Traditional Chinese medicine meridian data dynamic analysis method based on edge calculation

Abstract

The invention belongs to the technical field of meridian analysis, and provides a traditional Chinese medicine meridian data dynamic analysis method based on edge calculation. The method comprises the following steps: setting edge nodes, dynamically collecting acupoint electric signals in combination with a midnight-noon ebb-flow rule, calculating a splitting sensitive coefficient of adjacent nodes, and constructing a splitting sensitive group; determining a maximum connected domain through percolation critical flow analysis, and performing hour rhythm analysis to judge whether hour-rhythm splitting exists or not; if the dissection exists, analyzing the relevance between the topological characteristics of the topological weak nodes and the rhythm dissection characteristics, and judging whether the temporal qi and blood streamer is hindered by the topological weak nodes or not; and optimizing the layout of the monitoring nodes through stream-oriented symmetric configuration based on an analysis result. According to the method, edge calculation and the traditional Chinese medicine midnight-noon ebb-flow theory are combined, dynamic analysis of meridian data and optimization of monitoring layout are achieved, and pertinence and adaptability of meridian data monitoring are improved.

Description

Technical Field

[0001] This invention belongs to the field of meridian analysis technology, specifically a dynamic analysis method for TCM meridian data based on edge computing. Background Technology

[0002] As a vital channel for the circulation of Qi and blood in the human body, the meridians in Traditional Chinese Medicine (TCM) offer crucial insights into their dynamic changes. Accurate analysis of these dynamic patterns is essential for verifying TCM theories, assessing health status, and optimizing related interventions. However, existing TCM meridian data analysis technologies still have significant limitations in adapting to the characteristics of TCM theories, quantifying correlation strength, tracing the causes of fragmentation, and optimizing monitoring layouts, making it difficult to meet the demands of dynamic analysis.

[0003] At the level of global network and circadian rhythm analysis, existing technologies mostly focus on local signals of discrete nodes, lacking a global perspective that transforms node data into a node-edge network model, making it difficult to capture the overall correlation of the meridian system. Furthermore, the analysis of circadian rhythm lacks integration with complex network theory, making it difficult to determine whether there is a circadian rhythm split in the core meridians that is not coordinated when it should be vigorous by quantifying the difference in correlation between double and single double segments, resulting in a strong subjectivity in the judgment of the split.

[0004] In terms of monitoring layout optimization, the node layout of existing meridian monitoring equipment is mostly fixed and does not dynamically adapt to the path characteristics and weak areas of qi and blood flow based on the meridian data analysis results. Furthermore, it lacks the integration of edge computing technology, making it difficult to achieve a closed loop from data analysis to monitoring optimization, thus limiting the pertinence and adaptability of meridian data monitoring.

[0005] Therefore, this invention provides a method for dynamic analysis of TCM meridian data based on edge computing. Summary of the Invention

[0006] In order to overcome the shortcomings of the prior art, at least one technical problem raised in the background art is solved.

[0007] The technical solution adopted by this invention to solve its technical problem is: a method for dynamic analysis of TCM meridian data based on edge computing, comprising: A dynamic analysis method for TCM meridian data based on edge computing includes the following steps: Edge nodes for collecting data from TCM meridians are defined. Based on the meridian flow time sequence, the acupoint electrical signals of the edge nodes are dynamically collected to perform a severance sensitivity analysis, obtain the severance sensitivity coefficients of adjacent edge nodes, and construct a severance sensitivity group.

[0008] Percolation critical flow analysis was performed on the segmentation sensitive group to determine the maximum connected domain. Chronological rhythm analysis was performed on the maximum connected domain to obtain the mean correlation degree of different overlapping segments, and it was determined whether there was chronological-rhythmic segmentation in the maximum connected domain.

[0009] If there is a time-rhythm break, perform topological break analysis on the largest connected component, extract the topological break features and rhythm break features of the topologically weak nodes, and perform correlation analysis to determine whether the topologically weak nodes are hindering the flow of Qi and blood during the time.

[0010] If a weak node in the topology obstructs the flow of Qi and blood during the time of birth, the associated transition region corresponding to the weak node is extracted. By performing a symmetrical analysis of the flow guidance configuration of the associated transition region, the layout ratio between the auxiliary monitoring node and the original node is determined.

[0011] Furthermore, the method for performing the aforementioned split sensitivity analysis is as follows: Obtain the meridian correlation degree of the sub-segment and divide the sub-segment into active period sub-segment and stable period sub-segment; The mean of the meridian correlation degree of all active period segments is extracted as the mean correlation degree; Define a split sensitivity mapping relationship to map the mean correlation degree to the split sensitivity coefficient.

[0012] Furthermore, the method for obtaining the meridian correlation degree of the sub-segment is as follows: Based on the acupoint electrical signals of edge nodes, the rate of change of the acupoint electrical signal intensity of each edge node at adjacent monitoring times is calculated as the intensity change rate; Obtain the intensity change rate of adjacent edge nodes A and B, and construct two sets of intensity change sequences, A and B; The intensity change sequence is segmented to obtain intensity change sequences of different sub-segments; Calculate the Pearson correlation coefficients of the intensity change sequences of the corresponding sub-segments for groups A and B, and use them as the meridian correlation coefficients of the sub-segments; Coherence analysis was performed on the intensity change sequences of the two groups in corresponding sub-segments to obtain the meridian coherence coefficient; The meridian correlation coefficient and meridian coherence coefficient are standardized, and the standardized meridian correlation coefficient and meridian coherence coefficient are summed to obtain the meridian correlation degree of the sub-segment.

[0013] Furthermore, the method for determining whether the largest connected component has a time-rhythm break is as follows: Obtain the meridian affiliation of network nodes within the largest connected component, and calculate the frequency of occurrence of different meridians based on the meridian affiliation; The meridians that appear most frequently are identified as the core meridians. Based on the core meridians and combined with the meridian flow pattern, the active and inactive times of the core meridians are determined. Determine whether there is any overlap between the active time and active period sub-segments. If there is overlap, extract the double-overlapping sub-segments of the active time and active period sub-segments and obtain the mean correlation degree of the double-overlapping sub-segments. Determine whether there is any overlap between inactive time periods and active period segments. If there is overlap, extract the single overlapping segments between inactive time periods and active period segments, and extract the mean correlation of the single overlapping segments. The time-meridian synergy criterion is set. If the mean correlation degree of double-syndrome segments and the mean correlation degree of single-syndrome segments satisfy the time-meridian synergy criterion, then it conforms to the time-meridian synergy law.

