A method and system for calculating fractional flow reserve based on a frustum-type cross-sectional microelement

Abstract

The application discloses a kind of based on the calculation of blood flow reserve fraction method and system of circular truncated cone cross section microelement. Method includes a kind of lumen segmentation method: the real area and perimeter of the cross section lumen obtained by extraction segmentation are extracted;A kind of cross section microelement construction method: the hydraulic equivalent area of each cross section lumen is obtained;A kind of fluid mechanics calculation method: the total pressure loss of target coronary artery is obtained;A kind of blood flow reserve fraction calculation method: the blood flow reserve fraction of target coronary artery is obtained.The application solves the problem that the pressure difference calculation of diffuse lesion is not accurate;Without relying on pressure guide wire and hyperemic drug, only through coronary three-dimensional image and extracorporeal blood pressure measurement can be completed, simple operation, short time-consuming, can satisfy clinical actual decision needs.

Description

Technical Field

[0001] This invention belongs to the field of medical image processing, and specifically relates to a method and system for calculating the fractional blood flow reserve based on a frustum-shaped cross-section micro-element. Background Technology

[0002] Coronary artery disease (CAD) is the leading cause of cardiovascular disease morbidity and mortality in adults worldwide. Its core pathological mechanism involves coronary atherosclerosis leading to stenosis or occlusion of the coronary arteries, resulting in myocardial ischemia and hypoxia, and in severe cases, myocardial necrosis. While coronary angiography can visually demonstrate the anatomical severity of vascular stenosis, it cannot reflect the impact of the stenosis on cardiovascular function, i.e., it cannot determine whether the stenosis causes myocardial ischemia. Fractional flow reserve (FFR), as the "gold standard" for assessing whether coronary artery stenosis causes myocardial ischemia, is valuable in accurately evaluating coronary artery stenosis with unknown functional effects, providing objective evidence for clinical intervention.

[0003] Traditional fractional flow reserve (FFR) measurement relies on pressure guidewire intervention, which has significant limitations: it requires invasive surgery to deliver the pressure sensor to the distal end of the blood vessel and the injection of vasoconstrictive drugs such as adenosine to induce maximum myocardial flow. This process is complex, and the equipment and consumables are expensive, not only prolonging the procedure but also causing significant individual variability in vasoconstrictive drug responses—some patients may not tolerate the drug's side effects, or the drug may not effectively induce maximum flow, leading to inaccurate FFR measurements. These limitations have resulted in a consistently low clinical application rate of traditional FFR, failing to meet large-scale clinical needs. To overcome these limitations, image-based computational FFR technology has emerged, currently mainly divided into two categories:

[0004] Computational fluid dynamics numerical simulation of fractional flow reserve (FVR): This method reconstructs the three-dimensional geometry of blood vessels using medical images, solves the Navier-Stokes equations to simulate hemodynamic states, and then calculates the FVR. However, this technique requires high precision in vascular geometry reconstruction, and the numerical calculation process is complex, with a single analysis typically taking 30 minutes to several hours, making it difficult to meet the real-time requirements of clinical practice and unsuitable for immediate decision-making during surgery. Analytical fluid dynamics simplified calculation of FVR: This method simplifies the blood flow model (e.g., dividing the vessel into "normal segment-stenotic segment-normal segment") and calculates the pressure loss in the vessel based on the classical fluid dynamics formula for pipe pressure loss, ultimately obtaining the FVR. This technique is widely used in clinical practice due to its fast calculation process and ease of operation, but the simplified model also introduces limitations in computational accuracy.

[0005] Current methods for calculating blood flow pressure changes based on geometric parameters such as coronary artery diameter or cross-sectional area have a core technical flaw: they cannot accurately distinguish the differential impact of geometric parameter changes on blood flow pressure under different degrees of stenosis, especially in the assessment of moderate diffuse coronary artery lesions, where the error is more significant. Specifically, existing methods typically simplify stenotic vessels into a cylindrical structure connecting "normal segment - uniformly stenotic segment - normal segment." This simplified model completely ignores key characteristics of diffuse stenosis—such as long transition sections at the stenosis inlet and outlet, wide vascular extent affected by stenosis, and complex luminal morphology. Furthermore, existing computational fractional flow reserve (FLR) methods only calculate the pressure loss along the stenotic segment (based on Poiseuille's law) and the local pressure loss due to sudden expansion at the outlet, without considering the effects of gradual contraction at the inlet, blood flow disturbance within the stenotic segment, and the expansion angle on blood flow separation. Simultaneously, coronary arteries are subject to periodic compression by the myocardium, resulting in irregular luminal cross-sections. Existing models directly assume a circular lumen, ignoring the influence of non-circular structures on blood flow resistance, ultimately leading to deviations in pressure loss calculations and affecting the accuracy of FLR assessment.

[0006] The technical solution disclosed in patent CN112535466A proposes using an artificial intelligence model to segment target blood vessels. Specifically, the target blood vessel is divided into several morphologically similar segments, with the morphological characteristics of "contraction followed by expansion," "simple contraction," or "simple expansion," and the length of each segment has adaptive variation characteristics. After completing the blood vessel segmentation, this technical solution further uses an artificial intelligence model to predict blood vessel-related parameters. The network structure of the artificial intelligence model 3 satisfies two core conditions: first, it contains at least one convolutional layer; second, it has at least one fully connected layer. This artificial intelligence model can predict the pressure loss coefficient along the blood vessel and the local pressure loss coefficient. After obtaining the pressure loss coefficient, it is combined with a conventional fluid model for calculation, ultimately obtaining the pressure loss and blood flow reserve fraction of the blood vessel. However, this existing technology has obvious technical defects: On the one hand, because the segment lengths formed by segmenting blood vessels are not fixed, it is necessary to collect and construct a large amount of lumen pressure loss data with different lengths, different diameter ratios, and different expansion angles when building the training dataset of the above-mentioned artificial intelligence model. This undoubtedly increases the difficulty and cost of dataset construction, and also has an adverse effect on the efficiency and stability of model training. On the other hand, this technology only relies on the artificial intelligence model to predict the pressure loss coefficient of each segment, and fails to fully consider the inherent hydrodynamic characteristics of the coronary artery stenosis area itself. This makes it difficult for the prediction results to accurately reflect the real hydrodynamic state of the stenotic blood vessel, which may affect the reliability and accuracy of subsequent pressure loss and fractional flow reserve calculation results.

[0007] Therefore, there is an urgent need for a method and system for calculating fractional flow reserve that can accurately capture the microscopic geometric features of blood vessels and comprehensively calculate blood flow pressure loss, so as to improve the assessment accuracy of special cases such as diffuse coronary artery disease and provide more reliable functional basis for clinical diagnosis and treatment. Summary of the Invention

[0008] To address the problems existing in the background art, this invention provides a method and system for calculating fractional flow reserve based on a frustum-shaped cross-section micro-element. This solves the technical problem of the lack of a method and system for calculating fractional flow reserve that can accurately capture the microscopic geometric features of blood vessels and comprehensively calculate blood flow pressure loss. This invention improves the accuracy of assessment for special cases such as diffuse coronary artery disease, providing more reliable functional evidence for clinical diagnosis and treatment.

[0009] The technical solution adopted in this invention is:

[0010] I. A method for calculating fractional blood flow reserve based on a frustum-shaped cross-section micro-element:

[0011] S1. A lumen segmentation method: Based on the three-dimensional image of the target coronary artery, the main vessels in the target coronary artery are segmented along the cross-section, and the actual area and perimeter of the lumen of each segmented section are extracted.

