart attack patient. In step S144, target category signal segments have been filtered out, namely signal segments representing abnormal heart rate events during a heart attack. Now, it's necessary to find signal segments related to these target category signal segments within the selected category of sensor signals (such as heart rate signals, blood pressure signals, etc.) and store them in association.

[0233] In one implementation, step S140, which involves associating and storing signal segments from the selected category of sensor signals related to the sensor signal to be processed, further includes:

[0234] Step S146: Identify unselected category signal segments in the set of signal segments to be processed, excluding those of the target category.

[0235] Step S147: Retrieve signal segments related to signal segments of unselected categories from sensor signals of the selected category to obtain a set of candidate signal segments;

[0236] Step S148: Select M signal segments from the candidate signal segment set and store them together, where M is a positive integer greater than or equal to 1.

[0237] In step S146, the computer system identifies signal segments from the set of signal segments to be processed that do not belong to the target category. These signal segments are called unselected category signal segments. Unselected category signal segments may contain information related to the sensor signals to be processed but not explicitly classified into the target category, and therefore they also have potential analytical value.

[0238] If each signal segment in the set of signal segments to be processed has been assigned a corresponding label (such as "target category", "non-target category", etc.), the computer system can simply identify signal segments without a selected category by checking the label.

[0239] If signal segments lack explicit labels, computer systems can distinguish whether they belong to the target category by analyzing their characteristics (such as mean, variance, kurtosis, skewness, etc.). Signal segments whose characteristics significantly differ from those of the target category are considered unselected category segments. In some cases, computer systems can use machine learning models to predict whether each signal segment belongs to the target category. This model can be trained on historical data and learn how to distinguish between target category and unselected category signal segments.

[0240] For example, suppose we are analyzing the blood glucose data of a diabetic patient. The target category signal segments are those with abnormally high or low blood glucose levels. In the set of signal segments to be processed, in addition to the target category signal segments, there may be other signal segments with smaller fluctuations in blood glucose levels. These signal segments are considered as unselected category signal segments.

[0241] In step S147, after identifying the signal segments of the unselected categories, the computer system retrieves signal segments from the sensor signals of the selected categories that are related to these signal segments. Correlation is typically assessed based on the similarity or association between signal segments, resulting in a set of candidate signal segments that may contain similar information or patterns to the unselected category signal segments.

[0242] The computer system can calculate the similarity (such as cosine similarity, Euclidean distance, etc.) between each signal segment in the selected category sensor signal and the signal segment in the unselected category, and select the signal segment with a similarity exceeding a certain threshold as the candidate signal segment.

[0243] For time-series data, such as heart rate and blood pressure, computer systems can examine the temporal overlap between signal segments from selected categories of sensors and those from unselected categories. If two signal segments significantly overlap in time and exhibit similar trends, they are, for example, correlated. In addition to similarity and temporal synchronization, computer systems can also extract features of the signal segments (such as mean and variance) and compare the consistency of these features to assess correlation. Higher feature consistency indicates a greater likelihood of correlation between the two signal segments.

[0244] For example, let's continue with the example of blood glucose data from diabetic patients. Assume the selected categories of sensor signals include diet records, exercise records, etc. For each unselected category signal segment (i.e., a segment with relatively small fluctuations in blood glucose levels), the computer system can search the diet and exercise records for signal segments that are time-synchronized with and have similar characteristics to these signal segments as candidate signal segments. For instance, if a patient's diet and exercise habits are relatively stable within the time period corresponding to a certain unselected category signal segment, the computer system can find similar signal segments in these records as candidate signal segments.

[0245] In step S148, after obtaining the candidate signal segment set, the final step of the computer system is to select M signal segments for associated storage. The value of M can be determined according to actual needs and analysis objectives. The selection process may involve comprehensive consideration of various factors, such as the similarity of signal segments, time synchronization, and feature consistency.

[0246] The computer system can sort candidate signal segments based on indicators such as similarity, time synchronization, or feature consistency, and select the top M signal segments for associated storage.

