A four-dimensional EIT respiration monitoring method based on digital twin assistance

Abstract

The application discloses a four-dimensional EIT respiratory monitoring method based on digital twin assistance, which comprises the following steps: first, constructing an individualized digital twin model containing accurate three-dimensional geometric structure and electrical characteristic parameters and covering the whole organ system of the thoracic cavity including double lungs, a heart, central axial skeleton, main bronchus and esophagus and the like based on the chest medical image CT of a patient; second, generating a stable and standardized theoretical reference voltage on the model through electromagnetic field simulation; then, fusing the theoretical reference voltage with the voltage signal collected in real time in the clinic and the reference voltage sequence measured in the clinic; and finally, reconstructing a four-dimensional conductivity change image based on the fused dynamic voltage difference. The application effectively suppresses reference drift, fuses individualized anatomical information and realizes semi-absolute quantitative evaluation of lung ventilation function.

Description

TECHNICAL FIELD

[0001] The present application belongs to the field of biomedical engineering and intelligent medical technology, and specifically relates to a four-dimensional electrical impedance tomography (4D-EIT) respiratory monitoring method based on digital twin assistance, aiming to realize high-precision visualization and pathological quantitative analysis of the whole cycle of respiratory function by fusing dynamic time series data and individualized digital twin models. BACKGROUND

[0002] As a non-invasive and real-time medical imaging technology, four-dimensional electrical impedance tomography (4D-EIT) has important application value in intensive care, lung function assessment and other fields. Currently, the differential imaging method is a widely used core technology in the field of EIT respiratory monitoring. It calculates the difference between the voltage at other times in the respiratory cycle and the voltage at the end of exhalation as a reference baseline to reconstruct the lung ventilation change image.

[0003] Although the differential imaging method has core advantages in eliminating baseline interference and suppressing noise, it still has the following inherent and significant limitations, which seriously restrict the depth and breadth of its clinical application:

[0004] 1) Reference drift problem: The end-of-exhalation voltage as the reference baseline is easily affected by factors such as rapid breathing, respiratory pattern fluctuations in tidal volume, pathological factors such as pleural effusion, and atelectasis, resulting in instability of the reference baseline itself, and thus reducing the imaging accuracy and reliability.

[0005] 2) Lack of individualization: Traditional methods do not consider patient-specific anatomical structures such as thoracic deformity, heart displacement or postoperative organ displacement, which can significantly affect the electric field distribution in EIT measurements, thus limiting the accuracy of imaging.

[0006] 3) Lack of absolute reference and loss of static information: The images generated by the differential imaging method are only relative changes in impedance, completely losing static anatomical information that can reflect individual thoracic geometry, organ position and resting tissue conductivity distribution. This not only leads to the inability to perform absolute or semi-absolute quantitative assessment of lung ventilation function, but also greatly reduces the interpretability of the images.

[0007] Therefore, there is an urgent need in the art for a new four-dimensional EIT respiratory monitoring method that can effectively supplement key static anatomical information, correct the deviation of the measured reference frame, and improve the quantitative ability of imaging while retaining the core advantages of the differential imaging method. SUMMARY

[0008] The technical problem to be solved by this invention is to overcome the defects of existing EIT respiratory monitoring technology based on differential imaging, such as inaccurate and unreliable imaging and lack of quantitative ability due to reference drift, insufficient individualization and loss of static information.

[0009] To solve the above-mentioned technical problems, the present invention provides the following technical solution:

[0010] This invention provides a four-dimensional EIT respiratory monitoring method based on digital twin assistance. The core idea of ​​this method is to generate a stable and accurate theoretical voltage reference by constructing a high-fidelity individualized digital twin model. This data is then fused with real-time clinical data to create an enhanced reference that combines anatomical stability with dynamic physiological adaptability.

