A ventilation state recognition method and system based on three-dimensional electrical impedance tomography

Abstract

The present application relates to a kind of ventilation state identification method and system based on three-dimensional electrical impedance tomography, method includes: the double-row flexible electrode array band is respectively installed in the 3 intercostal space and the 5 intercostal space of patient, high-frequency safe alternating current is generated by micro-current excitation module, cross injection human thoracic cavity is passed through double-row flexible electrode array band, corresponding boundary voltage is collected to patient body surface;Obtain the three-dimensional thoracic finite element model of human body matched with patient, and is dispersed into multiple space voxel, in combination with the boundary voltage collected, the conductivity variation of each space voxel on continuous breathing cycle is solved, and is converted into impedance, obtain voxel impedance time sequence matrix;Further determine global inspiration time window, carry out phase polarity determination and reverse flow detection, to generate Z-axis ventilation delay gradient distribution diagram, swing air three-dimensional vector and three-dimensional swing air absolute volume.Compared with prior art, the present application breaks through spatial dimension, is helpful to eliminate hidden missed diagnosis.

Description

Technical Field

[0001] This invention relates to the field of respiratory monitoring technology, and in particular to a method and system for identifying ventilation status based on three-dimensional electrical impedance tomography. Background Technology

[0002] Currently, in clinical critical care respiratory monitoring, the assessment of pulmonary ventilation heterogeneity and swing gas mainly relies on two-dimensional electrical impedance tomography (2D-EIT) systems. Existing methods typically place a single-row electrode strip containing 16 electrodes in a specific layer of the patient's chest (such as the 4th or 5th intercostal space). When assessing ventilation delay and swing gas, existing algorithms estimate gas exchange only within this single two-dimensional plane by identifying the difference between the time difference of local pixel impedance reaching its lowest point and the global curve. For example, the critical care patient monitoring method and system based on EIT technology disclosed in invention publication number CN118436888A uses an EIT-based approach to analyze chest images at different PEEP levels, divide the lung recruitment area and heterogeneous area, calculate local PEEP reference weights, and redetermine the final optimal PEEP level to accommodate the pulmonary heterogeneity characteristics of ARDS patients.

[0003] The existing methods described above have the following drawbacks: 1. The 2D-EIT electric field diverges vertically in a spindle shape, exhibiting severe partial volume effects and out-of-band artifacts. Heartbeats and vertical displacement of the diaphragm can severely interfere with and contaminate the ventilation signal in the central region of the 2D image.

[0004] 2. The real abnormal airflow redistribution in the lungs is often a three-dimensional airflow transfer across lobes from gravity-dependent areas to non-gravity-dependent areas, or even along the head-to-tail direction. Existing 2D methods completely lose the spatial transfer characteristics along the Z-axis, assuming that all abnormal flows occur only in the same plane, resulting in a serious underestimation of the measured absolute volume of oscillating air.

[0005] 3. The existing regional ventilation delay index cannot reflect the heterogeneity of the whole lung's mechanical gradient along the Z-axis, and doctors cannot fully assess the overall stress distribution risk of intratidal re-expansion of the whole lung. Summary of the Invention

[0006] The purpose of this invention is to overcome the shortcomings of the existing technology, which has the physical characteristics of a single row of electrodes that make it unable to distinguish electric field disturbances on the Z-axis and can only be passively projected, and to provide a ventilation status identification method and system based on three-dimensional electrical impedance tomography.

[0007] The objective of this invention can be achieved through the following technical solutions: A ventilation status identification method based on three-dimensional electrical impedance tomography includes: The dual-row flexible electrode array strips are installed in the 3rd and 5th intercostal spaces of the patient, respectively. A high-frequency safe alternating current is generated by the microcurrent excitation module and injected into the human thoracic cavity through the dual-row flexible electrode array strips to collect the boundary voltage signal of the patient's body surface. A three-dimensional finite element model of the human chest cavity matching the patient is obtained and discretized into multiple spatial voxels. Combined with the acquired boundary voltage, the conductivity change of each spatial voxel in the continuous respiratory cycle is calculated and converted into impedance to obtain the voxel impedance time series matrix. The global inhalation time window is determined based on the voxel impedance timing matrix. By determining the global inhalation time window and performing phase polarity judgment and reverse flow detection, a Z-axis ventilation delay gradient distribution map, a three-dimensional vector of oscillating air, and the absolute volume of three-dimensional oscillating air are generated. The combination of the Z-axis ventilation delay gradient, the three-dimensional vector of the oscillating air, and the absolute volume of the three-dimensional oscillating air helps doctors optimize ventilation parameters.