[0012] S2. A method for constructing cross-sectional micro-element: The whole between two adjacent cross-sectional cavities is regarded as a frustum-shaped cross-sectional micro-element. The actual area and perimeter of the cross-sectional cavity obtained in step S1 are resampled according to the preset thickness of the frustum-shaped cross-sectional micro-element. Based on the resampled actual area and perimeter, the hydraulic equivalent area corresponding to each cross-sectional cavity is obtained.

[0013] S3. A fluid dynamics calculation method: obtain the reference blood flow velocity based on the hydraulic equivalent area corresponding to the cross-section cavity, obtain the pressure loss along the cross-section and the local pressure loss of each frustum-shaped cross-section micro-element based on the structure of the frustum-shaped cross-section micro-element and the hydraulic equivalent area corresponding to the cross-section cavity and the reference blood flow velocity, and then obtain the total pressure loss of the target coronary artery.

[0014] S4. A method for calculating fractional flow reserve: The fractional flow reserve of the target coronary artery is obtained by combining the mean aortic pressure and the total pressure loss of the target coronary artery.

[0015] Step S1 specifically involves:

[0016] S11. Acquire a three-dimensional image of the target coronary artery; the acquisition of the three-dimensional image of the target coronary artery adopts one of the following medical imaging techniques: intravascular ultrasound, intravascular optical coherence tomography, and computed tomography.

[0017] S12. Based on the three-dimensional image, a lumen segmentation algorithm based on image processing or neural network is used to segment the main blood vessel in the target coronary artery along the cross section to obtain the contour data of the lumen of each segmented section.

[0018] S13. Based on the contour data of the cross-sectional cavity, the pixel area and perimeter of the cross-sectional cavity are statistically analyzed. Each pixel is converted to the corresponding actual physical length, thereby obtaining the real area and perimeter of each cross-sectional cavity.

[0019] The frustum-shaped cross-section micro-element is divided into a contracting cross-section micro-element and a expanding cross-section micro-element; the contracting cross-section micro-element has an inlet hydraulic equivalent area greater than or equal to the outlet hydraulic equivalent area; the expanding cross-section micro-element has an inlet hydraulic equivalent area less than the outlet hydraulic equivalent area; the inlet hydraulic equivalent area of ​​the frustum-shaped cross-section micro-element is the hydraulic equivalent area corresponding to the previous cross-section lumen along the blood flow direction; the outlet hydraulic equivalent area of ​​the frustum-shaped cross-section micro-element is the hydraulic equivalent area corresponding to the next cross-section lumen along the blood flow direction.

[0020] The hydraulically equivalent area corresponding to the cross-sectional cavity is set according to the following formula:

[0021] A i =πd i 2 / 4;d i =4A ri / C ri

[0022] Where i represents the index; A i d represents the hydraulically equivalent area of ​​the i-th cross-section of the pipe; i A represents the hydraulic diameter of the i-th cross-section of the pipe. ri C represents the actual area corresponding to the i-th cross-section of the tube; ri This represents the actual perimeter of the lumen corresponding to the i-th cross-section.

[0023] Step S3 specifically involves:

[0024] S31. Based on the hydraulic equivalent area of ​​all cross-section lumens, obtain the reference blood flow velocity corresponding to each cross-section lumen using either Method 1 or Method 2.

[0025] S32. Obtain the friction loss and local pressure loss corresponding to the frustum-shaped cross-section micro-element based on the reference blood flow velocity, hydraulic equivalent area, and type of the frustum-shaped cross-section micro-element for each cross-section lumen.

[0026] S33. The sum of the pressure loss along the friction and the local pressure loss corresponding to all the frustum-shaped cross-section micro-elements is the total pressure loss of the target coronary artery.

[0027] The first method specifically involves: selecting the area of ​​the cross-sectional lumen of a normal coronary artery as the initial reference lumen area of ​​the first cross-sectional lumen of the main vessel in the target coronary artery; obtaining the reference blood flow of the main vessel based on the reference area of ​​the first cross-sectional lumen and the hydraulic equivalent area corresponding to all cross-sectional lumens; determining whether the target coronary artery has branch vessels based on the three-dimensional image of the target coronary artery using a branch detection algorithm based on a neural network; obtaining the reference blood flow at the location of each cross-sectional lumen based on the presence or absence of branch vessels; and obtaining the reference blood flow velocity corresponding to each cross-sectional lumen based on the reference blood flow of the main vessel, the hydraulic equivalent area corresponding to each cross-sectional lumen, and the reference blood flow at the location.

[0028] Method 1 is set according to the following formula to obtain the reference blood flow velocity corresponding to each cross-section lumen:

[0029] u i =Q i / A i Q1=Q

[0030] When there are no branch vessels at the i-th cross-section lumen: Q i =Q i-1

[0031] When a branch vessel exists at the lumen of the i-th cross-section: Q i =Q i-1 ·(d i ') 3 / ((d i ') 3 +(d bi ) 3 )

[0032] Q=u·(Σ i=1 n A refi ') / n

[0033] A refi '=A refi +(max(A j -A refj ,n')) / n',j=1,2,…,n

[0034] A refi =A ref1 β i-1 ;β=200(Σ i=1 n A i ) / n+0.995

[0035] Where i represents the index; u i Q represents the reference blood flow velocity corresponding to the lumen at the i-th cross-section;i and Q i-1 A represents the reference blood flow rate at the location of the lumen in the i-th and (i-1)-th cross-sections, respectively; i d represents the hydraulically equivalent area of ​​the lumen at the i-th cross-section; Q1 represents the reference blood flow rate at the location of the lumen at the 1-th cross-section; Q represents the reference blood flow rate of the main trunk of the target coronary artery; d i ' represents the diameter of the main blood vessel at the i-th cross-section; d bi represents the diameter of the branch vessel at the lumen of the i-th cross-section; Q represents the reference blood flow rate of the main coronary artery; u represents the clinically experienced blood flow velocity at rest; A refi ' represents the final reference cavity area of ​​the i-th cross-section cavity; A refi Let represent the initial reference lumen area of ​​the i-th cross-section; j represents the index; n represents the total number of cross-section lumens; max(A j -A refj (n') represents the vector A j -A refj A function that selects the first n' values ​​after descending order; A j A represents the hydraulically equivalent area of ​​the j-th cross-section of the pipe cavity; refj A represents the initial reference cavity area of ​​the j-th cross-section cavity; ref1 β represents the initial reference lumen area of ​​the first cross-section; β is the lumen taper.

[0036] The second method is specifically as follows:

[0037] D1. Construct a sequence of cross-sectional cavities in the order of arrangement of all cross-sectional cavities. Using each cross-sectional cavity as the center of the sliding window, slide the sequence of cross-sectional cavities according to the preset sliding window size with a sliding step of one. Obtain the initial reference cavity area of ​​the cross-sectional cavity sequence within each sliding window to get the initial reference cavity area corresponding to the cross-sectional cavity at the center of the sliding window, and finally obtain the initial reference cavity area corresponding to each cross-sectional cavity sequence.

[0038] The specific process for obtaining the initial reference cavity area is as follows: Calculate the N% quantile of the hydraulic equivalent area corresponding to all cross-section cavities within the sliding window; take all cross-section cavities with a hydraulic equivalent area greater than the N% quantile as candidate normal cross-section cavities; calculate the M% quantile of the hydraulic equivalent area corresponding to all candidate normal cross-section cavities; and take the hydraulic equivalent area of ​​the candidate normal cross-section cavities corresponding to the M% quantile as the initial reference cavity area corresponding to the cross-section cavity at the center of the current sliding window.

