Orthopedic postoperative limb swelling monitoring method and system

Abstract

The invention relates to the technical field of medical monitoring, and discloses an orthopedic postoperative limb swelling monitoring method and system, and the method comprises the steps: carrying out the interaction relation mining of multi-source heterogeneous data in a target limb region, and obtaining a coupling relation group of the target limb region; performing multi-scale physiological signal separation on the multi-source heterogeneous data to obtain an integrated physiological state parameter of the target limb region; analyzing local microcirculation impedance characteristics of the target limb region, and performing spatial topological mapping on the local microcirculation impedance characteristics to obtain an impedance distribution diagram of the target limb region; performing path deduction on the individualized swelling trend of the target limb region to obtain a reference track of the individualized swelling trend; performing feedback type dynamic correction on the reference track to obtain a dynamic synchronous track of the individualized swelling trend; performing multi-modal joint evaluation on the individualized swelling trend to obtain a swelling report of the target limb region; the orthopedic postoperative limb swelling monitoring efficiency can be improved.

Description

Technical Field

[0001] This invention relates to the field of medical monitoring technology, and in particular to a method and system for monitoring limb swelling after orthopedic surgery. Background Technology

[0002] Current methods for monitoring postoperative limb swelling in orthopedic surgery primarily rely on intermittent measurements of single physiological parameters or on empirical palpation assessments by clinicians. These methods can only obtain single-dimensional information such as limb circumference or local stiffness, failing to comprehensively and synchronously reflect the coordinated changes and spatial heterogeneity of subcutaneous tissue in multidimensional physiological processes such as perfusion, metabolism, and interstitial fluid dynamics. Due to the lack of fusion analysis of multi-source physiological data and decoupling from individual physiological characteristics, existing technologies struggle to establish comprehensive parameters that characterize the true state of the tissue, resulting in biased, delayed monitoring results and an inability to quantify the dynamic evolution of swelling.

[0003] The aforementioned technological limitations mean that clinical early warning relies on static thresholds or cross-sectional comparisons, making it impossible to establish an individualized dynamic assessment baseline tailored to each patient's unique anatomical structure, surgical trauma characteristics, and physiological response patterns. This further leads to the difficulty in timely identification of subtle early pathophysiological deviations, insufficient early warning sensitivity, and an inability to accurately locate and predict the origin, spatial diffusion pattern, and subsequent evolution direction of swelling. Consequently, this limits the timeliness and precision of clinical intervention. Therefore, improving the efficiency of monitoring postoperative limb swelling in orthopedic surgery has become an urgent problem to be solved. Summary of the Invention

[0004] This invention provides a method and system for monitoring limb swelling after orthopedic surgery to solve the problems mentioned in the background art.

[0005] To achieve the above objectives, the present invention provides a method for monitoring limb swelling after orthopedic surgery, comprising: S01. Perform interaction relationship mining on the cross-modal association patterns of multi-source heterogeneous data in the target limb region to obtain the coupling relationship group of the target limb region; S02. Based on the coupling relationship group, multi-scale physiological signal separation is performed on multi-source heterogeneous data to obtain integrated physiological state parameters of the target limb region; S03. Based on the integrated physiological state parameters, analyze the local microcirculation impedance characteristics of the target limb region, and perform spatial topological mapping on the local microcirculation impedance characteristics to obtain the impedance distribution map of the target limb region. S04. Based on the impedance distribution map, the individualized swelling trend of the target limb region is deduced to obtain the baseline trajectory of the individualized swelling trend. S05. Based on the integrated physiological state parameters and impedance distribution map, the baseline trajectory is dynamically corrected by feedback to obtain a dynamic synchronous trajectory of individualized swelling trend. S06. Based on impedance distribution maps and dynamic synchronization trajectories, a multimodal joint assessment of individualized swelling trends is performed to obtain a swelling report for the target limb region.

[0006] In a preferred embodiment, interaction relationship mining is performed on the cross-modal association patterns of multi-source heterogeneous data in the target limb region to obtain a set of coupling relationships in the target limb region, including: Extract preliminary correlation features from multi-source heterogeneous data in the target limb region; Based on preliminary association features, a collaborative evolution analysis of cross-modal association patterns of multi-source heterogeneous data is performed to obtain an initial association pattern set for the target limb region. Cross-scale interaction relationships are constructed on the initial association pattern group to obtain the nonlinear metric relationship of the initial association pattern group; The physiological interpretability of the nonlinear metric relationship was verified, and the verified relationship was fused and screened to obtain the coupling relationship group of the target limb region.

[0007] In a preferred embodiment, based on coupling relationship groups, multi-scale physiological signal separation is performed on multi-source heterogeneous data to obtain integrated physiological state parameters of the target limb region, including: Based on the coupling relationship group, dynamic physiological rhythm analysis is performed on multi-source heterogeneous data to obtain the temporal physiological elements of multi-source heterogeneous data; Regional attribution analysis was performed on temporal physiological elements to obtain their spatial attribution characteristics. Multi-level information gain fusion of spatial attribution features yields physiological state primitives of the target limb region; By performing state regularization on the physiological state primitives, integrated physiological state parameters of the target limb region are obtained.

[0008] In a preferred embodiment, based on integrated physiological state parameters, the local microcirculatory impedance characteristics of the target limb region are analyzed, and spatial topological mapping is performed on the local microcirculatory impedance characteristics to obtain an impedance distribution map of the target limb region, including: Based on integrated physiological state parameters, regional differences in tissue perfusion and metabolic load in the target limb region were traced to obtain the trend of relative impedance changes in the target limb region. Based on the coupling relationship group, impedance characteristic inversion is performed on the relative impedance change trend to obtain a quantitative impedance characteristic map of local microcirculation in the target limb region. Physiological constraint interpolation is performed on the quantized impedance characteristic map to obtain the basic impedance spatial distribution map of the target limb region; Anisotropic smoothing is performed on the basic impedance spatial distribution map to obtain the impedance distribution map of the target limb region.

[0009] In a preferred embodiment, based on the coupling relationship group, impedance characteristic inversion is performed on the relative impedance change trend to obtain a quantitative impedance characteristic map of the local microcirculation in the target limb region, including: Spatial synchronous weight allocation is performed on the relative impedance change trend and coupling relationship group to obtain the relationship distribution of the relative impedance change trend; Based on the coupling relationship group, the relationship distribution is solved point by point iteratively to obtain the impedance characteristic value of the relative impedance change trend. The formula for calculating the impedance characteristic value is as follows: ; In the formula, For the first Position after the next iteration The impedance characteristic value at that point, Position in the coupling relationship group With position The strength of interaction between them For position The trend of relative impedance change at the location, The trend of interaction intensity and relative impedance variation in the target limb region The weighted integral on, Interaction intensity in the target limb area The points on the top The preset regularization adjustment coefficient is used. For position The mean of the interaction intensity at that location. For the first Position after the next iteration The impedance characteristic value at that point, For the first Position after the next iteration The impedance characteristic value at that point, The deviation in interaction strength and the difference in impedance characteristics in the target limb region Cooperative correction integrals on; By extending the impedance characteristic values ​​to physiological continuity, a quantitative impedance characteristic map of local microcirculation in the target limb region is obtained.

[0010] In a preferred embodiment, based on the impedance distribution map, the individualized swelling trend of the target limb region is deduced to obtain the baseline trajectory of the individualized swelling trend, including: Based on the limb anatomy and postoperative suture tension distribution in the target limb region, morphological features of local gradient changes in the impedance distribution map are extracted to obtain an initial diffusion feature set of individualized swelling trends in the target limb region. Based on integrated physiological state parameters, specific morphological corrections are performed on the initial diffusion feature set to obtain an individualized diffusion feature set of swelling trends; Based on the historical recovery rhythm characteristics of the target limb region, the temporal evolution simulation of the diffusion feature set is performed to obtain the temporal development path set of individualized swelling trends; Based on the physiological characteristics of the target limb region, consistency screening is performed on the temporal development path set to obtain the baseline trajectory of individualized swelling trends.