[0247] For example, let's continue using blood glucose data from diabetic patients as an example. Assume we set M=3, meaning we want to associate and store three candidate signal segments. The computer system can sort these candidate signal segments based on their similarity and temporal synchronization, and select the top three segments for association and storage. These signal segments may represent the most relevant and important dietary or exercise records to the unselected signal segments, helping users to gain a deeper understanding of the potential causes and influencing factors of blood glucose fluctuations.

[0248] In one implementation, step S141 involves retrieving multiple signal segments related to the sensor signal to be processed from the selected category of sensor signals, including:

[0249] Step S1411: Obtain the first and second signal segments with adjacent signal coordinates from the sensor signal to be processed;

[0250] Step S1412: Retrieve the third signal segment associated with the first signal segment and the fourth signal segment associated with the second signal segment from the sensor signals of the selected category.

[0251] Step S142, determining continuity evaluation parameters between multiple signal segments by using the signal coordinates of multiple signal segments in the selected category of sensor signals, including:

[0252] Step S1421: Determine the continuity evaluation parameters between the third signal segment and the fourth signal segment based on the signal coordinates of the third signal segment in the selected category of sensor signals and the signal coordinates of the fourth signal segment in the selected category of sensor signals.

[0253] Step S143: When the continuity evaluation parameters between multiple signal segments reach a preset standard, the signal segments related to the sensor signal to be processed in the selected category of sensor signals are associated and stored, including:

[0254] Step S1431: When the continuity evaluation parameter between the third signal segment and the fourth signal segment reaches the preset standard, the signal segments related to the sensor signal to be processed are selected from the selected category of sensor signals, and the selected signal segments are associated and stored.

[0255] In step S1411, the computer system finds two signal segments with adjacent signal coordinates in the sensor signal to be processed, referred to as the first signal segment and the second signal segment, respectively. These two signal segments should be closely connected so that the corresponding signal segments can be found in the selected category of sensor signals.

[0256] If the sensor signal to be processed is time-series data (such as heart rate, blood pressure, etc.), the computer system can set a fixed-length time window to divide the signal into multiple signal segments. Then, signal segments within two adjacent time windows are selected as the first and second signal segments. In some cases, the sensor signal to be processed may contain specific event markers (such as outliers, peaks, etc.). The computer system can divide the signal into different segments based on these event markers and select adjacent event segments as the first and second signal segments. If the characteristics of the sensor signal to be processed (such as mean, variance, etc.) change significantly within a short period of time, the computer system can divide the signal into different segments based on these points of change and select adjacent characteristic change segments as the first and second signal segments.

[0257] For example, suppose the sensor signal to be processed is a continuous heart rate data segment spanning one hour. The computer system sets a 5-minute time window, dividing the heart rate data into 12 signal segments. Now, two adjacent signal segments (such as the 5th and 6th signal segments) are selected as the first and second signal segments.

[0258] In step S1412, after determining the first and second signal segments, the next step of the computer system is to retrieve the signal segments that are respectively related to these two signal segments from the selected category of sensor signals, referred to as the third and fourth signal segments. Correlation is typically assessed based on the similarity or association between signal segments.

[0259] The computer system can calculate the similarity (such as cosine similarity, Euclidean distance, etc.) between each signal segment in the selected category of sensor signals and the first and second signal segments, and select signal segments with similarity exceeding a certain threshold as the third and fourth signal segments.

[0260] For time-series data, a computer system can examine the degree of temporal overlap between signal segments in selected categories of sensor signals and first and second signal segments. If two signal segments have significant temporal overlap and similar trends, they are, for example, correlated.

[0261] In addition to similarity and time synchronization, computer systems can also extract features of signal segments (such as mean, variance, etc.) and compare the consistency of these features to assess correlation. The higher the feature consistency, the more likely the two signal segments are to be correlated.

[0262] For example, let's continue with heart rate data. Suppose the selected categories of sensor signals include blood pressure and step count signals. For the first signal segment (the 5th heart rate signal segment), the computer system might find a similar time period in the blood pressure signal as the third signal segment; for the second signal segment (the 6th heart rate signal segment), the computer system might find a similar time period in the step count signal as the fourth signal segment.

[0263] In step S1421, the computer system evaluates the continuity between the third and fourth signal segments to determine whether they are closely connected, thus forming a meaningful sequence of signal segments. The continuity evaluation parameter is a quantitative indicator used to measure the closeness and continuity between signal segments.

