A non-invasive method to calculate CFR

Abstract

The present invention belongs to the field of coronary artery medical technology, and discloses a non-invasive method for calculating CFR, comprising S1, reconstructing a three-dimensional model of coronary arteries and generating a grid; S2, setting boundary conditions of the model; S3, simulating CFD configuration of hemodynamics; S4, calculating CFR; the present invention realizes non-invasive CFR calculation based on coronary artery CTA images, does not require invasively measured blood flow (such as blood pressure, flow velocity) data, only requires CTA images, and analyzes the CTA images through a proposed algorithm and indirectly calculates the CFR. The method for calculating CFR realizes the unity of economy and safety, and the obtained CFR value is accurate, can effectively evaluate the coronary artery vascular function, and is helpful for medical staff to specify the best treatment plan.

Description

Technical Field

[0001] The present invention relates to the technical field of coronary artery medicine, and in particular to a non-invasive method for calculating CFR. Background Art

[0002] CFR refers to the ratio of coronary blood flow in the heart's congested state to that in the resting state, and is mainly used to quantitatively evaluate the regulatory capacity of the heart's coronary arteries. Generally speaking, a normal CFR value should be above 2.5. If the CFR value is low, it means that the heart is not getting enough blood, which may be caused by coronary artery stenosis, coronary artery spasm, or coronary microvascular disease. Therefore, accurate measurement of the CFR value can effectively evaluate coronary vascular function and help medical staff specify the best treatment plan.

[0003] Currently, there are many methods for measuring CFR in clinical practice, including invasive methods and non-invasive methods. The invasive methods include Doppler guidewire method and thermal dilution method, and the non-invasive methods include PET-CT and CMR detection methods.

[0004] Among them, Doppler guidewire is a common method for measuring CFR. It mainly uses a Doppler sensor that can sense blood flow velocity to be sent to the narrow part of the coronary artery through a guide tube, first measuring the blood flow velocity at rest, then injecting vasodilators into the patient to measure the blood flow velocity in the congestive state, and then calculating the ratio of the two to obtain CFR. However, the Doppler guidewire method is an invasive method for measuring CFR, which involves inserting a catheter into a blood vessel and guiding it to the heart. Patients have to bear higher risks. In addition, this method is expensive and needs to be performed simultaneously with coronary angiography, so its clinical application is limited.

[0005] Thermodilution is currently widely used in clinical measurement of CFR. It mainly uses two temperature sensors placed on the guide tube and the pressure-temperature guide wire to measure CFR. During the test, normal temperature saline is quickly injected into the coronary artery. The temperature sensor senses the temperature difference between blood and saline and measures the time from saline leaving the guide tube to the temperature sensor at the tip of the guide wire, i.e., the mean conduction time at rest. Then, the mean conduction time at hyperemia is measured by injecting vasodilators. Finally, the CFR is obtained by dividing the mean conduction time at rest by the mean conduction time at hyperemia. However, the CFR measured by the thermodilution method is susceptible to influence. During the measurement, great attention should be paid to the influence of the guide catheter, sensor position, injection technique, hyperemia reaction and branch vessels on the measurement results. Specifically, the guide catheter with side holes should not be used. The temperature sensor on the guide wire tip should be placed as far as possible at the distal end of the stenosis. The saline should be injected quickly and evenly. When there are branch vessels proximal to the measured stenosis, the CFR value will be overestimated. Under the influence of the above variables, only experienced doctors can correctly implement the thermodilution method. The accuracy of CFR measured by ordinary doctors using the thermodilution method is limited.

[0006] PET-CT testing requires patients to be injected with radioactive tracers, which will emit positrons in the body. When they meet electrons in the body, they will produce electron annihilation effects and release gamma rays. Then, the patient is scanned with PET to obtain the distribution of the radioactive tracer in the body. Then, the patient is scanned with CT to obtain a CT reconstructed image. Finally, the computer fuses the image data of PET and CT to obtain an image with blood flow information to directly calculate CFR. However, the radiation dose of PET-CT testing is relatively large, and artifacts are prone to occur, which can affect the doctor's judgment.