[0011] A four-dimensional EIT respiratory monitoring method based on digital twin assistance includes the following steps:

[0012] Step S1: Construction of a personalized digital twin: Based on the chest medical imaging data of the target object, construct a personalized digital twin model that includes the three-dimensional geometric structure and electrical characteristic parameters of the main organs in the chest cavity of the target object;

[0013] Step S2, Theoretical Reference Voltage Generation: On the individualized digital twin model, simulate the electromagnetic field distribution of the target object under end-expiratory hypoventilation steady state, and solve for the standardized theoretical reference voltage. );

[0014] Step S3, Clinical Data Acquisition: Using an electrode array deployed on the target subject's body surface, the surface voltage signal of the target subject is acquired in real time during the respiratory cycle. The surface voltage signal includes a real-time voltage signal (…). ) and clinically measured reference voltage sequences that vary with the respiratory cycle ( );

[0015] Step S4, Virtual-Real Fusion Calculation: The theoretical reference voltage ( The clinically measured reference voltage sequence ( ) and the real-time voltage signal ( Dynamic fusion is performed to calculate the dynamic voltage difference between the virtual and real fusion. The calculation formula is as follows:

[0016]

[0017] in, This is the virtual-real fusion scaling factor, and its value is based on the theoretical reference voltage ( ) and the clinically measured reference voltage sequence ( ) between the peak-to-peak differences is adaptively adjusted;

[0018] Step S5, four-dimensional inversion and imaging: based on the dynamic voltage difference (Vd) , reconstruct a four-dimensional conductivity change image (σ) reflecting the dynamic change of lung ventilation by solving an optimization problem. ).

[0019] Preferably, the individualized digital twin construction step specifically comprises:

[0020] Data input: acquire chest CT image or MRI image data of the target object;

[0021] Geometric modeling: use a medical image processing tool to segment the image data, extract organ contours of both lungs, heart, central axis skeleton, main bronchus and esophagus, and construct a three-dimensional geometric model, wherein the central axis skeleton includes sternum, rib and thoracic vertebrae;

[0022] Electrical property assignment: assign electrical property parameters at a preset EIT working frequency to each organ in the three-dimensional geometric model;

[0023] Sensor integration: integrate an EIT electrode array model including multiple electrodes at the corresponding positions on the surface of the three-dimensional geometric model.

[0024] Preferably, the electrical property assignment step further comprises:

[0025] Establish a dynamic motion model of organs within a respiratory cycle, and dynamically adjust the conductivity and other time-varying electrical property parameters of lung tissue according to the inspiration phase and expiration phase within the respiratory cycle.

[0026] Preferably, the clinically measured reference voltage sequence (Vd ) is a sequence composed of body surface voltage signals collected at the end of expiration of each respiratory cycle.

[0027] Preferably, the virtual-real fusion proportion factor adaptive adjustment logic includes:

[0028] Calculate the peak-to-peak difference between the theoretical reference voltage (Vth ) and the clinically measured reference voltage sequence (Vd ) within one or more respiratory cycles ;

[0029] When the peak-to-peak difference is greater than a preset pathological state threshold, increase the value of to enhance the correction weight of the theoretical reference voltage (Vth ), and at this time The value range is from 0.8 to 1.0;

[0030] When the peak-to-peak difference When the value of k is not greater than the preset pathological state threshold, the value of k is reduced to prioritize the adoption of clinically measured data. At this time, the value of k ranges from 0.3 to 0.5.

[0031] Preferably, in the four-dimensional inversion and imaging step, the mathematical expression of the optimization problem is:

[0032]

[0033] in, for The system's three-dimensional sensitive field matrix, For regularization parameters, These are the prior constraints used to ensure the stability and smoothness of the solution.

[0034] A four-dimensional EIT respiratory monitoring system based on digital twin assistance, comprising:

[0035] The model building module is used to construct an individualized digital twin model based on the chest medical imaging data of the target object, which includes the three-dimensional geometric structure and electrical characteristic parameters of the major organs in the chest cavity of the target object, including at least the two lungs, heart, axial skeleton, main bronchus and esophagus.