[0264] For time series data, a computer system can calculate the time interval between the third and fourth signal segments. A shorter time interval indicates better continuity between the two segments. In addition to the time interval, the computer system can also consider the reference value of the signal coordinates. For example, if both the third and fourth signal segments contain important feature changes or event markers, then their continuity may be more important. The computer system can combine the time interval and signal coordinate reference value, calculating continuity evaluation parameters through weighted summation or multiplication. The weights can be determined based on actual needs and analytical objectives.

[0265] For example, let's continue with heart rate data. Suppose the third signal segment is a time interval within the blood pressure signal, and the fourth signal segment is a time interval within the step count signal. The computer system can calculate the time interval between these two time segments and consider whether they both contain significant characteristic changes (such as a sudden rise in blood pressure, a sudden decrease in steps, etc.). Then, based on the time interval and signal coordinate reference, continuity assessment parameters are calculated.

[0266] In step S1431, the computer system determines whether the continuity evaluation parameters between the third and fourth signal segments meet the preset standard. If the standard is met, the signal segments related to the sensor signal to be processed in the selected category of sensor signals are associated and stored; otherwise, further filtering or adjustment of the signal segments may be necessary.

[0267] For example, continuing with heart rate data. Suppose the continuity evaluation parameters between the third and fourth signal segments meet a preset standard, indicating that these two signal segments are closely connected and form a meaningful signal segment sequence. At this point, the computer system can associate and store the signal segments (such as the third and fourth signal segments) related to the sensor signal to be processed from the selected category of sensor signals.

[0268] Through the above steps, the computer system can accurately retrieve multiple signal segments related to the sensor signal to be processed from the selected category of sensor signals, and determine which signal segments should be associated and stored by evaluating the continuity between these signal segments.

[0269] In one implementation, step S1421, based on the signal coordinates of the third signal segment in the selected category of sensor signals and the signal coordinates of the fourth signal segment in the selected category of sensor signals, determines the continuity evaluation parameters between the third signal segment and the fourth signal segment, including:

[0270] Step S14211: Determine the signal spacing between the third signal segment and the fourth signal segment based on the signal coordinates of the third signal segment in the selected category of sensor signals and the signal coordinates of the fourth signal segment in the selected category of sensor signals;

[0271] Step S14212: Determine the continuity evaluation parameters between the third and fourth signal segments based on the signal coordinate reference of the third and fourth signal segments and the spacing between the third and fourth signal segments.

[0272] In step S14211, the computer system calculates the signal spacing between the third and fourth signal segments. Signal spacing is a quantitative indicator used to measure the distance between two signal segments in time or space. For time-series data, signal spacing typically represents the interval between two signal segments on the time axis; for non-time-series data, signal spacing may represent the distance between two signal segments in the feature space.

[0273] If the third and fourth signal segments are two time periods in a time series data set, the computer system can calculate the difference between the start and end times of these two time periods to obtain the signal interval. For example, if the start time of the third signal segment is t1 and the end time is t2, and the start time of the fourth signal segment is t3 and the end time is t4, then the signal interval can be calculated as |t3 + t2| (assuming the two signal segments are continuous in time and do not overlap).

[0274] For non-time-series data, such as images and text, computer systems can first convert signal segments into feature vectors, and then use metrics such as Euclidean distance, Manhattan distance, or cosine similarity to calculate the signal spacing. For example, if the third and fourth signal segments are converted into feature vectors v1 and v2 respectively, then the signal spacing can be calculated as ‖v1 v2‖ (Euclidean distance).

[0275] For example, suppose we are analyzing the heart rate and blood glucose data of a diabetic patient. The third signal segment is a time period within the heart rate data, representing rapid changes in the patient's heart rate over a given time period; the fourth signal segment is a time period within the blood glucose data, representing corresponding changes in the patient's blood glucose level. If both signal segments are time-series data, the computer system can calculate the time interval between them as the signal gap. For example, if the third signal segment starts at 9:00 AM and ends at 9:10 AM, and the fourth signal segment starts at 9:05 AM and ends at 9:15 AM, then the signal gap is 5 minutes.