[0011] In a preferred embodiment, based on integrated physiological state parameters and impedance distribution maps, a feedback-based dynamic correction is performed on the baseline trajectory to obtain a dynamically synchronized trajectory for individualized swelling trends, including: Based on the impedance distribution map, causal correlation analysis was performed on the integrated physiological state parameters to obtain the hidden physiological guidance characteristics of individualized swelling trends. Dynamic consistency matching between the baseline trajectory and the concealed physiological guidance features is performed to obtain the potential deviation segments of the baseline trajectory; Multi-objective collaborative transitions are performed on potential deviation segments to obtain a cluster of candidate transition trajectories for individualized swelling trends; By traversing and selecting candidate transition trajectory clusters, dynamic synchronous trajectories with individualized swelling trends are obtained.

[0012] In a preferred embodiment, multi-objective cooperative transitions are performed on potential deviation segments to obtain a cluster of candidate transition trajectories for individualized swelling trends, including: Multidimensional conflict analysis is performed on the potential deviation section to obtain the conflict constraint set of the potential deviation section; The global interaction strength of the coupling relationship group is quantized by topological structure strength quantization to obtain the interaction strength norm of the coupling relationship group; The spatial dispersion of the numerical distribution of the impedance distribution is evaluated to obtain the spatial heterogeneity entropy of the impedance distribution. The temporal change rate of the concealed physiological guidance features was statistically calibrated to obtain the sensitivity coefficient of the concealed physiological guidance features; The interaction strength of the coupling relationship group is extracted to obtain the feature relaxation scale of the coupling relationship group; Based on the coupling relationship group, dynamic priority weighting is applied to the conflict constraint set to obtain the time-varying weight distribution of the conflict constraint set; Based on the time-varying weight distribution, the transition threshold of the potential deviation segment is calculated, where the formula for calculating the transition threshold is: ; In the formula, For the transition threshold, As a preset baseline transition potential, It is an exponential function. For the interaction strength norm, The characteristic relaxation scale, It is the hyperbolic tangent function. This represents the net change in the concealed physiological guidance characteristics. Sensitivity coefficient The Gaussian error function is... The average deviation of the potential deviation segment. The preset deviation reference value, The standard deviation of the preset deviation is... For spatial heterogeneity entropy, is the negative correlation decay coefficient of spatial heterogeneity entropy; Based on the transition threshold, morphological transitions are performed on potential deviation segments to obtain a cluster of transition trajectory segments for potential deviation segments; Clustering and integrating the transition trajectory segments yields candidate transition trajectory clusters with individualized swelling trends.

[0013] In a preferred embodiment, based on impedance distribution maps and dynamic synchronization trajectories, a multimodal joint assessment of individualized swelling trends is performed to obtain a swelling report for the target limb region, including: Based on the dynamic synchronization trajectory, spatial features are extracted from the impedance distribution map to obtain a multi-phase feature set of the impedance distribution map; Cross-phase feature transmission simulation was performed on a multi-phase feature set to obtain the feature transmission law of individualized swelling trend; The matching degree between the feature transmission law and the dynamic synchronization trajectory is quantized collaboratively to obtain the trajectory matching degree of the feature transmission law; Based on trajectory matching degree and preset clinical risk level, clinical consistency assessment of feature transmission pattern is performed, and the assessment results are semantically encapsulated to obtain swelling report of target limb area.

[0014] To address the aforementioned problems, the present invention also provides a postoperative limb swelling monitoring system for orthopedic surgery, the system comprising: The association mining module is used to mine the interaction relationship of cross-modal association patterns of multi-source heterogeneous data in the target limb region, and obtain the coupling relationship group of the target limb region; The signal integration module is used to separate multi-scale physiological signals from multi-source heterogeneous data based on coupling relationship groups, so as to obtain integrated physiological state parameters of the target limb region. The impedance mapping module is used to analyze the local microcirculation impedance characteristics of the target limb region based on integrated physiological state parameters, and to perform spatial topological mapping on the local microcirculation impedance characteristics to obtain the impedance distribution map of the target limb region. The trend extrapolation module is used to extrapolate the individualized swelling trend of the target limb region based on the impedance distribution map, and obtain the baseline trajectory of the individualized swelling trend. The dynamic correction module is used to perform feedback dynamic correction on the baseline trajectory based on integrated physiological state parameters and impedance distribution map to obtain a dynamic synchronous trajectory of individualized swelling trend. The assessment report module is used to perform multimodal joint assessment of individualized swelling trends based on impedance distribution maps and dynamic synchronization trajectories, and obtain a swelling report for the target limb area.

[0015] Compared with the prior art, the present invention has the following beneficial effects: 1. This invention achieves deep fusion and precise analysis of multi-dimensional physiological data into individualized intrinsic state parameters. By autonomously mining the unique physiological coupling relationships from real-time multi-source patient data, and using this as a guide for multi-scale signal separation and reconstruction, highly integrated state parameters reflecting the true physiological dynamics within the tissue can be generated. This process overcomes the limitations of relying on a single or pre-set physiological model, enabling the monitoring of postoperative swelling to be rooted in the patient's own physiological characteristics and immediate responses, thereby significantly improving the individual consistency and physiological interpretability of state assessment.

[0016] 2. This invention establishes a dynamic data-driven capability for tracking and predictively warning about swelling progression. By combining impedance distribution reflecting spatial heterogeneity with individualized temporal trajectory projections and introducing a real-time feedback correction mechanism, it can continuously generate dynamic trajectories synchronized with the patient's actual pathological progression. This enables monitoring not only to quantify the current swelling state but also to prospectively simulate its evolution path and to identify and spatially locate minor deviations early, thereby providing clinical decision support with both temporal lead and spatial orientation, improving the timeliness and targeted nature of interventions. Attached Figure Description

[0017] Figure 1 This is a flowchart illustrating a method for monitoring limb swelling after orthopedic surgery, provided in one embodiment of the present invention. Figure 2 This is a functional block diagram of a postoperative limb swelling monitoring system for orthopedic surgery provided in one embodiment of the present invention; The realization of the objective, functional features and advantages of the present invention will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0018] It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.

[0019] This application provides a method for monitoring limb swelling after orthopedic surgery. The executing entity of this method includes, but is not limited to, at least one of the following electronic devices that can be configured to execute the method provided in this application: a server, a terminal, etc. In other words, the method for monitoring limb swelling after orthopedic surgery can be executed by software or hardware installed on a terminal device or a server device. The server includes, but is not limited to, a single server, a server cluster, a cloud server, or a cluster of cloud servers. The server can be a standalone server or a cloud server that provides basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communication, middleware services, domain name services, security services, content delivery networks (CDN), and big data and artificial intelligence platforms.

[0020] Reference Figure 1 The diagram shown is a flowchart illustrating a method for monitoring limb swelling after orthopedic surgery according to an embodiment of the present invention. In this embodiment, the method for monitoring limb swelling after orthopedic surgery includes: S01. Perform interaction relationship mining on the cross-modal association patterns of multi-source heterogeneous data in the target limb region to obtain the coupling relationship group of the target limb region; In this embodiment of the invention, interaction relationship mining is performed on the cross-modal association patterns of multi-source heterogeneous data in the target limb region to obtain a coupling relationship group of the target limb region, including: Extract preliminary correlation features from multi-source heterogeneous data in the target limb region; Based on preliminary association features, a collaborative evolution analysis of cross-modal association patterns of multi-source heterogeneous data is performed to obtain an initial association pattern set for the target limb region. Cross-scale interaction relationships are constructed on the initial association pattern group to obtain the nonlinear metric relationship of the initial association pattern group; The physiological interpretability of the nonlinear metric relationship was verified, and the verified relationship was fused and screened to obtain the coupling relationship group of the target limb region.

[0021] Surface electromyography (EMG) signals, local temperature distribution data, optical contour scanning information, and bioelectrical impedance measurements synchronously acquired from multiple sensors in the target limb region were subjected to time window alignment and standardization. The processed data constituted multi-source heterogeneous data. Subsequently, a method based on statistical correlation and mutual information analysis was used to traverse the paired combinations of different modalities, calculating their linear correlation coefficients and nonlinear dependencies. Each calculated high correlation strength value and its corresponding data modality pair were recorded as an independent preliminary correlation feature. All recorded correlation strength values ​​and data pairs together constituted the preliminary correlation feature set of the multi-source heterogeneous data.

