A method for rapid calculation of cerebral tissue oxygen consumption level

Abstract

The application discloses a method for rapidly calculating oxygen consumption level of brain tissue, belongs to the field of hemodynamic numerical simulation, and is a method for rapidly and noninvasively calculating oxygen consumption level of brain tissue. The method comprises the following steps: determining a cerebral microcirculation anatomical structure model based on a brain D-K partition atlas and cerebral blood vessel blood supply relationship, establishing a brain tissue microcirculation hemodynamic lumped parameter model, reversely reconstructing neural activity by scalp electroencephalogram, and establishing a blood flow-oxygen coupling model. The method can rapidly calculate oxygen extraction fraction OEF and cerebral metabolic rate of oxygen CMRO2 which reflect physiological conditions of the brain.

Description

Technical Field

[0001] This invention belongs to the field of hemodynamic numerical simulation and is a system and method for rapidly and non-invasively calculating the oxygen consumption level of brain tissue. Background Technology

[0002] The brain is the organ with the highest oxygen demand in the human body, and its oxygen supply comes from oxygen carried by the cerebral bloodstream. In ischemic stroke, insufficient blood supply to local brain tissue leads to cerebral hypoxia, ultimately resulting in brain tissue necrosis. While cerebral oxygen consumption is primarily measured using magnetic resonance imaging (MRI), the diagnosis of acute stroke is time-sensitive and requires rapid diagnosis. MRI scans are time-consuming, which cannot meet the diagnostic time requirements. Therefore, currently, the primary diagnostic examination for cerebrovascular diseases in clinical practice is computed tomography angiography (CTA) to assess cerebral vascular conditions. For patients with functional impairments, electroencephalography (EEG) is used to assess brain function. This invention proposes a method for constructing a brain tissue oxygen metabolism model based on carotid artery CTA images and scalp EEG signals to assess the level of whole-brain oxygen consumption. This method has reference value for rapidly assessing the physiological condition of the brain in patients with ischemic cerebrovascular diseases. Summary of the Invention

[0003] This invention provides a method for rapidly calculating the oxygen consumption level of brain tissue. This method can be applied to calculate the oxygen consumption of brain tissue in patients with ischemic stroke, analyze the extent of brain tissue functional impairment, and also to analyze changes in brain tissue oxygen supply in patients with carotid artery stenosis. The method includes the following steps: construction of the cerebral microcirculatory vascular structure, establishment of a lumped parameter model of cerebral hemodynamics, reverse reconstruction of neural activity via scalp electroencephalography, and establishment of a blood flow-oxygen coupling model.

[0004] To achieve the above objectives, the present invention employs the following technical solution:

[0005] Step A1: Divide and name brain regions based on the brain DK partition map, and determine the anatomical model of brain microcirculation based on the blood supply relationship of brain blood vessels;

[0006] Step A2: Calculate the microcirculatory resistance of brain tissue based on the flow rate determined by the human head model partition, and establish a lumped parameter model of brain tissue microcirculatory hemodynamics by combining the anatomical model of brain microcirculation from Step A1.

[0007] Step A3: Based on the 64-lead scalp EEG signal, the cortical region current dipoles are fitted using Brainstorm software to calculate the local field potential (LFP). Correlation analysis of the LFP signal is performed to obtain the brain functional connectivity matrix. Based on the functional connectivity matrix, an inverse neuronal cluster network is established to calculate the excitatory postsynaptic potential (EPSP) and inhibitory postsynaptic potential (IPSP) of each brain region. The neural activity is obtained by solving the simultaneous equations.

[0008] Step A4: Establish a blood flow-oxygen coupling model. The microcirculatory blood flow calculated by the lumped parameter model of brain tissue microcirculation established in Step A2 and the brain neural activity signal obtained in Step A3 are used as the calculation parameters of the blood flow-oxygen coupling model. Based on the relationship between oxygen concentration change and oxygen concentration in blood flow, the indicators reflecting the brain oxygen metabolism level, oxygen extraction fraction (OEF) and brain oxygen metabolism rate (CMRO2), can be calculated.