[0039] D2. According to the arrangement order of the cross-sectional cavities, the final reference cavity area is obtained for each cross-sectional cavity in sequence, so as to obtain the final reference cavity area corresponding to all cross-sectional cavities.

[0040] D3. Based on the final reference lumen area corresponding to all cross-section lumens, the reference blood flow velocity is obtained sequentially for each cross-section lumen according to the arrangement order of the cross-section lumens, thereby obtaining the reference blood flow velocity corresponding to each cross-section lumen.

[0041] The final reference cavity area acquisition process is set according to the following formula to obtain the final reference cavity area corresponding to the current i-th cross-section cavity:

[0042] When A ref1 <7mm 2 Time: A refi '=A refi ''+(7–A ref1 '')

[0043] When A ref1 ≥7mm 2 Time: A refi '=A refi ''

[0044] A refi ''=A refi (1) –min(A refj (1) –A j ,n'') / n'',j=1,2,…,n

[0045] When K Ai >0:

[0046] A reft (1) =A reft +K Ai ΔL, t=1,2,…,i-1; A reft (1) =A reft t=i,i+1…,n

[0047] When -0.03 ≤ K Ai When ≤0:

[0048] A reft (1) =A reft t=1,2,…,n

[0049] When K Ai When <-0.03:

[0050] A reft (1) =A reft , t=1,2,…,i-1; A reft (1) =A reft+ΔL(-0.03–K Ai ), t=i,i+1…,n

[0051] K A1 =0;K Ai =(A refi –A refi-1 ) / ΔL

[0052] Where i, j, and t all represent indices; A refi ' represents the final reference cavity area corresponding to the i-th cross-section cavity; A ref1 '' and A refi '' represents the intermediate reference cavity area corresponding to the 1st and i-th cross-section cavities, respectively; A refi (1) A refj (1) and A reft (1) Let A represent the adjusted initial reference cavity areas corresponding to the i-th, j-th, and t-th cross-section cavities, respectively; n represents the total number of cross-section cavities; A j Min(A) represents the hydraulically equivalent area of ​​the j-th cross-section of the pipe. refj (1) –A j ,n'') represents the vector A refj (1) –A j A function that selects the first n'' values ​​after ascending order; in specific implementation, n'' <n;K A1 and K Ai K represents the slopes corresponding to the 1st and 1st cross-sections of the tube, respectively; A1 and K Ai A represents the slope corresponding to the 1st and 1st cross-sections of the tube, respectively; reft A refi and A refi-1 ΔL represents the initial reference cavity area corresponding to the t-th, i-th, and i-1-th cross-section cavities, respectively; ΔL is the preset thickness of the frustum-shaped cross-section micro-element; n represents the total number of cross-section cavities.

[0053] The reference blood flow velocity acquisition process is set according to the following formula to obtain the reference blood flow velocity corresponding to the current i-th cross-section lumen:

[0054] u i =u'A refi '''' / A i

[0055] When there are no branch vessels at the current i-th cross-section lumen:

[0056] A reft ''''=A reft',t=i,i+1,…,n

[0057] When there is a branch vessel at the lumen of the current i-th cross-section:

[0058] A reft ''''=A reft '–(A refi-1 '-((A refi-1 ') 3 / 2 -(A bi ) 3 / 2 ) 2 / 3 ), t=i,i+1,…,n

[0059] A bi =4πA rbi 2 / C rbi 2

[0060] Where i and t are both indices; u i U is the reference blood flow velocity corresponding to the lumen at the i-th cross-section; u' is the clinically experienced blood flow velocity under congestion; A i Let A be the hydraulically equivalent area of ​​the i-th cross-section of the pipe. refi '''' and A reft '''' represents the final reference cavity area after adjustment for the i-th and t-th cross-section cavities, respectively; A reft 'and A refi-1 'A' represents the final reference cavity area corresponding to the t-th and (i-1)-th cross-sections, respectively; bi Let A be the hydraulically equivalent area of ​​the branch vessel at the i-th cross-section lumen; rbi C represents the actual area of ​​the branch vessel at the lumen of the i-th cross-section; rbi This represents the actual circumference of the branch vessel at the i-th cross-section.

[0061] The friction loss and local pressure loss corresponding to the frustum-shaped cross-section micro-element in step S32 are set according to the following formula:

[0062] When the i-th frustum-shaped section element is a tapered section element:

[0063] dP fi =ρλ i u i+1 2 (1-n c 2 ) / (16sin(θ i / 2));dP ji =(0.0348sin(θ i / 2)(1-n c 2.5 )ξi 2 +(ξ i -1) 2 / 2)ρu i+1 2

[0064] When the i-th frustum-shaped section element is an expanding section element:

[0065] dP fi =ρλ i u i 2 (1-n e 2 ) / (16sin(θ i / 2));dP ji =4.15(tan(θ i / 2)) 2 (1-n e ) 2 ρu i 2

[0066] λ i =16νπ 1 / 2 / (ρu i A i 1 / 2 )+16νπ 1 / 2 / (ρu i+1 A i+1 1 / 2 )

[0067] n c =A i+1 / A i ;n e =A i / A i+1 ξ i =1+0.622(θ i / 180) 0.8 (1-0.215n c -0.785n c 2.5 )

[0068] Where i represents the index; dP fi dP represents the friction loss along the path corresponding to the i-th frustum-shaped cross-section element; ji λ represents the local pressure loss corresponding to the i-th frustum-shaped cross-section element; ρ represents the blood density; i u represents the average friction loss coefficient corresponding to the i-th cross-section of the pipe. i and u i+1 These represent the reference blood flow velocities corresponding to the ith and (i+1)th cross-sections of the lumen, respectively; nc This represents the ratio of the hydraulically equivalent area of ​​the outlet and inlet of a truncated cone-shaped tapered section element; n e A represents the ratio of the hydraulically equivalent area of ​​the inlet and outlet of a frustum-shaped, gradually expanding cross-section element; i and A i+1 θ represents the hydraulically equivalent area corresponding to the i-th and (i+1)-th cross-sections of the pipe, respectively; i ξ represents the shrinking angle of a tapered cross-section element or the expanding angle of a dilatant cross-section element; i ν represents the jet velocity ratio; ν represents blood viscosity.

[0069] II. A system for calculating fractional flow reserve suitable for methods of calculating fractional flow reserve:

[0070] The lumen segmentation module is used to acquire three-dimensional images of the target coronary artery and extract the true area and perimeter of the segmented lumen based on the three-dimensional images.

[0071] The cross-section micro-element construction module is used to construct frustum-shaped cross-section micro-elements and obtain the hydraulic equivalent area corresponding to each cross-section cavity after resampling.

[0072] The fluid dynamics calculation module is used to obtain the pressure loss along the friction and the local pressure loss of each frustum-shaped cross-section micro-element, and then obtain the total pressure loss of the target coronary artery.

[0073] The fractional flow reserve (FVR) calculation module is used to obtain the FVR of the target coronary artery.

[0074] The beneficial effects of this invention are:

[0075] This invention extracts the area and perimeter of the lumen from three-dimensional coronary artery images. Through hydraulic equivalent area conversion, it eliminates the bias in calculating blood flow resistance caused by non-circular lumens, making it particularly suitable for irregular lumens compressed by the myocardium. The blood vessel is modeled as a series of frustum-shaped cross-sectional micro-elements, and pressure loss along the vessel and local pressure loss are obtained separately. It fully considers factors such as inlet constriction, outlet expansion angle, and stenosis disturbance, solving the problem of inaccurate pressure difference calculation in diffuse lesions. It does not rely on pressure guidewires and congestion drugs, and can be completed solely through three-dimensional coronary artery images and external blood pressure measurement. The operation is simple and time-saving, meeting the actual clinical decision-making needs. Attached Figure Description

[0076] Figure 1 This is a flowchart of the method of the present invention.