[0022] Based on the obtained preliminary set of association features, the dynamic community discovery approach from graph theory is employed. Each data modality is treated as a node in the graph, and weighted connections are established between nodes according to the association strength described by the preliminary association features. By simulating the transmission and competition of influence between nodes, the community affiliation of nodes is iteratively adjusted, ultimately stably dividing the nodes into several groups with tight internal connections and sparse external connections. Each stable group represents a stable pattern formed by the synergistic effect of multimodal data. The complete list of these groups constitutes the initial association pattern group for the target limb region.

[0023] A hierarchical analysis was performed on the initially grouped association patterns. First, at the intra-group scale, the synchronicity and dominance of data modal changes within each pattern were analyzed. Then, at the inter-group scale, the temporal order of different patterns and their relative strengths were analyzed. By establishing quantitative scores for intra-group tightness and inter-group interaction strength, and repeatedly adjusting the association weights between pattern members and patterns based on the scores, a multi-layered relationship network was ultimately formed that describes the entire process from micro-level data fluctuations to macro-level pattern evolution. The final stable weight values ​​of each node and connecting edge in this network constitute the nonlinear metric relationship of the initial association pattern group.

[0024] Each connection in the aforementioned nonlinear measurement relationships is reviewed based on a well-known knowledge base of human anatomy and pathophysiology. The review process involves determining whether the direction and magnitude of the coordinated changes in the data modes or patterns connected by the relationship are consistent with known physiological mechanisms, such as whether increased tissue perfusion is accompanied by specific changes in local temperature and electrical impedance. Relationships that pass the review are marked as valid, while contradictory relationships are removed. Subsequently, all valid relationships undergo redundancy removal and fusion, merging multiple similar relationships describing the same physiological process into a single, more generalized strong relationship. Finally, all the strong relationships retained after fusion and screening are systematically organized into a relationship set; this complete relationship set is defined as the coupling relationship group for the target limb region.

[0025] The beneficial effects of this method are as follows: after synchronizing multi-source heterogeneous sensor data, statistical and information theory methods are used to extract the correlation features between data modes. Then, the dynamic community discovery idea is used to divide different modes into internally co-stable mode groups. Through cross-scale interaction analysis, a multi-layer nonlinear relationship network that characterizes micro-fluctuations and macro-evolution is constructed. Finally, the relationship of this network is verified and fused based on physiological knowledge, thereby generating a coupled relationship group that closely fits the actual physiological mechanism of individuals and has clear physiological interpretation. This provides a highly reliable and individualized physiological interaction map foundation for all subsequent analysis steps.

[0026] S02. Based on the coupling relationship group, multi-scale physiological signal separation is performed on multi-source heterogeneous data to obtain integrated physiological state parameters of the target limb region; In this embodiment of the invention, based on the coupling relationship group, multi-scale physiological signal separation is performed on multi-source heterogeneous data to obtain integrated physiological state parameters of the target limb region, including: Based on the coupling relationship group, dynamic physiological rhythm analysis is performed on multi-source heterogeneous data to obtain the temporal physiological elements of multi-source heterogeneous data; Regional attribution analysis was performed on temporal physiological elements to obtain their spatial attribution characteristics. Multi-level information gain fusion of spatial attribution features yields physiological state primitives of the target limb region; By performing state regularization on the physiological state primitives, integrated physiological state parameters of the target limb region are obtained.

[0027] Based on the interaction strength and dominant direction among the data modalities defined in the coupling relationship group, time-frequency decomposition and pattern matching are performed on synchronously acquired multi-source heterogeneous data streams. This process identifies the signal components most synchronous with specific physiological rhythms, such as vasomotor cycles or muscle activity cycles, based on the coupling relationships, and then isolates and reconstructs these components from the original mixed data stream into independent signal sequences. These separated independent signal sequences with clear rhythmic attributes collectively constitute the temporal physiological elements of the multi-source heterogeneous data.

[0028] Using known spatial coordinates of sensors on the limb surface and anatomical atlases, each signal sequence in a temporal physiological element is mapped to its most likely deep tissue origin region. This mapping process is based on signal amplitude attenuation models and tissue conduction characteristics, assigning a specific spatial location label to each signal sequence. All temporal physiological elements assigned spatial location labels are collectively described as having spatial attribution characteristics.

[0029] The stability and repeatability of each signal sequence in the spatial attribution features are evaluated over a continuous period on a temporal scale. The differences and complementarities between signal sequences from different anatomical regions are evaluated on a spatial scale. An iterative weighting process assigns higher fusion weights to signal sequences with high stability and strong complementarity, while suppressing redundant or significantly noisy sequences. All weighted signal sequences are superimposed and synthesized according to their spatial attribution to form a deredundant and information-enhanced comprehensive feature set, which represents the physiological state primitives of the target limb region.

[0030] All comprehensive features in the physiological state primitive set are subjected to dimensional unification and amplitude normalization to eliminate numerical differences caused by different sensor types and units. Then, based on the physiological meaning represented by these comprehensive features, they are organized into a pre-defined multi-dimensional vector framework, with each dimension corresponding to a specific physiological state dimension such as metabolic intensity or perfusion level. The complete vector containing multi-dimensional quantitative indicators, integrated through this framework, is defined as the integrated physiological state parameter of the target limb region.

[0031] The beneficial effect is that this method analyzes and separates time-series signal elements synchronized with specific physiological rhythms from multi-source data based on individualized coupling relationships. Then, it uses sensor locations and anatomical atlases to accurately locate these elements to their tissue origin regions, forming spatial attribution features. Through cross-scale evaluation and iterative weighted fusion, these features generate redundant and information-enhanced physiological state primitives. Finally, through dimensional unification and multi-dimensional vector framework integration, a high-value integrated physiological state parameter that systematically represents tissue dynamics is obtained.

[0032] S03. Based on the integrated physiological state parameters, analyze the local microcirculation impedance characteristics of the target limb region, and perform spatial topological mapping on the local microcirculation impedance characteristics to obtain the impedance distribution map of the target limb region. In this embodiment of the invention, based on integrated physiological state parameters, the local microcirculatory impedance characteristics of the target limb region are analyzed, and spatial topological mapping is performed on the local microcirculatory impedance characteristics to obtain an impedance distribution map of the target limb region, including: Based on integrated physiological state parameters, regional differences in tissue perfusion and metabolic load in the target limb region were traced to obtain the trend of relative impedance changes in the target limb region. Based on the coupling relationship group, impedance characteristic inversion is performed on the relative impedance change trend to obtain a quantitative impedance characteristic map of local microcirculation in the target limb region. Physiological constraint interpolation is performed on the quantized impedance characteristic map to obtain the basic impedance spatial distribution map of the target limb region; Anisotropic smoothing is performed on the basic impedance spatial distribution map to obtain the impedance distribution map of the target limb region.

[0033] Based on the coupling relationship group, impedance characteristic inversion is performed on the relative impedance change trend to obtain a quantitative impedance characteristic map of local microcirculation in the target limb region, including: Spatial synchronous weight allocation is performed on the relative impedance change trend and coupling relationship group to obtain the relationship distribution of the relative impedance change trend; Based on the coupling relationship group, the relationship distribution is solved point by point iteratively to obtain the impedance characteristic value of the relative impedance change trend. The formula for calculating the impedance characteristic value is as follows: ; In the formula, For the first Position after the next iteration The impedance characteristic value at that point, Position in the coupling relationship group With position The strength of interaction between them For position The trend of relative impedance change at the location, The trend of interaction intensity and relative impedance variation in the target limb region The weighted integral on, Interaction intensity in the target limb area The points on the top The preset regularization adjustment coefficient is used. For position The mean of the interaction intensity at that location. For the first Position after the next iteration The impedance characteristic value at that point, For the first Position after the next iteration The impedance characteristic value at that point, The deviation in interaction strength and the difference in impedance characteristics in the target limb region Cooperative correction integrals on; By extending the impedance characteristic values ​​to physiological continuity, a quantitative impedance characteristic map of local microcirculation in the target limb region is obtained.