[0014] Furthermore, the method for obtaining the maximum connected component is as follows: Obtain all edge nodes as network nodes and construct a network node set; In the set of network nodes, edges are created based on adjacent edge nodes, and the weight of the edge is the split sensitivity coefficient. Using the meridian correlation degree as the edge connectivity probability of network nodes, a percolation critical analysis model is constructed to determine the maximum connected region size of network nodes, and the maximum connected region is determined based on the maximum connected region size.

[0015] Furthermore, the percolation critical analysis model is constructed as follows: Generate networks with different connectivity states based on edge connectivity probability, and determine whether to retain all edges; Based on repeated simulations of the grid corresponding to each edge, the maximum connected component size obtained from the repeated simulations is extracted. After M repeated simulations, obtain the largest connected subset in the grid, and extract the number of network nodes in the largest connected subset and the total number of nodes; The ratio of the number of network nodes in the largest connected subset to the total number of nodes is used as the size of the largest connected component. If the size of the largest connected component is higher than the preset connected component size threshold, then the network formed by the largest connected subset is taken as the largest connected component.

[0016] Furthermore, the method to determine whether the flow of Qi and blood during the time of birth is obstructed by weak topological nodes is as follows: Extract the path overlap rate of topological fragmentation features for spatial correlation verification; The correlation between the connection strength values ​​of topological fragmentation features and the abnormally poor correlation between rhythmic fragmentation features is used to verify the strength correlation. Temporal correlation verification was performed on the proportion of dual zygotic segments with topologically weak nodes and rhythmic fragmentation. If the topological fragmentation feature blocks the flow of Qi and blood during the time of birth, and there is a spatial correlation, a negative correlation between the connection strength value and the correlation anomaly, and a temporal correlation, then it is determined that the weak topological node blocks the flow of Qi and blood during the time of birth.

[0017] Furthermore, the method for extracting the topological segmentation features is as follows: Obtain the initial screening coefficient and order value of the edge nodes within the largest connected component, and determine the weak nodes in the topology based on the initial screening coefficient and order value; Topological fragmentation features are obtained by extracting fragmented topological features from weak nodes. The method for extracting fragmented topological features is as follows: The degree of overlap between weak nodes in the topology and the flow path of Qi and blood in the core meridians is analyzed to obtain the node flow value; The average value of the cut sensitivity coefficient of the edges associated with weak nodes in the topology is collected and used as the connection strength value. Connectivity strength and node streamer values ​​are used as topology splitting features.

[0018] Furthermore, the method for determining the layout ratio of the auxiliary monitoring nodes to the original nodes is as follows: Obtain the average cut sensitivity coefficient and association anomaly difference of the associated edges of weak nodes in the topology, and sum the average cut sensitivity coefficient and association anomaly difference to obtain the weak proportion coefficient; Different levels are defined based on the weak point ratio coefficient, and the layout ratio of auxiliary monitoring nodes to original nodes is determined based on the different levels.

[0019] Furthermore, the method for determining the auxiliary monitoring node is as follows: A meridian coordinate system was constructed with the suprasternal notch as the origin to obtain the coordinates of topologically weak nodes; The core meridian flow is digitally processed and transformed into a continuous path curve on the meridian coordinate system; Extract the streamer direction vector from the continuous path curve; A symmetric configuration rule is established based on the stream direction vector to determine the auxiliary monitoring nodes.

[0020] The beneficial effects of this invention are as follows: 1. By selecting key locations such as branch points and intersection points of the twelve regular meridians in Traditional Chinese Medicine as edge nodes, and combining the time-meridian correspondence law of the meridian flow, the electrical signals of acupoints are dynamically collected. This is conducive to the high degree of consistency between node selection and TCM meridian theory. The dynamic characteristics of Qi and blood flow are captured through a time-matched acquisition scheme. By calculating quantitative indicators such as intensity change rate, Pearson correlation coefficient, and coherence coefficient of characteristic frequency, the abstract meridian association is transformed into a calculable separation sensitivity coefficient, realizing the objective quantification of association strength. This provides standardized basic data for subsequent network analysis and separation judgment, which helps to reduce the problems of blind data collection and ambiguity in association description in traditional meridian analysis.

[0021] 2. By transforming the sensitive group of ruptures into a network model through percolation critical flow analysis, and using Monte Carlo simulation to dynamically adjust the edge connectivity threshold to determine the maximum connected domain, the transformation from discrete node data to the overall network structure is realized, which is conducive to capturing the global correlation of the meridian system. Combined with the meridian flow law, the maximum connected domain is analyzed by time rhythm. By extracting the mean correlation degree of double-syndrome segments and single-syndrome segments, the time-rhythm rupture of the core meridians that should be vigorous is quantitatively judged. The integration of complex network analysis with the time theory of traditional Chinese medicine is conducive to making up for the deficiency of simple network analysis that ignores the temporal characteristics, reducing the problem of lack of quantitative verification in traditional time theory, and improving the scientificity of rupture judgment.

[0022] 3. When time-rhythm disruption exists, topologically weak nodes are located by calculating the initial screening coefficient and betweenness centrality. Their spatial distribution topological features are extracted, and a multi-dimensional correlation analysis is performed with rhythmic features (spatial, intensity, and temporal dimensions) to clarify the causal relationship between topological weak points and rhythm disruption. This facilitates the discovery of the correlation between meridian structure and function, provides a clear basis for the causes of disruption through quantitative correlation verification, reduces subjective speculation about the causes of disruption, and determines the direction for subsequent monitoring layout optimization.

