A cardiac pacemaker multi-source heterogeneous clinical data fusion method and system

Abstract

The application discloses a kind of cardiac pacemaker multi-source heterogeneous clinical data fusion method and system, it is related to cardiac pacemaker technical field, and its technical solution points include the following steps: obtaining the multimodal clinical data of patient cardiac surgery postoperative temporary pacemaker, from the fusion feature information for parameter uncertainty quantification analysis in multimodal clinical data Extraction;Sampling is carried out based on cloud model to the key input parameter, and simulation is run to quantify the dynamic influence of the uncertainty distribution range of each key input parameter on pacemaker system output index to obtain uncertainty quantification result;Uncertainty quantification result is based on the sorting result of key input parameter sequencing, according to the sorting result, generate decision support information for guiding clinical adjustment temporary pacemaker parameter, effect is to guarantee the safety and stability of postoperative cardiac rhythm management of patient.

Description

Technical Field

[0001] This invention relates to the field of cardiac pacemaker technology, and more specifically, to a method and system for fusing multi-source heterogeneous clinical data of cardiac pacemakers. Background Technology

[0002] Post-cardiac surgery patients often require temporary pacemakers to maintain stable heart rhythm. However, current clinical practice lacks quantitative analysis and guidance for adjusting the parameters of temporary pacemakers, resulting in insufficient precision and efficiency. This makes it impossible to fully capture the correlation between the patient's post-operative dynamic physiological changes and the pacemaker's operational status. Some methods rely on simple statistical approaches to assess parameter impact, failing to quantify the dynamic impact of parameter uncertainty on pacemaker output indicators. Furthermore, they cannot construct a two-way mapping relationship between the qualitative state and quantitative fluctuations of parameters, leading to a lack of clinical interpretability in the analysis results. In addition, existing methods mostly remain at the data processing and analysis level, failing to form a complete closed loop from data fusion and uncertainty quantification to clinical decision support, thus unable to directly provide physicians with structured guidance for parameter adjustment. Summary of the Invention

[0003] To address the shortcomings of existing technologies, the present invention aims to provide a method and system for fusing multi-source heterogeneous clinical data of cardiac pacemakers.

[0004] To achieve the above objectives, the present invention provides the following technical solution: A method for fusing multi-source heterogeneous clinical data from cardiac pacemakers includes the following steps: Acquire multimodal clinical data of temporary pacemakers after cardiac surgery in patients, and extract fusion feature information from the multimodal clinical data for parameter uncertainty quantification analysis; Based on the fusion of feature information, the key input parameters affecting the working state of the temporary pacemaker and their uncertainty distribution range are determined, and a cloud model is established to quantify the uncertainty of the key input parameters. Based on the cloud model, the key input parameters are sampled and simulations are run to quantify the dynamic impact of the uncertainty distribution range of each key input parameter on the output index of the pacemaker system, and uncertainty quantification results are obtained. The key input parameters are sorted based on the uncertainty quantification results to obtain the sorting results. Decision support information is then generated based on the sorting results to guide clinical adjustments of temporary pacemaker parameters.

[0005] Preferably, the multimodal clinical data includes patient information, temporary pacemaker information, and time-series physiological signal data.

[0006] Preferably, the patient information includes the patient's baseline data and surgical data; the temporary pacemaker information includes the brand and model of the temporary pacemaker, initial parameter settings, and adjustment records.

[0007] Preferably, the key input parameters affecting the working state of the temporary pacemaker and their uncertainty distribution range are determined based on fused feature information, specifically as follows: By performing sensitivity analysis on the fusion feature information of the cardiac pacemaker in-loop simulation system, the key input parameters affecting the working state of the temporary pacemaker and their uncertainty distribution range are determined.