[0077] Figure 2 The image shows the original cross-sectional view of the target coronary artery under intravascular optical coherence tomography.

[0078] Figure 3The image shows the segmentation of the lumen in a cross-sectional image of the target coronary artery under intravascular optical coherence tomography.

[0079] Figure 4 This diagram illustrates the measurement of the diameter of the main coronary artery and its branches.

[0080] Figure 5 This is a schematic diagram of modeling a narrowed coronary artery based on a frustum-shaped cross-section micro-element.

[0081] Figure 6 The curves show the results of the fractional flow reserve calculated by the method of this invention (solid line) and the existing fractional flow reserve calculation method (dashed line). Detailed Implementation

[0082] The present invention will now be described in more detail with reference to the accompanying drawings and embodiments. However, the present invention is not limited thereto. For those skilled in the art, several improvements and modifications can be made without departing from the principles of the present invention, and these improvements and modifications are also considered to be within the scope of protection of the present invention. Contents not described in detail in this specification are prior art known to those skilled in the art.

[0083] like Figure 1 As shown, the method for calculating the fractional blood flow reserve based on a frustum-shaped cross-section element in this embodiment includes the following steps:

[0084] S1. A lumen segmentation method: Based on the three-dimensional image of the target coronary artery, the main vessels in the target coronary artery are segmented along the cross-section, and the actual area and perimeter of the lumen of each segmented section are extracted.

[0085] S11. Acquire three-dimensional images of the target coronary artery according to clinical standards and store them in a computer database; the acquisition of three-dimensional images of the target coronary artery uses one of the following medical imaging techniques: intravascular ultrasound (IVUS), intravascular optical coherence tomography (IVOCT), and computed tomography (CT).

[0086] S12. Based on the three-dimensional images in the computer database, a lumen segmentation algorithm based on image processing or neural networks is used to segment the main coronary artery along the cross-section to obtain the contour data of the lumen of each segmented section.

[0087] S13. Based on the contour data of the cross-sectional cavity, the pixel area and perimeter of the cross-sectional cavity are statistically analyzed. Each pixel is converted to the corresponding actual physical length, thereby obtaining the real area and perimeter of each cross-sectional cavity.

[0088] In this embodiment, IVOCT images of the target coronary artery are obtained using clinical standards, and the original image of one cross-section of the lumen is shown below. Figure 2 As shown, the PM-UNet network was used to segment the lumen of the IVOCT image to obtain the morphological features of the blood vessels. Figure 3 As shown, the pixel area and perimeter corresponding to the morphological features of each coronary artery section are calculated. Combined with the imaging depth and number of sampling points of the IVOCT system, the physical length corresponding to each pixel is calculated, and then the true area and true perimeter of the cross-section lumen are obtained.

[0089] S2. A method for constructing a cross-sectional micro-element: The entire area between two adjacent cross-sectional cavities is considered as a frustum-shaped cross-sectional micro-element. The actual area and perimeter of the cross-sectional cavity obtained in step S1 are resampled according to a preset frustum-shaped cross-sectional micro-element thickness. This ensures that the cross-sectional interval between two adjacent cross-sectional cavities after resampling matches the preset frustum-shaped cross-sectional micro-element thickness, eliminating errors caused by differences in IVOCT equipment acquisition parameter settings. The hydraulic equivalent area corresponding to each cross-sectional cavity is obtained based on the resampled actual area and perimeter. In this embodiment, the micro-element thickness is 0.2 mm.

[0090] The frustum-shaped cross-section element is divided into a contracting cross-section element and a expanding cross-section element. For a contracting cross-section element, the hydraulic equivalent area at the inlet is greater than or equal to the hydraulic equivalent area at the outlet. For a expanding cross-section element, the hydraulic equivalent area at the inlet is less than the hydraulic equivalent area at the outlet. The hydraulic equivalent area at the inlet of the frustum-shaped cross-section element is the hydraulic equivalent area corresponding to the preceding cross-section lumen along the blood flow direction. The hydraulic equivalent area at the outlet of the frustum-shaped cross-section element is the hydraulic equivalent area corresponding to the following cross-section lumen along the blood flow direction. The preceding cross-section lumen is located upstream, and the following cross-section lumen is located downstream.

[0091] like Figure 4 As shown, using these two types of frustum cross-section micro-elements, a segment of a target coronary artery with narrowing is modeled as a series of lumens formed by sequentially connecting tapered and expanded cross-section micro-elements, which are highly similar to the original lumen morphology.

[0092] The hydraulically equivalent area corresponding to the cross-sectional cavity is set according to the following formula:

[0093] A i =π(d i ) 2 / 4;d i =4A ri / C ri

[0094] Where i represents the index; A i d represents the hydraulically equivalent area of ​​the i-th cross-section of the pipe; i A represents the hydraulic diameter of the i-th cross-section of the pipe. riC represents the actual area of ​​the lumen corresponding to the i-th cross-section after resampling; ri This represents the true perimeter of the lumen at the i-th cross-section after resampling.

[0095] S3. A fluid dynamics calculation method: obtain the reference blood flow velocity based on the hydraulic equivalent area corresponding to the cross-section cavity, obtain the pressure loss along the cross-section and the local pressure loss of each frustum-shaped cross-section micro-element based on the structure of the frustum-shaped cross-section micro-element and the hydraulic equivalent area corresponding to the cross-section cavity and the reference blood flow velocity, and then obtain the total pressure loss of the target coronary artery.

[0096] S31. Based on the hydraulic equivalent area of ​​all cross-section lumens, obtain the reference blood flow velocity corresponding to each cross-section lumen using either Method 1 or Method 2.

[0097] Method 1 specifically involves: selecting the area of ​​the lumen of a normal coronary artery cross-section verified by clinical experience as the initial reference lumen area of ​​the first cross-section lumen of the main vessel in the target coronary artery; obtaining the reference blood flow of the main vessel of the target coronary artery based on the reference area of ​​the first cross-section lumen and the hydraulic equivalent area corresponding to all cross-section lumens; determining whether the target coronary artery has branch vessels based on the three-dimensional image of the target coronary artery using a branch detection algorithm based on a neural network; obtaining the reference blood flow at the location of each cross-section lumen based on the presence or absence of branch vessels; and obtaining the reference blood flow velocity corresponding to each cross-section lumen based on the reference blood flow of the main vessel, the hydraulic equivalent area corresponding to each cross-section lumen, and the reference blood flow at the location of each cross-section lumen.

[0098] In practice, if there are branch vessels at the lumen of the i-th cross section, the diameter of the branch vessels and the diameter of the main vessel at the lumen of the i-th cross section will be measured.

[0099] like Figure 5 As shown, IVOCT imaging can detect the branching status of coronary arteries. It searches for the presence of branch vessels in a segment of coronary artery imaging cross-section sequence obtained by IVOCT, uses an AI neural network to measure the pixel length of the diameter of the branch vessels and the main vessels, and combines the imaging depth and the number of sampling points of the IVOCT system to calculate the physical diameter of the main vessels and branch vessels. The location of the branch vessels is recorded, and the reference blood flow of the target coronary artery is based on the main vessels and branch vessels.