[0276] After determining the signal interval, the computer system considers the signal coordinate reference value of the third and fourth signal segments, as well as the signal interval itself, to comprehensively evaluate the continuity between these two signal segments. The computer system can assign a reference score to each signal segment, reflecting its importance in the analysis. The reference score can be determined based on the characteristics of the signal segments (such as mean, variance, kurtosis, etc.), patterns (such as periodicity, trends, etc.), or outliers (such as extreme values, abrupt changes, etc.). For example, if the third signal segment contains a significant heart rate peak, and the fourth signal segment also shows a corresponding blood glucose peak, then the reference scores for both signal segments can be high.

[0277] When calculating continuity evaluation parameters, the computer system can weight the signal spacing to take into account the reference value of the signal coordinates. Specifically, if both signal segments have high reference values, the continuity evaluation parameter may be high even if the signal spacing between them is slightly large; conversely, if the signal spacing is small but the reference value is low, the continuity evaluation parameter can be low.

[0278] The continuity assessment parameters can be calculated using the following formula:

[0279] ;

[0280] The reciprocal of the signal spacing is used to convert the spacing into a quantity positively correlated with continuity (because a smaller spacing indicates better continuity), while the sum of the reference scores reflects the importance of the two signal segments in the analysis. This formula can be adjusted according to actual needs, for example, by introducing a weighting factor to balance the influence of the signal spacing and the reference score.

[0281] For example, let's continue using heart rate and blood glucose data from diabetic patients. Assume the reference scores for the third and fourth signal segments are 8 and 9 (out of 10), respectively, with a signal interval of 5 minutes. Based on the formula above, the continuity assessment parameter can be calculated as follows:

[0282] ;

[0283] This value indicates good continuity between the third and fourth signal segments. Although the time interval between them is 5 minutes, the continuity evaluation parameter is still high because their reference scores are high (indicating that these two signal segments are very important in the analysis).

[0284] Through the above steps, the computer system can accurately calculate the signal spacing between the third and fourth signal segments, and determine the continuity evaluation parameters by comprehensively considering the reference value of the signal coordinates and the signal spacing. This not only helps to evaluate the tightness and continuity between the two signal segments, but also provides strong support for subsequent data analysis and pattern recognition.

[0285] This application provides a computer system, such as... Figure 2 As shown, the computer system 100 includes a processor 101 and a memory 103. The processor 101 and the memory 103 are connected, for example, via a bus 102. Optionally, the computer system 100 may also include a transceiver 104. It should be noted that in practical applications, the transceiver 104 is not limited to one type, and the structure of this computer system 100 does not constitute a limitation on the embodiments of this application.

[0286] This application provides a computer system, which includes: one or more processors; a memory; and one or more computer programs, wherein the one or more computer programs are stored in the memory and configured to be executed by the one or more processors, and when the one or more programs are executed by the processors, the above-mentioned machine learning-based medical monitoring sensor data association storage method is implemented.

[0287] The above description is only a partial embodiment of this application. It should be noted that for those skilled in the art, several improvements and modifications can be made without departing from the principle of this application, and these improvements and modifications should also be considered within the scope of protection of this application.

Claims

1. A method for storing medical monitoring sensor data based on machine learning, characterized in that, The method comprises: acquiring a target sensor signal related to a to-be-processed sensor signal, the target sensor signal covering different categories of sensor signals; acquiring, according to an influence factor of the different categories of sensor signals, a correlation degree between the different categories of sensor signals covered by the target sensor signal and the to-be-processed sensor signal; screening a selected category of sensor signals from the different categories of sensor signals according to the correlation degree between the different categories of sensor signals and the to-be-processed sensor signal; associatively storing a signal segment related to the to-be-processed sensor signal in the selected category of sensor signals; wherein the associatively storing the signal segment related to the to-be-processed sensor signal in the selected category of sensor signals comprises: retrieving a plurality of signal segments related to the to-be-processed sensor signal from the selected category of sensor signals; determining a continuity evaluation parameter between the plurality of signal segments through signal coordinates of the plurality of signal segments in the selected category of sensor signals; when the continuity evaluation parameter between the plurality of signal segments reaches a preset standard, associatively storing the signal segment related to the to-be-processed sensor signal in the selected category of sensor signals; the acquiring, according to the influence factor of the different categories of sensor signals, the correlation degree between the different categories of sensor signals covered by the target sensor signal and the to-be-processed sensor signal comprises: performing signal segment splitting on the to-be-processed sensor signal to obtain a to-be-processed signal segment; acquiring a basic correlation degree between the different categories of sensor signals and the to-be-processed sensor signal according to the to-be-processed signal segment; acquiring the correlation degree between the different categories of sensor signals and the to-be-processed sensor signal according to the basic correlation degree between the different categories of sensor signals and the to-be-processed sensor signal and the influence factor of the different categories of sensor signals.