[0007] CMR uses a strong magnetic field to excite water molecules in the body to obtain images, and different information is obtained by combining different scanning sequences during the detection process. Specifically, CMR accurately measures coronary blood flow and obtains clear coronary artery imaging through a combination of specific sequences, and finally calculates CFR directly from blood flow; however, CMR is expensive and requires strong magnetic field excitation, so it cannot be used by some specific groups of people (such as patients with pacemakers, heart stents or other metal implants). In addition, CMR requires a relatively long detection time, which may cause discomfort to some patients.

[0008] In summary, the two invasive methods are difficult to be widely used in clinical practice due to their economic, invasive and difficult operation. Moreover, since they require the placement of sensors in the coronary arteries, they bring additional physiological risks to patients. In addition, these methods require sensor measurements, have relatively more operation steps, and the detection process is cumbersome. Although the two non-invasive methods can directly obtain blood flow information from the reconstructed images and directly calculate CFR, the high cost of PET-CT and CMR examinations is not conducive to widespread clinical use.

[0009] In addition, the CFR calculation method based on coronary angiography does not require invasive Doppler guidewires or pressure guidewires, and can calculate CFR through computational fluid dynamics simulation methods. Specifically, the computer reconstructs the coronary artery morphology in three dimensions from multiple coronary angiography images from different angles, and then the pressure guidewire measures blood pressure data at the entrance and exit of the blood vessel to adjust the boundary conditions. CFD simulation is then used to calculate the pressure and blood flow in the coronary artery, gradually converging to its numerical solution. Finally, the CFR is calculated by the average transit time of the coronary artery in the hyperemic state and the resting state. However, this calculation method requires invasive coronary angiography in the catheterization room, which increases the patient's medical expenses and risk of complications. Summary of the invention

[0010] The present invention aims to provide a non-invasive method for calculating CFR, which is based on coronary artery CTA images to achieve non-invasive CFR calculation. It does not require invasively measured blood flow (such as blood pressure, flow velocity) data, but only requires CTA images. The CTA images are analyzed by the proposed algorithm and CFR is indirectly calculated. This method achieves the unity of economy and safety in calculating CFR. The obtained CFR value is accurate, can effectively evaluate coronary vascular function, and helps medical staff to specify the best treatment plan.

[0011] In order to achieve the above object, the present invention provides the following technical solutions:

[0012] A non-invasive method for calculating CFR comprises the following steps:

[0013] S1. Reconstruct the coronary artery 3D model and generate mesh

[0014] Based on the coronary artery CTA image data, the coronary artery CTA image is segmented, the geometric information of the coronary artery is extracted and the coronary artery three-dimensional model is reconstructed. The reconstructed three-dimensional model includes the coronary artery and its side branches, excluding the aorta;

[0015] During the model construction process, the distal small blood vessels were cut off, and the three-dimensional model was divided into tetrahedral meshes using an automated adaptive meshing method;

[0016] S2. Set the boundary conditions of the model

[0017] The inlet boundary condition is set to the mean aortic pressure MAP, and the outlet boundary condition is set to the outlet resistance R of each outlet of the model. i ;

[0018] S3. CFD configuration for simulating hemodynamics

[0019] By adjusting the parameters, CFD simulations were performed on the coronary blood flow in the resting state and the hyperemic state, respectively, and the pressure and flow velocity distribution of the entire coronary artery in the resting state and the hyperemic state were obtained;

[0020] S4. Calculate CFR

[0021] CFR was calculated by dividing the mean transit time in the hyperemic state by the mean transit time in the resting state.

[0022] Furthermore, in S1, the coronary artery CTA image is segmented using an automatic or semi-automatic segmentation algorithm.