[0100] Method 1: Set the parameters according to the following formula to obtain the reference blood flow velocity for each cross-section of the lumen:

[0101] u i =Q i / A i Q1=Q

[0102] When there are no branch vessels at the i-th cross-section lumen: Q i =Q i-1

[0103] When a branch vessel exists at the lumen of the i-th cross-section: Q i =Q i-1 ·(d i ') 3 / ((d i ') 3 +(d bi ) 3 )

[0104] Q=u·(Σ i=1 n A refi ') / n

[0105] A refi '=A refi +(max(A j -A refj ,n')) / n',j=1,2,…,n

[0106] A refi =A ref1 β i-1 ;β=200(Σ i=1 n A i ) / n+0.995

[0107] Where i represents the index; u i Q represents the reference blood flow velocity corresponding to the lumen at the i-th cross-section; i and Q i-1 A represents the reference blood flow rate at the location of the lumen in the i-th and (i-1)-th cross-sections, respectively; i d represents the hydraulically equivalent area corresponding to the lumen at the i-th cross-section; Q1 represents the reference blood flow rate at the location of the lumen at the 1-th cross-section; Q represents the reference blood flow rate of the main trunk of the target coronary artery; d i ' represents the diameter of the main blood vessel at the i-th cross-section; d bi represents the diameter of the branch vessel at the lumen of the i-th cross-section; Q represents the reference blood flow rate of the main coronary artery; u represents the clinically experienced blood flow velocity at rest; A refi ' represents the final reference cavity area of ​​the i-th cross-section; A refi Let represent the initial reference lumen area of ​​the i-th cross-section; j represents the index; n represents the total number of cross-section lumens; max(A j -A refj (n') represents the vector A j -A refjA function that selects the first n' values ​​after descending order; in specific implementation, n' <n;A j A represents the hydraulically equivalent area of ​​the j-th cross-section of the pipe cavity; refj A represents the initial reference cavity area of ​​the j-th cross-section cavity; ref1 β represents the initial reference lumen area of ​​the first cross-section lumen; i-1 β represents β raised to the power of i-1; β is the taper of the lumen of the main blood vessel. In this embodiment, the clinically empirical blood flow velocity u is set to 20 cm / s.

[0108] Method two is as follows:

[0109] D1. Construct a sequence of cross-sectional lumens of the target coronary artery according to the arrangement order of the cross-sectional lumens on the target coronary artery. Using each cross-sectional lumen as the center of the sliding window, slide the cross-sectional lumen sequence from beginning to end according to the preset sliding window size, with a sliding step size of one. Obtain the initial reference lumen area of ​​the cross-sectional lumen sequence within each sliding window, and obtain the initial reference lumen area corresponding to the cross-sectional lumen at the center of the sliding window. Thus, the initial reference lumen area corresponding to each cross-sectional lumen sequence in the target coronary artery is finally obtained.

[0110] The initial reference cavity area acquisition process is as follows: Calculate the N% quantile of the hydraulic equivalent area corresponding to all cross-section cavities within the sliding window; select cross-section cavities with a hydraulic equivalent area greater than the N% quantile as candidate normal cross-section cavities; calculate the M% quantile of the hydraulic equivalent area corresponding to all candidate normal cross-section cavities; and use the hydraulic equivalent area of ​​the candidate normal cross-section cavities corresponding to the M% quantile as the initial reference cavity area corresponding to the cross-section cavity at the center of the current sliding window. In practice, the quantile values ​​are sorted from smallest to largest by default.

[0111] D2. According to the arrangement order of the cross-sectional cavities, the final reference cavity area is obtained for each cross-sectional cavity in sequence, so as to obtain the final reference cavity area corresponding to all cross-sectional cavities.

[0112] The specific process for obtaining the final reference cavity area is as follows: obtain the slope corresponding to the current cross-section cavity, adjust the initial reference cavity area corresponding to all cross-section cavities according to the slope corresponding to the current cross-section cavity, and use the adjusted initial reference cavity area as the new initial reference cavity area of ​​the corresponding cross-section cavity. The final reference cavity area corresponding to the current cross-section cavity is obtained based on the new initial reference cavity area corresponding to all cross-section cavities and the hydraulic equivalent area corresponding to all cross-section cavities.

[0113] The final reference lumen area is obtained by setting the following formula to get the final reference lumen area corresponding to the current i-th cross-section:

[0114] When Aref1 <7mm 2 Time: A refi '=A refi ''+(7–A ref1 '')

[0115] When A ref1 ≥7mm 2 Time: A refi '=A refi ''

[0116] A refi ''=A refi (1) –(min(A refj (1) –A j ,n'')) / n'',j=1,2,…,n

[0117] When K Ai >0:

[0118] A reft (1) =A reft +K Ai ΔL, t=1,2,…,i-1; A reft (1) =A reft t=i,i+1…,n

[0119] When -0.03 ≤ K Ai When ≤0:

[0120] A reft (1) =A reft t=1,2,…,n

[0121] When K Ai When <-0.03:

[0122] A reft (1) =A reft , t=1,2,…,i-1; A reft (1) =A reft +ΔL(-0.03–K Ai ), t=i,i+1…,n

[0123] K A1 =0;K Ai =(A refi –A refi-1 ) / ΔL

[0124] Where i, j, and t all represent indices; A refi' represents the final reference cavity area corresponding to the i-th cross-section cavity; A ref1 '' and A refi '' represents the intermediate reference cavity area corresponding to the 1st and i-th cross-section cavities, respectively; A refi (1) A refj (1) and A reft (1) Let $\mathbf{i}$, $\mathbf{j}$, and $\mathbf{t}$ represent the adjusted initial reference cavity areas corresponding to the ith, j, and t-th cross-section cavities, respectively; $\mathbf{mean()}$ represents the average value; $\mathbf{n}$ represents the total number of cross-section cavities; $A$ j Min(A) represents the hydraulically equivalent area of ​​the j-th cross-section of the pipe. refj (1) –A j ,n'') represents the vector A refj (1) –A j A function that selects the first n'' values ​​after ascending order; in specific implementation, n'' <n;K A1 and K Ai K represents the slopes corresponding to the 1st and 1st cross-sections of the tube, respectively; A1 and K Ai A represents the slope corresponding to the 1st and 1st cross-sections of the tube, respectively; reft A refi and A refi-1 ΔL represents the initial reference cavity area corresponding to the t-th, i-th, and i-1-th cross-section cavities, respectively; ΔL is the preset thickness of the frustum-shaped cross-section micro-element; n represents the total number of cross-section cavities.

[0125] In specific implementation, when the final reference cavity area is obtained for each cross-section cavity in the order of arrangement of cross-section cavities: the adjusted initial reference cavity area is used as the new initial reference cavity area for the corresponding cross-section cavity, and is used when the final reference cavity area is obtained for the next cross-section cavity.

[0126] D3. Based on the final reference lumen area corresponding to all cross-section lumens, the reference blood flow velocity is obtained sequentially for each cross-section lumen according to the arrangement order of the cross-section lumens, thereby obtaining the reference blood flow velocity corresponding to each cross-section lumen.

[0127] The reference blood flow velocity acquisition process is as follows: Based on the three-dimensional image of the target coronary artery, a branch detection algorithm based on a neural network is used to determine whether there are branch vessels in the current cross-section lumen. Based on the presence of branch vessels and the final reference lumen area corresponding to all cross-section lumens, the adjusted final reference lumen area corresponding to the current and subsequent cross-section lumens is obtained, and the adjusted final reference lumen area is used as the new final reference lumen area for the corresponding cross-section lumen. Based on the hydraulic equivalent area corresponding to the current cross-section lumen and the corresponding new final reference lumen area, the reference blood flow velocity corresponding to the current cross-section lumen is obtained.