[0034] Based on the physiological significance of each dimension of the integrated physiological state parameters, a subset of indicators directly related to tissue fluid accumulation and capillary blood flow is selected. Within this subset, the indicator values ​​of each pre-divided anatomical region within the target limb area are compared horizontally. By calculating the relative deviation of each region's indicator value from the overall average, abnormal regions with significantly deviated perfusion or metabolic levels are identified. The relative deviation of these regions is correlated with spatial location and assigned a sign to indicate the direction of deviation: above the average is positive, below the average is negative. The resulting deviation map with spatial location and direction attributes represents the relative impedance change trend of the target limb region.

[0035] The deviation value of the relative impedance change trend at each spatial location is used as the initial input. The interaction strength between any two spatial locations defined in the coupling relationship group is used as the influence weight. Iterative updates begin from a preset initial impedance distribution guess. In each iteration, for each spatial location, its own initial deviation value is calculated and weighted by averaging the current impedance estimates of all other spatial locations according to the interaction strength weight. This calculation ensures that neighboring locations with high interaction strength have a greater influence on the impedance value of the current location. The calculation result is used as the new impedance estimate for that location. After completing the above calculation for all spatial locations in the region, one iteration is completed. This iterative process is repeated until the change in the impedance estimates of all spatial locations in two consecutive iterations is less than a preset small threshold, at which point the impedance value is considered to have reached stable convergence. These final stable values ​​at each spatial location are the impedance characteristic values ​​of the relative impedance change trend.

[0036] After obtaining the impedance characteristic values ​​at discrete spatial locations, processing is performed according to the principle of continuity of known limb anatomical structures. Based on the orientation of fascial compartments and muscle bundles in the limb tissue, it is determined which discrete points are physiologically directly connected, and virtual connection paths are established between these connected points. Along these virtual paths, a distance-weighted numerical diffusion method is used to extend and fill the impedance characteristic values ​​of discrete points to their neighboring, unmeasured areas. This process ensures that the impedance values ​​also exhibit a smooth transition in areas of physiological structural continuity, ultimately generating a numerically continuous image covering the entire target limb region. This image is a quantitative impedance characteristic map of local microcirculation in the target limb region.

[0037] Based on the known discrete impedance values ​​in the quantized impedance characteristic map, operations are performed at the pixel level according to the anatomical structure map of the target limb region. Regions corresponding to different physiological tissue boundaries are identified in the image, such as the junction of skin and subcutaneous fat or the fascia layer between muscles. Hard constraints are applied at these boundaries to prevent the interpolation process from mixing impedance characteristics of different tissues across the boundaries. Within the same tissue defined by the boundary constraints, an anisotropic distance function is used for interpolation calculations, assigning shorter weights to distances along muscle fibers or major blood vessels to make the interpolation results more consistent with the physiological conduction characteristics of the tissue. In this way, sparse discrete impedance values ​​are filled to generate a high-resolution, anatomically oriented continuous impedance image covering the entire target region; this image is the basic impedance spatial distribution map of the target limb region.

[0038] Image processing is performed on the basic impedance spatial distribution map. The processing is based on the known structural orientation characteristics of different physiological tissues, such as the longitudinal texture of smooth muscle or the isotropic properties of adipose tissue. At each pixel in the image, the pre-defined dominant direction of its respective tissue is analyzed. A stronger smoothing filter is applied along this dominant direction to eliminate noise while maintaining the physiological continuity of that direction. A weaker smoothing filter is applied in directions perpendicular to the dominant direction to preserve possible physiological boundaries or abrupt changes. This direction-dependent, non-uniform smoothing process traverses the entire image, ultimately yielding an image that is both clear and noise-removed while perfectly preserving the boundaries of important physiological structures. This final image is the impedance distribution map of the target limb region.

[0039] The interaction strength between locations in the formula is directly derived from the coupling relationship group mined and verified from multi-source heterogeneous data. The relative impedance change trend at a location in the formula is directly derived from the output results of regional difference tracing based on integrated physiological state parameters. The regularization adjustment coefficient in the formula is a preset fixed value used to control the strength of subsequent correction terms. The mean interaction strength at a location in the formula is calculated by arithmetically averaging the interaction strengths between that location and all other locations within the coupling relationship group. The impedance characteristic values ​​of the location and the location after the previous iteration in the formula are derived from the intermediate state of the current iterative calculation process.

[0040] The significance of this formula lies in solving for the impedance characteristic value of each spatial location through an iteratively updated mathematical rule. The rule consists of two parts working together. The first part calculates a weighted average, using the interaction strength as weight to average the relative impedance change trends of all locations within the region, thus providing an initial estimate of the current location based on the overall regional trend. The second part introduces a co-correction term, which measures the degree to which the interaction strength between the current location and other locations within the region deviates from the average level, while also considering the differences in the current impedance estimates between these locations. This correction term drives adaptive adjustments to the impedance values ​​between neighboring regions with stronger or weaker interactions to reduce inconsistencies. The entire formula determines the new impedance characteristic value for that location in the next iteration by adding the basic estimate from the first part and the spatial co-correction from the second part.

[0041] The formula shows a trend where the impedance characteristics at each spatial location gradually stabilize with increasing iterations. In the early stages of iteration, the impedance values ​​at each location may vary significantly as the trend-based estimation and spatially coordinated correction seek a balance. As iterations continue, the impedance values ​​between regions with high interaction strength tend to converge under the influence of the correction term, while regions with abnormal interaction strength or deviations from the average may develop unique impedance characteristics. Eventually, the impedance values ​​at all locations will no longer change significantly with further iterations; at this point, the formula reaches a stable state. The impedance distribution in this stable state represents the most consistent solution under this mathematical rule with the given interaction strength relationship and relative trend of change. This characterizes the quantitative distribution of the microcirculatory impedance characteristics within the tissue.

[0042] The beneficial effects of this method are as follows: by identifying the relative deviation between regional perfusion and metabolism from integrated physiological state parameters to form an impedance change trend, and then using individualized coupling relationship groups as weight constraints, the precise impedance characteristic values ​​of discrete spatial locations are solved through iterative updates. Based on physiological continuity, the discrete values ​​are extended into a continuous characteristic map. Then, anisotropic interpolation is performed with anatomical structure as a guide to generate a basic spatial distribution map. Finally, structure-guided anisotropic smoothing is performed to obtain a high-fidelity impedance distribution map that clearly represents the microcirculation state inside the tissue.

[0043] This calculation formula directly utilizes the interaction strength relationship mined from multi-source data and the regional change trend traced from physiological state parameters as core input parameters. Its mathematical rules consist of two parts: a weighted average estimation based on the overall trend and a correction term that drives spatial coordinated adjustment. Through iterative updates, the impedance values ​​at each location tend to coordinate with the neighboring regions under the guidance of interaction strength and eventually reach a stable self-consistent distribution, thereby achieving accurate and stable quantitative inversion of the microcirculation impedance characteristics within the tissue.

[0044] S04. Based on the impedance distribution map, the individualized swelling trend of the target limb region is deduced to obtain the baseline trajectory of the individualized swelling trend. In this embodiment of the invention, based on the impedance distribution map, the individualized swelling trend of the target limb region is deduced to obtain the baseline trajectory of the individualized swelling trend, including: Based on the limb anatomy and postoperative suture tension distribution in the target limb region, morphological features of local gradient changes in the impedance distribution map are extracted to obtain an initial diffusion feature set of individualized swelling trends in the target limb region. Based on integrated physiological state parameters, specific morphological corrections are performed on the initial diffusion feature set to obtain an individualized diffusion feature set of swelling trends; Based on the historical recovery rhythm characteristics of the target limb region, the temporal evolution simulation of the diffusion feature set is performed to obtain the temporal development path set of individualized swelling trends; Based on the physiological characteristics of the target limb region, consistency screening is performed on the temporal development path set to obtain the baseline trajectory of individualized swelling trends.

[0045] Based on the reconstructed 3D anatomical structure map of the target limb region from medical images and the suture direction and tension data recorded in the surgical record, the impedance distribution map is subjected to pixel-by-pixel gradient calculation and analysis. This process identifies spatial locations where the impedance gradient amplitude is significantly higher than the surrounding area, and determines the steep direction of tissue characteristic change based on the gradient direction of these points. These high-gradient points and their steep directions are correlated with the overlying anatomical layers, such as intermuscular or fascial planes, and adjacent areas of concentrated suture tension to determine the most likely initial anatomical path and restricted direction of swelling fluid diffusion. All identified initial diffusion paths and their spatial orientation descriptions are summarized into a set, which constitutes the initial diffusion feature set of individualized swelling trends in the target limb region.