[0008] Furthermore, the process of determining the global inhalation time window includes: Extract the sum of impedances of all spatial voxels within the patient's total lung volume as the global impedance; Based on the curve of global impedance changing over time, the global inhalation start time is determined. and the end point of global inhalation And based on this, the global inhalation time window is calculated.

[0009] Furthermore, the calculation process for the Z-axis ventilation delay gradient distribution map includes: Within the global inhalation time window, the impedance time curves of each spatial voxel are traversed to locate the time point when it rises to 40% of the local maximum amplitude, thereby calculating the voxel-level time constant of each spatial voxel. The three-dimensional finite element model of the human thoracic cavity is divided into cranial, central and caudal regions along the Z-axis, and the mean voxel-level time constant of the corresponding regions is calculated to generate a Z-axis ventilation delay gradient distribution map.

[0010] Furthermore, the phase polarity determination process includes: Within each of the aforementioned global inhalation time windows, calculate the partial derivative of the impedance of each spatial voxel with time. If the partial derivative of a spatial voxel remains negative throughout the global inspiratory phase, it is labeled as a three-dimensional anti-phase voxel.

[0011] Furthermore, the calculation process for the absolute volume of the three-dimensional oscillating gas includes: For all three-dimensional anti-phase voxels, the negative impedance change component within the global inhalation time window is extracted, and the absolute value of the negative impedance change component is accumulated to obtain the absolute amount of reverse impedance variation. After summing the absolute values ​​of the reverse impedance variations of all three-dimensional anti-phase voxels, impedance-volume calibration is performed by combining the total tidal volume output synchronously by the ventilator with the global tidal impedance variation to obtain the absolute volume of the whole lung in three-dimensional oscillation.

[0012] Furthermore, the calculation process of the oscillating air three-dimensional vector includes: The spatial connectivity domain clustering of three-dimensional anti-phase voxels is used to form anti-phase collapsed regions; Voxels that meet the positive rapid inflation condition within the global inhalation time window are selected from non-three-dimensional anti-phase voxels, and spatial connectivity domain clustering is performed to form an extremely positive rapid inflation region. The volume-weighted geometric centroid coordinates of the anti-phase depression region and the extremely positive phase rapid inflation region are calculated respectively, and the two volume-weighted geometric centroid coordinates are connected to generate a three-dimensional vector of oscillating gas.

[0013] Furthermore, the frequency of the high-frequency safe AC power is within the range of 40-60kHz and the current is less than 1mA.

[0014] Furthermore, the dual-row flexible electrode array includes an upper row of electrodes and a lower row of electrodes, each comprising 16 to 32 electrodes.

[0015] The present invention also provides a ventilation status identification system for implementing the ventilation status identification method based on three-dimensional electrical impedance tomography as described above, comprising: A dual-row flexible electrode array is installed in the patient's 3rd and 5th intercostal spaces, respectively; The microcurrent excitation module is used to generate high-frequency safe alternating current, which is injected into the human chest cavity in a jumping and cross manner through the double-row flexible electrode array. A high-frequency data acquisition module is used to acquire the boundary voltage on the patient's body surface; An image reconstruction processor is used to acquire a three-dimensional finite element model of the human thoracic cavity that matches the patient, and discretize it into multiple spatial voxels; The spatiotemporal feature fusion and core calculation module is used to calculate the conductivity change of each spatial voxel in a continuous respiratory cycle, convert it into impedance, generate a voxel impedance time series matrix, and generate the Z-axis ventilation delay gradient, the three-dimensional vector of the oscillating air and the absolute volume of the three-dimensional oscillating air based on this matrix. The three-dimensional visualization and clinical early warning terminal device is used to visualize the Z-axis ventilation delay gradient distribution map, the three-dimensional vector of the swing gas and the absolute volume of the three-dimensional swing gas, and generate early warning information to assist doctors in optimizing ventilation parameters.