[0009] As a further technical solution of the present invention, the feature described in step A1 is to determine the anatomical model of cerebral microcirculation based on the DK partition map and the blood supply relationship of cerebral blood vessels. The DK partition map is a method of dividing the brain into regions based on brain function, with the entire brain divided into 68 regions. Based on physiological blood supply relationships, the small arteries supplying each brain region are determined to be microcirculatory vessels. The blood vessels of the cerebral microcirculation are generally composed of arterioles, capillaries, and venules. Each brain region's blood supply unit is composed of these three types of vessels. The modeling method in this invention uses a lumped parameter model, which is composed of basic electronic components such as inductors and resistors. Assuming a blood vessel length of l and a cross-sectional area of ​​A, then according to Frank's elastic cavity theory, the quantitative relationship between flow rate, pressure, and related resistance in this blood vessel segment can be expressed as:

[0010] ΔP=Q*R′#(1)

[0011]

[0012] In the formula, ΔP represents the pressure drop generated in the segment of the blood vessel, Q represents the blood flow of the branch of the blood vessel, R' represents the viscous resistance generated in the segment of the blood vessel; μ is the blood viscosity value under normal physiological conditions, l represents the length of the segment of the blood vessel, and A represents the cross-sectional area of ​​the segment of the blood vessel.

[0013] The self-inductance of inductance is equivalent to the flow inertia of blood, and the resistance effect of resistance is equivalent to the microcirculatory resistance of each branch of the coronary artery. In the lumped parameter model, the values ​​of resistance R and inductance L are:

[0014]

[0015]

[0016] In the formula, P 末 This indicates the pressure at the distal end of the blood vessel, and ρ represents the blood flow density value.

[0017] As a further technical solution of the present invention, the feature described in step A2 involves establishing a lumped parameter model of brain tissue microcirculation, wherein the standard average flow rate values ​​of 68 cortical regions are calculated using the open-source software Brainstorm. First, the surface area of ​​the 68 cortical regions can be extracted using a standard head model. Simultaneously, the cortical thickness of the 68 cortical regions is determined according to literature. The volume of the 68 cortical regions is calculated based on the thickness and surface area. The average gray matter density is obtained from a literature review, and the mass of the 68 cortical regions is calculated by combining the volume of each region. Based on a gray matter perfusion rate of 60-80 mL / 100 g / min, the flow rate is allocated according to the mass of each cortical region, and the average flow rate of the 68 cortical regions is calculated. Microcirculatory resistance is obtained based on the flow rate and vascular pressure. Parameters are assigned to the lumped parameter model based on the microcirculatory resistance, and the model is corrected based on the calculation results.

[0018] As a further technical solution of the present invention, the feature described in step A3 involves deriving the calculation of neural activity from scalp EEG. First, EEG signals are obtained through standard 64-lead measurement and preprocessed to obtain relatively pure EEG signals. The source reconstruction method is used to fit the cortical region current dipoles using Brainstorm software, and the local field potential (LFP) is solved using dipole theory. The brain functional connectivity matrix is ​​obtained from the LFP signal through Pearson correlation. Based on the brain functional matrix, a neuronal cluster network model is established, and the corresponding excitatory postsynaptic potentials (EPSP) and inhibitory postsynaptic potentials (IPSP) of the corresponding brain regions are calculated. The neural activity signal u(t) is obtained using the neural signal calculation formula.