[0046] The subset of values ​​from the integrated physiological state parameters, reflecting real-time tissue metabolism and perfusion levels, are compared one by one with the expected physiological change intensity for each path in the initial diffusion feature set. This comparison process calculates the numerical deviation between the current actual physiological state and the expected state under the initial path assumptions. Based on the direction and magnitude of the deviation, the estimated diffusion intensity for each path in the initial diffusion feature set is weighted and adjusted; paths with large deviations have their weights reduced, while those with good deviations have their weights increased. Simultaneously, the possible dominant direction of diffusion is fine-tuned based on the real-time dominant physiological processes revealed by the physiological parameters. This weighted and directionally adjusted set of diffusion path descriptions, more closely reflecting the current instantaneous physiological state, constitutes the individualized diffusion feature set of swelling trends.

[0047] The system retrieves a template of time-related key physiological indicator changes—the historical recovery rhythm features—recorded during similar recovery processes in patients corresponding to the target limb region. Each path in the diffusion feature set is mapped onto this time template. Based on the typical physiological activity intensity and pattern at different recovery stages in the template, a trend intensity coefficient that changes over time is assigned to the diffusion feature. By simulating a clock, the cumulative state of each diffusion path under the influence of the corresponding trend coefficient is calculated at multiple consecutive future time points, thereby generating a state evolution curve for each path extending from the current moment to a future moment. The set of evolution curves generated for all paths constitutes the temporal development path set of the individualized swelling trend.

[0048] Based on the patient's inherent physiological characteristic profile corresponding to the target limb region, which includes stable characteristic indicators such as basal metabolic rate and circulatory system function level, the rationality of each evolution curve in the time-series development path set is assessed using these indicators. The assessment process examines whether the overall amplitude, rate of change, and possible inflection points of the curve are within a reasonable range supported by the patient's physiological function. The assessment results are quantified into a consistency score, curves with scores below a preset threshold are eliminated, and the evolution curve with the highest consistency score is selected from the remaining curves. This finally selected evolution curve, which best conforms to the individual patient's physiological tolerance and development pattern, is defined as the baseline trajectory of the individualized swelling trend.

[0049] The beneficial effect is that this method extracts the morphological characteristics of the initial spread of swelling from the impedance distribution by combining individual anatomical and surgical tension data, calibrates these characteristics using real-time physiological parameters, simulates multiple possible paths of its evolution over time based on the patient's historical recovery rhythm, and finally selects the most reasonable path based on the patient's inherent physiological characteristics, thereby generating a highly personalized and physiologically interpretable baseline trajectory for predicting the swelling development trend.

[0050] S05. Based on the integrated physiological state parameters and impedance distribution map, the baseline trajectory is dynamically corrected by feedback to obtain a dynamic synchronous trajectory of individualized swelling trend. In this embodiment of the invention, based on integrated physiological state parameters and impedance distribution maps, a feedback-based dynamic correction is performed on the baseline trajectory to obtain a dynamically synchronized trajectory of individualized swelling trends, including: Based on the impedance distribution map, causal correlation analysis was performed on the integrated physiological state parameters to obtain the hidden physiological guidance characteristics of individualized swelling trends. Dynamic consistency matching between the baseline trajectory and the concealed physiological guidance features is performed to obtain the potential deviation segments of the baseline trajectory; Multi-objective collaborative transitions are performed on potential deviation segments to obtain a cluster of candidate transition trajectories for individualized swelling trends; By traversing and selecting candidate transition trajectory clusters, dynamic synchronous trajectories with individualized swelling trends are obtained.

[0051] Multi-objective collaborative transitions are performed on potential deviation segments to obtain a cluster of candidate transition trajectories for individualized swelling trends, including: Multidimensional conflict analysis is performed on the potential deviation section to obtain the conflict constraint set of the potential deviation section; The global interaction strength of the coupling relationship group is quantized by topological structure strength quantization to obtain the interaction strength norm of the coupling relationship group; The spatial dispersion of the numerical distribution of the impedance distribution is evaluated to obtain the spatial heterogeneity entropy of the impedance distribution. The temporal change rate of the concealed physiological guidance features was statistically calibrated to obtain the sensitivity coefficient of the concealed physiological guidance features; The interaction strength of the coupling relationship group is extracted to obtain the feature relaxation scale of the coupling relationship group; Based on the coupling relationship group, dynamic priority weighting is applied to the conflict constraint set to obtain the time-varying weight distribution of the conflict constraint set; Based on the time-varying weight distribution, the transition threshold of the potential deviation segment is calculated, where the formula for calculating the transition threshold is: ; In the formula, For the transition threshold, As a preset baseline transition potential, It is an exponential function. For the interaction strength norm, The characteristic relaxation scale, It is the hyperbolic tangent function. This represents the net change in the concealed physiological guidance characteristics. Sensitivity coefficient The Gaussian error function is... The average deviation of the potential deviation segment. The preset deviation reference value, The standard deviation of the preset deviation is... For spatial heterogeneity entropy, is the negative correlation decay coefficient of spatial heterogeneity entropy; Based on the transition threshold, morphological transitions are performed on potential deviation segments to obtain a cluster of transition trajectory segments for potential deviation segments; Clustering and integrating the transition trajectory segments yields candidate transition trajectory clusters with individualized swelling trends.

[0052] Based on the dynamic change sequence of impedance values ​​in different regions of the impedance distribution map, a lag and lead correlation analysis was performed on the temporal changes of integrated physiological state parameters. This analysis, using a sliding time window, calculated the correlation strength between impedance changes and physiological parameter changes under specific time lag conditions, identifying specific physiological parameter fluctuation patterns that consistently precede changes in macroscopic swelling indicators. These identified, forward-looking indicative fluctuation patterns were extracted from the original physiological parameter sequences and reconstructed into new time series; these new sequences represent the latent physiological guiding features of individualized swelling trends.

[0053] The predicted swelling intensity curve of the baseline trajectory and the intensity curve of the concealed physiological guidance feature are aligned on the same time axis, and point-by-point difference calculations are performed. This calculation process is continuous, and whenever the difference between multiple consecutive time points exceeds a preset tolerance range, the consecutive time period is marked as an inconsistency interval. Simultaneously, the slope change points of the guidance feature curve are monitored; if the baseline trajectory curve does not show the expected slope change at the corresponding point, the area surrounding that point is also marked as an inconsistency interval. All marked inconsistency intervals are connected chronologically to form several consecutive time periods; the set of these time periods constitutes the potential deviation segment of the baseline trajectory.

[0054] This paper analyzes several desired correction objectives within the potential deviation section, including conforming to the guidance characteristic curve, achieving a smooth transition, and naturally connecting with the historical trajectory. It analyzes the potential contradictions and conflicts that may arise when these objectives are simultaneously satisfied; for example, excessive conformity to the guidance characteristic may lead to an uneven trajectory. These contradictions are explicitly expressed as a series of conditions that cannot be fully satisfied simultaneously; the set of these conditions constitutes the conflict constraint set of the potential deviation section.

[0055] The coupling group is viewed as a network, where each spatial location is a node and the interaction strength is the weight of the connecting edge. The sum of squares of all edge weights in the network is calculated, and then the square root of this sum is taken. This result is a scalar measure characterizing the overall connectivity of the interaction network; this scalar measure is the interaction strength norm of the coupling group.

[0056] A statistical distribution analysis of the impedance values ​​of all pixels in the impedance distribution map is performed to calculate the information entropy of this distribution. Specifically, the range of impedance values ​​is divided into multiple equally wide intervals. The proportion of pixels falling into each interval is counted relative to the total number of pixels. These proportions are then used to calculate the information entropy using the definition of information entropy. The calculated information entropy value quantifies the degree of disorder and non-uniformity in the spatial distribution of impedance values; this entropy value is the spatial heterogeneity entropy of the impedance distribution map.