[0128] The reference blood flow velocity is obtained by setting the following formula to get the reference blood flow velocity corresponding to the lumen at the current i-th cross-section:

[0129] u i =u'A refi '''' / A i

[0130] When there are no branch vessels at the current i-th cross-section lumen:

[0131] A reft ''''=A reft ',t=i,i+1,…,n

[0132] When there is a branch vessel at the lumen of the current i-th cross-section:

[0133] A reft ''''=A reft '–(A refi-1 '-((A refi-1 ') 3 / 2 -(A bi ) 3 / 2 ) 2 / 3 ), t=i,i+1,…,n

[0134] A bi =4π(A rbi ) 2 / (C rbi ) 2

[0135] Where i and t are both indices; u i U is the reference blood flow velocity corresponding to the lumen at the i-th cross-section; u' is the clinically experienced blood flow velocity under congestion; A i Let A be the hydraulically equivalent area of ​​the i-th cross-section of the pipe. refi '''' and A reft '''' represents the final reference cavity area after adjustment for the i-th and t-th cross-section cavities, respectively; A reft 'and A refi-1'A' represents the final reference cavity area corresponding to the t-th and (i-1)-th cross-sections, respectively; bi Let A be the hydraulically equivalent area of ​​the branch vessel at the i-th cross-section lumen; rbi C represents the actual area of ​​the branch vessel at the lumen of the i-th cross-section; rbi This represents the actual circumference of the branch vessel at the i-th cross-section.

[0136] In practice, when the reference blood flow velocity is obtained for each cross-section lumen in the order of the cross-section lumen arrangement, the adjusted final reference lumen area is used as the new final reference lumen area for the corresponding cross-section lumen, and is used when obtaining the reference blood flow velocity for the next cross-section lumen.

[0137] S32. Obtain the friction loss and local pressure loss corresponding to the frustum-shaped cross-section micro-element based on the reference blood flow velocity, hydraulic equivalent area, and type of the frustum-shaped cross-section micro-element for each cross-section lumen.

[0138] The friction loss and local pressure loss corresponding to the frustum-shaped cross-section micro-element in step S32 are set according to the following formulas:

[0139] When the i-th frustum-shaped section element is a tapered section element:

[0140] dP fi =ρλ i (u i+1 ) 2 (1-(n c ) 2 ) / (16sin(θ i / 2));dP ji =(0.0348sin(θ i / 2)(1-(n c ) 2.5 (ξ) i ) 2 +(ξ i -1) 2 / 2)ρ(u i+1 ) 2

[0141] When the i-th frustum-shaped section element is an expanding section element:

[0142] dP fi =ρλ i (u i ) 2 (1-(n e ) 2 ) / (16sin(θ i / 2));dP ji =4.15(tan(θi / 2)) 2 (1-(n e )) 2 ρ(u i ) 2

[0143] λ i =16νπ 1 / 2 / (ρu i (A i ) 1 / 2 )+16νπ 1 / 2 / (ρu i+1 (A i+1 ) 1 / 2 )

[0144] n c =A i+1 / A i ;n e =A i / A i+1 ξ i =1+0.622(θ i / 180) 0.8 (1-0.215n c -0.785(n c ) 2.5 )

[0145] Where i represents the index; dP fi dP represents the friction loss along the path corresponding to the i-th frustum-shaped cross-section element; ji λ represents the local pressure loss corresponding to the i-th frustum-shaped cross-section element; ρ represents the blood density; i u represents the average friction loss coefficient corresponding to the i-th cross-section of the pipe. i and u i+1 These represent the reference blood flow velocities corresponding to the ith and (i+1)th cross-sections of the lumen, respectively; n c This represents the ratio of the hydraulically equivalent area of ​​the outlet and inlet of a truncated cone-shaped tapered section element; n e A represents the ratio of the hydraulically equivalent area of ​​the inlet and outlet of a frustum-shaped, gradually expanding cross-section element; i and A i+1 θ represents the hydraulically equivalent area corresponding to the i-th and (i+1)-th cross-sections of the pipe, respectively; i ξ represents the shrinkage angle of the i-th tapered section element or the expansion angle of the i-th expanding section element; i ν represents the velocity ratio of the i-th jet; ν represents the blood viscosity. In specific implementation, in the calculation of local pressure loss, n c and n e All are directly fixed at 0.7.

[0146] Furthermore, the following methods can be used to obtain the friction loss and local pressure loss corresponding to each frustum-shaped cross-section element:

[0147] A database was established containing a large number of historical reference blood flow velocities and their corresponding friction loss and local pressure loss for frustum-shaped cross-section elements at different angles. The friction loss coefficient was determined experimentally or through simulation fitting based on the reference blood flow velocities and corresponding friction losses in the database. Similarly, the local pressure loss coefficient was determined experimentally or through simulation fitting based on the reference blood flow velocities and corresponding local pressure losses in the database. Finally, the friction loss and local pressure loss corresponding to the frustum-shaped cross-section element were obtained based on the reference blood flow velocity, friction loss coefficient, and local pressure loss coefficient.

[0148] The friction loss and local pressure loss corresponding to the frustum-shaped cross-section micro-element are set according to the following formula:

[0149] dP fi =αu i dP ji =β(u i ) 2

[0150] Wherein dP fi dP represents the friction loss along the path corresponding to the i-th frustum-shaped cross-section element; ji u represents the local pressure loss corresponding to the i-th frustum-shaped cross-section element; i α represents the reference blood flow velocity corresponding to the lumen at the i-th cross section; α and β are the friction loss coefficient and the local pressure loss coefficient, respectively.

[0151] S33. The sum of the pressure loss along the friction and the local pressure loss corresponding to all the frustum-shaped cross-section micro-elements is the total pressure loss of the target coronary artery.

[0152] The total pressure loss of the target coronary artery is set according to the following formula:

[0153] ΔP=Σ(dP fi +dP ji )

[0154] Where ΔP represents the total pressure loss in the target coronary artery; dP fi and dP ji These represent the friction loss and local pressure loss corresponding to the i-th cross-section of the pipe, respectively.

[0155] S4. A method for calculating fractional flow reserve: The fractional flow reserve of the target coronary artery is obtained by combining the mean aortic pressure and the total pressure loss of the target coronary artery.

[0156] The fractional flow reserve of the target coronary artery is set according to the following formula:

[0157] FFR=(P a –ΔP) / P a

[0158] Where FFR represents the fractional flow reserve of the target coronary artery; P a Mean aortic pressure is represented by ΔP; ΔP represents the total pressure loss in the target coronary artery. Mean aortic pressure is obtained through non-invasive external measurement methods such as the radial or femoral artery.

[0159] like Figure 6 The figure shows a comparison between the fractional flow reserve calculated by the method of this embodiment and the fractional flow reserve calculated by the prior art.

[0160] The fractional flow reserve calculation system of this embodiment includes:

[0161] The lumen segmentation module is used to acquire three-dimensional images of the target coronary artery and extract the true area and perimeter of the segmented lumen based on the three-dimensional images.

[0162] The cross-section micro-element construction module is used to construct frustum-shaped cross-section micro-elements and obtain the hydraulic equivalent area corresponding to each cross-section cavity after resampling.

[0163] The fluid dynamics calculation module is used to obtain the pressure loss along the friction and the local pressure loss of each frustum-shaped cross-section micro-element, and then obtain the total pressure loss of the target coronary artery.

[0164] The fractional flow reserve (FVR) calculation module is used to obtain the FVR of the target coronary artery.

[0165] This invention solves the problem of inaccurate calculation of pressure difference in diffuse lesions; it does not rely on pressure guidewires and congestion drugs, and can be completed solely through three-dimensional coronary artery imaging and external blood pressure measurement. It is simple to operate, time-saving, and can meet the actual clinical decision-making needs.