[0023] 4. Next, by extracting the flow direction vector, a symmetrical configuration rule is established. Combined with the weak point ratio coefficient from the previous analysis, the layout ratio of auxiliary monitoring nodes to the original nodes is determined, ultimately optimizing the sampling points of the meridian instrument. This achieves dynamic matching between the monitoring layout and the flow pattern of Qi and blood. The symmetrical quantitative configuration guided by the flow helps reduce the randomness of traditional monitoring point layouts. Furthermore, the quantitative analysis based on previous data ensures the targeted nature of the layout optimization, enabling the monitoring equipment to capture changes in Qi and blood in weak areas, thus improving the adaptability and effectiveness of meridian data monitoring. Attached Figure Description

[0024] The invention will now be further described with reference to the accompanying drawings.

[0025] Figure 1 This is a flowchart of a method for dynamic analysis of TCM meridian data based on edge computing, according to the present invention.

[0026] Figure 2 This is a flowchart of the edge connectivity determination method of the present invention.

[0027] Figure 3 This is a module architecture diagram of a dynamic analysis system for TCM meridian data based on edge computing, as described in an embodiment of the present invention. Detailed Implementation

[0028] To make the technical means, creative features, objectives and effects of this invention easier to understand, the invention will be further described below in conjunction with specific embodiments.

[0029] Example 1 Please see Figure 1 As shown in the embodiment of the present invention, a method for dynamic analysis of TCM meridian data based on edge computing includes the following steps: S1. Set the edge nodes for TCM meridian data collection, and perform a severing sensitivity analysis on the acupoint electrical signals of the edge nodes based on the meridian flow time sequence to obtain the severing sensitivity coefficients of adjacent edge nodes and construct a severing sensitivity group.

[0030] The method for setting the edge nodes of TCM meridian data collection is as follows: Preferably, based on the pathways of the twelve regular meridians and eight extraordinary meridians in Traditional Chinese Medicine, meridian branch points, intersections of different meridians, and meridian start and end points are selected as edge nodes for TCM meridian data collection. Based on the correspondence between the time of day and the meridians in the Meridian Flow Theory, acupoint electrical signals are collected at the edge nodes of the monitoring time.

[0031] The method for obtaining the cut sensitivity coefficient of adjacent edge nodes through cut sensitivity analysis is as follows: Based on the acupoint electrical signals of edge nodes, the rate of change of the acupoint electrical signal intensity of each edge node at adjacent monitoring times is calculated as the intensity change rate; Obtain the intensity change rate of adjacent edge nodes A and B, and construct two sets of intensity change sequences, A and B; The intensity change sequence is segmented to obtain intensity change sequences of different sub-segments.

[0032] It should be explained that when segmenting, it is necessary to ensure that the time range of each sub-segment corresponds one-to-one with the individual time periods of the meridian flow, and to reduce the segmentation across time periods in order to ensure the temporal matching of subsequent time period rhythm analysis.

[0033] Calculate the Pearson correlation coefficients of the intensity change sequences of the corresponding sub-segments for groups A and B, and use them as the meridian correlation coefficients of the sub-segments.

[0034] Coherence analysis was performed on the intensity change sequences of the two groups in corresponding sub-segments to obtain the meridian coherence coefficient.

[0035] Preferably, frequency domain decomposition is performed on the two sets of intensity change sequences in corresponding sub-segments to extract the characteristic frequencies corresponding to the meridian flow time. .

[0036] Those skilled in the art will understand that when performing frequency domain decomposition on the intensity change sequences of two corresponding segments, Fourier transform is used to convert the intensity change signal in the time domain into a frequency domain signal to obtain the energy distribution of different frequency components. Based on this, and combining the corresponding rhythms of the twelve two-hour periods and the waxing and waning of Qi and blood in the twelve meridians in the Meridian Flow Theory, the dominant frequency (the frequency with the most concentrated energy or that matches the theoretical rhythm) that matches the Qi and blood flow characteristics of the meridian corresponding to the current analysis time is extracted from the power spectrum in the frequency domain, and used as the characteristic frequency corresponding to that Meridian Flow Time. .

[0037] Through the formula: Obtain feature frequencies The meridian coherence coefficient below .

[0038] in, The squared modulus of the cross-power spectrum. For the signal at node A at the characteristic frequency The self-power spectral density estimation at point A reflects the energy distribution of node A at frequency.

[0039] Node B signal at characteristic frequency Estimation of the self-power spectral density at the location; Signals at nodes A and B at characteristic frequencies The cross-power spectral density estimation at a certain frequency reflects the cross-power spectral density of the two signals at their characteristic frequencies. The co-energies include amplitude and phase relationships.

[0040] The meridian correlation coefficients are standardized, and the standardized meridian correlation coefficients and meridian coherence coefficients are then made consistent. Finally, the consistent meridian correlation coefficients and meridian coherence coefficients are summed to obtain the meridian correlation degree of the sub-segment.

[0041] Based on the meridian correlation of sub-segments, sub-segments are divided into active period sub-segments and stable period sub-segments.

[0042] It is understandable that the meridian correlation coefficients are normalized by min-max normalization to map them to the [0,1] interval in order to eliminate the numerical magnitude deviation caused by the difference in signal strength at different nodes.

[0043] Since the meridian correlation coefficient, which reflects the correlation in the time domain, and the meridian coherence coefficient, which reflects the synergy in the frequency domain, have different dimensions and distribution characteristics, they need to be standardized. The standardization process converts both into positive indicators and scales them to the same numerical range with equal weights to achieve the additivity between the indicators and reduce the influence of a single indicator on the results due to its excessively large numerical range.

[0044] The meridian correlation degree obtained after summing is used as the threshold value by taking the mean of the correlation degree of all sub-segments plus 1 / 2 standard deviation. A correlation degree higher than the threshold value reflects an active period sub-segment with strong Qi and blood synergy, while a correlation degree lower than the threshold value reflects a stable period sub-segment with weak Qi and blood synergy.

[0045] The classification is based on the theory of the waxing and waning of Qi and blood in traditional Chinese medicine, which provides a classification basis for identifying the fragmented characteristics of inactivity when they should be active in subsequent circadian rhythm analysis.

[0046] The mean of the meridian correlation degree of all active period segments is extracted as the mean correlation degree.

[0047] Define a split sensitivity mapping relationship to map the mean correlation degree to the split sensitivity coefficient.