[0036] The theoretical benchmark generation module is used to simulate the electromagnetic field distribution of the target object under end-expiratory hypoventilation steady state on the individualized digital twin model, and solve for the standardized theoretical benchmark voltage. );

[0037] The data acquisition module is used to acquire the real-time voltage signal of the target object during the respiratory cycle via an electrode array. ) and clinically measured reference voltage sequence ( );

[0038] The fusion calculation module is used to convert the theoretical reference voltage ( The clinically measured reference voltage sequence ( ) and the real-time voltage signal ( Dynamic fusion is performed to calculate the dynamic voltage difference between the virtual and real fusion. The calculation formula is as follows: ,in This is a scaling factor that is adaptively adjusted based on the peak-to-peak difference between the theoretical reference voltage and the clinically measured reference voltage sequence;

[0039] Imaging module, used for imaging based on the dynamic voltage difference ( ), by solving an optimization problem to reconstruct a four-dimensional conductivity change image reflecting the dynamic changes of lung ventilation ].

[0040] Compared with the prior art, the present application has the following beneficial effects:

[0041] 1. The present application precisely matches the patient's thoracic geometry and organ distribution through a digital twin model constructed based on the patient's CT / MRI data, breaking through the limitations of traditional EIT that is not sensitive to anatomical differences.

[0042] 2. The present application dynamically fuses stable theoretical benchmarks and real-time clinical benchmarks through an adaptively adjusted scaling factor , effectively compensating for the measured benchmark deviation caused by unstable breathing or pathological interference, significantly improving the imaging robustness.

[0043] 3. Based on the individualized conductivity prior information provided by the digital twin, the present application first realizes semi-absolute quantitative evaluation of lung ventilation function, such as regional ventilation percentage calculation, breaking through the limitations of traditional methods that only rely on relative changes. BRIEF DESCRIPTION OF DRAWINGS

[0044] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the following will briefly introduce the drawings needed to be used in the embodiment description.

[0045] Figure 1 The figure is a schematic diagram of the geometric model establishment process of the digital twin in the present application;

[0046] Figure 2 The figure is a schematic diagram of the overall process of the four-dimensional EIT respiratory monitoring method assisted by the digital twin in the present application;

[0047] Figure 3 The figure is a detailed analysis diagram of the virtual-real fusion calculation method in the present application, where the left 'virtual space' shows the simulated normalized global impedance waveform and stable theoretical benchmark voltage ); the right 'physical space' shows the clinically measured normalized global impedance waveform, real-time voltage signal ) and clinically measured benchmark voltage sequence ); the 'virtual-real fusion' arrow in the middle represents the fusion calculation process of the three;

[0048] Figure 4 The figure is a schematic diagram of the four-dimensional EIT respiratory monitoring result assisted by the digital twin in the present application, showing the process from CT image to digital twin system, and finally generating a series of reconstructed conductivity change distribution maps at different breathing times. DETAILED DESCRIPTION

[0049] The technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by a person of ordinary skill in the art without creative effort belong to the protection scope of the present application.

[0050] The present embodiment is specifically the respiratory function monitoring of a chronic obstructive pulmonary disease (COPD) patient, and the present embodiment aims to illustrate the specific application of the present application in a clinical scenario.

[0051] Step S1: Construction of individualized digital twin

[0052] Please refer to Figure 1 The first step of the present application is the construction of an individualized digital twin, which starts from obtaining chest CT image data of a patient, as specifically shown in Figure 1 indicated in the upper left corner; then, the CT image is segmented by medical image processing software to accurately extract the contours of key organs such as the lungs, heart, axial skeleton, main bronchus and esophagus, wherein the axial skeleton includes the sternum, ribs and thoracic vertebrae; at the same time, the information of a double-layer 32-electrode EIT sensor array model is integrated; finally, all segmented organ models and sensor models are integrated in three-dimensional space to complete the establishment of a high-fidelity geometric model, as shown in Figure 1 indicated in the lower right corner, which accurately reproduces the individualized thoracic anatomy of the patient.