[0019] u(t)=E(t)-I(t)#(5)

[0020]

[0021] Where E(t) and I(t) are the excitatory and inhibitory postsynaptic potentials, respectively, κ is a parameter that determines the amplitude of the neural activity signal, and τ I It is a time constant;

[0022] As a further technical solution of the present invention, the feature described in step A4, based on the cerebral microcirculation regional blood flow values ​​obtained in step A2 and the neural activity signals obtained in step A3, establishes a blood flow-oxygen coupling model, and the specific implementation steps are as follows:

[0023] Step B1: Normalize the blood flow values ​​of the cerebral microcirculation regions obtained in Step A2. Divide the blood flow value calculated at each time step by the average blood flow value at rest. The average blood flow value at rest is defined as the time-averaged blood flow value obtained within one cardiac cycle. The blood flow value obtained at this time is defined as the input f of the blood flow-oxygen coupling model. in ;

[0024] Step B2: Establish the relationship between neural activity and changes in cerebral blood flow. Here, a flow-inducing signal s is defined, and its temporal differential relationship with the flow rate is as follows:

[0025]

[0026] This signal is generated from the neural activity signal u(t) obtained in step A3:

[0027]

[0028] Where ε, τ s τ f These are model parameters, which respectively represent the degree of signal increase caused by neuronal activity, the time constant of signal decay or complete disappearance, and the time constant of autoregulatory feedback from blood flow.

[0029] Step B3: Calculate the oxygen extraction fraction (OEF) and the average oxygen concentration in the capillaries. The time derivative relationships of the parameter g, representing the ratio of oxygen concentration in brain tissue to that in blood plasma, are as follows:

[0030]

[0031]

[0032]

[0033] Where T is the capillary transit time, which is set to 0.2 in this method. A parameter related to capillary transit time. Arterial oxygen saturation, J is a proportionality factor that takes into account the transport lag in tissue oxygen diffusion, and V tis / V cap It is the ratio of brain tissue blood volume to capillary blood volume, r represents the ratio of plasma oxygen concentration to blood oxygen concentration, K is a parameter that regulates the amplitude of neural activity signals, OEF0, g0 represents the initial values ​​of the three parameters in the resting state;

[0034] Step B4: Solve the differential equations established in Steps B2 and B3 simultaneously, E, g is used as the parameter required to solve the other two differential equations. The fourth-order Runge-Kutta method is used to solve the differential equations.

[0035] K1=hf(x i ,y i )#(12)

[0036]

[0037]

[0038] K4=hf(x i +h,y i +K3)#(15)

[0039]

[0040] Where y i+1 The value to be solved, y i It is the value at that moment, x i Here, h is the time step number at that moment, K1 is the slope at the beginning of the time interval, and K2 is the slope at the midpoint of the time interval. The slope K1 is used in the Euler method to determine the position of y at that point. The value of y is K3, which is also the slope at the midpoint, but here the slope K2 is used to determine the y value, and K4 is the slope at the end of the time interval, whose y value is determined by K3.

[0041] For example, to solve for OEF, then y i That is OEF; and so on, y i It is the average oxygen concentration in capillaries. Or a parameter g representing the ratio of oxygen concentration in brain tissue to that in blood plasma;

[0042] Step B5: Based on the cerebral microcirculation regional blood flow values ​​obtained in Step A2, convert them to the clinical standard blood flow value rCBF. Combined with the oxygen extraction fraction OEF obtained in Step B4, CMRO2 can be calculated.

[0043]

[0044]

[0045] Wherein CBF represents the blood flow value of the cerebral microcirculation region obtained in step A2. [Hb] represents the oxygen-carrying capacity of hemoglobin, while [Hb] represents the concentration of hemoglobin in the blood. region The mass of a brain region can be obtained by multiplying the brain region's surface area by its cortical thickness, and then multiplying it by the average density of the brain region.

[0046] Table 1 shows the parameter values ​​assigned in this invention within the model.