[0048] Those skilled in the art will understand that when setting the separation-sensitive mapping relationship, based on the logic that "the higher the mean correlation, the stronger the synergy between nodes and the more stable the correlation, and the lower the risk of separation; the lower the mean correlation, the worse the synergy and the more fragile the correlation, and the higher the risk of separation", the reverse mapping rule is adopted: the mean correlation of the active period segment is transformed into a separation sensitivity coefficient through a function, for example, through a linear mapping of separation sensitivity coefficient = 1 - mean correlation.

[0049] The purpose of mapping the mean correlation degree to the split sensitivity coefficient is as follows: Objective 1: To transform the mean correlation degree, which reflects the strength of the synergy between Qi and Blood in active segments, into an indicator that can directly quantify the vulnerability of the relationship between adjacent nodes. This would allow the risk of the relationship being broken to be reflected through specific numerical values, thus solving the problem that abstract synergy is difficult to use directly for subsequent analysis.

[0050] Objective 2: To provide a standardized basis for setting edge weights when constructing node-edge network models, so that the severance sensitivity coefficient can be used as the edge weight in percolation critical flow analysis, and to assist in the analysis of the global correlation and severance characteristics of the meridian system.

[0051] Obtain the cut sensitivity coefficients of N groups of adjacent edge nodes and construct the cut sensitivity group.

[0052] S2. Perform percolation critical flow analysis on the segmentation sensitive group to determine the maximum connected domain, perform chrono-rhythmic analysis on the maximum connected domain to obtain the mean correlation degree of different overlapping segments, and determine whether there is chrono-rhythmic segmentation in the maximum connected domain.

[0053] The method for performing percolation critical flow analysis on the fracture-sensitive group is as follows: Obtain all edge nodes as network nodes and construct a network node set; In the set of network nodes, edges are created based on adjacent edge nodes, and the weight of the edge is the split sensitivity coefficient. Using the meridian correlation degree as the edge connectivity probability of network nodes, a percolation critical analysis model is constructed through Monte Carlo simulation algorithm to determine the maximum connected region size of network nodes, and the maximum connected region is determined based on the maximum connected region size.

[0054] Preferably, the method for constructing the percolation critical analysis model is as follows: S201. Generate networks with different connectivity states based on edge connectivity probability, and determine whether to retain all edges.

[0055] It needs to be explained that obtaining the edge connectivity threshold... , The edge connectivity threshold is dynamically adjusted in steps of 0.01 to determine whether all edges should be retained.

[0056] like Figure 2 As shown, if the edge connectivity probability is higher than the dynamically adjusted edge connectivity threshold, the edge is retained, i.e., it is determined to be edge connected. Conversely, removing the edge is considered a cut.

[0057] S202. Based on the mesh corresponding to each edge, repeat the simulation and extract the maximum connected component size obtained from the repeated simulation.

[0058] It needs to be explained that, based on the grid corresponding to each edge, the simulation is repeated M times to obtain the largest connected subset in the grid after the M repeated simulations, and the number of network nodes in the largest connected subset and the total number of nodes are extracted.

[0059] The ratio of the number of network nodes in the largest connected subset to the total number of nodes is used as the size of the largest connected component.

[0060] If the size of the largest connected component is higher than the preset connected component size threshold, then the network formed by the largest connected subset is taken as the largest connected component.

[0061] The method for obtaining the mean correlation degree of different overlapping sub-segments by performing circadian rhythm analysis on the largest connected component is as follows: The method for conducting circadian rhythm analysis is as follows: Obtain the meridian affiliation of network nodes within the largest connected component, and calculate the frequency of occurrence of different meridians based on the meridian affiliation; The meridians that appear most frequently are identified as the core meridians. Based on these core meridians and the meridian flow pattern, the active and inactive times of these core meridians are determined.

[0062] It should be noted that in the network nodes of the largest connected domain, the meridians to which each node belongs (such as the Lung Meridian, Heart Meridian, etc.) are counted. The meridian that appears most frequently is the core meridian, because the core meridian has the highest proportion in the current connected domain and can better reflect the main operating characteristics of the overall meridian system.

[0063] Based on the core meridians and combined with the meridian flow pattern, active and inactive times are determined. The meridian flow pattern is the rule in traditional Chinese medicine that the twelve meridians correspond to the twelve two-hour periods, and the qi and blood alternate between waxing and waning at specific times (e.g., the lung meridian corresponds to the Yin hour, the large intestine meridian corresponds to the Mao hour, etc., and each meridian has the most vigorous qi and blood at the corresponding time). Therefore, by combining the affiliation of the core meridians, the time when its qi and blood are most vigorous can be determined according to this rule as the active time, and the other times are inactive times. This provides a temporal basis for subsequent analysis of whether the meridian should show stronger qi and blood synergy during the active time.

[0064] For example, if the core meridian is the Lung Meridian, the Yin hour (3-5 AM) is determined to be the active hour based on the meridian flow pattern.

[0065] Determine whether there is overlap between the active time segment and the active period segment. If there is overlap, extract the double-overlapping segments of the active time segment and the active period segment, and obtain the mean correlation degree Rz of the double-overlapping segments.

[0066] Determine whether there is overlap between inactive time periods and active period segments. If there is overlap, extract the single overlapping segments between inactive time periods and active period segments, and extract the mean correlation degree Rf of the single overlapping segments.

[0067] The method for determining whether the largest connected component has a time-rhythm break is as follows: If the mean correlation coefficient Rz and the mean correlation coefficient Rf satisfy If so, it can be determined that the synergy of Qi and blood in the core meridians should be significantly enhanced during their peak hours, which conforms to the time-meridian synergy rule.

[0068] in, To establish the time-meridian synergy criterion, This represents the minimum significant difference, meaning that the difference between the two is considered statistically interpretable only when the correlation between the mean of active hours and that of inactive hours exceeds this difference.

[0069] As will be understood by those skilled in the art, the minimum significance difference By simulating the difference distribution between active and inactive times using Bootstrap resampling when there is no real difference, the high-confidence quantile of this distribution (e.g., the 95th percentile, representing the largest spurious difference under random fluctuations) is extracted as... .