[0166] The above embodiments are merely preferred embodiments provided to fully illustrate the present invention, and the scope of protection of the present invention is not limited thereto. Equivalent substitutions or modifications made by those skilled in the art based on the present invention are all within the scope of protection of the present invention. The scope of protection of the present invention is defined by the claims.

Claims

1. A method for calculating fractional blood flow reserve based on a frustum-shaped cross-section micro-element, characterized in that, include: S1. Lumen Segmentation: Based on the three-dimensional image of the target coronary artery, the main vessels in the target coronary artery are segmented along the cross-section, and the true area and perimeter of the lumen of each segmented section are extracted. S2, Construction of cross-sectional micro-element: The whole between two adjacent cross-sectional cavities is regarded as a frustum-shaped cross-sectional micro-element. The actual area and perimeter of the cross-sectional cavity obtained in step S1 are resampled according to the preset thickness of the frustum-shaped cross-sectional micro-element. Based on the resampled actual area and perimeter, the hydraulic equivalent area corresponding to each cross-sectional cavity is obtained. S3. Fluid dynamics calculation: Obtain the reference blood flow velocity based on the hydraulic equivalent area corresponding to the cross-section lumen. Based on the structure of the frustum-shaped cross-section micro-element and according to the hydraulic equivalent area corresponding to the cross-section lumen and the reference blood flow velocity, obtain the pressure loss along the friction and the local pressure loss of each frustum-shaped cross-section micro-element, and then obtain the total pressure loss of the target coronary artery. Step S3 specifically involves: S31. Based on the hydraulic equivalent area of ​​all cross-section lumens, obtain the reference blood flow velocity corresponding to each cross-section lumen using either Method 1 or Method 2. S32. Obtain the friction loss and local pressure loss corresponding to the frustum-shaped cross-section micro-element based on the reference blood flow velocity, hydraulic equivalent area, and type of the frustum-shaped cross-section micro-element for each cross-section lumen. S33. The sum of the friction loss and local pressure loss corresponding to all frustum-shaped cross-section micro-elements is the total pressure loss of the target coronary artery. S4. Calculation of fractional flow reserve: The fractional flow reserve of the target coronary artery is obtained by combining the mean aortic pressure and the total pressure loss of the target coronary artery.

2. The method for calculating fractional blood flow reserve based on a frustum-shaped cross-section micro-element according to claim 1, characterized in that, Step S1 specifically involves: S11. Acquire a three-dimensional image of the target coronary artery; the acquisition of the three-dimensional image of the target coronary artery adopts one of the following medical imaging techniques: intravascular ultrasound, intravascular optical coherence tomography, and computed tomography. S12. Based on the three-dimensional image, a lumen segmentation algorithm based on image processing or neural network is used to segment the main blood vessel in the target coronary artery along the cross section to obtain the contour data of the lumen of each segmented section. S13. Based on the contour data of the cross-sectional cavity, the pixel area and perimeter of the cross-sectional cavity are statistically analyzed. Each pixel is converted to the corresponding actual physical length, thereby obtaining the real area and perimeter of each cross-sectional cavity.

3. The method for calculating fractional blood flow reserve based on a frustum-shaped cross-section micro-element according to claim 1, characterized in that: The frustum-shaped cross-section micro-element is divided into a contracting cross-section micro-element and a expanding cross-section micro-element; the contracting cross-section micro-element has an inlet hydraulic equivalent area greater than or equal to the outlet hydraulic equivalent area; the expanding cross-section micro-element has an inlet hydraulic equivalent area less than the outlet hydraulic equivalent area; the inlet hydraulic equivalent area of ​​the frustum-shaped cross-section micro-element is the hydraulic equivalent area corresponding to the previous cross-section lumen along the blood flow direction; the outlet hydraulic equivalent area of ​​the frustum-shaped cross-section micro-element is the hydraulic equivalent area corresponding to the next cross-section lumen along the blood flow direction. The hydraulically equivalent area corresponding to the cross-sectional cavity is set according to the following formula: A i =πd i 2 / 4;d i =4A ri / C ri Where i represents the index; A i d represents the hydraulically equivalent area of ​​the i-th cross-section of the pipe; i A represents the hydraulic diameter of the i-th cross-section of the pipe. ri C represents the actual area corresponding to the i-th cross-section of the tube; ri This represents the actual perimeter of the lumen corresponding to the i-th cross-section.

4. The method for calculating fractional blood flow reserve based on a frustum-shaped cross-section micro-element according to claim 1, characterized in that, The first method is specifically as follows: The area of ​​the lumen of a normal coronary artery is selected as the initial reference lumen area of ​​the first cross-section of the main vessel in the target coronary artery. The reference blood flow of the main vessel is obtained based on the reference area of ​​the first cross-section lumen and the hydraulic equivalent area of ​​all cross-section lumens. The presence of branch vessels in the target coronary artery is determined by a branch detection algorithm based on a neural network using the three-dimensional image of the target coronary artery. The reference blood flow at the location of each cross-section lumen is obtained based on the presence of branch vessels. The reference blood flow velocity for each cross-section lumen is obtained based on the reference blood flow of the main vessel, the hydraulic equivalent area of ​​each cross-section lumen, and the reference blood flow at the location.

5. The method for calculating fractional blood flow reserve based on a frustum-shaped cross-section micro-element according to claim 1, characterized in that: Method 1 is set according to the following formula to obtain the reference blood flow velocity corresponding to each cross-section lumen: u i =Q i / A i ;Q1=Q When there are no branch vessels at the i-th cross-section lumen: Q i =Q i-1 When a branch vessel exists at the lumen of the i-th cross-section: Q i =Q i-1 ·(d i ') 3 / ((d i ') 3 +(d bi ) 3 ) Q=u·(Σ i=1 n A refi ') / n A refi '=A refi +(max(A j -A refj ,n')) / n',j=1,2,…,n A refi =A ref1 b i-1 ;β=200(Σ i=1 n A i ) / n+0.995 Where i represents the index; u i Q represents the reference blood flow velocity corresponding to the lumen at the i-th cross-section; i and Q i-1 A represents the reference blood flow rate at the location of the lumen in the i-th and (i-1)-th cross-sections, respectively; i d represents the hydraulically equivalent area corresponding to the lumen at the i-th cross-section; Q1 represents the reference blood flow rate at the location of the lumen at the 1-th cross-section; Q represents the reference blood flow rate of the main trunk of the target coronary artery; d i ' represents the diameter of the main blood vessel at the i-th cross-section; d bi represents the diameter of the branch vessel at the lumen of the i-th cross-section; Q represents the reference blood flow rate of the main coronary artery; u represents the clinically experienced blood flow velocity at rest; A refi ' represents the final reference cavity area of ​​the i-th cross-section; A refi Let represent the initial reference lumen area of ​​the i-th cross-section; j represents the index; n represents the total number of cross-section lumens; max(A j -A refj (n') represents the vector A j -A refj A function that selects the first n' values ​​after descending order; A j A represents the hydraulically equivalent area of ​​the j-th cross-section of the pipe cavity; refj A represents the initial reference lumen area of ​​the j-th cross-section; ref1 β represents the initial reference lumen area of ​​the first cross-section; β is the lumen taper.