[0053] On the basis of the geometric model, electrical property parameters of each organ in the model under a specific operating frequency of, for example, 125 kHz are assigned, and the key parameters can be set as follows: the lung tissue conductivity under the end-inspiration inflation state is 0.109 S / m; and the lung tissue conductivity under the end-expiration low-ventilation state is 0.275 S / m.

[0054] In order to accurately simulate the dynamic deformation of the lungs during the breathing process, the present application constructs a biomechanical model based on multi-physics coupling, which specifically includes the following mechanical model and electrical and mechanical coupling model:

[0055] (1) Mechanical model

[0056] In the normal breathing process, the lungs reach the maximum inflation state at the end of inspiration, at which time the alveoli are fully expanded, the lung tissue tension is large, and the thoracic volume is at a high level; as the expiration begins, the diaphragm gradually rises, the intrathoracic pressure rises, the lung tissue elastically retracts, the alveoli gradually collapse, and until the end of expiration, the lung volume is reduced to the lowest, and during the whole process from the end of inspiration to the end of expiration, the lung tissue undergoes significant three-dimensional deformation, accompanied by regional ventilation differences and changes in tissue density, which directly affect the electrical conductivity distribution.

[0057] To accurately simulate the dynamic deformation process of the lungs during a complete respiratory cycle, this invention constructs a biomechanical model based on finite element analysis and uses solid mechanics methods to describe the mechanical response of lung tissue under the drive of periodic breathing.

[0058] In the modeling, we assume that lung tissue conforms to an isotropic linear elastic material model, and its stress-strain relationship satisfies Hooke's Law:

[0059]

[0060] in For the Cauchy stress tensor; It is a linear strain tensor; , The Lamé parameters are calculated using Young's modulus E and Poisson's ratio v, respectively, and the calculation formula is as follows:

[0061]

[0062] Considering the large deformation characteristics of lung tissue during respiration, geometric nonlinear analysis was enabled in the model, using a more general deformation description variable—the deformation gradient tensor:

[0063]

[0064] And define the Green-Lagrange strain tensor:

[0065]

[0066] Within this framework, stress can be expressed using the second Piola-Kirchhoff stress tensor. Express:

[0067]

[0068] In terms of mechanical equilibrium, the momentum conservation equation under quasi-static conditions is:

[0069]

[0070] In a nonlinear configuration, this is transformed into the equilibrium equations under the reference configuration:

[0071]

[0072] Combining the anatomical structure and physiological characteristics of the lungs, the driving mechanism is designed with the periodic active contraction of the diaphragm and the dynamic changes in intrathoracic pressure as the main factors, simulating the expansion and deformation of lung tissue during inhalation and the passive recoil behavior during exhalation.

[0073] Specifically, in the inhalation phase, a preset boundary pressure load along the three directions of xyz is applied to the diaphragm region adjacent to the lung base (denoted as ), to simulate the mechanical extrusion of the diaphragm on the lung:

[0074]

[0075] At the same time, a gradually increasing negative pressure load condition is applied to the outer surface of the lung (denoted as ), to simulate the stretching effect caused by the decrease of intrathoracic pressure:

[0076]

[0077] This combined load induces nonlinear deformation of the lung tissue, resulting in an increase in overall volume, which approximates the physiological state of alveolar expansion. The volume of the lung changes over time:

[0078]

[0079] In the exhalation phase, the model reverses the mechanical boundary conditions to reproduce the deformation process of the lung from the inflated state to the end of exhalation. Specifically, the upward displacement load applied to the diaphragm region is gradually reduced until it is eliminated ( → 0), while the external negative pressure load is also released synchronously ( → 0), simulating the process of intrathoracic pressure recovery and diaphragm relaxation.