[0008] Preferably, the key input parameters are sampled based on a cloud model, and simulations are run to quantify the dynamic impact of the uncertainty of each key input parameter on the output indicators of the pacemaker system, thereby obtaining uncertainty quantification results. Specifically, this includes the following steps: Based on the cloud model, the Wilks method is used to determine the minimum sample size, and the key input parameters are sampled to generate a parameter sample set. The parameter sample set is input into the loop simulation system of the cardiac pacemaker hardware, and the dynamic impact of the output index is used to obtain the uncertainty quantification result.

[0009] Preferably, extracting fusion feature information from multimodal clinical data for parameter uncertainty quantification analysis specifically includes the following steps: The multimodal clinical data is normalized to obtain a normalized clinical data set; Feature mining is performed on a normalized clinical dataset to obtain a multidimensional set of related features; Cross-domain feature fusion is performed on a multidimensional associated feature set to generate a fused feature set; Feature filtering is performed on the fused feature set to obtain fused feature information.

[0010] Preferably, a cloud model is established to quantify the uncertainty of key input parameters, specifically including the following steps: Extract the discrete fluctuation patterns and continuous changing trends of key input parameters to obtain a set of parameter uncertainty characteristics; The set of parameter uncertainty features is normalized to obtain the standardized uncertainty characterization result; Based on the results of standardized uncertainty characterization, an uncertainty mapping relationship is constructed, and a two-way correspondence mechanism between the qualitative state of parameters and quantitative fluctuations is established to form a cloud model.

[0011] Preferably, the ranking of key input parameters based on the uncertainty quantification results is obtained by sorting the parameters, specifically including the following steps: Extract the strength and depth of the influence of key input parameters on the pacing system's operating status to obtain a set of parameter influence features; The standardized influence degree characterization result is obtained by calibrating the influence degree of the parameter influence feature set; Based on the standardized impact characterization results, a parameter correlation comparison system is constructed. Based on the parameter correlation comparison system, the differences in impact weights and the primary and secondary relationships among the parameters are clarified, and the parameter comparison analysis results are obtained. The results of the parameter comparison and analysis were prioritized and screened to obtain the initial screening results of parameter priorities. The ranking results are obtained by sorting the key input parameters according to the initial screening results based on parameter priority.

[0012] Preferably, the decision support information generated based on the ranking results to guide clinical adjustments of temporary pacemaker parameters specifically includes the following steps: Clinical fit analysis based on the ranking results yields a set of parameter clinical fit features. Clinical risk association labeling was performed on the set of clinically adaptable features of parameters to obtain parameter risk association characterization results; Based on the parameter risk association characterization results, a clinical adjustment logic system is constructed. Based on the clinical adjustment logic system, the priority and adaptation direction of parameter adjustment under different risk levels are clarified, and the parameter adjustment logic results are formed. A set of feasible adjustment schemes is obtained by performing clinical feasibility verification on the results of parameter adjustment logic. Integrate a set of feasible adjustment options to generate structured decision support information.

[0013] A multi-source heterogeneous clinical data fusion system for cardiac pacemakers includes: Extraction module: Acquires multimodal clinical data of temporary pacemakers after cardiac surgery in patients, and extracts fusion feature information from the multimodal clinical data for parameter uncertainty quantification analysis; Establishment Module: Based on fused feature information, determine the key input parameters affecting the working state of the temporary pacemaker and their uncertainty distribution range, and establish a cloud model for quantifying the uncertainty of key input parameters; Processing module: Based on the cloud model, the key input parameters are sampled and simulation is run to quantify the dynamic impact of the uncertainty distribution range of each key input parameter on the output index of the pacemaker system to obtain uncertainty quantification results; Generation module: Based on the uncertainty quantification results, the key input parameters are sorted to obtain the sorting results, and decision support information is generated based on the sorting results to guide the clinical adjustment of temporary pacemaker parameters.