6. The method for calculating fractional blood flow reserve based on a frustum-shaped cross-section micro-element according to claim 1, characterized in that, The second method is specifically as follows: D1. Construct a sequence of cross-sectional cavities in the order of arrangement of all cross-sectional cavities. Take each cross-sectional cavity as the center of the sliding window and slide the sequence of cross-sectional cavities according to the preset sliding window size. The sliding step size is one. The initial reference cavity area of ​​the cross-sectional cavity sequence in each sliding window is obtained. The initial reference cavity area corresponding to the cross-sectional cavity at the center of the sliding window is obtained. Thus, the initial reference cavity area corresponding to each cross-sectional cavity sequence is finally obtained. The specific process for obtaining the initial reference cavity area is as follows: calculate the N% quantile of the hydraulic equivalent area corresponding to all cross-section cavities within the sliding window, and take the cross-section cavities with a hydraulic equivalent area greater than the N% quantile as candidate normal cross-section cavities. Calculate the M% quantile of the hydraulic equivalent area corresponding to all candidate normal cross-section cavities, and take the hydraulic equivalent area of ​​the candidate normal cross-section cavities corresponding to the M% quantile as the initial reference cavity area corresponding to the cross-section cavities at the center of the current sliding window. D2. According to the arrangement order of the cross-section cavities, the final reference cavity area is obtained for each cross-section cavity in sequence, so as to obtain the final reference cavity area corresponding to all cross-section cavities. D3. Based on the final reference lumen area corresponding to all cross-section lumens, the reference blood flow velocity is obtained sequentially for each cross-section lumen according to the arrangement order of the cross-section lumens, thereby obtaining the reference blood flow velocity corresponding to each cross-section lumen.

7. The method for calculating fractional blood flow reserve based on a frustum-shaped cross-section micro-element according to claim 6, characterized in that: The final reference cavity area acquisition process is set according to the following formula to obtain the final reference cavity area corresponding to the current i-th cross-section cavity: When A ref1 ’’<7mm 2 : A refi ’ = A refi ’’ + (7 – A ref1 ’’) This A ref1 ≥7mm 2 Time: A refi '=A refi `` HAS refi ''=A refi (1) –min(A refj (1) -HAS j ,n'') / n'',j=1,2,…,n When K Ai >0: TO reft (1) =A reft +K Ai ΔL, t=1,2,…,i-1; A reft (1) =A reft ,t=i,i+1…,n When -0.03 ≤ K Ai When ≤0: A reft (1) =A reft ,t=1,2,…,n When K Ai When <-0.03: TO reft (1) =A reft ,t=1,2,…,i-1;A reft (1) =A reft +ΔL(-0.03–K Ai ),t=i,i+1…,n K A1 =0;K Ai =(A refi -TO refi-1 ) / ΔL Where i, j, and t all represent indices; A refi ' represents the final reference cavity area corresponding to the i-th cross-section cavity; A ref1 '' and A refi '' represents the intermediate reference cavity area corresponding to the 1st and i-th cross-section cavities, respectively; A refi (1) A refj (1) and A reft (1) Let A represent the adjusted initial reference cavity areas corresponding to the i-th, j-th, and t-th cross-section cavities, respectively; n represents the total number of cross-section cavities; A j Min(A) represents the hydraulically equivalent area of ​​the j-th cross-section of the pipe. refj (1) –A j ,n'') represents the vector A refj (1) –A j A function that selects the first n'' values ​​after ascending order; K A1 and K Ai K represents the slopes corresponding to the 1st and 1st cross-sections of the tube, respectively; A1 and K Ai A represents the slope corresponding to the 1st and 1st cross-sections of the tube, respectively; reft A refi and A refi-1 ΔL represents the initial reference cavity area corresponding to the t-th, i-th, and i-1-th cross-section cavities, respectively; ΔL is the preset thickness of the frustum-shaped cross-section micro-element; n represents the total number of cross-section cavities; The reference blood flow velocity acquisition process is set according to the following formula to obtain the reference blood flow velocity corresponding to the current i-th cross-section lumen: u i =u’A refi ’’’’ / A i When there are no branch vessels at the current i-th cross-section lumen: A reft ’’’’=A reft ’,t=i,i+1,…,n When there is a branch vessel at the lumen of the current i-th cross-section: A reft ’’’’=A reft ’–(A refi-1 ’-((A refi-1 ’) 3 / 2 -(A bi ) 3 / 2 ) 2 / 3 ),t=i,i+1,…,n A bi =4πA rbi 2 / C rbi 2 Where i and t are both indices; u i U is the reference blood flow velocity corresponding to the lumen at the i-th cross-section; u' is the clinically experienced blood flow velocity under congestion; A i Let A be the hydraulically equivalent area of ​​the i-th cross-section of the pipe. refi '''' and A reft '''' represents the final reference cavity area after adjustment for the i-th and t-th cross-section cavities, respectively; A reft 'and A refi-1 'A' represents the final reference cavity area corresponding to the t-th and (i-1)-th cross-sections, respectively; bi Let A be the hydraulically equivalent area of ​​the branch vessel at the i-th cross-section lumen; rbi C represents the actual area of ​​the branch vessel at the lumen of the i-th cross-section; rbi This represents the actual circumference of the branch vessel at the i-th cross-section.

8. The method for calculating fractional blood flow reserve based on a frustum-shaped cross-section micro-element according to claim 1, characterized in that: The friction loss and local pressure loss corresponding to the frustum-shaped cross-section micro-element in step S32 are set according to the following formula: When the i-th frustum-shaped section element is a tapered section element: dP fi =rλ i you i+1 2 (1-n c 2 ) / (16sin(θ i / 2));dP ji =(0.0348sin(θ i / 2)(1-n c 2.5 )ξ i 2 +(ξ i -1) 2 / 2)ru i+1 2 When the i-th frustum-shaped section element is an expanding section element: dP fi =rλ i you i 2 (1-n e 2 ) / (16sin(θ i / 2));dP ji =4.15(tan(θ i / 2)) 2 (1-n e ) 2 ru i 2 l i =16np 1 / 2 / (ru i A i 1 / 2 )+16np 1 / 2 / (ru i+1 A i+1 1 / 2 ) n c =A i+1 / A i ;n e =A i / A i+1 ;x i =1+0.622(θ i / 180) 0.8 (1-0.215n c -0.785n c 2.5 ) Where i represents the index; dP fi dP represents the friction loss along the path corresponding to the i-th frustum-shaped cross-section element; ji λ represents the local pressure loss corresponding to the i-th frustum-shaped cross-section element; ρ represents the blood density; i u represents the average friction loss coefficient corresponding to the i-th cross-section of the pipe. i and u i+1 These represent the reference blood flow velocities corresponding to the ith and (i+1)th cross-sections of the lumen, respectively; n c This represents the ratio of the hydraulically equivalent area of ​​the outlet and inlet of a truncated cone-shaped tapered section element; n e A represents the ratio of the hydraulically equivalent area of ​​the inlet and outlet of a frustum-shaped, gradually expanding cross-section element; i and A i+1 θ represents the hydraulically equivalent area corresponding to the i-th and (i+1)-th cross-sections of the pipe, respectively; i ξ represents the shrinking angle of a tapered cross-section element or the expanding angle of a dilatant cross-section element; i ν represents the jet velocity ratio; ν represents blood viscosity.

9. A system for calculating fractional flow reserve suitable for the method described in any one of claims 1-8, characterized in that, include: The lumen segmentation module is used to acquire three-dimensional images of the target coronary artery and extract the true area and perimeter of the segmented lumen based on the three-dimensional images. The cross-section micro-element construction module is used to construct frustum-shaped cross-section micro-elements and obtain the hydraulic equivalent area corresponding to each cross-section cavity after resampling; The fluid dynamics calculation module is used to obtain the pressure loss along the friction and local pressure loss of each frustum-shaped cross-section micro-element, and then obtain the total pressure loss of the target coronary artery. The fractional flow reserve (FVR) calculation module is used to obtain the FVR of the target coronary artery.

Images