[0047]

[0048]

[0049] This invention can calculate the oxygen extraction fraction (OEF) and cerebral oxygen metabolism rate (CMRO2), which reflect the physiological condition of the brain, within 6 minutes. This has certain reference value for the rapid clinical assessment of the physiological condition of patients with ischemic stroke. Attached image description:

[0050] Figure 1 Flowchart of the method of this invention

[0051] Figure 2(1) Nomenclature of DK brain regions

[0052] Figure 2(2) Brain vascular structure

[0053] Figure 2(3) Microcirculatory vascular structure of the left brain 1

[0054] Figure 2(4) Microcirculatory vascular structure of the left brain 2

[0055] Figure 2(5) Microcirculatory vascular structure of the left brain 3

[0056] Figure 2(6) Blood supply relationships of different brain regions

[0057] Figure 3 Brain microcirculation blood flow model structure

[0058] Figure 4 EEG diagram illustrating neural activity

[0059] Figure 5 Schematic diagram of oxygen consumption in brain microcirculation

[0060] Figure 6 OEF and CMRO2 calculation results Detailed Implementation

[0061] The present invention will be further described below with reference to the embodiments, but the present invention is not limited to the following embodiments.

[0062] Example 1

[0063] Based on the features described in step A1: a model of the cerebral microcirculation anatomy was determined based on CTA scan results, brain DK region atlas, and cerebral vascular supply relationships. The small arteries supplying blood to each brain region were identified according to physiological blood supply relationships.

[0064] Based on the characteristics described in step A2: A lumped-parameter model of brain tissue microcirculation is established. First, the surface area of ​​68 cortical regions is extracted using a human head model. Simultaneously, the cortical thickness of these 68 regions is determined according to literature. The volume of each of the 68 cortical regions is calculated based on the thickness and surface area. The average gray matter density is obtained from a literature review, and the mass of each of the 68 cortical regions is calculated by combining this with the region volume. Based on a gray matter perfusion rate of 60-80 mL / 100 g / min, the flow rate is allocated according to the mass of each cortical region, and the average flow rate of the 68 cortical regions is calculated. Microcirculatory resistance is obtained based on the flow rate and vascular pressure. Parameters are assigned to the lumped-parameter model based on the microcirculatory resistance, and the model is calibrated based on the calculation results to ensure that the calculated results are equal to the clinically measured flow rates of the internal carotid artery and basilar artery.

[0065] Table 2 Calculation results of microcirculation flow in each brain region

[0066]

[0067]

[0068] Step A3: Based on the 64-lead scalp EEG signal of the case, the cortical regional current dipoles were fitted using Brainstorm software to calculate the local field potential (LFP). Correlation analysis of the LFP signals was performed to obtain the brain functional connectivity matrix. Based on the functional connectivity matrix, an inverse neuronal cluster network was established to calculate the excitatory postsynaptic potential (EPSP) and inhibitory postsynaptic potential (IPSP) of each brain region. The neural activity was obtained by solving the simultaneous equations.

[0069] Step A4: Establish a blood flow-oxygen coupling model. The microcirculatory blood flow calculated from the lumped parameter model of brain tissue microcirculation established in Step A2 and the brain neural activity signals obtained in Step A3 are used as input parameters for the model. Based on the relationship between oxygen concentration changes and oxygen concentration in the blood flow, solve for the oxygen extraction fraction (OEF) and the average oxygen concentration in the capillaries. And g is a parameter representing the ratio of oxygen concentration in brain tissue to that in blood plasma.

[0070] Step A5: Based on the cerebral microcirculation regional blood flow values ​​obtained in Step A2, convert them to the clinical standard blood flow value rCBF. Combined with the oxygen extraction fraction OEF obtained in Step A4, the CMRO2 can be calculated.

[0071] The objective of this invention is to rapidly calculate oxygen consumption in brain tissue. Based on the brain's physiological and anatomical structure and blood flow in the left and right internal carotid arteries and basilar arteries, combined with the size of the cortical layer, a flow distribution model of the cerebral microcirculation hemodynamics is established. Then, Brainstorm software is used to inversely reconstruct EEG signals, and an inverse neuronal cluster model is established based on the brain functional matrix. The corresponding excitatory postsynaptic potentials (EPSP) and inhibitory postsynaptic potentials (IPSP) in the corresponding brain regions are calculated to deduce neural activity signals. Finally, a cerebral microcirculation blood flow-oxygen coupling model is established to calculate the oxygen extraction fraction (OEF) and cerebral oxygen metabolism rate (CMRO2), parameters reflecting the oxygen supply from cerebral blood vessels to brain tissue, to assess the physiological condition of the brain in patients with ischemic cerebrovascular diseases.