[0014] Compared with the prior art, the present invention has the following beneficial effects: This invention establishes fused feature information by organizing, mining, cross-domain fusion, and filtering patient information, pacemaker data, and physiological signals. A cloud-based parameter uncertainty quantification method is then used to identify and evaluate key parameters. Sensitivity analysis determines key parameters and their distribution range. Cloud model construction and simulation sampling quantify the dynamic impact of parameter fluctuations on pacemaker operation, ensuring statistical reliability of the analysis results while intuitively presenting the clinical significance of parameter changes. This addresses the problems of traditional parameter adjustment relying on experience and lacking quantitative evidence. This solution constructs a complete clinical decision support system, generating structured parameter adjustment guidance information through parameter priority ranking, risk association calibration, and feasibility verification. It clarifies the priority, adaptation direction, and safety boundaries of parameter adjustments, effectively improving the accuracy and efficiency of temporary pacemaker parameter adjustments, reducing the risk of complications such as arrhythmias and pacing failure caused by improper parameter adjustments, and ensuring the safety and stability of postoperative cardiac rhythm management for patients. Attached Figure Description

[0015] Figure 1 This is a flowchart illustrating a method for fusing multi-source heterogeneous clinical data of a cardiac pacemaker, as provided in an embodiment of the present invention. Figure 2 This is a schematic diagram of a multi-source heterogeneous clinical data fusion system for cardiac pacemakers provided in an embodiment of the present invention. Detailed Implementation

[0016] To make the above-mentioned objects, features and advantages of the present invention more apparent and understandable, the specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings.

[0017] Many specific details are set forth in the following description in order to provide a full understanding of the invention. However, the invention may also be practiced in other ways different from those described herein, and those skilled in the art can make similar extensions without departing from the spirit of the invention. Therefore, the invention is not limited to the specific embodiments disclosed below.

[0018] Secondly, the term "an embodiment" or "embodiment" as used herein refers to a specific feature, structure, or characteristic that may be included in at least one implementation of the present invention. The phrase "in one embodiment" appearing in different places throughout this specification does not necessarily refer to the same embodiment, nor is it a single embodiment or an embodiment selectively excluded from other embodiments.

[0019] Reference Figures 1-2 As shown.

[0020] The embodiments further illustrate the multi-source heterogeneous clinical data fusion method and system for cardiac pacemakers proposed in this invention.

[0021] A method for fusing multi-source heterogeneous clinical data from cardiac pacemakers includes the following steps: Acquire multimodal clinical data of temporary pacemakers after cardiac surgery in patients, and extract fusion feature information from the multimodal clinical data for parameter uncertainty quantification analysis; Based on the fusion of feature information, the key input parameters affecting the working state of the temporary pacemaker and their uncertainty distribution range are determined, and a cloud model is established to quantify the uncertainty of the key input parameters. Based on the cloud model, key input parameters are sampled and simulations are run to quantify the dynamic impact of the uncertainty distribution range of each key input parameter on the output indicators of the pacemaker system, and uncertainty quantification results are obtained. The key input parameters are sorted based on the uncertainty quantification results to obtain the sorting results. Decision support information is then generated based on the sorting results to guide clinical adjustments of temporary pacemaker parameters.

[0022] Multimodal clinical data includes patient information, temporary pacemaker information, and time-series physiological signal data.

[0023] Patient information includes the patient's baseline data and surgical data; temporary pacemaker information includes the brand and model of the temporary pacemaker, initial parameter settings, and adjustment records.

[0024] Patient information includes baseline data and surgical data. Baseline data includes the patient's age, gender, and underlying disease information, which determines the patient's basic physiological state and the basic conditions for pacemaker parameter matching. Surgical data records the patient's surgical procedure, operation duration, and cardiopulmonary bypass time, which directly reflects the impact of the surgery on the patient's cardiac physiological state. Temporary pacemaker information includes the brand and model of the temporary pacemaker, initial parameter settings, and adjustment records. The brand and model determine the pacemaker's hardware performance and adjustable parameter range. The initial parameter settings are the baseline parameters for the pacemaker's first operation, and the adjustment records fully present the process of adjusting the pacemaker parameters in clinical practice.