[0016] Furthermore, the dual-row flexible electrode array strips are respectively attached to and tied to the 3rd and 5th intercostal spaces of an ARDS patient preparing to undergo a spontaneous breathing test.

[0017] Compared with the prior art, the present invention has the following advantages: (1) The physical characteristics of a single-row electrode determine that it cannot distinguish the disturbance of the electric field on the Z-axis and can only be passively projected. This invention uses a double-row array to construct a three-dimensional electric field distribution, and accurately discretizes the physical space into three-dimensional voxels; through the time-series partial derivative analysis of the voxel impedance curve, the forward inflation signal and the reverse deflation signal are accurately separated at the mathematical level. Based on the principles of integration and clustering, the complex micro-fluid dynamics are reduced to a macroscopic three-dimensional physical vector with a clear direction and absolute capacity value.

[0018] This invention breaks through the spatial dimension, which helps to eliminate hidden missed diagnoses: it completely breaks the limitations of the 2D plane, eliminates the contamination of data by out-of-plane motion artifacts, accurately captures the abnormal transfer of longitudinal airflow across the lung lobes, and the measured swing air volume has extremely high physiological fidelity.

[0019] (2) Pioneering Z-axis mechanical gradient target: The quantification of the Z-axis ventilation delay gradient ensures that doctors can take into account the uniformity of the time constant from the apex to the base of the lung when performing prone turning or adjusting PEEP, thus avoiding local severe tearing caused by neglecting one aspect.

[0020] (3) Minimalist 3D visualization early warning: The time-series operation of tens of thousands of complex voxels is reduced to an intuitive 3D vector arrow and absolute volume value, which greatly reduces the cognitive threshold of critical care physicians for high-dimensional images, enabling them to judge the risk level of spontaneous lung injury in seconds. Attached Figure Description

[0021] Figure 1 This is a flowchart illustrating a ventilation status identification method based on three-dimensional electrical impedance tomography provided in an embodiment of the present invention. Figure 2 This is a schematic diagram of the output interface of a ventilation status identification method based on three-dimensional electrical impedance tomography provided in an embodiment of the present invention. Detailed Implementation

[0022] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. The components of the embodiments of the present invention described and shown in the accompanying drawings can generally be arranged and designed in various different configurations.

[0023] Therefore, the following detailed description of the embodiments of the invention provided in the accompanying drawings is not intended to limit the scope of the claimed invention, but merely to illustrate selected embodiments of the invention. All other embodiments obtained by those skilled in the art based on the embodiments of the invention without inventive effort are within the scope of protection of the invention.

[0024] It should be noted that similar labels and letters in the following figures indicate similar items. Therefore, once an item is defined in one figure, it does not need to be further defined and explained in subsequent figures.

[0025] Example 1 like Figure 1 As shown, this embodiment provides a ventilation status identification method based on three-dimensional electrical impedance tomography, including: S1: The double-row flexible electrode array strips are installed in the 3rd and 5th intercostal spaces of the patient, respectively. The microcurrent excitation module generates high-frequency safe AC current, which is injected into the human thoracic cavity in a jumping and cross manner through the double-row flexible electrode array strips. Then, the corresponding boundary voltage on the patient's body surface is collected. S2: Obtain a three-dimensional finite element model of the human chest cavity that matches the patient, and discretize it into multiple spatial voxels. Combined with the acquired boundary voltage, calculate the conductivity change of each spatial voxel in a continuous respiratory cycle, and convert it into impedance to obtain the voxel impedance time series matrix. S3: Determine the global inspiratory time window based on the voxel impedance time sequence matrix. Use the global inspiratory time window to determine and perform phase polarity judgment and reverse flow detection to generate the Z-axis ventilation delay gradient distribution map, the three-dimensional vector of the oscillating air and the absolute volume of the three-dimensional oscillating air. Combine the Z-axis ventilation delay gradient, the three-dimensional vector of the oscillating air and the absolute volume of the three-dimensional oscillating air to help doctors optimize ventilation parameters.

[0026] Specifically, in step S1, the frequency of the high-frequency safe AC power is within the range of 40-60kHz and the current is less than 1mA.

[0027] The dual-row flexible electrode array includes an upper row of electrodes and a lower row of electrodes, each containing 16 to 32 electrodes.