[0070] By deriving the Bootstrap resampling algorithm, The inherent variation level of the meridian correlation is closely aligned with the data, which filters out noise interference while retaining the real difference signals of Qi and blood gain during active periods, thus supporting the statistical rigor and physiological scenario adaptability of the conclusions.

[0071] Conversely, this does not conform to the time-meridian synergy rule, meaning that when the core meridians should be at their peak, the synergy of Qi and blood is not enhanced, posing a potential risk of rhythmic disruption.

[0072] Those skilled in the art will understand that if the time-meridian synergy pattern is followed, it indicates that the synergy of Qi and blood in the largest connected domain is consistent with the time-time pattern in traditional Chinese medicine, and the risk of disruption mainly comes from the weak links in the topological structure.

[0073] If the time-meridian synergy pattern is not followed, it indicates that there is a time-rhythm disconnect in the largest connected domain. The core meridians do not show the expected synergistic enhancement during peak hours, and the risk of disconnection includes rhythmic issues.

[0074] Understandably, the purpose of determining whether the largest connected region conforms to the time-meridian synergy pattern is: One function is to transform the qualitative description of the waxing and waning of Qi and blood in the meridians during the time of day in the theory of meridian flow in traditional Chinese medicine into a quantifiable and verifiable standard based on the difference in correlation, thereby reducing the limitation of traditional analysis that only makes empirical judgments on the patterns of time.

[0075] Secondly, it defines a targeted scope for subsequent analysis, allowing for in-depth investigation of the causes of disconnection only in connected components that do not conform to the rules, reducing redundancy in indiscriminate analysis of all connected components, and improving analysis efficiency.

[0076] Thirdly, it provides a temporal reference for optimizing the layout of monitoring nodes, enabling the symmetrical configuration of subsequent flow-guided systems to adapt to spatial paths and conform to the time-based activity characteristics of core meridians. This allows the layout optimization to take into account both temporal and spatial dimensions, reflecting an improvement from static layout to spatiotemporal collaborative optimization.

[0077] Example 2 like Figure 1 As shown in the embodiment of the present invention, a method for dynamic analysis of TCM meridian data based on edge computing further includes the following steps: S3. If there is a time-rhythm break, perform topological break analysis on the largest connected region, extract the topological break features of the topological weak nodes, perform correlation analysis on the rhythm break features and the topological break features, and determine whether the topological weak nodes are hindering the flow of Qi and blood during the time.

[0078] The method for performing topological splitting analysis on the largest connected component is as follows: S301. Obtain the initial screening coefficient and order value of the edge nodes in the largest connected domain, and determine the weak nodes in the topology based on the initial screening coefficient and order value.

[0079] Preferably, the mean and standard deviation of the split sensitivity coefficients of all adjacent edge nodes in the split sensitivity group are obtained, and the mean and standard deviation are summed to obtain the weak initial screening coefficient.

[0080] It should be explained that the weak initial screening coefficient combines the overall average level and individual differences of the split sensitivity coefficients of all adjacent edge nodes in the split sensitivity group to form a quantitative threshold for initially judging whether the vulnerability of an edge requires special attention. The weak initial screening coefficient reflects the upper limit of the normal fluctuation of the vulnerability of adjacent nodes. When the weight (split sensitivity coefficient) of an edge is higher than this coefficient, it indicates that the vulnerability of the edge significantly exceeds the overall average level and the normal fluctuation range, and it can be initially marked as a weak initial screening edge, providing a basis for subsequent screening to locate topologically weak nodes in conjunction with betweenness centrality.

[0081] If the weight of an edge within the largest connected component is higher than the weak initial screening coefficient, then the edge is marked as a weak initial screening edge.

[0082] Obtain the order values ​​of betweenness centrality of all edge nodes in the largest connected component. Based on the order values ​​and the weak initial screening edge pairs, establish a topological weakness criterion for the edge nodes to determine the topologically weak nodes in the largest connected component.

[0083] It is understandable that the higher the betweenness centrality, the more important the node is as a key hub for the flow of energy in the network, such as the intersection of meridians. If the edge node of the largest connected domain is connected to ≥2 weak initial screening edges at the same time, and the order value of betweenness centrality is 0.2, which is lower than the preset 0.3, then the edge node is marked as a topologically weak node.

[0084] S302. Extract the topological features of weak nodes to obtain topological fragmentation features.

[0085] Preferably, the degree of overlap between weak nodes in the topology and the flow path of Qi and blood in the core meridian is analyzed to obtain the node flow value.

[0086] The average value of the cut sensitivity coefficient of the edges associated with weak nodes in the topology is collected and used as the connection strength value. Connectivity strength and node streamer values ​​are used as topology splitting features.

[0087] It is understandable that the node stream flow value corresponds to the spatial distribution characteristics, and the connection strength value corresponds to the connection strength characteristics.

[0088] The difference between the mean correlation degree Rz of the doublet segment and the mean correlation degree Rf of the single doublet segment is obtained to obtain the correlation anomaly.

[0089] Obtain the percentage of double-overlapping sub-segments that overlap with the active period sub-segment within the active time period; The correlation anomaly and the proportion of double zygotic segments were used as rhythmic features; Understandably, the smaller the correlation anomaly, the more severe the rhythm disruption; the lower the proportion of double zygotic segments, the smaller the chance of Qi and blood coordination during active time.

[0090] Among them, the method of performing correlation analysis on topological fragmentation features to determine whether the flow of Qi and blood during the time is hindered by weak topological nodes is as follows: S311. Extract the path overlap rate of topological fragmentation features for spatial association verification.

[0091] For example, the method for performing spatial correlation verification is as follows: if the path overlap rate of the topological weak nodes is >60%, that is, concentrated on the Qi and blood flow path of the core meridian, and the location of the topological weak nodes is the transition point of the core meridian from the inactive area to the active area, such as the turning point of the Lung Meridian from the chest to the arm, then the spatial correlation of the Qi and blood flow when the topological dissection feature blocks the Qi and blood flow is strong, and it is determined that there is a spatial correlation when the topological dissection feature blocks the Qi and blood flow.

[0092] Conversely, if weak nodes are randomly distributed, i.e., the path overlap rate is less than 30%, then the spatial correlation is weak.