[0025] Time-series physiological signal data are continuous physiological records of patients after surgery, including electrocardiogram signals, heart rate changes, blood pressure fluctuations, and blood oxygen saturation changes. This type of data dynamically reflects the patient's real-time physiological state in the form of a continuous time series, and can intuitively show the matching between pacemaker operating parameters and the patient's physiological state, providing dynamic physiological basis for parameter uncertainty analysis and adjustment. By complementing data from the dimensions of patient baseline conditions, device operation status, and real-time physiological state, a comprehensive and reliable data foundation is provided for feature extraction, parameter analysis, and decision support.

[0026] Based on fused feature information, the key input parameters affecting the working state of the temporary pacemaker and their uncertainty distribution range are determined, specifically: By performing sensitivity analysis on the fusion feature information of the cardiac pacemaker in-loop simulation system, the key input parameters affecting the working state of the temporary pacemaker and their uncertainty distribution range are determined.

[0027] The fused feature information is input into a cardiac pacemaker-in-the-loop simulation system, and sensitivity analysis is used to systematically evaluate the correlation between different input parameters and the pacemaker's operating state. Sensitivity analysis quantifies the impact of changes in each parameter on the pacemaker's core operating indicators, thereby identifying parameters with significant impacts as key input parameters. Simultaneously, during the analysis, the parameter fluctuations recorded in clinical data are combined with the simulation results to determine the reasonable fluctuation range of each key input parameter in actual clinical application, i.e., its uncertainty distribution range. This allows for the identification of parameters with the greatest impact on the pacemaker's operating state, while clarifying the adjustable range of these parameters in actual use, providing a clear analytical object and boundary conditions for uncertainty quantification analysis.

[0028] Based on the cloud model, key input parameters are sampled, and simulations are run to quantify the dynamic impact of the uncertainty of each key input parameter on the output indicators of the pacemaker system, thereby obtaining the uncertainty quantification results. The specific steps include: Based on the cloud model, the Wilks method is used to determine the minimum sample size, and the key input parameters are sampled to generate a parameter sample set. The parameter sample set is input into the loop simulation system of the cardiac pacemaker hardware, and the dynamic impact of the output index is used to obtain the uncertainty quantification result.

[0029] Based on a cloud model, the Wilks method is used to determine the minimum sample size required to meet statistical confidence requirements. This effectively controls the computational cost and time consumption of the simulation while ensuring the reliability of the analysis results, avoiding resource waste due to excessively large sample sizes or result bias due to insufficient sample sizes. Following the determined minimum sample size, key input parameters are sampled based on the distribution characteristics of the cloud model, generating parameter sample sets containing multiple parameter combinations. These samples comprehensively cover the uncertainty distribution range of key input parameters, realistically reflecting the fluctuation characteristics of parameters in clinical scenarios. The parameter sample sets are then input one by one into the hardware-in-the-loop simulation system for a cardiac pacemaker. The simulation process is run for each parameter combination, simulating the pacemaker's operating state under different parameter settings and collecting dynamic change data of various output indicators of the pacemaker system. By statistically analyzing all simulation results, the uncertainty distribution range of each key input parameter and its dynamic impact on the pacemaker system's output indicators can be quantified, ultimately yielding uncertainty quantification results. These results will intuitively present the intensity and pattern of the impact of different parameter fluctuations on pacemaker performance, providing reliable data for parameter ranking and clinical decision support.

[0030] Extracting fusion feature information from multimodal clinical data for parameter uncertainty quantification analysis includes the following steps: The multimodal clinical data is normalized to obtain a normalized clinical data set; Feature mining is performed on a normalized clinical dataset to obtain a multidimensional set of related features; Cross-domain feature fusion is performed on a multidimensional associated feature set to generate a fused feature set; Feature filtering is performed on the fused feature set to obtain fused feature information.