[0023] Furthermore, in S2, the inlet boundary condition is set as follows:

[0024] Mean aortic pressure (MAP) is the average blood pressure in the aorta during one cardiac cycle. It reflects the combined effect of the amount of blood output per minute by the heart and the peripheral vascular resistance. It is an important indicator for evaluating cardiovascular function. The mean aortic pressure (MAP) at rest is calculated based on the pressure of the brachial cuff. rest :

[0025] MAP rest =0.4×(SBP-DBP)+DBP

[0026] In the formula, SBP and DBP are the systolic and diastolic blood pressures respectively;

[0027] When simulating the hyperemia state, select the average aortic pressure in the hyperemia state as the inlet boundary condition MAP hyp , converted from the resting mean aortic pressure:

[0028] MAP hyp =1 / 1.1*MAP rest

[0029] The method for setting the outlet boundary condition includes the following steps:

[0030] M1. Calculate the left ventricular blood flow Q at rest and in hyperemic state by using multiple specific physiological characteristics of each patient based on multivariate linear regression analysis;

[0031] For the resting state, multiple linear regression analysis showed that the resting coronary blood flow was strongly linearly correlated with the highest pulse pressure tertile value, left ventricular mass index, gender and diabetes status; HPPT was the highest pulse pressure tertile value, Mass LV is the left ventricular mass index, Sex is the gender, Diabetes is the diabetes status, then the resting blood flow Q rest for:

[0032] Q rest =a*HPPT+b*Mass LV +c*Sex+d*Diabetes+e

[0033] In the formula, a, b, c, d, and e are coefficients;

[0034] For the hyperemic state, multiple linear regression analysis showed that coronary blood flow during hyperemia was strongly correlated with the highest pulse pressure tertile, diabetes status, gender, age, highest systolic blood pressure tertile, and left ventricular mass index; HPPT was the highest pulse pressure tertile, Diabetes was the diabetes status, Sex was the gender, Age was the age, HSBPT was the highest systolic blood pressure tertile, Mass LV is the left ventricular mass index, then the blood flow in the hyperemic state Q hyp for:

[0035] Q hyp =f*HPPT+g*Diabetes+h*Sex+i*Age+j*HSBPT+k*Mass LV +l

[0036] In the formula, f, g, h, i, j, k, and l are coefficients;

[0037] M2, distribute the left and right coronary blood flow Q according to Murray's law in ; The coronary artery diameter D and the blood flow Q of the left and right coronary arteries are determined by the reconstructed three-dimensional model in for:

[0038]

[0039] In the formula, D1 and D2 are the diameters of the left and right coronary arteries, respectively, and β is 2.55;

[0040] M3. Distribute the blood flow Q at each outlet of the coronary artery according to Murray's law out ; The diameter D of each coronary artery branch outlet and the blood flow Q of the i-th outlet are determined by reconstructing the three-dimensional model out,i for:

[0041]

[0042] Where D i is the average diameter near the ith outlet, N is the total number of outlets, and β is 2.55;

[0043] M4. Determine the resting state outlet resistance of the i-th outlet and congestion state outlet resistance Outlet boundary conditions for simulating the intravascular dynamics in resting and hyperemic states, respectively:

[0044]

[0045]

[0046] P v The reference venous pressure was set to 5 mmHg.

[0047] Furthermore, in S3, the CFD method for simulating hemodynamics is:

[0048] Assume that the blood is a nearly Newtonian fluid, the tube wall is rigid, and there is no wall slip; by solving the three-dimensional incompressible Navier-Stokes equations, for the fluid Ω with a boundary of Г, solve the velocity and pressure

[0049]

[0050]

[0051] In the formula, is the velocity field, is the pressure field, t, μ, and ρ are time, fluid viscosity, and fluid density, respectively; the fluid viscosity and density of blood are set to 0.0035 Pa·s and 1050 Kg / m, respectively. 3 .

[0052] Furthermore, in S4, the method for calculating CFR is:

[0053] The hemodynamics in the resting state and the hyperemic state are simulated respectively, and this purpose is achieved by changing the boundary conditions; among them, the boundary condition of the resting state fluid domain is rest Set as the mean aortic pressure MAP at rest rest and exit speed Boundary conditions of fluid domain in congested state Г hyp Set to mean aortic pressure MAP at inlet under hyperemia hyp and outlet resistance After two simulations, the CFR is calculated based on the obtained pressure field and flow field:

[0054]

[0055] Where, T rest T is the mean transit time of coronary blood flow in the resting state, hyp The average transit time of coronary blood flow in a hyperemic state.