[0028] In step S2, under the conditions of fixed excitation mode, fixed finite element mesh generation, and neglecting the influence of reactance, each spatial voxel can be regarded as an equivalent conductive element. For the v-th spatial voxel, its equivalent impedance at time t can be expressed as: Z v(t) =k v / σ v(t) Among them, Z v(t) Let σ represent the equivalent impedance of the v-th spatial voxel at time t. v(t) Let k represent the equivalent conductivity of the v-th spatial voxel at time t. v Represents the geometric constants related to the geometry of the space voxel and the equivalent conduction path.

[0029] Furthermore, when the change in voxel conductivity during the respiratory cycle is a small perturbation relative to the baseline state, the impedance change of the vth spatial voxel can be expressed by a first-order approximation as follows: ΔZ v(t) ≈-(k v / σ̄ v^2 )·Δσ v(t) Among them, σ̄ v Δσ represents the baseline conductivity of the v-th spatial voxel during the analyzed respiratory cycle. v(t) ΔZ represents the change in conductivity relative to a reference time. v(t) This represents the change in impedance relative to the reference time.

[0030] Therefore, in this embodiment, the conductivity changes of each spatial voxel are first obtained through three-dimensional electrical impedance inversion, and then the corresponding voxel impedance changes are obtained according to the above equivalent conversion relationship. This leads to the construction of a voxel impedance time series matrix, which is used for subsequent global inhalation time window identification, phase polarity determination, reverse flow detection, and oscillating gas volume calculation. Step S3, the determination process of the global inhalation time window, includes: Extract the sum of impedances of all spatial voxels within the patient's total lung volume as the global impedance; Based on the curve of global impedance changing over time, the start and end times of global inhalation are determined and used as the global inhalation time window.

[0031] In this embodiment, the sum of the conductivities of all voxels within the entire lung volume is extracted, and a global impedance-time curve is plotted to accurately identify the global inspiratory initiation point. and the end point Construct a global inhalation time window.

[0032] The calculation process for the Z-axis ventilation delay gradient distribution map includes: Within the global inhalation time window, the impedance time curves of each spatial voxel are traversed to locate the time point when it rises to 40% of the local maximum amplitude, thereby calculating the voxel-level time constant of each spatial voxel. The three-dimensional finite element model of the human thoracic cavity is divided into cranial, central and caudal regions along the Z-axis, and the mean voxel-level time constant of the corresponding regions is calculated to generate a Z-axis ventilation delay gradient distribution map.

[0033] In this embodiment, within the global inspiratory time window, the time curve of each voxel v(x,y,z) is independently traversed, and the time point when its own curve rises to 40% of the local maximum amplitude is located, and the voxel-level time constant (3D-RVD) is calculated. The three-dimensional mesh is divided into cranial, mid- and caudal sides along the Z-axis, and the mean 3D-RVD of each plane is calculated to generate a Z-axis ventilation delay gradient distribution map.

[0034] The process of phase polarity determination and reverse flow detection includes: Within each global inhalation time window, the partial derivative of the impedance of each spatial voxel with time is calculated, and if the partial derivative of the impedance of a spatial voxel with time is continuously negative within the global inhalation time window, it is marked as a three-dimensional anti-phase voxel.

[0035] In this embodiment, within the global inspiratory time window, the system calculates the impedance partial derivative dZ(x,y,z,t) / dt of each spatial voxel as a function of time. If a voxel's partial derivative remains negative throughout the global inspiratory period (i.e., anomalous expiratory collapse), it is precisely labeled as a three-dimensional anti-phase voxel.

[0036] The calculation process for the absolute volume of a three-dimensional oscillating gas includes: For all three-dimensional anti-phase voxels, the negative impedance change component opposite to the global inspiratory direction within the global inspiratory time window is extracted, and the absolute value of the negative impedance change component is accumulated and integrated to obtain the absolute amount of reverse impedance variation for each three-dimensional anti-phase voxel. Then, the absolute amounts of reverse impedance variation corresponding to all three-dimensional anti-phase voxels are summed, and impedance-volume calibration conversion is performed in combination with the total tidal volume output synchronously by the ventilator to obtain the absolute volume of the three-dimensional oscillating air of the whole lung in three-dimensional plane.