Claims

1. A method for rapidly calculating the oxygen consumption level of brain tissue, characterized in that, Includes the following steps: Step A1: Divide and name brain regions based on the brain DK partition map, and determine the anatomical model of brain microcirculation based on the blood supply relationship of brain blood vessels; Step A2: Calculate the microcirculatory resistance of brain tissue based on the flow rate determined by the human head model partition, and establish a lumped parameter model of brain tissue microcirculatory hemodynamics by combining the anatomical model of brain microcirculation from Step A1. Step A3: Based on the 64-lead scalp EEG signal, the cortical region current dipoles are fitted using Brainstorm software to calculate the local field potential (LFP); correlation analysis is performed on the LFP signal to obtain the brain functional connectivity matrix; an inverse neuronal cluster network is established based on the functional connectivity matrix; the excitatory postsynaptic potential (EPSP) and inhibitory postsynaptic potential (IPSP) of each brain region are calculated; and the neural activity is obtained by solving the simultaneous equations. Step A4: Establish a blood flow-oxygen coupling model. The microcirculatory blood flow calculated by the lumped parameter model of brain tissue microcirculation established in Step A2 and the brain neural activity signal obtained in Step A3 are used as the calculation parameters of the blood flow-oxygen coupling model. Based on the relationship between oxygen concentration change and oxygen concentration in blood flow, the oxygen extraction fraction (OEF) and brain oxygen metabolism rate (CMRO2), which reflect the level of brain oxygen metabolism, are calculated.

2. The method for rapidly calculating the oxygen consumption level of brain tissue according to claim 1, characterized in that, The features described in step A1 are used to determine the anatomical model of the brain's microcirculation based on the DK partition map and the blood supply relationship of the brain's blood vessels. The DK partition map is a method of dividing the brain into regions based on brain function, with the entire brain divided into 68 regions. Based on physiological blood supply relationships, the small arteries supplying each brain region are identified as microcirculatory vessels. The blood vessels in the brain's microcirculation consist of arterioles, capillaries, and venules. Each brain region's blood supply unit is composed of these three types of vessels. The modeling method uses a lumped parameter model, which is composed of basic electronic components such as inductors and resistors. Based on Frank's elastic cavity theory, the quantitative relationship between flow rate, pressure, and related resistance in this segment of the blood vessel is expressed as follows: ； ； in the formula This indicates the pressure drop generated in that segment of the blood vessel. This indicates the blood flow rate of that blood vessel. This indicates the viscous resistance generated in this segment of the blood vessel; This is the normal physiological blood viscosity value, which is [value missing]. s; This indicates the length of the blood vessel segment. This indicates the cross-sectional area of ​​the blood vessel segment; The self-inductance of inductance is equivalent to the flow inertia of blood, and the resistance effect of resistance is equivalent to the microcirculatory resistance of each branch of the coronary artery; in the lumped parameter model, resistance... and inductance The value is: ； ； in the formula This indicates the pressure at the end of that blood vessel segment. This represents the blood flow density value, which is defined as... .

3. A method for rapidly calculating the oxygen consumption level of brain tissue according to claim 1, characterized in that, The features described in step A2 involve establishing a lumped parameter model of brain tissue microcirculation. The standard average flow rate values ​​for 68 cortical regions were calculated using the open-source software Brainstorm. First, the surface area of ​​the 68 cortical regions was extracted using a standard head model. Simultaneously, the cortical thickness of the 68 cortical regions was determined based on literature. The volume of the 68 cortical regions was calculated based on the thickness and surface area. The average gray matter density was obtained from literature review, and the mass of the 68 cortical regions was calculated by combining the volume of each region. Based on a gray matter perfusion rate of 60-80 mL / 100 g / min, the flow rate was allocated according to the mass of each cortical region, and the average flow rate of the 68 cortical regions was calculated. Microcirculatory resistance was obtained based on the flow rate and vascular pressure. Parameters were assigned to the lumped parameter model based on the microcirculatory resistance, and the model was corrected based on the calculation results.