[0080] In this phase, the lung tissue spontaneously shrinks under the action of the elastic potential energy accumulated after the previous expansion, showing obvious nonlinear elastic recovery behavior. Its dynamics is as follows:

[0081]

[0082] During this process, the alveolar volume gradually decreases, the tissue tension decreases, and the overall volume decreases, which approximates the passive rebound mechanism of lung tissue in the real exhalation process.

[0083] (2) Electrical and mechanical coupling model

[0084] In terms of material properties, lung tissue is considered as a flexible, biologically responsive elastic body with certain compressibility. To accurately reflect its actual mechanical properties, the Poisson's ratio of lung tissue is set to 0.1, which indicates that lung tissue has small transverse deformation during compression or stretching, and has obvious volume compression ability, which can better reflect the changes in tissue density and geometric changes caused by gas exchange during the breathing process.

[0085] ​During the whole simulation, the multi-time transient analysis method is used to dynamically model the whole respiratory cycle.

[0086] At each time point, the model outputs the three-dimensional displacement field, volume change rate and stress distribution information of lung tissue, reflecting the continuous deformation characteristics and compressible response of the lung from the end of inspiration to the end of expiration. Through the calculation results of the mechanical model, the dynamic geometric change characteristics of the lung during the respiratory process can be intuitively obtained. These deformation information not only reflects the real lung movement law, but also provides important input parameters for the EIT simulation module.

[0087] Specifically, according to the local volume change of lung tissue at each time point, the change trend of its electrical conductivity with time can be further deduced, and then the dynamic updating and regulation of the electrical parameters in the EIT image are realized.

[0088] The lung tissue deforms under the driving of respiration, resulting in local volume change and readjustment of gas distribution. These mechanical deformations directly lead to the dynamic distribution changes of lung electrical conductivity in space and time, thereby affecting the distribution of electric field and measurement results.

[0089] Therefore, in lung electrical impedance tomography (EIT), in order to accurately describe the dynamic changes of lung electrical properties during respiration, an electrical and mechanical coupling model must be constructed.

[0090] The mechanical deformation of the lung during respiration can be represented by the displacement field , and its corresponding local volume strain is defined as:

[0091]

[0092] The deformation of the lung satisfies the dynamic equilibrium equation:

[0093]

[0094] wherein is the stress tensor, is the external driving force, is the tissue density; by solving the above equation, the deformation state of the lung tissue at different respiratory times can be obtained.

[0095] The electrical conductivity of the lung tissue can be defined as a function of deformation, that is, the electrical conductivity dynamically adjusts with the change of lung volume:

[0096]

[0097] wherein is the static electrical conductivity at the end of expiration in the reference state, and the function describes the sensitivity of electrical conductivity to deformation; the commonly used linear equation is approximately:

[0098]

[0099] Where parameters >0 indicates the degree to which conductivity decreases with volume expansion.

[0100] The electric field is solved by satisfying the steady-state current continuity equation:

[0101]

[0102] in, This indicates the distribution of electric field and potential.

[0103] This coupling mechanism ensures that the electrical response is driven by the actual deformation of the lungs, significantly improving the physical consistency and predictive ability of the simulation model for the real breathing process.

[0104] Steps S2 to S4: Monitoring of virtual-real fusion based on digital twins

[0105] Please see Figure 2 The core monitoring process of this invention integrates the advantages of physical space and virtual space.

[0106] Please see Figure 3 First, in step S2, the theoretical reference voltage is generated. Using the digital twin constructed in step S1, we perform a multiphysics coupling simulation in virtual space. Through this simulation, we can calculate the sensor output voltage signal generated by the virtual EIT excitation at end-expiratory steady state. Figure 3 As shown in the 'virtual space' on the left, the voltage signal obtained from this simulation is the stable and standardized theoretical reference voltage. .

[0107] Secondly, in step S3, clinical data acquisition, we used an EIT device at the patient's bedside in the physical space to continuously monitor the patient's respiration through a physically deployed double-layer 32-electrode array; we acquired two data sequences in real time: one was the real-time voltage signal at each sampling point within the respiratory cycle. Secondly, the clinically measured baseline voltage sequence at the end of each respiratory cycle. ;like Figure 3 As shown in the 'physical space' on the right, the measured global impedance waveform may exhibit baseline drift due to the patient's unstable breathing.