[0057] Calculate the absolute value sequence of the first-order difference of the time series of concealed physiological guidance characteristics per unit time. Calculate the mean and standard deviation of this absolute value sequence. The ratio of the mean to the standard deviation is used as a coefficient, reflecting the relative stability and significance level of the guidance characteristic's variation amplitude; this coefficient is the sensitivity coefficient of the concealed physiological guidance characteristics.

[0058] Analyze the distribution range of all interaction strength values ​​in the coupling group and calculate the standard deviation of these strength values. This standard deviation reflects the typical amplitude of the interaction strength fluctuations around its mean, and this typical amplitude of fluctuation is defined as the characteristic relaxation scale of the coupling group.

[0059] Based on the average interaction strength between the spatial locations involved in each constraint condition defined in the coupling relationship group, a priority weight is dynamically assigned to each constraint condition in the conflict constraint set. The higher the interaction strength, the higher the initial weight of the associated constraint condition. During the time evolution of the potential deviation segment, these weights are fine-tuned in real time according to the satisfaction status of each constraint condition at the previous moment, and the weights of constraints that are continuously unsatisfied are appropriately increased. This series of weight values ​​that change over time constitutes the time-varying weight distribution of the conflict constraint set.

[0060] A time-varying weight distribution is used to weight and integrate the conflict constraint set. The process involves quantifying the degree of non-compliance for each constraint into a score, multiplying this score by the constraint's current time-varying weight, and finally summing the weighted scores of all constraints. This weighted total score reflects the overall degree to which a specific set of trajectory correction schemes violates all current weighted constraints. By searching for the correction scheme that minimizes this weighted total score, the minimum weighted total score value corresponding to this scheme is defined as the transition threshold for the potential deviation segment.

[0061] Guided by the transition threshold, the original trajectory shape of the potential deviation segment is corrected. The operation involves attempting various local adjustment schemes on the time-swelling intensity two-dimensional plane of the segment, such as translating a curve, changing the local curvature, or inserting new inflection points. The degree to which each adjustment scheme achieves the transition threshold target is evaluated. All specific adjustment schemes that achieve an evaluation result at or above the transition threshold generate a corresponding corrected local trajectory segment. The set of these locally corrected trajectory segments constitutes the transition trajectory segment cluster of the potential deviation segment.

[0062] All trajectory segments within the transition trajectory cluster are grouped based on their morphological similarity. Similarity is measured by comparing the shape, key point locations, and time span of the trajectory segments. Trajectory segments with highly similar morphologies are grouped together, and then all trajectory segments within each group are merged and averaged to generate a new trajectory segment representing the consensus of that group. Finally, all consensus trajectory segments are connected in chronological order to form several complete candidate correction trajectories covering the entire potential deviation range. The set of these complete candidate correction trajectories constitutes the candidate transition trajectory cluster for individualized swelling trends.

[0063] The portions of the baseline trajectory not marked as potential deviations are left unchanged. At each potential deviation location, the corresponding candidate modified trajectory from the candidate transition trajectory cluster is used for replacement. This generates multiple complete possible trajectories covering the entire monitoring time range. These complete trajectories are traversed, and the overall smoothness, consistency with hidden physiological guidance characteristics, and fit with the patient's historical recovery pattern are evaluated. The complete trajectory with the highest overall score across all evaluation dimensions is selected, and this trajectory is defined as the dynamic synchronization trajectory for the individualized swelling trend.

[0064] In the formula, the baseline transition potential is directly derived from a preset fixed constant, representing the basic energy level required to initiate correction. The interaction strength norm is a scalar result obtained after topological quantization of the coupled network structure. The feature relaxation scale is a measure of the fluctuation range obtained after statistical analysis of all interaction strength values ​​in the coupled group. The net change of the concealed physiological guidance feature is the numerical difference between the start and end points of the feature's time series in the potential deviation segment. The sensitivity coefficient is a stability ratio obtained after statistical calculation of the temporal change rate of the concealed physiological guidance feature. The average deviation of the potential deviation segment is the arithmetic mean of the differences between the predicted baseline trajectory value and the actual guidance feature value at all time points within the segment. The baseline value and standard deviation of the deviation are both derived from preset reference constants. The spatial heterogeneity entropy is a scalar result obtained after calculating the information entropy of the pixel values ​​of the impedance distribution map. The negative correlation decay coefficient of the spatial heterogeneity entropy is calculated by dividing a preset proportional constant by the value of the spatial heterogeneity entropy.

[0065] The significance of this formula lies in its systematic fusion of multiple key factors influencing trajectory correction decisions through a product structure, outputting a unified quantitative threshold. The formula uses an exponential saturation function to process the ratio of the interaction strength norm to the feature relaxation scale, expressing the saturating promoting effect of the overall strength of the interaction network on correction initiation. It uses a hyperbolic tangent function to process the ratio of the net change in guiding features to the sensitivity coefficient, expressing the smoothing effect of the standardized change amplitude of the guiding signal on the correction direction. It uses a Gaussian error function to process the standardized difference of the average deviation relative to a preset benchmark, expressing the hierarchical impact on the urgency of correction when the current deviation is statistically typical or extreme. Finally, it uses an exponential decay function to process the ratio of spatial heterogeneity entropy to its decay coefficient, expressing the damping and suppressing effect of spatial inhomogeneity on correction behavior. The product of all factors constitutes the transition threshold, which comprehensively quantifies the permissible strength for trajectory morphological transitions under the current multidimensional conditions.

[0066] The formula exhibits a nonlinear dynamic response in the transition threshold as different input factors change. The threshold increases with the interaction strength norm, but the rate of increase gradually slows and eventually approaches an upper limit. Similarly, the threshold monotonically increases from zero when the net change in guiding features increases, also eventually reaching an upper limit. When the average deviation exceeds a preset baseline, the threshold rises rapidly from a stable value as the degree of excess increases. The threshold decays exponentially as spatial heterogeneity entropy increases. Overall, the formula ensures that the transition threshold reaches a high level only under the combined conditions of a sufficiently strong interaction network, sufficiently significant guiding signals, sufficiently prominent current deviations, and low spatial heterogeneity, thus allowing for more significant trajectory corrections; conversely, it suppresses aggressive correction behavior under these conditions.

[0067] The beneficial effect is that this method analyzes the causal relationship between impedance distribution and real-time physiological parameters to identify prospective hidden physiological guidance features, and then performs dynamic consistency matching with the baseline trajectory to locate potential deviation segments. Then, a modified candidate trajectory cluster is generated through a multi-objective collaborative transition process that includes conflict constraint analysis, spatial heterogeneity quantification, sensitivity coefficient calibration and dynamic weighting. Finally, a dynamic trajectory that is synchronized with the patient's actual pathophysiological evolution in real time is obtained through comprehensive selection.

[0068] This calculation formula integrates quantitative indicators generated from multi-dimensional preceding steps. Based on a nonlinear fusion mechanism, it unifies the influence of interactive network strength, guidance signal significance, real-time deviation degree, and spatial heterogeneity into a dynamic decision threshold. Its output value can adaptively adjust according to the changes in each input factor, thereby activating state correction when comprehensive conditions are met and suppressing over-adjustment when conditions are insufficient, thus achieving precise quantitative control of the timing and intensity of trajectory correction.

[0069] S06. Based on impedance distribution map and dynamic synchronization trajectory, perform multimodal joint assessment of individualized swelling trend to obtain swelling report of target limb area; In this embodiment of the invention, based on impedance distribution maps and dynamic synchronization trajectories, a multimodal joint assessment of individualized swelling trends is performed to obtain a swelling report for the target limb region, including: Based on the dynamic synchronization trajectory, spatial features are extracted from the impedance distribution map to obtain a multi-phase feature set of the impedance distribution map; Cross-phase feature transmission simulation was performed on a multi-phase feature set to obtain the feature transmission law of individualized swelling trend; The matching degree between the feature transmission law and the dynamic synchronization trajectory is quantized collaboratively to obtain the trajectory matching degree of the feature transmission law; Based on trajectory matching degree and preset clinical risk level, clinical consistency assessment of feature transmission pattern is performed, and the assessment results are semantically encapsulated to obtain swelling report of target limb area.