[0031] Regularization of multimodal clinical data yields a regularized clinical dataset. Since multimodal clinical data varies in source and format, and sampling rates, units, and storage methods differ across data types, regularization standardizes the data format and sampling frequency, appropriately fills in missing values, filters and denoises noise signals, and eliminates data heterogeneity. This provides a unified processing foundation for data from different sources, resulting in a structured and reliable regularized clinical dataset. Feature mining is then performed on the regularized clinical dataset to obtain a multidimensional set of related features. Adaptive feature extraction methods are employed for different data types. Static features related to pacemaker adaptation are extracted from patient information, parameter variation patterns are extracted from pacemaker operation data, and time-frequency and dynamic change features are extracted from time-series physiological signals. Simultaneously, the correlations between different data types are explored, resulting in a multi-dimensional set of correlated features covering patient status, device operation, and physiological changes. Cross-domain feature fusion is then performed on this multi-dimensional correlated feature set to generate a fused feature set. This integrates features from patient information, pacemaker information, and physiological signal data across domains, establishing correlation mapping relationships between features to achieve complementarity and fusion of features from different dimensions. This generates a fused feature set that simultaneously reflects the patient's baseline status, pacemaker operation, and real-time physiological response, eliminating the limitations of single-data-domain features. The fused feature set undergoes feature filtering to obtain fused feature information. Redundant and irrelevant features are removed, retaining features relevant to subsequent parameter analysis and decision-making, thus forming the fused feature information.

[0032] And establish a cloud model for quantifying the uncertainty of key input parameters, specifically including the following steps: Extract the discrete fluctuation patterns and continuous changing trends of key input parameters to obtain a set of parameter uncertainty characteristics; The set of parameter uncertainty features is normalized to obtain the standardized uncertainty characterization result; Based on the results of standardized uncertainty characterization, an uncertainty mapping relationship is constructed, and a two-way correspondence mechanism between the qualitative state of parameters and quantitative fluctuations is established to form a cloud model.

[0033] Discrete fluctuation data of key input parameters during actual use are extracted from clinical data and simulation records, including single adjustment values ​​and value distributions in different scenarios. Simultaneously, the continuous trend of parameter changes with the patient's physiological state is captured, recording continuous change curves of parameters at different time points. These discrete and continuous features are integrated to form a feature set comprehensively reflecting parameter uncertainty. This feature set is then normalized to obtain standardized uncertainty representation results. Since the value ranges, units, and fluctuation amplitudes of different key input parameters vary, the parameter uncertainty feature set is normalized to eliminate the influence of dimensions and numerical ranges, mapping the uncertainty features of all parameters to a unified numerical interval, resulting in standardized uncertainty representation results. This ensures the comparability of uncertainty features of different parameters. Based on the standardized uncertainty representation results, uncertainty mapping relationships are constructed, establishing a two-way correspondence mechanism between the qualitative state and quantitative fluctuation of parameters, forming a cloud model. Based on the standardized uncertainty characterization results, a mapping relationship between the quantitative values ​​of parameters and their qualitative states is constructed. The specific fluctuation values ​​of parameters are established in a two-way correspondence with qualitative descriptions such as high, low, and moderate. This allows the quantitative changes of parameters to be transformed into intuitive qualitative states. At the same time, the reasonable quantitative fluctuation range of parameters can be inferred from the qualitative states, ultimately forming a cloud model.

[0034] The ranking results are obtained by sorting the key input parameters based on the uncertainty quantification results, specifically including the following steps: Extract the strength and depth of the influence of key input parameters on the pacing system's operating status to obtain a set of parameter influence features; The standardized influence degree characterization result is obtained by calibrating the influence degree of the parameter influence feature set; Based on the standardized impact characterization results, a parameter correlation comparison system is constructed. Based on the parameter correlation comparison system, the differences in impact weights and the primary and secondary relationships among the parameters are clarified, and the parameter comparison analysis results are obtained. The results of the parameter comparison and analysis were prioritized and screened to obtain the initial screening results of parameter priorities. The ranking results are obtained by sorting the key input parameters according to the initial screening results based on parameter priority.