[0093] S312. Extract the connection strength value of the topological split feature and verify the strong correlation between the abnormal correlation of the rhythmic split feature and the connection strength value.

[0094] Preferably, the Pearson correlation coefficient between the connection strength value and the correlation anomaly is calculated. If the Pearson correlation coefficient is lower than -0.5, there is a negative correlation between the connection strength value and the correlation anomaly. That is, the higher the correlation coefficient, the smaller the correlation anomaly, indicating that the topological connection is more fragile and the rhythm fragmentation is more severe.

[0095] S313. Perform time-series correlation verification on the proportion of topologically weak nodes and rhythmically fragmented dual-synonyms.

[0096] Preferably, the time range for obtaining the high-risk edge split of topologically weak nodes and the abnormal time range for the bizygotic segments of rhythmic splitting are obtained.

[0097] Calculate the time coincidence coefficient of the time range of high-risk edge break and the abnormal time range of double synaptic segments; The temporal correlation between the proportion of dual synchrotron segments of topologically weak nodes and rhythmic fragmentation is determined based on the time coincidence coefficient.

[0098] For example, if the time overlap coefficient is greater than 70%, and a weak edge breaks at 3-4 AM, which corresponds to the period when the mean correlation degree Rz decreases abnormally, then there is a temporal correlation; otherwise, there is no correlation.

[0099] If the topological fragmentation feature blocks the flow of Qi and blood during the time of birth, and there is a spatial correlation, a negative correlation between the connection strength value and the correlation anomaly, and a temporal correlation, then it is determined that the weak topological node blocks the flow of Qi and blood during the time of birth.

[0100] It is understandable that determining the role of weak topological nodes in obstructing the flow of Qi and blood during specific times is as follows: Function 1: Through multi-dimensional correlation analysis of space, intensity, and time, a traceable causal relationship is formed between the path overlap rate and connection strength value of weak nodes in the topology and the functional anomalies of rhythmic breakage (such as poor correlation and the proportion of double zygotic segments). This provides a specific and quantitative explanation for why time rhythm breakage occurs, reduces vague speculation about the cause of breakage, and determines the direction for subsequent analysis.

[0101] Secondly, by focusing on the direct correlation between weak nodes in the topology and obstruction of Qi and blood flow, it can locate key nodes with weak structure and abnormal function in the global network of the meridian system. This provides a judgment criterion for distinguishing flow obstruction caused by topological defects from rhythm abnormalities caused by other non-structural factors, reducing interference factors in the analysis and making the analysis of meridian abnormalities more focused on the core contradictions.

[0102] Thirdly, it provides specific spatial anchor points for subsequent flow-oriented monitoring node layout optimization. After identifying the topologically weak nodes that obstruct the flow of Qi and blood, the delineation of the associated transition area and the symmetrical configuration of auxiliary monitoring nodes can be directly carried out around the flow path of these nodes. This makes the layout optimization no longer dependent on the generalized coverage of the entire meridian, and improves the adaptability of the monitoring layout to the Qi and blood flow pattern.

[0103] S4. If a weak node in the topology obstructs the flow of Qi and blood during the time, the associated transition region corresponding to the weak node in the topology is extracted. By performing a symmetrical analysis of the flow guidance configuration of the associated transition region, the layout ratio between the auxiliary monitoring node and the original node is determined.

[0104] The method for extracting the transitional region associated with the core meridian flow path corresponding to the weak topological node is as follows: Construct a meridian coordinate system with the suprasternal notch as the origin to obtain the coordinates of topologically weak nodes. ; The core meridian flow is digitized and transformed into a continuous path curve on the meridian coordinate system. .

[0105] Where t is a path parameter used to reflect the direction of blood and gas flow. These are the coordinates of the continuous path curve.

[0106] If a weak node in the topology is located in the transition section between inactive and active areas, then extend the flow path upstream and downstream by the distance between each adjacent acupoint by K, forming a rectangular area and serving as the associated transition area.

[0107] Preferably, K=1 or 2; Those skilled in the art will understand that acupoints in the meridians of traditional Chinese medicine do not exist in isolation. The distance between adjacent acupoints is essentially a basic synergistic unit for the flow of Qi and blood. The flow of Qi and blood within the same unit is synchronous, while the transmission across units has functional boundaries.

[0108] When a weak node in the topology obstructs the flow of Qi and blood, its influence will not be limited to a single point, but will spread upstream and downstream along the flow path. However, due to the synergy of Qi and blood, the spread range usually does not exceed two basic functional segments, that is, the distance between two adjacent acupoints.

[0109] If the extension is less than 1 interval, the influence of the weak point on the upstream adjacent acupoints may be missed. For example, the Qi and blood at the weak point 0.5 intervals upstream may be disordered due to obstruction.

[0110] If the extension exceeds two intervals, areas that are not directly related to the weak point will be included. For example, the correlation between the changes in Qi and blood at acupoints outside the third interval and the weak point is significantly reduced, resulting in an overly wide transition area and increased redundant monitoring.

[0111] If a weak node in the topology is located at a turning point in a continuous path curve, then the path containing the distance between one adjacent acupoint before and after the turning point forms a fan-shaped region, which is used as the associated transition region.

[0112] It should be explained that if the turning angle is greater than twice the average turning angle of the straight segment of the meridian, it is considered a significant turning point, reflecting a significant change in the direction of Qi and blood flow. At the same time, the included angle of the fan-shaped area is the turning angle plus the conventional buffer angle at similar turning points of the meridian, which is used to adapt to the diffusion characteristics of Qi and blood diversion.

[0113] The method for determining the layout ratio of auxiliary monitoring nodes to original nodes by performing a symmetrical analysis of the flow-guided configuration in the associated transition region is as follows: S401. Extract the streamer direction vector of the continuous path curve.

[0114] Preferably, if the associated transition region is a rectangular region, then the tangent direction vector of the topologically weak node is calculated on the continuous path curve L(t). .

[0115] The tangent direction vector is obtained by differentiating the continuous path curve L(t). First, the parametric equation of curve L(t) is defined. Let the coordinates of the points change with t as x(t) and y(t). The tangent components in the x and y directions are then calculated. .

[0116] If the associated transition region is a fan-shaped region, then extract the direction vector before the transition. and the direction vector after the turn .