[0037] In this embodiment, the end-expiratory time t of the analyzed respiratory cycle is used as the reference point. ref Using the reference time as a baseline, the impedance change of the v-th space voxel at time t is defined as: ΔZv (t) =Z v(t) -Z v(tref) Wherein, ΔZv (t) Represents time t relative to the reference time. ref The change in quantity.

[0038] For any three-dimensional anti-phase voxel v, within the global inspiratory time window [T] start ,T end Within [the specified range], the absolute value of its reverse impedance variation is defined as: ΔZ rev,v =∑max(0,-(ΔZ v[n] -ΔZv [n-1] )) Where, ΔZ rev,v Z_rev represents the cumulative reverse impedance change of the v-th three-dimensional anti-phase voxel within the global inhalation time window; rev in Z_rev stands for reverse, used to characterize the abnormal reverse impedance change of this voxel during the global inhalation phase; n represents the discrete sampling point number within the inhalation time window.

[0039] Let Ω be the set of all three-dimensional antiphase voxels. oop The total variation in reverse impedance of the whole lung can be expressed as: ΔZ rev,total =∑(v∈Ω oop )ΔZ rev,v Here, Ωoop represents the set of all three-dimensional antiphase voxels.

[0040] Furthermore, take the global inhalation starting point T. start With the global inspiratory endpoint T end The corresponding global impedance difference is taken as the global moisture impedance change: ΔZ global =Z global(Tend) -Z global(Tstart) Synchronously read the tidal volume VT output by the ventilator during the corresponding respiratory cycle and establish the impedance-volume conversion factor K. V : K V =VT / ΔZ global Then the absolute volume V of the three-dimensional oscillating gas pend,3D It can be represented as: V pend,3D =K V ·ΔZ rev,total Preferably, VT and ΔZ global The average value of several consecutive stable respiratory cycles is taken to reduce the impact of noise and fluctuations in a single breath on the calibration results.

[0041] Therefore, in this embodiment, by accumulating the reverse impedance change of the three-dimensional anti-phase voxel during the global inhalation phase and using the tidal volume synchronized with the ventilator to complete the impedance-volume conversion, the absolute volume of the three-dimensional oscillating gas in milliliters can be obtained.

[0042] The calculation process of the three-dimensional vector of the oscillating air includes: Based on the spatial distribution of three-dimensional anti-phase voxels, a spatial connected domain clustering algorithm is used to cluster adjacent or near-neighbor three-dimensional anti-phase voxels, and isolated noisy connected domains with voxel numbers or volumes less than a preset threshold are filtered out, thereby forming anti-phase shrinkage regions. In the non-three-dimensional anti-phase voxels, spatial voxels that meet the positive rapid inflation conditions within the global inhalation time window are further screened out, and spatial connected domain clustering is performed on the screened spatial voxels to form an extremely positive rapid inflation zone. The volume-weighted geometric centroid coordinates of the anti-phase depression region and the extremely positive phase rapid inflation region are calculated separately. The two volume-weighted geometric centroid coordinates are connected to generate the oscillating gas three-dimensional vector. The oscillating gas three-dimensional vector includes the vector amplitude, three-dimensional yaw angle and pitch angle.

[0043] In this embodiment, the anti-phase collapse region is the gas supply region formed by spatial clustering of the above-mentioned three-dimensional anti-phase voxels. It is not the same as a simple set of a single discrete anti-phase voxel, but a spatial region after filtering by connected components.

[0044] In this embodiment, the extremely positive-phase rapid inflation region does not refer to all voxels except for the three-dimensional anti-phase voxels, but rather to the abnormally rapid inflation region further selected from non-three-dimensional anti-phase voxels. Preferably, the selection criteria include: (1) Within the global inhalation time window, the partial derivative of the voxel impedance with time is positive; (2) The maximum positive impedance change rate of the voxel within the global inhalation time window is greater than the preset rate threshold θ1; (3) The cumulative positive impedance increment of the voxel within the global inhalation time window is greater than the preset amplitude threshold θ2; (4) The difference between the moment when the voxel is rapidly inflated and the moment when the reverse change peak of the anti-phase depression region is less than the preset timing threshold τ.

[0045] After voxels that meet the above conditions are clustered by spatial connected domains, they form a rapidly inflating region exhibiting extremely positive phase, i.e., the gas receiving region.