[0108] Next, in the crucial step S4, the calculation of the dynamic voltage difference between the virtual and real systems, we perform... Figure 3 The 'virtual-real fusion' calculation is shown in the center. The processing unit performs the following operations:

[0109] Difference Calculation: Real-time Calculation of Theoretical Benchmark Compared with clinical benchmarks The peak-to-peak difference δ between them, in this case, due to the unstable breathing of the COPD patient, is detected to reach 15%, exceeding the preset 10% pathological state threshold.

[0110] k value adaptive adjustment: according to the judgment result of δ>10%, the virtual-real fusion ratio factor k is dynamically adjusted to 0.8, so as to enhance the correction effect of the theoretical baseline .

[0111] Fusion calculation: the core formula is applied to calculate the corrected dynamic voltage difference , and the fusion process uses to correct the drift of , so as to obtain a corrected dynamic voltage difference which not only retains the true dynamic information but also eliminates the baseline drift, laying a solid foundation for subsequent high-precision imaging.

[0112] Step S5: four-dimensional conductivity image reconstruction and analysis

[0113] The calculated is taken as input and substituted into the reconstruction algorithm, as shown in Figure 2 , the reconstruction process needs to combine the pre-calculated three-dimensional spatial EIT sensitivity field, that is, the sensitivity field matrix, to solve the inverse problem, and finally realize three-dimensional image reconstruction.

[0114] The final imaging result is shown in Figure 4 , which shows a series of three-dimensional images arranged in time sequence, corresponding to key moments such as end-expiration, mid-inspiration, end-inspiration, etc. in the respiratory cycle, which effectively verifies the effective suppression of baseline drift and the visualization ability of the ventilation dynamic process of the method.

[0115] Through post-processing of the reconstructed four-dimensional image, specifically, the four-dimensional image is a time sequence of a series of three-dimensional images, it can be quantitatively analyzed that the left lung ventilation of the patient accounts for 62.3% of the total ventilation, and the right lung accounts for only 37.7%, showing obvious ventilation heterogeneity, and this quantitative result can provide accurate basis for the clinician to formulate personalized body position treatment or ventilator parameter setting.

[0116] As can be seen through this embodiment, the method of the present application effectively overcomes the defects of traditional EIT, and provides more accurate, more robust, and more clinically valuable quantitative monitoring results.

[0117] Although embodiments of the present application have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made therein without departing from the principles and spirit of the present application, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A method of four-dimensional EIT respiration monitoring based on digital twin assistance, characterized in that, The method comprises the following steps: Step S1, individualized digital twin construction: based on the chest medical image data of the target object, an individualized digital twin model containing the three-dimensional geometric structure and electrical characteristic parameters of the main organs in the chest cavity of the target object is constructed; Step S2, theoretical reference voltage generation: on the individualized digital twin model, simulate the electromagnetic field distribution of the target object under the end-expiratory low-tidal volume steady state, and solve the standardized theoretical reference voltage (Vth) ) Step S3, clinical data acquisition: acquiring, in real time, a body surface voltage signal of the target object in a respiratory cycle through an electrode array arranged on a body surface of the target object, the body surface voltage signal including a real-time voltage signal ( ) and a clinically measured reference voltage sequence varying with the respiratory cycle ( ); Step S4, virtual-real fusion calculation: dynamically fusing the theoretical reference voltage ( ), the clinical measured reference voltage sequence ( ) and the real-time voltage signal ( ) to calculate the virtual-real fusion dynamic voltage difference ( ), and the calculation formula is: wherein, is a virtual-real fusion scaling factor, which is adaptively adjusted according to a peak-to-peak difference between the theoretical reference voltage (Vth) ) and the clinically measured reference voltage sequence (Vmeas) ). Step S5, four-dimensional inversion and imaging: Based on the dynamic voltage differences (Vd) ), a four-dimensional conductivity change image reflecting the dynamic changes of lung ventilation is reconstructed by solving an optimization problem. ).