[0056] The beneficial effects of the technical solution are:

[0057] 1. The present invention provides a non-invasive method for calculating CFR. When calculating CFR, only non-invasive CTA imaging data is required, which avoids invasive examinations, reduces treatment costs and clinical risks, and has good clinical application value.

[0058] 2. The present invention provides a non-invasive method for calculating CFR, which can obtain the CFR value of each position in the target coronary artery model through one calculation and accurately locate the diseased blood vessel;

[0059] 3. The present invention provides a non-invasive method for calculating CFR, which can comprehensively evaluate the physiological function status of the entire coronary artery and help clinicians formulate the best treatment plan. BRIEF DESCRIPTION OF THE DRAWINGS

[0060] Figure 1 The present invention provides a method for non-invasively calculating CFR, and a flow chart of a non-invasively calculating CFR algorithm for determining boundary conditions. DETAILED DESCRIPTION

[0061] The present invention is further described in detail below in conjunction with the accompanying drawings and embodiments:

[0062] like Figure 1 As shown, a non-invasive method for calculating CFR includes the following steps:

[0063] S1. Reconstruct the coronary artery 3D model and generate mesh

[0064] Based on the coronary artery CTA image data, the coronary artery CTA image is segmented using an automatic or semi-automatic segmentation algorithm, the geometric information of the coronary artery is extracted and a coronary artery three-dimensional model is reconstructed. The reconstructed three-dimensional model includes the coronary artery and its collateral vessels, excluding the aorta;

[0065] During the model construction process, the distal small blood vessels were cut off, and the three-dimensional model was divided into tetrahedral meshes using an automated adaptive meshing method;

[0066] S2. Set the boundary conditions of the model

[0067] The inlet boundary condition is set to the mean aortic pressure MAP, and the outlet boundary condition is set to the outlet resistance R of each outlet of the model. i ;

[0068] The setting method of the inlet boundary condition is:

[0069] Mean aortic pressure (MAP) is the average blood pressure in the aorta during one cardiac cycle. It reflects the combined effect of the amount of blood output per minute by the heart and the peripheral vascular resistance. It is an important indicator for evaluating cardiovascular function. The mean aortic pressure (MAP) at rest is calculated based on the pressure of the brachial cuff. rest :

[0070] MAP rest =0.4×(SBP-DBP)+DBP

[0071] In the formula, SBP and DBP are the systolic and diastolic blood pressures respectively;

[0072] When simulating the hyperemia state, select the average aortic pressure in the hyperemia state as the inlet boundary condition MAP hyp , converted from the resting mean aortic pressure:

[0073] MAP hyp =1 / 1.1*MAP rest

[0074] The method for setting the outlet boundary condition includes the following steps:

[0075] M1. Calculate the left ventricular blood flow Q at rest and in hyperemic state by using multiple specific physiological characteristics of each patient based on multivariate linear regression analysis;

[0076] For the resting state, multiple linear regression analysis showed that the resting coronary blood flow was strongly linearly correlated with the highest pulse pressure tertile value, left ventricular mass index, gender and diabetes status; HPPT was the highest pulse pressure tertile value, Mass LV is the left ventricular mass index, Sex is the gender, Diabetes is the diabetes status, then the resting blood flow Q rest for:

[0077] Q rest =a*HPPT+b*Mass LV +c*Sex+d*Diabetes+e

[0078] In the formula, a, b, c, d, and e are coefficients;

[0079] For the hyperemic state, multiple linear regression analysis showed that coronary blood flow during hyperemia was strongly correlated with the highest pulse pressure tertile, diabetes status, gender, age, highest systolic blood pressure tertile, and left ventricular mass index; HPPT was the highest pulse pressure tertile, Diabetes was the diabetes status, Sex was the gender, Age was the age, HSBPT was the highest systolic blood pressure tertile, Mass LV is the left ventricular mass index, then the blood flow in the hyperemic state Q hyp for:

[0080] Q hyp =f*HPPT+g*Diabetes+h*Sex+i*Age+j*HSBPT+k*Mass LV +l

[0081] In the formula, f, g, h, i, j, k, and l are coefficients;

[0082] M2, distribute the left and right coronary blood flow Q according to Murray's law in ; The coronary artery diameter D and the blood flow Q of the left and right coronary arteries are determined by the reconstructed three-dimensional model in for:

[0083]

[0084] In the formula, D1 and D2 are the diameters of the left and right coronary arteries, respectively, and β is 2.55;

[0085] M3. Distribute the blood flow Q at each outlet of the coronary artery according to Murray's law out ; The diameter D of each coronary artery branch outlet and the blood flow Q of the i-th outlet are determined by reconstructing the three-dimensional model out,i for:

[0086]

[0087] Where D i is the average diameter near the ith outlet, N is the total number of outlets, and β is 2.55;

[0088] M4. Determine the resting state outlet resistance of the i-th outlet and congestion state outlet resistance Outlet boundary conditions for simulating the intravascular dynamics in resting and hyperemic states, respectively:

[0089]

[0090]

[0091] P vIt is the reference venous pressure, which is set to 5 mmHg;

[0092] S3. CFD configuration for simulating hemodynamics

[0093] Assume that the blood is a nearly Newtonian fluid, the tube wall is rigid, and there is no wall slip; by solving the three-dimensional incompressible Navier-Stokes equations, for the fluid Ω with a boundary of Г, solve the velocity and pressure

[0094]

[0095]

[0096] In the formula, is the velocity field, is the pressure field, t, μ, and ρ are time, fluid viscosity, and fluid density, respectively; the fluid viscosity and density of blood are set to 0.0035 Pa·s and 1050 Kg / m, respectively. 3 ;

[0097] By adjusting the parameters, CFD simulations were performed on the coronary blood flow in the resting state and the hyperemic state, respectively, and the pressure and flow velocity distribution of the entire coronary artery in the resting state and the hyperemic state were obtained;

[0098] S4. Calculate CFR

[0099] The hemodynamics in the resting state and the hyperemic state are simulated respectively, and this purpose is achieved by changing the boundary conditions; among them, the boundary condition of the resting state fluid domain is rest Set as the mean aortic pressure MAP at rest rest and exit speed Boundary conditions of fluid domain in congested state Г hyp Set to mean aortic pressure MAP at inlet under hyperemia hyp and outlet resistance After two simulations, the CFR is calculated based on the obtained pressure field and flow field:

[0100]

[0101] Where, T rest T is the mean transit time of coronary blood flow in the resting state, hyp The average transit time of coronary blood flow in a hyperemic state.

[0102] The above is only an embodiment of the present invention, and the common knowledge such as the known specific technical solutions or characteristics in the solution is not described in detail here. It should be pointed out that for those skilled in the art, several modifications and improvements can be made without departing from the technical solution of the present invention, which should also be regarded as the protection scope of the present invention, and these will not affect the effect of the implementation of the present invention and the practicality of the patent. The scope of protection required by this application shall be based on the content of its claims, and the specific implementation methods and other records in the specification can be used to interpret the content of the claims.

Claims

1. A non-invasive method for calculating CFR, characterized in that: The following steps are involved: S1. Reconstruct the coronary artery 3D model and generate mesh Based on the coronary artery CTA image data, the coronary artery CTA image is segmented, the geometric information of the coronary artery is extracted and the coronary artery three-dimensional model is reconstructed. The reconstructed three-dimensional model includes the coronary artery and its side branches, excluding the aorta; During the model construction process, the distal small blood vessels were cut off, and the three-dimensional model was divided into tetrahedral meshes using an automated adaptive meshing method; S2. Set the boundary conditions of the model The inlet boundary condition is set to the mean aortic pressure MAP, and the outlet boundary condition is set to the outlet resistance R of each outlet of the model. i ; S3. CFD configuration for simulating hemodynamics By adjusting the parameters, CFD simulations were performed on the coronary blood flow in the resting state and the hyperemic state, respectively, and the pressure and flow velocity distribution of the entire coronary artery in the resting state and the hyperemic state were obtained; Among them, the CFD method for simulating hemodynamics is: Assume that the blood is a nearly Newtonian fluid, the tube wall is rigid, and there is no wall slip; by solving the three-dimensional incompressible Navier-Stokes equations, for the fluid Ω with a boundary of Г, solve the velocity and pressure In the formula, is the velocity field, is the pressure field, t, μ, and ρ are time, fluid viscosity, and fluid density, respectively; the fluid viscosity and density of blood are set to 0.0035 Pa·s and 1050 Kg / m, respectively. 3 ; S4. Calculate CFR The CFR was calculated by dividing the mean transit time in the hyperemic state by the mean transit time in the resting state; The method for calculating CFR is: The hemodynamics in the resting state and the hyperemic state are simulated respectively, and this purpose is achieved by changing the boundary conditions; among them, the boundary condition of the resting state fluid domain is rest Set as the mean aortic pressure MAP at rest rest and outlet resistance Boundary conditions of fluid domain in congested state Г hyp Set to mean aortic pressure MAP at inlet under hyperemia hyp and outlet resistance After two simulations, the CFR is calculated based on the obtained pressure field and flow field: Where, T rest T is the mean transit time of coronary blood flow in the resting state, hyp The average transit time of coronary blood flow in a hyperemic state.