4. A method for rapidly calculating the oxygen consumption level of brain tissue according to claim 1, characterized in that, The features described in step A3, derived from scalp EEG, are used to calculate neural activity. First, EEG signals are obtained using a standard 64-lead measurement and preprocessed to obtain relatively pure signals. Then, the source reconstruction method is used with Brainstorm software to fit cortical regional current dipoles, and dipole theory is used to solve for local field potentials (LFPs). The brain functional connectivity matrix is ​​obtained from the LFP signals through Pearson correlation. Based on this matrix, an inverse neuronal cluster model is established, and the excitatory postsynaptic potentials (EPSPs) and inhibitory postsynaptic potentials (IPSPs) for corresponding brain regions are calculated. Finally, neural activity signals are obtained using neural signal calculation formulas. ； ； in These are excitatory postsynaptic sites and inhibitory postsynaptic potentials, respectively. It is a parameter that determines the amplitude of neural activity signals. It is a time constant.

5. A method for rapidly calculating the oxygen consumption level of brain tissue according to claim 1, characterized in that, Based on the features described in step A4, the blood flow-oxygen coupling model is established using the cerebral microcirculation regional blood flow values ​​obtained in step A2 and the neural activity signals obtained in step A3. The specific implementation steps are as follows: Step B1: The blood flow values ​​of the cerebral microcirculation regions obtained in Step A2 are normalized. The blood flow value calculated at each time step is divided by the average blood flow value at rest. The average blood flow value at rest is defined as the time-averaged blood flow value obtained within one cardiac cycle. The blood flow value obtained at this time is defined as the input of the blood flow-oxygen coupling model. ; Step B2: Establish the relationship between neural activity and changes in cerebral blood flow. Here, a flow-inducing signal is used, and its temporal differential relationship with blood flow is as follows: ； The flow-induced signal is a unit representation of the normalized rate of change in cerebral blood flow; this signal is the neural activity signal obtained in step A3. Generated; ； in These are model parameters, which respectively represent the degree of signal increase caused by neuronal activity, the time constant of signal decay or complete disappearance, and the time constant of autoregulatory feedback from blood flow. Step B3: Calculate the oxygen extraction fraction Average oxygen concentration in capillaries And the parameter representing the ratio of oxygen concentration in brain tissue to that in blood plasma. Their time differential relationships are as follows: ； ； ； in For capillary transit time A parameter related to capillary transit time. Arterial blood oxygen saturation, This is a proportionality factor that takes into account the transport lag in tissue oxygen diffusion. It is the ratio of brain tissue blood volume to capillary blood volume. Represents the ratio of oxygen concentration in plasma to that in blood. Parameters for regulating the amplitude of neural activity signals, These represent the initial values ​​of the three parameters in the resting state; Step B4: Solve the differential equations established in Steps B2 and B3 simultaneously. These parameters are used respectively as the parameters required to solve the other two differential equations. The fourth-order Runge-Kutta method is used to solve the differential equations. ； ； ； ； ； in It is the value to be solved. It is the value at that moment. It is the time step at that moment. It calculates the step size. It is the slope at the beginning of the time period. It is the slope at the midpoint of the time interval, and the slope is obtained using the Euler method. To decide At point The value, It is also the slope at the midpoint, but here we use the slope. Determine the value of y. It is the slope at the end of the time period, and its Value Decide; Step B5: Convert the cerebral microcirculation regional blood flow values ​​obtained in Step A2 into clinical standard blood flow values. Combined with the oxygen extraction fraction obtained in step B4 It can be solved to obtain : ； ； in This represents the blood flow values ​​of the cerebral microcirculation regions obtained in step A2. This represents the oxygen-carrying capacity of hemoglobin. Represents the concentration of hemoglobin in the blood. Represents arterial blood oxygen saturation. It refers to the quality of brain regions.

Images