2. The four-dimensional EIT respiration monitoring method based on digital twin assistance according to claim 1, characterized in that, The individualized digital twin construction step specifically comprises: Data input: obtaining the chest CT image or MRI image data of the target object; Geometric modeling: using a medical image processing tool to segment the image data, extract the organ contours of the double lungs, heart, central axis skeleton, main bronchus and esophagus, and construct a three-dimensional geometric model, wherein the central axis skeleton includes the sternum, rib and thoracic vertebrae; Electrical characteristic assignment: assigning electrical characteristic parameters at a preset EIT working frequency to each organ in the three-dimensional geometric model; Sensor integration: integrating an EIT electrode array model containing a plurality of electrodes at the corresponding positions of the body surface of the three-dimensional geometric model.

3. The four-dimensional EIT respiration monitoring method based on digital twin assistance according to claim 2, characterized in that, The electrical characteristic assignment step further comprises: Establishing a dynamic motion model of the organs within a respiratory cycle, and dynamically adjusting the time-varying electrical characteristic parameters such as the electrical conductivity of the lung tissue according to the inspiration phase and expiration phase within the respiratory cycle.

4. The four-dimensional EIT respiration monitoring method based on digital twin assistance of claim 1, wherein, said sequence of clinically measured reference voltages (Vref) ) is a sequence of body surface voltage signals acquired at the end of expiration time instants of each respiratory cycle.

5. The four-dimensional EIT respiration monitoring method based on digital twin assistance of claim 1, wherein, The virtual-real fusion proportion factor The adaptive adjustment logic comprises: calculating the peak-to-peak difference between the theoretical reference voltage (Vref) ) and the sequence of clinically measured reference voltages (Vref ) over one or more respiratory cycles ; when the peak-to-peak difference is greater than a preset pathological state threshold, the value of is increased to enhance the correction weight of the theoretical reference voltage , and at this time has a value ranging from 0.8 to 1.

0. when the peak-to-peak difference when the peak-to-peak difference is not greater than the preset pathological state threshold, the value of k is reduced to preferentially adopt the clinically measured data, and the value of k ranges from 0.3 to 0.

5.

6. The four-dimensional EIT respiration monitoring method based on digital twin assistance of claim 1, wherein, In the four-dimensional inversion and imaging step, the mathematical expression of the optimization problem is: wherein is a three-dimensional sensitivity field matrix of the system, is a regularization parameter, is a prior constraint term for ensuring stability and smoothness of the solution.

7. A four-dimensional EIT respiration monitoring system based on digital twin assistance, characterized in that, It comprises: A model construction module is configured to construct, based on chest medical image data of a target object, an individualized digital twin model containing three-dimensional geometric structure and electrical characteristic parameters of main organs in the chest cavity of the target object, including at least double lungs, heart, central axis skeleton, main bronchus and esophagus; a theoretical benchmark generating module, configured to simulate electromagnetic field distribution of the target object in the end-expiratory low-compliance steady state on the individualized digital twin model, and solve a standardized theoretical benchmark voltage (Vth) ) A data acquisition module is configured to acquire, through the electrode array, real-time voltage signals of the target object in a breathing cycle and a clinically measured reference voltage sequence . a fusion computing module, configured to dynamically fuse the theoretical reference voltage ( ), the clinical measured reference voltage sequence ( ), and the real-time voltage signal ( ) to calculate a dynamic voltage difference of virtual-real fusion ( ), and a calculation formula thereof is , wherein is a proportional factor adaptively adjusted according to a peak-to-peak difference between the theoretical reference voltage and the clinical measured reference voltage sequence. an imaging module for reconstructing a four-dimensional conductivity change image reflecting dynamic changes in lung ventilation based on the dynamic voltage difference (ΔV) by solving an optimization problem .

Images