[0117] The tangent direction vector and the direction vectors before and after the inflection are used as the stream direction vector.

[0118] S402. Establish symmetric configuration rules based on the stream direction vector to determine auxiliary monitoring nodes.

[0119] If the weak node in the topology is located in a rectangular region, then the vector along the tangent direction... upstream direction and Auxiliary monitoring nodes are symmetrically set up in the corresponding downstream direction.

[0120] For example, auxiliary monitoring nodes in the upstream direction Distance from weak topology nodes The distance is , Coordinates are , =1-2cm.

[0121] upstream auxiliary monitoring nodes Distance from weak topology nodes The distance is , = , Coordinates are .

[0122] If the weak node in the topology is located in the sector region, then along the direction vector and direction vector Auxiliary monitoring nodes are set up symmetrically.

[0123] For example, along Direction: Auxiliary monitoring nodes in the upstream direction (distance 1cm), downstream auxiliary monitoring nodes (distance 1cm).

[0124] along Direction: Auxiliary monitoring nodes in the upstream direction (distance 1cm), downstream auxiliary monitoring nodes (distance 1cm), auxiliary monitoring nodes , It forms a cross-shaped symmetrical layout.

[0125] Obtain the average cut sensitivity coefficient and the association anomaly difference of the associated edges of the weak nodes in the topology. Sum the average cut sensitivity coefficient and the association anomaly difference to obtain the weak proportion coefficient.

[0126] It should be explained that the weakness ratio coefficient quantifies the average severance sensitivity coefficient of the associated edges of the topologically weak node and the severity of rhythmic severance, reflecting the comprehensive weakness of the node and its associated region at the structural-functional level. The average severance sensitivity coefficient reflects the inherent vulnerability of the associated edges, while the abnormality in association reflects the degree of rhythmic function abnormality. The sum of these two directly reflects the superimposed strength of structural weakness and functional abnormality in the region, providing a quantitative benchmark for subsequently determining the layout ratio of auxiliary monitoring nodes to the original nodes. This allows the monitoring layout optimization to be adjusted according to the weakness intensity of different regions, adapting to the actual impact range of obstructed blood and qi flow.

[0127] Different levels are defined based on the weak point ratio coefficient, and the layout ratio of auxiliary monitoring nodes to original nodes is determined based on the different levels.

[0128] The sampling points of the meridian instrument were optimized and adjusted based on the layout ratio.

[0129] Those skilled in the art will understand that, based on the weakness ratio coefficient, the levels are divided according to their numerical range (e.g., the higher the value, the higher the level, reflecting the stronger the overall structural-functional weakness); and the corresponding auxiliary monitoring nodes and original nodes are matched to different levels. The higher the level, the higher the proportion of auxiliary nodes, such as 1:1 for high level and 1:2 for medium level.

[0130] Furthermore, by combining the symmetrical configuration rules of the flow guide: the rectangular area is symmetrical upstream and downstream along the flow direction, and the fan-shaped area is symmetrical along the direction before and after the turn. Auxiliary nodes are set up in the associated transition area according to this ratio, and the sampling density of the original nodes is adjusted synchronously to achieve targeted coverage of the meridian instrument sampling points to the weak areas and complete the optimization of sampling points.

[0131] Example 3 like Figure 3 As shown in the embodiment of the present invention, a dynamic analysis system for TCM meridian data based on edge computing includes the following modules: Sensitive Acquisition Module: Used to set the edge nodes for TCM meridian acquisition. Based on the meridian flow time sequence, the acupoint electrical signals of the edge nodes are dynamically acquired to perform segmentation sensitivity analysis, obtain the segmentation sensitivity coefficient of adjacent edge nodes, and construct segmentation sensitivity groups.

[0132] The segmentation analysis module is used to perform percolation critical flow analysis on segmentation sensitive groups to determine the maximum connected region, perform circadian rhythm analysis on the maximum connected region to obtain the mean correlation degree of different overlapping segments, and determine whether there is circadian rhythm segmentation in the maximum connected region.

[0133] Association Analysis Module: If there is a time-rhythm break, topological break analysis is performed on the largest connected component to extract the topological break features and rhythm break features of the topologically weak nodes and perform association analysis to determine whether the flow of Qi and blood during the time is blocked due to the topologically weak nodes.

[0134] Layout optimization module: If a weak node in the topology obstructs the flow of Qi and blood during the time, the associated transition region corresponding to the weak node is extracted. By performing a symmetrical analysis of the flow guidance configuration of the associated transition region, the layout ratio between the auxiliary monitoring node and the original node is determined.

[0135] The foregoing has shown and described the basic principles, main features, and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The embodiments and descriptions in the specification are merely illustrative of the principles of the invention. Various changes and modifications can be made to the invention without departing from its spirit and scope, and all such changes and modifications fall within the scope of the present invention as claimed. The scope of protection of the present invention is defined by the appended claims and their equivalents.

Claims

1. A method for dynamic analysis of TCM meridian data based on edge computing, characterized in that: Includes the following steps: Edge nodes for collecting data from TCM meridians are defined. Based on the meridian flow time sequence, the acupoint electrical signals of the edge nodes are dynamically collected and subjected to severance sensitivity analysis to obtain the severance sensitivity coefficients of adjacent edge nodes and construct severance sensitivity groups. Percolation critical flow analysis was performed on the segmentation sensitive group to determine the maximum connected domain. Chrono-rhythmic analysis was performed on the maximum connected domain to obtain the mean correlation degree of different overlapping segments, and it was determined whether there was chrono-rhythmic segmentation in the maximum connected domain. If there is a time-rhythm break, perform topological break analysis on the largest connected component, extract the topological break features and rhythm break features of the topologically weak nodes, and perform correlation analysis to determine whether the topologically weak nodes are hindering the flow of Qi and blood during the time. If a weak node in the topology obstructs the flow of Qi and blood during the time of birth, the associated transition region corresponding to the weak node is extracted. By performing a symmetrical analysis of the flow guidance configuration of the associated transition region, the layout ratio between the auxiliary monitoring node and the original node is determined.