2. The method of claim 1, wherein, the acquiring the basic correlation degree between the different categories of sensor signals and the to-be-processed sensor signal according to the to-be-processed signal segment comprises: retrieving a target signal segment related to the to-be-processed signal segment from the different categories of sensor signals; determining the basic correlation degree between the different categories of sensor signals and the to-be-processed sensor signal according to a priority of the to-be-processed signal segment and a priority of the target signal segment related to the to-be-processed signal segment in the different categories of sensor signals.

3. The method of claim 1, wherein: the target sensor signal is a plurality of target sensor signals, and the method further comprises: determining an influence factor of each target sensor signal according to a correlation mechanism between the to-be-processed sensor signal and each target sensor signal; acquiring the correlation degree between each target sensor signal and the to-be-processed sensor signal according to the influence factor of each target sensor signal and the correlation degree between the different categories of sensor signals covered by each target sensor signal and the to-be-processed sensor signal. According to the correlation between each target sensor signal and the to-be-processed sensor signal, the plurality of target sensor signals are sorted; The target sensor signal related to the to-be-processed sensor signal comprises: The to-be-processed sensor signal is segmented into signal segments to obtain a to-be-processed signal segment set; The signal segments in the to-be-processed signal segment set are aggregated to obtain a plurality of signal segment clusters; Respectively obtain the target sensor signal related to the plurality of signal segment clusters, and one signal segment cluster corresponds to one correlation mechanism.

4. The method of claim 1, wherein, The selected category of sensor signals is filtered from the different categories of sensor signals according to the correlation between the different categories of sensor signals and the to-be-processed sensor signal, comprising: Filtering target category sensor signals and unselected category sensor signals other than the target category from the different categories of sensor signals; According to the correlation between the different categories of sensor signals and the to-be-processed sensor signal, the to-be-determined category of sensor signals is filtered from the unselected category of sensor signals; The to-be-determined category of sensor signals and the target category of sensor signals are used as the selected category of sensor signals; The selected category of sensor signals related to the to-be-processed sensor signal is associatedly stored, and further comprising: Filtering target category signal segments from the to-be-processed signal segment set, wherein the to-be-processed signal segment set comprises signal segments obtained by segmenting the to-be-processed sensor signal into signal segments; The selected category of sensor signals related to the target category signal segment is associatedly stored.

5. The method of claim 4, wherein, The to-be-determined category of sensor signals is filtered from the unselected category of sensor signals according to the correlation between the different categories of sensor signals and the to-be-processed sensor signal, comprising: According to the correlation between the different categories of sensor signals and the to-be-processed sensor signal, the original to-be-determined category of sensor signals with a correlation greater than or equal to a preset threshold is selected from the unselected category of sensor signals; The first N sensor signals with the largest correlation in the original to-be-determined category of sensor signals with a correlation greater than or equal to a preset threshold are used as the to-be-determined category of sensor signals, wherein N is a positive integer greater than or equal to 1.

6. The method of claim 5, wherein, The selected category of sensor signals related to the to-be-processed sensor signal is associatedly stored, further comprising: Determine the unselected category signal segments in the to-be-processed signal segment set other than the target category signal segments; The selected category of sensor signals related to the unselected category signal segments is used as a candidate signal segment set; M signal segments are filtered from the candidate signal segment set for associated storage, wherein M is a positive integer greater than or equal to 1.

7. A computer system, characterized by Comprise: One or more processors; Memory; One or more computer programs; wherein the one or more computer programs are stored in the memory and configured to be executed by the one or more processors, and the one or more computer programs, when executed by the processors, implement the method according to any one of claims 1-6.

Images