[0046] Furthermore, if the volume-weighted geometric centroid coordinates of the anti-phase shrinkage region are C out =(x out , y out , z out The volume-weighted geometric centroid coordinates of the extremely positive phase rapid inflation region are C.in =(x in , y in , z in Then, the three-dimensional vector of the oscillating air can be expressed as: P 3D =C in - C out Among them, P 3D The modulus represents the migration scale of the oscillating gas in three-dimensional space, and its direction represents the main flow direction of the oscillating gas from the gas supply area to the gas receiving area.

[0047] The pathophysiological, physical, and algorithmic basis of this scheme is as follows: Pathophysiological basis: In the pathological state of ARDS, lung tissue exhibits significant spatial mechanical heterogeneity. When critically ill patients maintain spontaneous breathing, the strong contraction of the diaphragm generates a large local negative intrathoracic pressure at the base of the lungs. Due to uneven pressure transmission, this negative pressure not only generates gas flow in the anteroposterior direction but also induces a strong gas outflow in the superior-inferior cranial-to-clonic direction (Z-axis direction), causing the gravity-dependent area to expand instantaneously and draw air from the upper airway.

[0048] Physical and Algorithmic Basis: The physical characteristics of a single-row electrode mean it cannot distinguish electric field perturbations along the Z-axis and can only passively project them. This invention utilizes a double-row array to construct a three-dimensional electric field distribution, precisely discretizing the physical space into three-dimensional voxels. Through time-series partial derivative analysis of voxel impedance curves, this invention mathematically and precisely separates the forward inflation signal and the reverse deflation signal. Based on the principles of integration and clustering, the complex microscopic fluid dynamics are reduced to a macroscopic three-dimensional physical vector with a clear direction and absolute capacity value.

[0049] Example 2 This embodiment provides a ventilation status identification system that implements the ventilation status identification method based on three-dimensional electrical impedance tomography as described in Embodiment 1, comprising: A dual-row flexible electrode array is installed in the patient's 3rd and 5th intercostal spaces, respectively; The microcurrent excitation module is used to generate high-frequency safe alternating current, which is injected into the human chest cavity in a jumping and cross manner through a double-row flexible electrode array. A high-frequency data acquisition module is used to acquire the boundary voltage on the patient's body surface; An image reconstruction processor is used to acquire a three-dimensional finite element model of the human thoracic cavity that matches the patient and discretize it into multiple spatial voxels; The spatiotemporal feature fusion and core calculation module is used to calculate the conductivity change of each spatial voxel in a continuous respiratory cycle and convert it into impedance to obtain the voxel impedance time series matrix. Based on the voxel impedance time series matrix, the global inspiratory time window is determined, and phase polarity judgment and reverse flow detection are performed to generate the Z-axis ventilation delay gradient distribution map, the three-dimensional vector of the oscillating air and the absolute volume of the three-dimensional oscillating air. The 3D visualization and clinical early warning terminal device is used to visualize the Z-axis ventilation delay gradient distribution map, the three-dimensional vector of the swing gas and the absolute volume of the three-dimensional swing gas, and generate early warning information.

[0050] The processing procedure of the above system is as follows: Hardware deployment and initialization: Flexible silicone electrode strips, consisting of 16 electrodes in the upper row and 16 electrodes in the lower row, were used and attached to the 3rd and 5th intercostal spaces of an ARDS patient preparing for a spontaneous breathing test. The electrode strips were connected to the 3D-EIT monitoring host via the data bus, and the host's calibration interface was connected to the ventilator to acquire real-time tidal volume data for impedance-volume conversion.

[0051] Data Acquisition and 3D Reconstruction: Upon startup, the system acquires boundary interaction voltages at a rate of 50 frames per second. The built-in processor calls the individualized matched 3D finite element model of the human thoracic cavity, discretizes the thoracic cavity into 32×32×16 spatial voxels, and reconstructs and smooths the voxel conductivity time series matrix over five consecutive stable respiratory cycles in real time.