2. The method for dynamic analysis of TCM meridian data based on edge computing according to claim 1, characterized in that: The method for performing the aforementioned severance sensitivity analysis is as follows: Obtain the meridian correlation degree of the sub-segment and divide the sub-segment into active period sub-segment and stable period sub-segment; The mean of the meridian correlation degree of all active period segments is extracted as the mean correlation degree; Define a split sensitivity mapping relationship to map the mean correlation degree to the split sensitivity coefficient.

3. The method for dynamic analysis of TCM meridian data based on edge computing according to claim 2, characterized in that: The method for obtaining the meridian correlation degree of the sub-segment is as follows: Based on the acupoint electrical signals of edge nodes, the rate of change of the acupoint electrical signal intensity of each edge node at adjacent monitoring times is calculated as the intensity change rate. Obtain the intensity change rate of adjacent edge nodes A and B, and construct two sets of intensity change sequences, A and B; The intensity change sequence is segmented to obtain intensity change sequences of different sub-segments; Calculate the Pearson correlation coefficients of the intensity change sequences of the corresponding sub-segments for groups A and B, and use them as the meridian correlation coefficients of the sub-segments; Coherence analysis was performed on the intensity change sequences of the two groups in corresponding sub-segments to obtain the meridian coherence coefficient; The meridian correlation coefficient and meridian coherence coefficient are standardized, and the standardized meridian correlation coefficient and meridian coherence coefficient are summed to obtain the meridian correlation degree of the sub-segment.

4. The method for dynamic analysis of TCM meridian data based on edge computing according to claim 1, characterized in that: The method for determining whether the largest connected component has a time-rhythm break is as follows: Obtain the meridian affiliation of network nodes within the largest connected component, and calculate the frequency of occurrence of different meridians based on the meridian affiliation; The meridians that appear most frequently are identified as the core meridians. Based on the core meridians and combined with the meridian flow pattern, the active and inactive times of the core meridians are determined. Determine whether there is any overlap between the active time and active period sub-segments. If there is overlap, extract the double-overlapping sub-segments of the active time and active period sub-segments and obtain the mean correlation degree of the double-overlapping sub-segments. Determine whether there is any overlap between inactive time periods and active period segments. If there is overlap, extract the single overlapping segments between inactive time periods and active period segments, and extract the mean correlation of the single overlapping segments. The time-meridian synergy criterion is set. If the mean correlation degree of double synodic segments and the mean correlation degree of single synodic segments meet the time-meridian synergy criterion, then it conforms to the time-meridian synergy law.

5. The method for dynamic analysis of TCM meridian data based on edge computing according to claim 4, characterized in that: The method for obtaining the maximum connected component is as follows: Obtain all edge nodes as network nodes and construct a network node set; In the set of network nodes, edges are created based on adjacent edge nodes, and the weight of the edge is the split sensitivity coefficient. Using the meridian correlation degree as the edge connectivity probability of network nodes, a percolation critical analysis model is constructed to determine the maximum connected region size of network nodes, and the maximum connected region is determined based on the maximum connected region size.

6. The method for dynamic analysis of TCM meridian data based on edge computing according to claim 5, characterized in that: The percolation critical analysis model is constructed as follows: Generate networks with different connectivity states based on edge connectivity probability, and determine whether to retain all edges; Based on repeated simulations of the grid corresponding to each edge, the maximum connected component size obtained from the repeated simulations is extracted. After M repeated simulations, obtain the largest connected subset in the grid, and extract the number of network nodes in the largest connected subset and the total number of nodes; The ratio of the number of network nodes in the largest connected subset to the total number of nodes is used as the size of the largest connected component. If the size of the largest connected component is higher than the preset connected component size threshold, then the network formed by the largest connected subset is taken as the largest connected component.

7. The method for dynamic analysis of TCM meridian data based on edge computing according to claim 1, characterized in that: The method to determine whether the flow of Qi and blood during the hour is obstructed by weak topological nodes is as follows: Extract the path overlap rate of topological fragmentation features for spatial correlation verification; The correlation between the connection strength values ​​of topological fragmentation features and the abnormally poor correlation between rhythmic fragmentation features is used to verify the strength correlation. Temporal correlation verification was performed on the proportion of dual zygotic segments with topologically weak nodes and rhythmic fragmentation. If the topological fragmentation feature blocks the flow of Qi and blood during the time of birth, and there is a spatial correlation, a negative correlation between the connection strength value and the correlation anomaly, and a temporal correlation, then it is determined that the weak topological node blocks the flow of Qi and blood during the time of birth.

8. The method for dynamic analysis of TCM meridian data based on edge computing according to claim 7, characterized in that: The method for extracting the topological fragmentation features is as follows: Obtain the initial screening coefficient and order value of the edge nodes within the largest connected component, and determine the weak nodes in the topology based on the initial screening coefficient and order value; Topological fragmentation features are obtained by extracting fragmented topological features from weak nodes. The method for extracting fragmented topological features is as follows: The degree of overlap between weak nodes in the topology and the flow path of Qi and blood in the core meridians is analyzed to obtain the node flow value; The average value of the cut sensitivity coefficient of the edges associated with weak nodes in the topology is collected and used as the connection strength value. Connectivity strength and node streamer values ​​are used as topology splitting features.

9. The method for dynamic analysis of TCM meridian data based on edge computing according to claim 1, characterized in that: The method for determining the layout ratio of the auxiliary monitoring nodes to the original nodes is as follows: Obtain the average cut sensitivity coefficient and association anomaly difference of the associated edges of weak nodes in the topology, and sum the average cut sensitivity coefficient and association anomaly difference to obtain the weak proportion coefficient; Different levels are defined based on the weak point ratio coefficient, and the layout ratio of auxiliary monitoring nodes to original nodes is determined based on the different levels.

10. A method for dynamic analysis of TCM meridian data based on edge computing according to claim 9, characterized in that: The auxiliary monitoring node is determined as follows: A meridian coordinate system was constructed with the suprasternal notch as the origin to obtain the coordinates of topologically weak nodes; The core meridian flow is digitally processed and transformed into a continuous path curve on the meridian coordinate system; Extract the stream direction vector of a continuous path curve; A symmetric configuration rule is established based on the stream direction vector to determine the auxiliary monitoring nodes.

Images