2. A non-invasive method for calculating CFR according to claim 1, characterized in that: In S1, the coronary artery CTA images are segmented using an automatic or semi-automatic segmentation algorithm.

3. A non-invasive method for calculating CFR according to claim 1, characterized in that: In S2, the inlet boundary conditions are set as follows: Mean aortic pressure (MAP) is the average blood pressure in the aorta during one cardiac cycle. It reflects the combined effect of the amount of blood output per minute by the heart and the peripheral vascular resistance. It is an important indicator for evaluating cardiovascular function. The mean aortic pressure (MAP) at rest is calculated based on the pressure of the brachial cuff. rest : MAP rest =0.4×(SBP-DBP)+DBP In the formula, SBP and DBP are the systolic and diastolic blood pressures respectively; When simulating the hyperemia state, select the average aortic pressure in the hyperemia state as the inlet boundary condition MAP hyp , converted from the resting mean aortic pressure: MAP hyp =1 / 1.1*MAP rest The method for setting the outlet boundary condition includes the following steps: M1. Calculate the left ventricular blood flow Q at rest and in hyperemic state by using multiple specific physiological characteristics of each patient based on multivariate linear regression analysis; For the resting state, multiple linear regression analysis showed that the resting coronary blood flow was strongly linearly correlated with the highest pulse pressure tertile value, left ventricular mass index, gender and diabetes status; HPPT was the highest pulse pressure tertile value, Mass LV is the left ventricular mass index, Sex is the gender, Diabetes is the diabetes status, then the resting blood flow Q rest for: Q rest =a*HPPT+b*Mass LV +c*Sex+d*Diabetes+e In the formula, a, b, c, d, and e are coefficients; For the hyperemic state, multiple linear regression analysis showed that coronary blood flow during hyperemia was strongly correlated with the highest pulse pressure tertile, diabetes status, gender, age, highest systolic blood pressure tertile, and left ventricular mass index; HPPT was the highest pulse pressure tertile, Diabetes was the diabetes status, Sex was the gender, Age was the age, HSBPT was the highest systolic blood pressure tertile, Mass LV is the left ventricular mass index, then the blood flow in the hyperemic state Q hyp for: Q hyp =f*HPPT+g*Diabetes+h*Sex+i*Age+j*HSBPT+k*Mass LV +l In the formula, f, g, h, i, j, k, and l are coefficients; M2, distribute the left and right coronary blood flow Q according to Murray's law in ; The coronary artery diameter D and the blood flow Q of the left and right coronary arteries are determined by the reconstructed three-dimensional model in for: In the formula, D1 and D2 are the diameters of the left and right coronary arteries, respectively, and β is 2.55; M3. Distribute the blood flow Q at each outlet of the coronary artery according to Murray's law out ; The diameter D of each coronary artery branch outlet and the blood flow Q of the i-th outlet are determined by reconstructing the three-dimensional model out,i for: Where D i is the average diameter near the ith outlet, N is the total number of outlets, and β is 2.55; M4. Determine the resting state outlet resistance of the i-th outlet and congestion state outlet resistance Outlet boundary conditions for simulating the intravascular dynamics in resting and hyperemic states, respectively: P v The reference venous pressure was set to 5 mmHg.

Images