[0052] Parallel feature decomposition: The system identified the global inspiratory window from Tstart to Tend. Traversal calculations revealed a significant aggregation of voxels and their partial derivatives dZdt<0 in the upper left cranial region, indicating collapse during overall inspiration. Inverse integration of the DeltaZ decrease of this cranial out-of-phase voxel cluster was performed, and using the VT calibration parameter, the reverse exhaust volume during this period was calculated to be 55 mL. Simultaneously, a highly time-overlapping zone of rapid anomalous inflation was located in the lower right caudal region. The system extracted the centroids of the deflation and refill zones separately.

[0053] Interface rendering and output: such as Figure 2 As shown, on the 3D semi-transparent mesh model of the lungs on the host screen, the system directly renders a prominent red 3D arrow pointing from the cranial side of the left lung to the caudal ventral side of the right lung, and outputs the text prompt next to the arrow: "3D-Pendelluft: 55mL, accounting for 12.2% of tidal volume". At the same time, the sidebar displays the three-layer RVD delay gradient along the Z-axis, showing that the caudal side inflates 30% earlier than the cranial side.

[0054] The usage process of the above system is as follows: Patient access: At the bedside in the ICU, when the patient is on mechanical ventilation with spontaneous breathing or is being assessed for weaning, the nursing staff will put on the double-row electrode straps at the corresponding position on the patient's chest and turn on the system power.

[0055] Dynamic monitoring: The attending physician can directly observe the dynamically generated three-dimensional lung ventilation thermogram and Z-axis delay gradient distribution on the bedside terminal screen.

[0056] Clinical Early Warning and Intervention Decision-Making: Under normal or safe breathing conditions, the absence of prominent red vector arrows on the interface indicates uniform airflow distribution and the absence of harmful oscillations, allowing physicians to confidently continue the weaning procedure. When a patient experiences abnormal inspiratory work due to changes in airway resistance or severe respiratory distress, the interface immediately generates a large, three-dimensional oscillating air vector arrow. If the absolute value of the displayed vector volume exceeds the set safety threshold, the system will trigger an audible and visual warning.

[0057] Parameter-targeted tuning: Based on the direction of the arrows on the interface and the severity of the Z-axis delay gradient, the physician can immediately take intervention measures at the ventilator end, such as increasing pressure support PSV to reduce the patient's work, re-titering PEEP, or stopping SBT to resume controlled ventilation, until the three-dimensional swing air vector on the monitor is significantly reduced or disappears, thereby achieving highly refined whole-lung protective ventilation.

[0058] The preferred embodiments of the present invention have been described in detail above. It should be understood that those skilled in the art can make numerous modifications and variations based on the concept of the present invention without creative effort. Therefore, all technical solutions that can be obtained by those skilled in the art based on the concept of the present invention through logical analysis, reasoning, or limited experimentation on the basis of existing technology should be within the scope of protection defined by the claims.

Claims

1. A ventilation status identification method based on three-dimensional electrical impedance tomography, characterized in that, include: The dual-row flexible electrode array strips are installed in the 3rd and 5th intercostal spaces of the patient, respectively. A high-frequency safe alternating current is generated by the microcurrent excitation module and injected into the human thoracic cavity through the dual-row flexible electrode array strips to collect the boundary voltage signal of the patient's body surface. A three-dimensional finite element model of the human chest cavity matching the patient is obtained and discretized into multiple spatial voxels. Combined with the acquired boundary voltage, the conductivity change of each spatial voxel in the continuous respiratory cycle is calculated and converted into impedance to obtain the voxel impedance time series matrix. The global inhalation time window is determined based on the voxel impedance timing matrix. By determining the global inhalation time window and performing phase polarity judgment and reverse flow detection, a Z-axis ventilation delay gradient distribution map, a three-dimensional vector of oscillating air, and the absolute volume of three-dimensional oscillating air are generated. The combination of the Z-axis ventilation delay gradient, the three-dimensional vector of the oscillating air, and the absolute volume of the three-dimensional oscillating air helps doctors optimize ventilation parameters.

2. The ventilation status identification method based on three-dimensional electrical impedance tomography according to claim 1, characterized in that, The process of determining the global inspiratory time window includes: Extract the sum of impedances of all spatial voxels within the patient's total lung volume as the global impedance; Based on the curve of global impedance changing over time, the global inhalation start time is determined. and the end point of global inhalation And based on this, the global inhalation time window is calculated.