[0070] Based on the swelling evolution divided by the dynamic synchronization trajectory, the time window corresponding to each stage is mapped onto the time series of the impedance distribution map. For each time stage, the average value of all impedance distribution map pixels within that stage is calculated to form a feature image representing the typical spatial impedance distribution of that stage. For each feature image, geometric and statistical measures such as the location of extreme points in the spatial gradient distribution region and the area of ​​high-impedance connected regions are further extracted. The feature images corresponding to all time stages and their extracted measure data are organized chronologically; this set of feature data organized by time phase is the multi-phase feature set of the impedance distribution map.

[0071] Pixel-by-pixel comparative analysis is performed on feature images from adjacent time stages within a multi-phase feature set. For each pixel location, the sequence of impedance value changes over time is tracked, the amount of change in this sequence between adjacent stages is calculated, and the direction of change is recorded. Simultaneously, the spatial expansion or contraction patterns of high-impedance regions across different stages are analyzed to identify the trend of impedance changes propagating along specific anatomical pathways. These pixel-by-pixel changes are integrated with regional change patterns to construct a set of rules describing how impedance changes propagate temporally from early to late stages and spatially from the central region to the peripheral region. This set of rules constitutes the characteristic propagation law of individualized swelling trends.

[0072] The expected swelling propagation pattern described in the feature propagation law is compared with the actual swelling intensity change curve over time recorded by the dynamic synchronization trajectory. The comparison process first converts the predicted swelling spatial expansion rate in the feature propagation law into a predicted growth rate of the overall swelling intensity. Then, this predicted growth rate curve is aligned with the slope curve of the actual dynamic synchronization trajectory on the same time axis. The degree of difference between the two curves at each time point is calculated, and a weighted average is calculated over the entire time range, assigning lower weights to time points with larger differences to reduce the impact of random fluctuations. Finally, the calculated weighted average difference is normalized to obtain a value between zero and one; this value represents the trajectory matching degree of the feature propagation law.

[0073] The trajectory matching score is input into a pre-defined clinical risk grading table. This table defines the clinical significance of different matching score ranges; for example, a high matching score indicates that the swelling development is in line with expectations and the risk is low, while a low matching score indicates that the swelling development is abnormal and the risk is high. The clinical risk level of the current swelling state is determined based on the range in which the matching score falls. Simultaneously, a descriptive text describing the spatial location and direction of swelling is generated by combining the specific transmission path identified in the feature transmission pattern with the anatomical structure. The risk level and spatial description text are integrated, and information on the intensity and trend of the current swelling is added. Organized into coherent paragraphs according to the standard clinical report format, the final output is a structured text, i.e., the swelling report for the target limb region.

[0074] The beneficial effect is that this method extracts typical spatial features of multiple stages from the impedance distribution based on the time phase of the dynamic synchronous trajectory, and then simulates the transmission law of these features in the spatiotemporal dimension to reveal the intrinsic pattern of swelling diffusion. Then, by quantifying the degree of coordination between the transmission law and the actual trajectory, the consistency of the development process is evaluated. Finally, this degree of coordination is combined with clinical risk criteria and spatial positioning information to generate a structured swelling report, realizing the comprehensive transformation from multimodal data to quantitative assessment and qualitative decision support.

[0075] like Figure 2 The diagram shown is a functional block diagram of a postoperative limb swelling monitoring system provided in an embodiment of the present invention.

[0076] This invention discloses a postoperative limb swelling monitoring system 10 for orthopedic surgery, which can be installed in an electronic device. Depending on the functions implemented, the postoperative limb swelling monitoring system 10 may include an association mining module 11, a signal integration module 12, an impedance mapping module 13, a trend inference module 14, a dynamic correction module 15, and an evaluation report module 16. The modules of this invention can also be referred to as units, which are a series of computer program segments that can be executed by the processor of an electronic device and perform a fixed function, and are stored in the memory of the electronic device.

[0077] In this embodiment, the functions of each module / unit are as follows: The association mining module 11 is used to mine the interaction relationship of cross-modal association patterns of multi-source heterogeneous data in the target limb region, and obtain the coupling relationship group of the target limb region. Signal integration module 12 is used to perform multi-scale physiological signal separation on multi-source heterogeneous data based on coupling relationship groups to obtain integrated physiological state parameters of the target limb region; Impedance mapping module 13 is used to analyze the local microcirculation impedance characteristics of the target limb region based on the integrated physiological state parameters, and to perform spatial topological mapping on the local microcirculation impedance characteristics to obtain the impedance distribution map of the target limb region. Trend extrapolation module 14 is used to extrapolate the individualized swelling trend of the target limb region based on the impedance distribution map, and obtain the baseline trajectory of the individualized swelling trend. The dynamic correction module 15 is used to perform feedback dynamic correction on the baseline trajectory based on integrated physiological state parameters and impedance distribution map to obtain a dynamic synchronous trajectory of individualized swelling trend. The assessment report module 16 is used to perform multimodal joint assessment of individualized swelling trends based on impedance distribution maps and dynamic synchronization trajectories, and obtain a swelling report for the target limb area.

[0078] In the several embodiments provided by this invention, it should be understood that the disclosed methods and systems can be implemented in other ways. For example, the system embodiments described above are merely illustrative; for instance, the division of modules is only a logical functional division, and other division methods may be used in actual implementation.

[0079] The modules described as separate components may or may not be physically separate. The components shown as modules may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs.

[0080] Furthermore, the functional modules in the various embodiments of the present invention can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or in the form of hardware plus software functional modules.

[0081] It will be apparent to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above, and that the present invention can be implemented in other specific forms without departing from the spirit or essential characteristics of the present invention.

[0082] This application embodiment can acquire and process relevant data based on artificial intelligence technology. Artificial intelligence is the theory, method, technology, and application system that uses digital computers or machines controlled by digital computers to simulate, extend, and expand human intelligence, perceive the environment, acquire knowledge, and use that knowledge to obtain optimal results.

[0083] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention.

Claims

1. A method for monitoring limb swelling after orthopedic surgery, characterized in that, The methods include: S01. Perform interaction relationship mining on the cross-modal association patterns of multi-source heterogeneous data in the target limb region to obtain the coupling relationship group of the target limb region; S02. Based on the coupling relationship group, multi-scale physiological signal separation is performed on multi-source heterogeneous data to obtain integrated physiological state parameters of the target limb region; S03. Based on the integrated physiological state parameters, analyze the local microcirculation impedance characteristics of the target limb region, and perform spatial topological mapping on the local microcirculation impedance characteristics to obtain the impedance distribution map of the target limb region. S04. Based on the impedance distribution map, the individualized swelling trend of the target limb region is deduced to obtain the baseline trajectory of the individualized swelling trend. S05. Based on the integrated physiological state parameters and impedance distribution map, the baseline trajectory is dynamically corrected by feedback to obtain a dynamic synchronous trajectory of individualized swelling trend. S06. Based on impedance distribution maps and dynamic synchronization trajectories, a multimodal joint assessment of individualized swelling trends is performed to obtain a swelling report for the target limb region.

2. The method for monitoring limb swelling after orthopedic surgery as described in claim 1, characterized in that, Interaction relationship mining was performed on cross-modal association patterns of multi-source heterogeneous data in the target limb region to obtain coupling relationship groups of the target limb region, including: Extract preliminary correlation features from multi-source heterogeneous data in the target limb region; Based on preliminary association features, a collaborative evolution analysis of cross-modal association patterns of multi-source heterogeneous data is performed to obtain an initial association pattern set for the target limb region. Cross-scale interaction relationships are constructed on the initial association pattern group to obtain the nonlinear metric relationship of the initial association pattern group; The physiological interpretability of the nonlinear metric relationship was verified, and the verified relationship was fused and screened to obtain the coupling relationship group of the target limb region.

3. The method for monitoring limb swelling after orthopedic surgery as described in claim 1, characterized in that, Based on the coupling relationship group, multi-scale physiological signal separation is performed on multi-source heterogeneous data to obtain integrated physiological state parameters of the target limb region, including: Based on the coupling relationship group, dynamic physiological rhythm analysis is performed on multi-source heterogeneous data to obtain the temporal physiological elements of multi-source heterogeneous data; Regional attribution analysis was performed on temporal physiological elements to obtain their spatial attribution characteristics. Multi-level information gain fusion of spatial attribution features yields physiological state primitives of the target limb region; By performing state regularization on the physiological state primitives, integrated physiological state parameters of the target limb region are obtained.