[0035] The strength and depth of the influence of key input parameters on the pacing system's operational status are extracted to obtain a set of parameter influence features. The impact magnitude of each key input parameter's fluctuation on various output indicators of the pacing system is extracted from the uncertainty quantification results. Simultaneously, the degree of correlation between parameters and system operational status is determined, forming a set of parameter influence features that includes both influence strength and correlation depth. The influence degree of this set is then calibrated to obtain standardized influence degree representation results. Since the dimensions and numerical ranges of the influence features of different parameters vary, these features are normalized, transforming influence strength and correlation depth into influence degree values ​​of a uniform scale. This eliminates the influence of dimensional differences, resulting in standardized influence degree representation results that ensure direct comparison of the influence levels of different parameters. Based on the standardized influence degree representation results, a parameter correlation comparison system is constructed to clarify the differences in influence weights and primary / secondary relationships among parameters, yielding parameter comparison analysis results. A multi-parameter comparison framework is established based on standardized influence degrees. The weight percentage of each parameter in the overall influence is calculated, distinguishing between primary and secondary influencing parameters, determining the interactions and correlations among different parameters, forming parameter comparison analysis results, and clarifying the primary / secondary order of parameters. The subsequent priority screening of parameter comparison analysis results yielded initial priority results. Based on preset impact thresholds and clinical importance criteria, the parameter comparison analysis results were screened, eliminating parameters with impact below the threshold and retaining key parameters that significantly affect the pacing system's operational status. Simultaneously, secondary parameters that interfered with each other were excluded, resulting in the initial priority screening results. The key input parameters were then ranked according to their initial priority screening results, generating a ranking list. This list, arranged from highest to lowest impact, clearly presents the order of influence of each parameter on the pacing system, providing a priority basis for clinical adjustment decisions. This ensures that the parameters with the greatest impact are prioritized during clinical adjustments, improving the targeting and efficiency of parameter adjustments.

[0036] Based on the ranking results, decision support information is generated to guide clinical adjustments of temporary pacemaker parameters. This includes the following steps: Clinical fit analysis based on the ranking results yields a set of parameter clinical fit features. Clinical risk association labeling was performed on the set of clinically adaptable features of parameters to obtain parameter risk association characterization results; Based on the parameter risk association characterization results, a clinical adjustment logic system is constructed. Based on the clinical adjustment logic system, the priority and adaptation direction of parameter adjustment under different risk levels are clarified, and the parameter adjustment logic results are formed. A set of feasible adjustment schemes is obtained by performing clinical feasibility verification on the results of parameter adjustment logic. Integrate a set of feasible adjustment options to generate structured decision support information.

[0037] Clinical fit analysis based on the ranking results yields a set of clinically fit parameters. Combining patient clinical information, postoperative physiological status, and pacemaker operating environment, the ranking of key parameters is assessed for fit. The reasonable range of values ​​for each parameter in the current patient state, adjustment constraints, and compatibility with other clinical factors are evaluated, forming a set of clinically fit parameters that includes fit conditions, constraints, and clinical scenario requirements. Clinical risk association labeling is then applied to this set, resulting in parameter risk association representations. Adjustments to each parameter are correlated with clinical risk events to determine the potential risks of arrhythmias, sensory abnormalities, and pacing failure caused by different parameter values ​​or adjustment directions. The risk levels corresponding to different parameter adjustments are quantified, thus forming parameter risk association representations and clarifying the potential risks and safety boundaries of each parameter adjustment. Based on the parameter risk association representations, a clinical adjustment logic system is constructed, clarifying the priority and fit direction of parameter adjustments at different risk levels, resulting in parameter adjustment logic. A hierarchical adjustment logic is established by combining parameter priority ranking and risk correlation characterization results. Adjustment restrictions and early warning mechanisms are set for high-risk parameters; reasonable adjustment ranges and directional guidelines are established for low-risk parameters. The order of parameter adjustments for different risk levels is clearly defined to ensure the adjustment process complies with clinical safety guidelines, forming a complete parameter adjustment logic. Clinical feasibility verification is performed on the parameter adjustment logic to obtain a set of feasible adjustment schemes. Based on actual clinical operation procedures, pacemaker device performance limitations, and patient physiological tolerance, the feasibility of the parameter adjustment logic is verified, excluding adjustment schemes that exceed the device's adjustable range, violate clinical operation procedures, or may cause serious adverse reactions. A set of feasible adjustment schemes that both conform to the adjustment logic and are clinically operable is selected. Finally, the set of feasible adjustment schemes is integrated to generate structured decision support information. All feasible adjustment schemes are organized and presented in a structured manner, sorted according to parameter adjustment priority, risk level, and operation steps, forming structured decision support information that includes adjustment goals, parameter adjustment values, operation sequence, risk warnings, and expected effects.