3. The ventilation status identification method based on three-dimensional electrical impedance tomography according to claim 1, characterized in that, The calculation process of the Z-axis ventilation delay gradient distribution map includes: Within the global inhalation time window, the impedance time curves of each spatial voxel are traversed to locate the time point when it rises to 40% of the local maximum amplitude, thereby calculating the voxel-level time constant of each spatial voxel. The three-dimensional finite element model of the human thoracic cavity is divided into cranial, central and caudal regions along the Z-axis, and the mean voxel-level time constant of the corresponding regions is calculated to generate a Z-axis ventilation delay gradient distribution map.

4. The ventilation status identification method based on three-dimensional electrical impedance tomography according to claim 1, characterized in that, The process of determining phase polarity includes: Within each of the aforementioned global inhalation time windows, calculate the partial derivative of the impedance of each spatial voxel with time. If the partial derivative of a spatial voxel remains negative throughout the global inspiratory phase, it is labeled as a three-dimensional anti-phase voxel.

5. The ventilation status identification method based on three-dimensional electrical impedance tomography according to claim 4, characterized in that, The calculation process for the absolute volume of the three-dimensional oscillating gas includes: For all three-dimensional anti-phase voxels, the negative impedance change component within the global inhalation time window is extracted, and the absolute value of the negative impedance change component is accumulated to obtain the absolute amount of reverse impedance variation. After summing the absolute values ​​of the reverse impedance variations of all three-dimensional anti-phase voxels, impedance-volume calibration is performed by combining the total tidal volume output synchronously by the ventilator with the global tidal impedance variation to obtain the absolute volume of the whole lung in three-dimensional oscillation.

6. The ventilation status identification method based on three-dimensional electrical impedance tomography according to claim 4, characterized in that, The calculation process of the oscillating air three-dimensional vector includes: Spatial connectivity domain clustering is performed on three-dimensional anti-phase voxels to form anti-phase collapsed regions; Voxels that meet the positive rapid inflation condition within the global inhalation time window are selected from non-three-dimensional anti-phase voxels, and spatial connectivity domain clustering is performed to form an extremely positive rapid inflation region. The volume-weighted geometric centroid coordinates of the anti-phase depression region and the extremely positive phase rapid inflation region are calculated respectively, and the two volume-weighted geometric centroid coordinates are connected to generate a three-dimensional vector of oscillating gas.

7. The ventilation status identification method based on three-dimensional electrical impedance tomography according to claim 4, characterized in that, The frequency of the high-frequency safe AC power is within the range of 40-60kHz and the current is less than 1mA.

8. The ventilation status identification method based on three-dimensional electrical impedance tomography according to claim 1, characterized in that, The dual-row flexible electrode array includes an upper row of electrodes and a lower row of electrodes, each comprising 16 to 32 electrodes.

9. A ventilation status identification system for implementing the ventilation status identification method based on three-dimensional electrical impedance tomography as described in any one of claims 1-8, characterized in that, include: A dual-row flexible electrode array is installed in the patient's 3rd and 5th intercostal spaces, respectively; The microcurrent excitation module is used to generate high-frequency safe alternating current, which is injected into the human chest cavity in a jumping and cross manner through the double-row flexible electrode array. A high-frequency data acquisition module is used to acquire the boundary voltage on the patient's body surface; An image reconstruction processor is used to acquire a three-dimensional finite element model of the human thoracic cavity that matches the patient and discretize it into multiple spatial voxels; The spatiotemporal feature fusion and core calculation module is used to calculate the conductivity change of each spatial voxel in a continuous respiratory cycle, convert it into impedance, generate a voxel impedance time series matrix, and generate the Z-axis ventilation delay gradient, the three-dimensional vector of the oscillating air and the absolute volume of the three-dimensional oscillating air based on this matrix. The three-dimensional visualization and clinical early warning terminal device is used to visualize the Z-axis ventilation delay gradient distribution map, the three-dimensional vector of the swing gas and the absolute volume of the three-dimensional swing gas, and generate early warning information to assist doctors in optimizing ventilation parameters.

10. The system according to claim 9, characterized in that, The dual-row flexible electrode array strips are respectively attached to and tied to the 3rd and 5th intercostal spaces of an ARDS patient who is preparing to undergo a spontaneous breathing test.

Images