4. The method for monitoring limb swelling after orthopedic surgery as described in claim 1, characterized in that, Based on integrated physiological state parameters, the local microcirculatory impedance characteristics of the target limb region are analyzed, and spatial topological mapping is performed on the local microcirculatory impedance characteristics to obtain the impedance distribution map of the target limb region, including: Based on integrated physiological state parameters, regional differences in tissue perfusion and metabolic load in the target limb region were traced to obtain the trend of relative impedance changes in the target limb region. Based on the coupling relationship group, impedance characteristic inversion is performed on the relative impedance change trend to obtain a quantitative impedance characteristic map of local microcirculation in the target limb region. Physiological constraint interpolation is performed on the quantized impedance characteristic map to obtain the basic impedance spatial distribution map of the target limb region; Anisotropic smoothing is performed on the basic impedance spatial distribution map to obtain the impedance distribution map of the target limb region.

5. A method for monitoring limb swelling after orthopedic surgery as described in claim 4, characterized in that, Based on the coupling relationship group, impedance characteristic inversion is performed on the relative impedance change trend to obtain a quantitative impedance characteristic map of local microcirculation in the target limb region, including: Spatial synchronous weight allocation is performed on the relative impedance change trend and coupling relationship group to obtain the relationship distribution of the relative impedance change trend; Based on the coupling relationship group, the relationship distribution is solved point by point iteratively to obtain the impedance characteristic value of the relative impedance change trend. The formula for calculating the impedance characteristic value is as follows: ； In the formula, For the first Position after the next iteration The impedance characteristic value at that point, Position in the coupling relationship group With position The strength of interaction between them For position The trend of relative impedance change at the location, The trend of interaction intensity and relative impedance variation in the target limb region The weighted integral on, Interaction intensity in the target limb area The points on the top The preset regularization adjustment coefficient is used. For position The mean of the interaction intensity at that location. For the first Position after the next iteration The impedance characteristic value at that point, For the first Position after the next iteration The impedance characteristic value at that point, The deviation in interaction strength and the difference in impedance characteristics in the target limb region Cooperative correction integrals on; By extending the impedance characteristic values ​​to physiological continuity, a quantitative impedance characteristic map of local microcirculation in the target limb region is obtained.

6. The method for monitoring limb swelling after orthopedic surgery as described in claim 1, characterized in that, Based on impedance distribution maps, path extrapolation is performed on individualized swelling trends in the target limb region to obtain the baseline trajectory of the individualized swelling trend, including: Based on the limb anatomy and postoperative suture tension distribution in the target limb region, morphological features of the local gradient changes in the impedance distribution map are extracted to obtain an initial diffusion feature set of individualized swelling trends in the target limb region. Based on integrated physiological state parameters, specific morphological corrections are performed on the initial diffusion feature set to obtain an individualized diffusion feature set of swelling trends. Based on the historical recovery rhythm characteristics of the target limb region, the temporal evolution simulation of the diffusion feature set is performed to obtain the temporal development path set of individualized swelling trends; Based on the physiological characteristics of the target limb region, consistency screening is performed on the temporal development path set to obtain the baseline trajectory of individualized swelling trends.

7. The method for monitoring limb swelling after orthopedic surgery as described in claim 1, characterized in that, Based on integrated physiological state parameters and impedance distribution maps, a feedback-based dynamic correction is performed on the baseline trajectory to obtain a dynamic synchronous trajectory for individualized swelling trends, including: Based on impedance distribution maps, causal correlation analysis was performed on integrated physiological state parameters to obtain the hidden physiological guidance characteristics of individualized swelling trends. Dynamic consistency matching between the baseline trajectory and the concealed physiological guidance features is performed to obtain the potential deviation segments of the baseline trajectory; Multi-objective collaborative transitions are performed on potential deviation segments to obtain a cluster of candidate transition trajectories for individualized swelling trends; By traversing and selecting candidate transition trajectory clusters, dynamic synchronous trajectories with individualized swelling trends are obtained.

8. A method for monitoring limb swelling after orthopedic surgery as described in claim 7, characterized in that, Multi-objective collaborative transitions are performed on potential deviation segments to obtain a cluster of candidate transition trajectories for individualized swelling trends, including: Multidimensional conflict analysis is performed on the potential deviation section to obtain the conflict constraint set of the potential deviation section; The global interaction strength of the coupling relationship group is quantized by topological structure strength quantization to obtain the interaction strength norm of the coupling relationship group; The spatial dispersion of the numerical distribution of the impedance distribution is evaluated to obtain the spatial heterogeneity entropy of the impedance distribution. The temporal change rate of the concealed physiological guidance features was statistically calibrated to obtain the sensitivity coefficient of the concealed physiological guidance features; The interaction strength of the coupling relationship group is extracted to obtain the feature relaxation scale of the coupling relationship group; Based on the coupling relationship group, dynamic priority weighting is applied to the conflict constraint set to obtain the time-varying weight distribution of the conflict constraint set; Based on the time-varying weight distribution, the transition threshold of the potential deviation segment is calculated, where the formula for calculating the transition threshold is: ； In the formula, For the transition threshold, As a preset baseline transition potential, It is an exponential function. For the interaction strength norm, The characteristic relaxation scale, It is the hyperbolic tangent function. This represents the net change in the concealed physiological guidance characteristics. Sensitivity coefficient The Gaussian error function is... The average deviation of the potential deviation segment. The preset deviation reference value, The standard deviation of the preset deviation is... For spatial heterogeneity entropy, is the negative correlation decay coefficient of spatial heterogeneity entropy; Based on the transition threshold, morphological transitions are performed on potential deviation segments to obtain a cluster of transition trajectory segments for potential deviation segments; Clustering and integrating the transition trajectory segments yields candidate transition trajectory clusters with individualized swelling trends.

9. A method for monitoring limb swelling after orthopedic surgery as described in claim 1, characterized in that, Based on impedance distribution maps and dynamic synchronization trajectories, a multimodal joint assessment of individualized swelling trends is performed to obtain a swelling report for the target limb region, including: Based on the dynamic synchronization trajectory, spatial features are extracted from the impedance distribution map to obtain a multi-phase feature set of the impedance distribution map; Cross-phase feature transmission simulation was performed on a multi-phase feature set to obtain the feature transmission law of individualized swelling trend; The matching degree between the feature transmission law and the dynamic synchronization trajectory is quantized collaboratively to obtain the trajectory matching degree of the feature transmission law; Based on trajectory matching degree and preset clinical risk level, clinical consistency assessment of feature transmission pattern is performed, and the assessment results are semantically encapsulated to obtain swelling report of target limb area.

10. A postoperative limb swelling monitoring system for orthopedic surgery, characterized in that, A method for monitoring postoperative limb swelling in orthopedic surgery according to claim 1, the system comprising: The association mining module is used to mine the interaction relationship of cross-modal association patterns of multi-source heterogeneous data in the target limb region, and obtain the coupling relationship group of the target limb region; The signal integration module is used to separate multi-scale physiological signals from multi-source heterogeneous data based on coupling relationship groups, so as to obtain integrated physiological state parameters of the target limb region. The impedance mapping module is used to analyze the local microcirculation impedance characteristics of the target limb region based on integrated physiological state parameters, and to perform spatial topological mapping on the local microcirculation impedance characteristics to obtain the impedance distribution map of the target limb region. The trend extrapolation module is used to extrapolate the individualized swelling trend of the target limb region based on the impedance distribution map, and obtain the baseline trajectory of the individualized swelling trend. The dynamic correction module is used to perform feedback dynamic correction on the baseline trajectory based on integrated physiological state parameters and impedance distribution map to obtain a dynamic synchronous trajectory of individualized swelling trend. The assessment report module is used to perform multimodal joint assessment of individualized swelling trends based on impedance distribution maps and dynamic synchronization trajectories, and obtain a swelling report for the target limb area.

Images