[0038] A multi-source heterogeneous clinical data fusion system for cardiac pacemakers includes: Extraction module: Acquires multimodal clinical data of temporary pacemakers after cardiac surgery in patients, and extracts fusion feature information from the multimodal clinical data for parameter uncertainty quantification analysis; Establishment Module: Based on fused feature information, determine the key input parameters affecting the working state of the temporary pacemaker and their uncertainty distribution range, and establish a cloud model for quantifying the uncertainty of key input parameters; Processing module: Based on the cloud model, key input parameters are sampled and simulations are run to quantify the dynamic impact of the uncertainty distribution range of each key input parameter on the output indicators of the pacemaker system, and uncertainty quantification results are obtained. Generation module: Based on the uncertainty quantification results, the key input parameters are sorted to obtain the sorting results, and decision support information is generated based on the sorting results to guide the clinical adjustment of temporary pacemaker parameters.

[0039] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate. The components shown as units 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. Those skilled in the art can understand and implement this without any creative effort.

[0040] Through the above description of the embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by means of software plus necessary general-purpose hardware platforms, and of course, it can also be implemented by hardware. Based on this understanding, the above technical solutions, in essence or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute the methods described in the various embodiments or some parts of the embodiments.

[0041] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A cardiac pacemaker multi-source heterogeneous clinical data fusion method, characterized in that, Includes the following steps: Acquire multimodal clinical data of temporary pacemakers after cardiac surgery in patients, and extract fusion feature information from the multimodal clinical data for parameter uncertainty quantification analysis; Based on the fusion of feature information, the key input parameters affecting the working state of the temporary pacemaker and their uncertainty distribution range are determined, and a cloud model is established to quantify the uncertainty of the key input parameters. Based on the cloud model, the key input parameters are sampled and simulations are run to quantify the dynamic impact of the uncertainty distribution range of each key input parameter on the output index of the pacemaker system, and uncertainty quantification results are obtained. The key input parameters are sorted based on the uncertainty quantification results to obtain the sorting results. Decision support information is then generated based on the sorting results to guide clinical adjustments of temporary pacemaker parameters.

2. The cardiac pacemaker multi-source heterogeneous clinical data fusion method of claim 1, wherein, The multimodal clinical data includes patient information, temporary pacemaker information, and time-series physiological signal data.

3. The method for fusing multi-source heterogeneous clinical data of a cardiac pacemaker according to claim 2, characterized in that, The patient information includes the patient's baseline data and surgical data; the temporary pacemaker information includes the brand and model of the temporary pacemaker, initial parameter settings, and adjustment records.

4. The cardiac pacemaker multi-source heterogeneous clinical data fusion method of claim 1, wherein, Based on fused feature information, the key input parameters affecting the working state of the temporary pacemaker and their uncertainty distribution range are determined, specifically: By performing sensitivity analysis on the fusion feature information of the cardiac pacemaker in-loop simulation system, the key input parameters affecting the working state of the temporary pacemaker and their uncertainty distribution range are determined.

5. The cardiac pacemaker multi-source heterogeneous clinical data fusion method of claim 1, wherein, Based on the cloud model, the key input parameters are sampled, and simulations are run to quantify the dynamic impact of the uncertainty of each key input parameter on the output indicators of the pacemaker system, thereby obtaining the uncertainty quantification results. Specifically, the steps include: Based on the cloud model, the Wilks method is used to determine the minimum sample size, and the key input parameters are sampled to generate a parameter sample set. The parameter sample set is input into the loop simulation system of the cardiac pacemaker hardware, and the dynamic impact of the output index is used to obtain the uncertainty quantification result.

6. The cardiac pacemaker multi-source heterogeneous clinical data fusion method of claim 1, wherein, Extracting fusion feature information from multimodal clinical data for parameter uncertainty quantification analysis includes the following steps: The multimodal clinical data is normalized to obtain a normalized clinical data set; Feature mining is performed on a normalized clinical dataset to obtain a multidimensional set of related features; Cross-domain feature fusion is performed on a multidimensional associated feature set to generate a fused feature set; Feature filtering is performed on the fused feature set to obtain fused feature information.

7. The method of claim 6, wherein, And establish a cloud model for quantifying the uncertainty of key input parameters, specifically including the following steps: Extract the discrete fluctuation patterns and continuous changing trends of key input parameters to obtain a set of parameter uncertainty characteristics; The set of parameter uncertainty features is normalized to obtain the standardized uncertainty characterization result; Based on the results of standardized uncertainty characterization, an uncertainty mapping relationship is constructed, and a two-way correspondence mechanism between the qualitative state of parameters and quantitative fluctuations is established to form a cloud model.

8. The cardiac pacemaker multi-source heterogeneous clinical data fusion method of claim 7, wherein, The ranking results are obtained by sorting the key input parameters based on the uncertainty quantification results, specifically including the following steps: Extract the strength and depth of the influence of key input parameters on the pacing system's operating status to obtain a set of parameter influence features; The standardized influence degree characterization result is obtained by calibrating the influence degree of the parameter influence feature set; Based on the standardized impact characterization results, a parameter correlation comparison system is constructed. Based on the parameter correlation comparison system, the differences in impact weights and the primary and secondary relationships among the parameters are clarified, and the parameter comparison analysis results are obtained. The results of the parameter comparison and analysis were prioritized and screened to obtain the initial screening results of parameter priorities. The ranking results are obtained by sorting the key input parameters according to the initial screening results based on parameter priority.

9. The cardiac pacemaker multi-source heterogeneous clinical data fusion method of claim 8, wherein, Based on the ranking results, decision support information is generated to guide clinical adjustments of temporary pacemaker parameters. This includes the following steps: Clinical fit analysis based on the ranking results yields a set of parameter clinical fit features. Clinical risk association labeling was performed on the set of clinically adaptable features of parameters to obtain parameter risk association characterization results; Based on the parameter risk association characterization results, a clinical adjustment logic system is constructed. Based on the clinical adjustment logic system, the priority and adaptation direction of parameter adjustment under different risk levels are clarified, and the parameter adjustment logic results are formed. A set of feasible adjustment schemes is obtained by performing clinical feasibility verification on the results of parameter adjustment logic. Integrate a set of feasible adjustment options to generate structured decision support information.

10. A cardiac pacemaker multi-source heterogeneous clinical data fusion system applied to the cardiac pacemaker multi-source heterogeneous clinical data fusion method of any one of claims 1 to 9, characterized in that, include: Extraction module: Acquires multimodal clinical data of temporary pacemakers after cardiac surgery in patients, and extracts fusion feature information from the multimodal clinical data for parameter uncertainty quantification analysis; Establishment Module: Based on fused feature information, determine the key input parameters affecting the working state of the temporary pacemaker and their uncertainty distribution range, and establish a cloud model for quantifying the uncertainty of key input parameters; Processing module: Based on the cloud model, the key input parameters are sampled and simulation is run to quantify the dynamic impact of the uncertainty distribution range of each key input parameter on the output index of the pacemaker system to obtain uncertainty quantification results; Generation module: Based on the uncertainty quantification results, the key input parameters are sorted to obtain the sorting results, and decision support information is generated based on the sorting results to guide the clinical adjustment of temporary pacemaker parameters.

Images